Medicine
OPEN
Prognostic, chemotherapy and immunotherapy roles of GPR37/GPR37L1 in pan-cancer
Guoqiang Zhu, MSª, Jiliang He, BSª, Ningkun Shi, BSª, Zhongyao Cai, BSb, Jiannan Zhang, PhDa, Susanna Chau Yi Wang, BSc, Juan Li, PhDa, Mao Zhang, MDd, Yajun Wang, PhDa,*ID
Abstract
G protein-coupled receptor 37 (GPR37) and G protein-coupled receptor 37 like 1 (GPR37L1) are implicated in tumorigenesis; however, their prognostic significance and roles in chemotherapy and immunotherapy responses across diverse cancers is incompletely defined. Utilizing transcriptomic, genomic, pharmacogenomic, and clinical data, we mapped the molecular landscapes of GPR37 and GPR37L1 and assessed their potential clinical value in retrospective real-world cohorts. Our analysis revealed high genetic alteration rates for GPR37 (10.2%) and GPR37L1 (10.7%). Aberrant expression of GPR37 served as a predictive biomarker for survival outcomes in breast invasive carcinoma, lower grade gliomas, and lung adenocarcinoma, a pattern similarly observed for GPR37L1. Intriguingly, the 2 receptors exhibited divergent prognostic effects in adrenocortical carcinoma, cervical squamous cell carcinoma and endocervical adenocarcinoma, lung squamous cell carcinoma, stomach adenocarcinoma, uterine corpus endometrial carcinoma, and uveal melanoma. Evaluation of chemotherapy response identified significant correlations between the expression levels of both GPR37 and GPR37L1 and sensitivity to 21 common chemotherapeutic agents. For example, in lung adenocarcinoma, the elevated expression of either receptor was significantly associated with reduced sensitivity to cisplatin and gemcitabine. Analysis of 2 immunotherapy-treated melanoma cohorts demonstrated that high GPR37 or GPR37L1 expression correlated with inferior overall survival, a finding corroborated by immune infiltration and chemokine profiles. Our in vitro experiments demonstrated that the proposed ligand TX14A was unable to activate GPR37 and GPR37L1 through the cyclic adenosine monophosphate or extracellular signal-regulated kinase/mitogen-activated protein kinase pathways. Notably, this is the first systematic pan-cancer profiling of both GPR37 and GPR37L1 that integrates analyses of chemotherapy sensitivity, immunotherapy response, and immune landscapes. GPR37 and GPR37L1 may represent candidate biomarkers for prognostic stratification and for predicting chemotherapy and immunotherapy response.
Abbreviations: ACC = adrenocortical carcinoma, BLCA = bladder urothelial carcinoma, BRCA = breast invasive carcinoma, CAMP = cyclic adenosine monophosphate, CESC = cervical squamous cell carcinoma and endocervical adenocarcinoma, CHOL = cholangiocarcinoma, COAD = colon adenocarcinoma, CR = complete response, ERK = extracellular signal-regulated kinase, ESCA = esophageal carcinoma, GBM = glioblastoma multiforme, GEO = gene expression omnibus, GPR37 = G protein-coupled receptor 37, GPR37L1 = G protein-coupled receptor 37 like 1, HNSC = head and neck squamous cell carcinoma, HR = hazard ratio, IC50 = half-maximal inhibitory concentration, KICH = kidney chromophobe, KIRC = kidney renal clear cell carcinoma, KIRP = kidney renal papillary cell carcinoma, KM = Kaplan-Meier, LGG = brain lower grade glioma, LIHC = liver hepatocellular carcinoma, LUAD = lung adenocarcinoma, LUSC = lung squamous cell carcinoma, MESO = mesothelioma, mOS = median overall survival, PAAD = pancreatic adenocarcinoma, PCPG = pheochromocytoma and paraganglioma, PD = progressive disease, PR = partial response, PSAP = prosaposin, READ = rectum adenocarcinoma, SARC = sarcoma, SD = stable disease, SKCM = skin cutaneous melanoma, SRE = serum response element, STAD = stomach adenocarcinoma, TCGA = the cancer genome atlas, TGCT = testicular germ cell tumor, THCA = thyroid carcinoma, THYM = thymoma, UCEC = uterine corpus endometrial carcinoma, UCS = uterine carcinosarcoma, UVM = uveal melanoma.
