1 Introduction

Analgesia and pain management are essential components of cancer treatment, and there is growing evidence that pain can promote cancer progression and metastasis [1]. Opioids are commonly used for pain control due to their ability to modulate pain perception in the central and peripheral nervous systems. As one of the three pillars of anesthesia (alongside sedation and muscle relaxation), opioids are an integral part of clinical analgesia and anesthesia. They also influence cellular and humoral immune responses, as well as the expression of chemokines and chemokine receptors [2]. In current clinical practice, opioids are the most commonly prescribed analgesics during the perioperative period of cancer surgery, and they are often administered to patients with moderate to severe pain associated with advanced cancer or treatment-related complications [3].

To exert their effects, opioids must bind to opioid receptors, which are present on both neuronal and non-neuronal cells. The opioid system has been extensively studied for its roles in pain, reward, and addiction. Currently, there are four known ORs: μ opioid receptor (MOR, encoded by the OPRM1), δ opioid receptor (DOR, encoded by the OPRD1), κ opioid receptor (KOR, encoded by the OPRK1), and the nociception/orphanin FQ peptide receptor (NOPR, encoded by the OPRL1) [4, 5].

The effect of opioids on cancer progression is still a matter of debate. Several experimental studies have suggested that morphine promotes proliferation, migration, and invasion in human breast cancer cells [6], clear cell renal cell carcinoma cells [7] and colorectal cancer cells [8], as well as stimulated angiogenesis in mouse breast cancer models [9, 10]. One small clinical cohort for naltrexone provided a proof of principle that opioid receptor antagonists have the potential to improve outcomes in breast cancer [11]. Preclinical studies also suggest that opioids may be associated with tumor progression and recurrence, which can have negative prognostic implications for cancer patients [12, 13]. A retrospective study found that opioid requirement remained an independent predictor of shorter survival in non-small cell lung cancer [14]. However, other studies have arrived at opposite conclusions, with some showing that opioids may even inhibit several cancer cell lines and animal models. For example, morphine significantly decreased the adhesion, invasion, and metastasis of metastatic colon cancer cells [15]; while fentanyl inhibited cell viability and invasion in lung [16] and gastric cancer [17]. Heterogeneity between existing studies, including differences in type of opioid, dose, and route of administration, different tumor cell lines and cancer models used may account for the varied effects of opioids. Also, there is a lack of guidance from the clinical trials on opioids and disease progression [18].

Studies have found that opioid receptors and peptides are widely distributed in various types of tumor cells and the expression levels differ according to the tumor tissue type [19]. Meanwhile, ORs can influence the release and function of cytokines and chemokines [20]. Some studies suggest that ORs may contribute to cancer metastasis since they are generally upregulated in metastatic samples compared to non-metastatic ones [21], and the ORs expression levels correlate to reduced overall survival and progression free survival in several cancers [22, 23]. There is also a connection between ORs and tumor immunity, although the available research evidence is insufficient. The tumor microenvironment (TME), which exerts a significant impact on cancer progression, constitutes a complex system consisting of cancer cells, endothelial and stromal cells, leukocytes, fibroblasts, and neurons that release soluble mediators triggering neuroinflammation [24]. The cellular components of the TME express MOR, which initiates signaling events that induce neuroinflammation in the presence of morphine. The activation of MOR not only affects pain and the quality of life for cancer patients, but also directly and/or indirectly influences tumor growth and metastasis [20]. Opioids may modify the infiltration of immune cells into the TME [25]. In comparison to MOR-knockout mice, long-term administration of the same morphine to wild-type mice led to the atrophy of lymphoid organs and a reduced ratio of CD4+/CD8+ T cells in the thymus [26]. Additionally, MOR affects immunity through the hypothalamic-pituitary-adrenal axis, resulting in adrenocortical dysfunction and immunosuppression [27]. In this way, the overall effect of opioid agonism on cancer is far from trivial.