Keywords: GPR37, GPR37L1, immunotherapy, pan-cancer, prognosis
The authors have no funding and conflicts of interest to disclose. The datasets generated during and/or analyzed during the current study are publicly available.
a Key Laboratory of Bio-Resources and Eco-Environment of Ministry of Education, College of Life Sciences, Sichuan University, Chengdu, China, b Sheng Yushou Center of Cell Biology and Immunology, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai, China, ” University of New South Wales, School of Biological, Earth and Environmental Sciences, Sydney, Australia, ª Division of Vascular Surgery, Sichuan Academy of Medical Sciences and Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, China.
* Correspondence: Yajun Wang, Key Laboratory of Bio-Resources and Eco-Environment of Ministry of Education, College of Life Sciences, Sichuan University, Chengdu 610065, China (e-mail: cdwyjhk@163.com).
Copyright @ 2026 the Author(s). Published by Wolters Kluwer Health, Inc. This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial License 4.0 (CCBY-NC), where it is permissible to download, share, remix, transform, and buildup the work provided it is properly cited. The work cannot be used commercially without permission from the journal.
How to cite this article: Zhu G, He J, Shi N, Cai Z, Zhang J, Wang SCY, Li J, Zhang M, Wang Y. Prognostic, chemotherapy and immunotherapy roles of GPR37/GPR37L1 in pan-cancer. Medicine 2026;105:10(e47813).
Received: 10 August 2025 / Received in final form: 8 January 2026 / Accepted: 3 February 2026
http://dx.doi.org/10.1097/MD.0000000000047813
1. Introduction
The orphan receptor G protein-coupled receptor 37 (GPR37) and its homologous receptor G protein-coupled receptor 37 like 1 (GPR37L1), have only been studied partially in terms of their prognostic value in cancer. Existing evidence reveals striking context-dependent roles of GPR37 across malignancies, where it can be tumor-suppressive in some cases, and oncogenic in oth- ers. In human hepatocellular carcinoma, low GPR37 expression correlates with disease progression and poor patient survival.[1] In lung adenocarcinoma (LUAD), high GPR37 expression relates to aggressive growth, enhanced malignancy via the TGF-ß/Smad pathway or binding to CDK6, and increased metastatic poten- tial via the competing endogenous RNA (ceRNA) network, all contributing to unfavorable prognosis.[2-4] In gliomas, GPR37 overexpression, driven by promoter hypomethylation, correlates with poor overall and disease-specific survival.[5,6] In colorectal cancer liver metastases, elevated GPR37 promotes metastatic progression and correlates with poor patient outcomes.[7] In gastric cancer, GPR37 mediates peritoneal metastasis, a process associated with worse patient prognosis.[8] In multiple myeloma, GPR37 affects cancer cell biological behavior, which in turn impacts prognostic outcomes.[9] Our previous comparative tran- scriptomic analysis of humans and hens identified that GPR37 can stratify ovarian cancer patients into distinct subgroups with different clinical outcomes.[10] To date, GPR37’s prognostic sig- nificance has been investigated in the 7 aforementioned cancers, while no studies have reported GPR37L1’s prognostic relevance in cancer, leaving its prognostic role completely uncharacter- ized. At present, research on GPR37’s prognostic significance is dispersed across individual cancer types, making systematic pan-cancer analysis essential to integrate GPR37’s prognostic roles across different cancers and address the knowledge gap of GPR37L1.
Beyond their prognostic roles, GPR37 and GPR37L1 reg- ulate cancer treatment responses, including chemotherapy efficacy and tumor immunity, but related research remains fragmented. GPR37 is only reported to promote cisplatin resis- tance in non-small cell lung cancer by activating the PI3K/Akt pathway,[11] and protects against chemotherapy-induced periph- eral neuropathy caused by drugs like paclitaxel.[12] In tumor immunity, GPR37 enhances macrophage phagocytosis, resolves inflammation, and protects against infection-induced sepsis,[13,14] and it also aids inflammation resolution via extracellular vesi- cle release mediated by prosaposin (PSAP)-GPR37 signaling.[15] Besides these, there are no other relevant studies on the roles of GPR37 in cancer treatment responses. GPR37L1, by contrast, has no reported data on its role in chemotherapy and immuno- modulation. How these 2 receptors influence the chemotherapy sensitivity, immune infiltration, and immunotherapy response across diverse malignancies is an interesting puzzle that awaits unraveling.
While the roles of GPR37 and GPR37L1 in cancer have been partially elucidated, research into their ligands remains impeded by conflicting findings. Such inconsistencies represent a critical barrier in deciphering their biological mechanisms in tumor progression. Among the list of proposed ligands, PSAP is considered the most likely endogenous ligand for GPR37 and GPR37L1.[16,17] Together with its synthetic peptide frag- ment TX14A, they have been the most extensively studied. Activation of GPR37 and GPR37L1 by PSAP and TX14A could reduce cyclic adenosine monophosphate (cAMP) levels and enhance extracellular signal-regulated kinase (ERK) phos- phorylation in the cultured HEK293T cell lines.[18-22] However, other studies challenge such activation potential. Notably, one study found that this activation could not be replicated in het- erologous cells overexpressing GPR37 or GPR37L1.[23] And another study explicitly classified GPR37L1 as an unliganded orphan receptor due to inconsistent validation of PSAP/TX14A- mediated activation.[24] The activation potential of human
GPR37 and GPR37L1 remains controversial, in which divided reports of success and failure are made by multiple independent research teams with equally robust evidence.[18-27] GPR37 and GPR37L1 exhibit cross-species conservation in their nucleo- tide sequences.[16,28] Additionally, current research is confined to humans and a small number of mouse models, with no studies conducted in other species. Instead of replicating the activation assays performed in human models, we strategized the design of cross-species ligand-receptor activation experiments will pro- vide new insights on the molecular mechanism behind GPR37 and GPR37L1 ligand binding and functionality.
Pan-cancer analyses focus on the performance of genes across all types of tumors. This study conducts comprehensive pan-cancer analyses of GPR37 and GPR37L1 using the cancer genome atlas (TCGA), gene expression omnibus (GEO), and GDSC2 data. To determine whether GPR37 and GPR37L1 are associated with patient survival and treatment response across cancers, we characterize their genetic variants and expression profiles, and present the first systematic investigation of che- motherapy sensitivity, immune infiltration, and immunother- apy response across diverse malignancies. Additionally, given chicken is a demonstrated animal model for human disease, [29-32] we conducted in vitro experiments in HEK293T cells using the putative ligand TX14A for chicken GPR37 and GPR37L1. This integrated approach reveals novel prognostic and therapeutic insights for both receptors.
2. Materials and methods
2.1. Basic genetic information of GPR37 and GPR37L1
Genomic location information for GPR37 was obtained from Ensembl (link: https://www.ensembl.org/Homo_sapiens/ Gene/Summary?g=ENSG00000170775). The plot of chro- mosome location was downloaded from Genecard (link: https://www.genecards.org/cgi-bin/carddisp.pl?gene=GPR37). And the plot containing the exons and topology was down- loaded from the cbioportal (link: https://www.cbioportal.org/ results/mutations?case_set_id=all&gene_list=GPR37&can- cer_study_list=5c8a7d55e4b046111fee2296). Genomic loca- tion information for GPR37L1 was obtained from Ensembl (link: https://www.ensembl.org/Homo_sapiens/Gene/ Summary?g=ENSG00000170075). The plot of chromosome location was downloaded from Genecard (link: https://www. genecards.org/cgi-bin/carddisp.pl?gene=GPR37L1). And the plot containing the exons and topology was downloaded from the cbioportal (link: https://www.cbioportal.org/results/muta- tions?case_set_id=all&gene_list=GPR37L1&cancer_study_ list=5c8a7d55e4b046111fee2296).