Contemporary research predominantly concentrates on μ or δ receptors in specific cancers, whereas the effects and functions of KOR and NOPR remain unclear in tumorigenesis. Additionally, the variances in expression levels, distribution characteristics, and connections with tumor immunity of the four ORs across diverse cancers have not been elucidated. Therefore, comprehending the overall perspective of the four ORs and their association with tumor development is crucial for providing anesthesia to surgical oncology patients and administering drugs to tumor patients. This study aims to acquire a comprehensive and systematically structured understanding of the correlation between the four ORs and the prognosis of different tumors. It primarily highlights the potential influence of the ORs on the tumor immune microenvironment (TIM). Furthermore, through this study, we have tentatively identified that NOPR, which has received less attention previously, may play a pivotal role in regulating the TIM and may represent a novel therapeutic target for tumors.

2 Methods

2.1 Expression and distribution of ORs in normal tissues and immune cells

The expression levels of four ORs in different normal tissues and immune cells derived from peripheral blood were acquired from three datasets: the HPA RNA-seq dataset [28, 29] (https://www.proteinatlas.org/), the Monaco dataset [30], and the Consensus dataset [28, 29, 31]. A total of 51 human tissues and 18 blood cell types were analyzed by RNA-seq to estimate the transcript abundance of each protein-coding gene in the HPA dataset. For the Monaco dataset, 29 immune cell types of healthy donors were characterized using RNA-seq and flow cytometry. The Consensus dataset comprised normalized expression levels for 55 tissue types, generated by merging the HPA and the Genotype-Tissue Expression dataset using the internal normalization pipeline [31]. All data was downloaded from the websites of the HPA and the Monaco dataset.

2.2 Expression of four ORs in cancer and paired adjacent tissues

To investigate the expression levels and variations of four ORs across various cancer types, a pan-cancer analysis of OR expression levels was conducted using TCGA dataset (https://portal.gdc.cancer.gov/). TCGA is a pioneering program in cancer genomics that encompasses over 20,000 primary cancer and matched normal samples spanning 33 cancer types, and has been extensively utilized for tumor research. Additionally, graph visualization was performed using TIMER 2.0 (http://timer.comp-genomics.org/) [32] for graph drawing. The detailed full names of different cancers are supplied in the Supplementary Table 1.

2.3 sCNA and gene mutation of four ORs across TCGA

As two important gene structure variations, the sCNA and gene mutation data were acquired and collected from the TCGA dataset and graph drawing was completed by using the tool TIMER 2.0. Increasing evidence has shown that copy number alteration (CNA) plays an important role in the pathogenesis and prognosis of cancers [33], and we mainly compared five types of CNA in cancer: deep deletion, arm-level deletion, diploid/normal, arm-level gain and high amplication [34]. Similarly, gene mutation levels reflect cancer genomic stability and gene function deletion [33].

2.4 Survival analyses of four ORs in different types of cancers

A pan-cancer survival analysis was performed to explore effects of four ORs on the overall survival (OS) of patients with tumor using the Kaplan-Meier survival analysis, and all survival analyses were completed by using the tool Kaplan-Meier Plotter (http://kmplot.com/analysis/) [35]. The optimal cut-off value of four ORs genes were selected and the Log-rank test was used for comparing differences between low/high expression groups.

2.5 Co-expression genes analysis and functional enrichment analyses

The co-expression genes of ORs were obtained from the cBio Cancer Genomics Portal (cBioPortal, http://cbioportal.org) [36]. The cBioPortal website is a publicly accessible platform for interactive exploration of complex cancer genomics datasets. It offers access to data from over 30,000 tumor samples derived from 334 distinct cancer studies. Furthermore, specific cancer types which exhibited statistically significant survival differences were chosen for co-expression gene analyses. Specifically, the TCGA dataset (sourced from GDAC Firehose) [37] was utilized for identification and analysis of co-expression genes associated with these cancers.

To investigate the biological functions of ORs in various cancers, Gene Ontology (GO) enrichment analyses and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were conducted using ORs and their co-expression genes. The 300 most highly correlated co-expression genes were selected for functional enrichment analyses based on Spearman's correlation score, and the top 20 enriched pathways were visually represented in figures. GO can be accessed at http://geneontology.org/, and KEGG at https://www.kegg.jp/. All functional enrichment analyses were performed using the Sangerbox tools, a free online platform for data analysis (http://www.sangerbox.com/tool).

2.6 Correlations between ORs and immune cell infiltration levels

The correlations between ORs expression and immune cell infiltration levels were analyzed by the Tumor IMmune Estimation Resource (TIMER 1.0) web server (https://cistrome.shinyapps.io/timer/) [32, 38]. TIMER 1.0 is a comprehensive resource that facilitates systematic analysis of immune infiltrates across diverse cancer types. The abundance of six immune infiltrates (B cells, CD4+ T cells, CD8+ T cells, Neutrophils, Macrophages, and Dendritic cells) are estimated by the TIMER algorithm. For further analysis of immune cell infiltration, specific cancers exhibiting statistically significant survival disparities were selected.

2.7 Expression distributions of ORs in TME

To investigate the expression patterns of ORs within the TME, the authors utilized a tool called the Tumor Immune Single-cell Hub (TISCH) that facilitates exploration of TME across different single-cell transcriptomic analyses (http://tisch.comp-genomics.org/home/) [39]. TISCH enables detailed cell-type annotation at the single-cell level, providing insights into TME composition in various cancer types. Four datasets were chosen for analysis: BLCA_GSE145281, LIHC_GSE140228, BRCA_GSE114727, and PAAD_GSE111672.

2.8 Cell culture and drugs treatments

In total 18 cell lines were used for fundamental experiments: breast cancer (BRCA) cell lines, T47D and MDA-MB-231; bladder cancer (BLCA) cell lines,5637 and T24; Liver hepatocellular carcinoma (LIHC) cell lines, HCCLM3, MHCC-97H, MHCC-97L, HepG2, PLC, Huh-7 and Hep3B; Kidney renal clear cell carcinoma (KIRC) cell line, OS-RC-2; Kidney renal papillary cell carcinoma (KIRP) cell line, ACHN; Uterine Corpus Endometrial Carcinoma (UCEC) cell line, AN3CA; Pancreatic adenocarcinoma (PAAD) cell line, PANC-1; Monocyte cell line, THP-1; T cell line, Jurkat; Normal hepatic cell line, LO-2. All cell lines were purchased from the FuHeng Cell Center (Shanghai, China) and authenticated by short tandem repeat tests before cell culture. All cell lines were cultured in high glucose Dulbecco’s modified Eagle’s medium (DMEM; Gibco BRL, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; Gibco) except AN3CA. AN3CA was cultured with McCOY’s 5A medium (Genomcell.bio, Zhejiang, China) supplemented with 10% FBS (Gibco). Also, 50 U/ml penicillin and 50 µg/ml streptomycin (Yeasen, Shanghai, China) were added into the medium regularly. Cells were culture in a 37 ℃ humidified incubator containing 5% CO2. When cells reached 80%–90% confluency, they were detached with trypsin–EDTA (0.25%) (Yeasen) and used for subsequent experiments.

To test effects of ORs on cell proliferation, cell migration and invasion, specific agonists of ORs were used during cell culture: 1.[D-Ala2, N-Me-Phe4, Gly5-ol]-enkephalin (DAMGO), MOR agonist (Cat. No. ab120674; Abcam, Boston, MA, USA); 2. [D-Pen2,D-Pen5]-enkephalin, DOR agonist (Cat. No. ab120675; Abcam); 3. U69593, KOR agonist (Cat. No.ab141703; Abcam); 4. Nociceptin, NOPR agonist (Cat. No. ab120070; Abcam). The concentration of all four agonists for the experiments is 10 μM. DAMGO, DPDPE and nociceptin were diluted in phosphate buffered saline (PBS), and U69593 was diluted in dimethyl sulfoxide (DMSO) based on specifications. The same volume of PBS or DMSO were added to the control group. U69593 is a potent and selective KOR agonist, which is commonly used as a tool drug [40, 41].

2.9 Cell counting kit (CCK-8) assay

Cells were seeded into the 96-well plate at the density of 5 × 103 cells/well with 100 μl culture medium and corresponding agonists or not. After 48 h culture, 10 μl CCK-8 (Cat. No. CK04; Dojindo, Japan) was added into cells for 30 min incubation. Then, optical density (OD) at 450 nm was detected by the microplate reader (Bio-Rad, Hercules, CA, USA). The value of 450 nm OD represents the number of cells.

2.10 Transwell migration and invasion assays

Cell migration and invasion ability were detected by Transwell chambers (8-μm pore size, Corning, New York, USA) and Matrigel (only for invasion ability detection, Corning) according to previous protocols [42]. In migration assays, 1 × 105 cells/ml in 200 μl medium without FBS were added to the upper chamber. In invasion assays, the upper chambers were initially coated with 0.5 mg/L Matrigel for 30 min to form a membrane. The lower chambers contained 600 μl DMEM (with 10% FBS). Plates were incubated for 24 h, and then the chambers were fixed with 4% paraformaldehyde (Servicebio, Wuhan, China) for 20 min and stained with 0.5% crystal violet for 20 min. For both assays, three randomly selected fields per chamber were quantified using ImageJ 1.51 software to calculate averages reflecting the migration or invasion ability for each sample. All experiments were performed in triplicate. Digital images were captured using an Olympus fluorescent microscope equipped with bright field function.

2.11 RNA isolation and RT-qPCR analysis

Total RNA was isolated using EZ-press RNA Purification Kit (EZ Bioscience, USA) following the manufacturers’ protocol. Reverse transcription of 0.5 μg of total RNA was performed using the reverse kit (EZ Bioscience). 2 × SYBR Green qPCR Master Mix (EZ Bioscience, USA) was used for RT-qPCR, and the reaction conditions were as follows: pre-denaturation at 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 10 s and 60 °C for 30 s. The Ct values obtained from the samples were calculated using the 2−ΔΔCt method. The primers of genes were as follow (all human derived): OPRM1, forward 5’- GCCCTTCCAGAGTGTGAATTAC-3’, reverse 5’- GTGCAGAGGGTGAATATGCTG-3’; OPRD1, forward 5’- CGTCCGGTACACTAAGATGAAGA-3’, reverse 5’- GCCACGTCTCCATCAGGTA-3’; OPRK1, forward 5’- ATCATCACGGCGGTCTACTC -3’, reverse 5’- ACTCTGAAAGGGCATGGTTGTA -3’; OPRL1, forward 5’- TTCTGGGAGGTTATCTACGGC-3’, reverse 5’- GGATGACGTACATGACAAGGC-3’; GAPDH, forward 5’- GGAGCGAGATCCCTCCAAAAT-3’, reverse 5’- GGCTGTTGTCATACTTCTCATGG-3’.

2.12 Statistical analysis

Statistical analyses were conducted using the IBM SPSS Statistics 26.0 software (SPSS Inc., Armonk, New York, USA) and GraphPad Prism 7 software. All data were expressed as the mean ± standard deviation (SD), and differences were analyzed through independent samples t-test. P value < 0.05 is considered statistically significantly different.

3 Results

3.1 Expression and distribution heterogeneity of four ORs in different tissues and immune cells

Through the examination of ORs expression levels in diverse tissues from the Consensus dataset, it is indicated that ORs are predominantly expressed in the central nervous system, encompassing regions such as the cerebellum, basal ganglia, cerebral cortex, and amygdala (Fig. 1A). Furthermore, OPRM1 displays high expression levels in the testis, OPRD1 is highly expressed in skeletal muscle and granulocytes, OPRK1 is highly expressed in the prostate and testis, while OPRL1 exhibits high expression levels in monocytes, testis, granulocytes, and dendritic cells (Fig. 1A & Supplementary Fig. 1A).

Fig. 1
figure 1

Variations in expression and distribution patterns of four ORs across distinct tissues and immune cells. A: The Consensus dataset reveals that gene expression levels of the four ORs vary among different tissues, with ORs predominantly expressed in the central nervous system, while OPRL1 displays high expression specifically in monocytes. B: Analysis of immune cells in the Consensus dataset shows that the four ORs are widely expressed and distributed. The expression level of OPRL1 is significantly higher than that of the other three ORs, and it is mainly expressed in monocytes. ORs, Opioid receptors; OPRM1, opioid receptor mu 1 gene; OPRD1, opioid receptor delta 1gene; OPRK1, Opioid receptor kappa 1 gene; OPRL1, opioid-related nociceptin receptor 1 gene; PBMC, peripheral blood mononuclear cell; TPM, transcripts per million

Full size image

Subsequently, expression levels of four ORs across various immune cell types were comprehensively compared in the Consensus dataset, HPA dataset, and Monaco dataset (Fig. 1B & Supplementary Fig. 1B–C). The expression of ORs was found to be ubiquitous and present across several immune cell types. Furthermore, the findings indicate that OPRL1 was primarily expressed in monocytes, eosinophils, neutrophils, and myeloid dendritic cells (Fig. 1B & Supplementary Fig. 1B–C). The expressions of OPRL1 in classical monocyte, non-classical monocyte and intermediate monocyte were 3.7, 11.9 and 5.6 respectively, while expressions of OPRM1 were 0.1, 0 and 0 respectively, expressions of OPRK1 were 0.2, 0.1 and 0 respectively, expressions of OPRD1 were 0.5, 0.3 and 0.3 respectively.

3.2 Expression changes and gene structure variations of four ORs in pan-cancer

The expression levels of four ORs were analyzed between tumor and normal tissue samples across the TCGA dataset. Pan-cancer analysis revealed that OPRL1 could be detected across all types of cancers (Fig. 2). However, the expression levels of OPRM1 and OPRD1 were too low to be detected in multiple cancer types (Fig. 2). This analysis also showed significant changes in gene expression levels, with notable downregulation or upregulation of the four ORs observed across multiple cancers.

Fig. 2
figure 2

The expression levels of four ORs between tumor and normal samples across TCGA cancer dataset. Pan-cancer analysis revealed that OPRL1 exhibited a significantly higher abundance than the other three ORs across different types of cancers. In addition, the expression levels of OPRM1 and OPRD1 were undetectable in multiple cancer subtypes. Collectively, these four ORs displayed significant downregulation or upregulation across various types of cancers. All data was shown as mean ± SD, *P < 0.05, **P < 0.01, ***P < 0.001. TCGA, the cancer genome atlas; SD, standard deviation

Full size image

Furthermore, the sCNA and gene mutation levels of four ORs across TCGA cancer dataset were analyzed (Supplementary Fig. 2). The sCNA analysis showed contrasting patterns in the composition of the four ORs, with more arm-level deletions observed in OPRM1 and OPRD1, while more arm-level gains and high amplifications were found in OPRK1 and OPRL1 (Supplementary Fig. 2A). Additionally, the gene mutations of the four ORs were relatively low in cancers, ranging from 0 to 6.4% (Supplementary Fig. 2B).

3.3 Survival analysis of four ORs in the pan-cancer

Based on survival analyses, higher expression levels of OPRM1 were associated with worse prognosis in BRCA and KIRC, as demonstrated by Figure 3A. Conversely, for OPRD1, higher expression levels were associated with worse prognosis in KIRC and LIHC, but better prognosis in BLCA and PAAD, as shown in Figure 3B. In the case of OPRK1, higher expression levels were linked to worse prognosis in KIRP, thymoma, and UCEC, as displayed in Figure 3C. Finally, for OPRL1, higher expression levels were associated with worse prognosis in BLCA and LIHC, but better prognosis in BRCA and PAAD (Figure 3D).

Fig. 3
figure 3

The OS analyses of four ORs across TCGA cancer dataset. A. The OS analyses of OPRM1 across TCGA cancer dataset. B. The OS analyses of OPRD1 across TCGA cancer dataset. C. The OS analyses of OPRK1 across TCGA cancer dataset. D. The OS analyses of OPRL1 across TCGA cancer dataset. OS, overall survival; HR, hazard ratio

Full size image

3.4 Functional enrichment analyses of ORs

Co-expression genes of ORs in specific cancers were selected for functional enrichment analyses. For certain cancer types, we observed inadequate potential pathways enriched, with less than 10 for KEGG or GO-BP/GO-CC/GO-MF, and therefore did not present these results. Top enriched pathways of KEGG and GO in KIRP and UCEC for OPRK1 were shown in Fig. 4A–B. Metabolic pathways such as amino acid metabolism and glycine metabolism were significantly enriched in KIRP (Fig. 4A) while immune-related pathways, such as T cell activation and regulation of leukocyte activation, were dramatically enriched in UCEC (Fig. 4B).

Fig. 4
figure 4

Functional enrichment analyses and potential mechanism prediction of OPRK1 and OPRL1 in multiple cancers. A. Top enriched pathways of KEGG and GO in KIRP for OPRK1. Metabolic pathways such as amino acid metabolism and glycine metabolism were significantly enriched in KIRP. B. Top enriched pathways of KEGG and GO in UCEC for OPRK1. Immune-related pathways were dramatically enriched in UCEC. C. Analysis of LIHC revealed that OPRL1 was associated with top enriched pathways in KEGG and GO analyses, with immune-related pathways exhibiting significant enrichment. D. Top enriched pathways of KEGG and GO in PAAD for OPRL1. Immune-related pathways were dramatically enriched in PAAD. KEGG, Kyoto Encyclopedia of Genes and Genomes; GO, Gene Ontology; KIRP, Kidney renal papillary cell carcinoma; UCEC, Uterine Corpus Endometrial Carcinoma; LIHC, liver hepatocellular carcinoma; PAAD, pancreatic adenocarcinoma

Full size image

Interestingly, immune-related pathways were significantly enriched both in LIHC and PAAD for OPRL1, which strongly suggested that OPRL1 might play a key role in immune response and tumor immune microenvironment regulation. Functional enrichment analyses for OPRD1 in LIHC and PAAD revealed connections with cell cycle and proliferation-related pathways, as well as hormone and synapse-related pathways (Supplementary Fig. 3).

3.5 Correlations between tumor immune cell infiltration and ORs

Results revealed that OPRL1 had the highest correlations with immune cell infiltration in comparison to the other three ORs (Fig. 5 & Supplementary Fig. 4). Notably, there was a significant positive correlation between the expression level of OPRL1 and multiple types of immune cells infiltrating LIHC and PAAD, as depicted in Fig. 5B. Conversely, weak relationships were observed between immune cell infiltration and OPRM1/OPRD1, as presented in Supplementary Fig. 4. These outcomes provide additional evidence suggesting that OPRL1 may play a role in tumor immune response.

Fig. 5
figure 5

Relationships among OPRK1 and OPRL1 expression levels and immune cell infiltration levels. A. Scatter plots between OPRK1 expression levels and immune cell infiltration levels. B. Scatter plots between OPRL1 expression levels and immune cell infiltration levels. Regression analyses suggested that higher expression levels of OPRL1 correlated with higher infiltration levels of immune cell in four kinds of cancers especially in LIHC and PAAD

Full size image

3.6 Expression levels of OPRL1 in LIHC cell lines and immune cells

Next, the mRNA expression levels of the four different ORs were analyzed in a monocyte cell line (THP-1) and a T cell line (Jurkat), respectively, to validate the bioinformatic analysis. As shown in Fig. 6A, the expression levels of OPRL1 were obviously higher than all other three ORs in both the two immune cell lines. Notably, Nociceptin, an agonist of NOPR, did not influence the growth of monocyte cell, as detected by viability assay (Fig. 6B). In a subsequent analysis of the mRNA expression levels of OPRL1 in seven LIHC cell lines and one normal hepatic cell line (LO-2), the results were also consistent with bioinformatic analyses (Fig. 2), which revealed that OPRL1 was highly expressed in all hepatic cancer and normal cell lines (Fig. 6C). In sharp contrast, the expression level of OPRD1 was modest, whereas the levels of OPRK1 and OPRM1 were almost undetectable in these cell lines.

Fig. 6
figure 6

mRNA expression levels of ORs and the effects of NOPR agonist on immune and HCC cell lines. A. mRNA expression levels of ORs in immune cell lines. B. Effect of NOPR agonist on the growth of THP-1. C. mRNA expression levels of ORs in normal hepatic cell line (LO-2) and HCC cell lines. D. Effects of NOPR agonist on the proliferation of BLCA, BRCA, LIHC and PAAD. E. Effects of NOPR agonist on the migration and invasion of BLCA, BRCA, LIHC and PAAD. *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant. NOPR, nociception/orphanin; FQ, peptide receptor; BRCA, breast cancer; BLCA, bladder cancer

Full size image

3.7 Effects of ORs agonists on the proliferation, migration and invasion of cancers in vitro

Experiments in vitro were performed to explore effects of ORs agonists on the proliferation, migration and invasion of specific types of cancers selected by the survival analyses (Fig. 3). CCK-8 assays demonstrated that the MOR agonist had negligible effects on cell proliferation in BRCA and KIRC (Supplementary Fig. 5A). DOR agonist could inhibit the growth of BLCA while promote the growth of KIRC and LIHC (Hep3B) (Supplementary Fig. 5B). KOR agonist significantly promoted the proliferation of KIRP and UCEC (Supplementary Fig. 5C). Conversely, the NOPR agonist had no significant effects on cell growth in BLCA, BRCA (MDA-MBA231), LIHC (HCCLM3) and PAAD (Fig. 6D). A slight inhibition was observed in T47D, while promotion was noted in Hep3B (Fig. 6D).

Transwell migration and invasion assays revealed that the MOR agonist could significantly increase the migration of BRCA and KIRC, as well as the invasion of BRCA (Supplementary Fig. 6A). The DOR agonist was observed to increase the migration and invasion of KIRC and LIHC, while opposite effects were observed in BLCA and PAAD (Supplementary Fig. 6B). Additionally, the KOR agonist significantly stimulated the invasion of KIRP and UCEC, as well as the migration of KIRP (Supplementary Fig. 6C). Moreover, the NOPR agonist was found to increase the migration and invasion of LIHC, while opposite effects were observed in BRCA and PAAD (Fig. 6E).

4 Discussion

This study is a general analysis of the four opioid receptors in pan-cancers, providing a prognostic and initial analysis, and is accompanied by preliminary in vitro experiments for validation. Based on the analysis of ORs expression levels in different normal tissues, we specifically found that OPRL1 was highly expressed in monocytes. Analysis of ORs expression in immune cells also revealed that OPRL1 was significantly higher than the other three ORs. These results indicated that OPRL1 may have an essential role in monocytes. According to the pan-cancer analysis, the highest abundance of OPRL1 was found and more arm-level gain and high amplication were found in OPRK1 and OPRL1. It is noteworthy that the expression level changes between normal and tumor that the TPMs are low and that the changes may be hitchhikers driven by rare tumor genotypes, rather than reflective of functionally consequential expression.

Prognostic analysis pointed that the ORs acted as protective factors and risk factors in different cancer types. Functional enrichment analysis and tumor immune cell infiltration correlation studies showed that immune-related pathways were significantly enriched for OPRK1 in UCEC, as well as for OPRL1 in LIHC and PAAD. In both tumors, OPRL1 expression levels were strongly and positively correlated with multiple immune cell infiltrations. In vitro cellular experiments using ORs agonists were validated and all results were consistent with the findings of the bioinformatic analysis. Together, these results demonstrated a general map of the relationship between four ORs and tumors. It also highlights the possible role of OPRL1, which has been less frequently mentioned before, in the regulation of the immune response and TIM.

Many studies over the past decade have found associations between ORs expression levels and the outcomes in specific cancer types [23, 43], which is consistent with the findings of our study. The implication is that at the molecular level, opioid agonism at the receptor has downstream effects that impact the oncological endpoint. To date, as MOR agonists, morphine and fentanyl are the leading analgesics in clinical treatment of moderate and severe pain, MOR has greater clinical relevance [44]. Thus, there are relatively more research on MOR and cancer [21, 45, 46]. DOR has been relatively well studied in breast cancer, with reports of promoting tumor metastasis and angiogenesis through epithelial-mesenchymal transition and activation of HIF-1α [47, 48]. Previous researches have revealed that KOR was increased in a variety of solid tumors, including liver cancer and esophageal squamous cell carcinoma, and the KOR expression is linked to cancer progression and bad prognosis [49, 50]. However, further studies are required to confirm the underlying mechanisms. In our study, the expression of OPRM1 and OPRD1 are very low in many cancers, and more arm-level deletion were found. For OPRK1, we reported the immune-related pathways were enriched in UCEC. In the endometrium, KOR is the most common opioid binding site [51]. This may provide a strong hint for subsequent research. There are few existing articles that synthesize multiple ORs for side-by-side comparisons or outline their roles comprehensively. Hashimi et.al detected transcripts for four ORs in whole human blood from healthy donors which contained full range of circulating immune cells. They found only NOPR mRNA, with no MOR, DOR or KOR in these samples [52]. This result complements our study and suggests a potential application of NOPR in immunity. Belltall et al. also explored the relationship between the expression of ORs (MOR, KOR, DOR) and oncologic outcomes across various solid tumors [53]. While they conducted a purely biochemical analysis, our study delved more deeply into the subject matter.

Current research on NOPR has primarily focused on neurology and psychiatry. Preclinical models have investigated the efficacy of NOPR agonists for anxiety, cough, substance abuse, migraine, pain (spinal and peripheral), and urinary incontinence, while NOPR antagonists have been studied for their potential in treating pain, depression, and motor symptoms in Parkinson's disease [54,55,56]. Another objective of research is to identify ligands and conduct analgesic studies to develop new anesthetic drugs with greater selectivity and safety. Regarding immunology research, NOPR mRNA has been identified in blood lymphocytes, monocytes, and granulocytes from healthy human donors using PCR or qPCR methods [52, 57], which is in line with our bioinformatic analysis results. However, few existing studies have examined the interaction between OPRL1 and tumors. Szalay et al. found that nociception was elevated in 18 hepatocellular carcinoma patients with a ten-fold increase compared to other liver diseases in a study of plasma nociception levels between liver diseases and healthy controls [58]. Similar results were shown in another study of 113 patients with different types of advanced cancer [59].

Additionally, some studies have explored the mechanisms underlying these phenomena. After Kuzumaki et al. reported upregulation of OPRL1 in Non-small cell lung cancer (NSCLC) for the first time [60], Wang and his colleagues found that NOPR is overexpressed in NSCLC and is inversely correlated with patients' postoperative survival, acting via the PI3K/Akt signaling pathway [61]. Studies on NOPR antagonist JTC-801 have shown inhibition of proliferation and metastasis in ovarian cancer cells [62], melanoma cells [63], and osteosarcoma cells [64] through the PI3K/AKT pathway. In our study, we discovered that OPRL1 is not only highly expressed in tumors, but also closely related to tumor immunity, which has not been previously reported. Since OPRL1 has the highest abundance in different cancers compared to other ORs, it may have greater clinical significance for oncology patients, and it can guide the development of new drugs targeting NOPR.

The limitations of the present study merit attention. Firstly, validation of immune-related signaling pathways enriched by functional analysis requires in vivo or in vitro experiments. While some validation was performed using cell lines in our study, this offers a promising avenue for further exploration by scientists. Secondly, additional research is imperative to elucidate the mechanism and clinical significance of modulated OR levels in specific tumors. Our group is embarking on the next phase of comprehensive investigation, utilizing animal models and single-cell RNA sequencing to unravel the relevant regulatory factors and mechanisms of action. This will establish the groundwork for improved treatment options for oncology patients.

5 Conclusion

By performing a pan-cancer bioinformatic analysis and fundamental validations in vitro, our study investigated the expression heterogeneity, tumor immune characteristics and the prognosis effects of four ORs on patients with tumors. OPRL1 exhibited the highest abundance across different types of cancers, while OPRM1 and OPRD1 were barely detectable in multiple cancer types. Notably, our findings highlight the significant clinical importance of OPRK1 and OPRL1, which may play a crucial role in tumor immunity. Furthermore, OPRL1 may play a vital role in monocytes and regulating the immune response and tumor-infiltrating macrophages. Consequently, targeting KOR and NOPR could represent a promising avenue for cancer treatment. Nonetheless, further studies are warranted to explore connections between OPRL1/OPRK1 and tumor immune regulation.