2.2. The mRNA profiles and clinical survival in pan-cancer
The transcriptome data of patients with different tumors (file name: EBPlusPlusAdjustPANCAN_IlluminaHiSeq_ RNASeqV2.geneExp.tsv) were downloaded from the official website of the TCGA Pan-Cancer Project at https://gdc.can- cer.gov/about-data/publications/pancanatlas. Paired clinical information including overall survival (file name: TCGA- clinical data resource outcome) was also obtained from the same web page. After pairing of GPR37 and GPR37L1 expression data with clinical information, a total of 10,163 tumor patients were included in our present study. There are 79 adrenocortical carcinomas (ACC), 407 bladder urothe- lial carcinomas (BLCA), 1094 breast invasive carcinomas (BRCA), 304 cervical squamous cell carcinomas and endo- cervical adenocarcinomas (CESC), 36 cholangiocarcinomas (CHOL), 448 colon adenocarcinomas (COAD), 48 lymphoid neoplasm diffuse large B-cell lymphomas, 184 esophageal carcinomas (ESCA), 160 glioblastoma multiformes (GBM),
519 head and neck squamous cell carcinomas (HNSC), 65 kidney chromophobes (KICH), 533 kidney renal clear cell carcinomas (KIRC), 289 kidney renal papillary cell carcino- mas (KIRP), 161 acute myeloid leukemias, 514 brain lower grade gliomas (LGG), 370 liver hepatocellular carcinomas (LIHC), 506 LUAD, 495 lung squamous cell carcinomas (LUSC), 86 mesotheliomas (MESO), 379 ovarian cancers, 178 pancreatic adenocarcinomas (PAAD), 179 pheochro- mocytoma and paraganglioma (PCPG), 497 prostate ade- nocarcinomas, 159 rectum adenocarcinomas (READ), 259 sarcomas (SARC), 454 skin cutaneous melanomas (SKCM), 409 stomach adenocarcinomas (STAD), 134 testicular germ cell tumors (TGCT), 505 thyroid carcinomas (THCA), 119 thymomas (THYM), 531 uterine corpus endometrial carci- noma (UCEC), 57 uterine carcinosarcoma (UCS) and 80 uveal melanoma (UVM), respectively. For prognostic assessment, we tested for overall survival differences between groups of patients with high and low GPR37 or GPR37L1 expres- sion using the Kaplan-Meier (KM) method. We employed the “surv_cutpoint” function of the survival package to identify the optimal cutoff value for continuous GPR37 or GPR37L1 expression.[33] For the visualization of overall survival, we generated forest plots and KM plots using the R packages forestplot and survminer, respectively.[33,34] This study used publicly available datasets (TCGA, GEO). No new human subjects were recruited, so institutional review board approval or informed consent was not required. This is consistent with standard practices for secondary analysis of public genomic data.
2.3. Single-cell transcriptomes of gliomas in GSE102130
Due to the high expression of GPR37 and GPR37L1 in GBM and LGG, we collected thirteen single-cell transcriptome data from gliomas in GSE102130.[35] The expression matrices of these datasets were downloaded from the gene expression omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) and were subjected to standardized processing utilizing the MAESTRO (minor-allele-enriched sequencing through recognition oligonu- cleotides) workflow on the platform TISCH (Tumor Immune Single-cell Hub, http://tisch.comp-genomics.org/).[36] Moreover, in order to mitigate potential batch effects, datasets exhibiting a median entropy below 0.7 were subjected to batch correction employing the R package Seurat. Visualizations were conducted on the TISCH.[36]
2.4. The sensitivity of common chemotherapy drugs
The genomics of drug sensitivity in cancer included tran- scriptomic data for different cell lines treated with 167 drugs and paired half-maximal inhibitory concentration (IC50) values.[37] Chemotherapeutic agents can be cate- gorized based on their mechanisms of action into several groups, including antimetabolites, alkylating agents, mitotic spindle inhibitors, topoisomerase inhibitors, and others.[38] The 21 most common chemotherapeutic drugs from the genomics of drug sensitivity in cancer database, including carmustine, cisplatin, cyclophosphamide, cytarabine, dacti- nomycin, docetaxel, epirubicin, fludarabine, gemcitabine, Irinotecan, mitoxantrone, nelarabine, oxaliplatin, pacli- taxel, temozolomide, teniposide, topotecan, vinblastine, vincristine, vinorelbine, vorinostat and 5-fluorouracil, were filtered. Ridge regression models were built to predict drug IC50 values for each patient in TCGA using the R pack- age oncoPredict.[39] The correlation between drug sensitivity and GPR37 or GPR37L1 expression was assessed by the Pearson correlation coefficient. For visual representation, we generated heatmaps and scatter plots using the R pack- ages pheatmap and ggpubr.[40,41]
2.5. The profiles of immune infiltration in pan-cancer
Macrophage M1, plasmacytoid B-cells, endothelial cells, and CD8 T-cells are all capable of inhibiting tumor growth and enhancing immune toxicity.[42-45] The degree of immune infil- tration was assessed using 4 widely used algorithms: TIMER, CIBERSORT, MCPCOUNTER, and XCELL.[46-49] The Pearson correlation between immune infiltration and GPR37 or GPR37L1 expression was computed, and heat maps were plot- ted using the R package pheatmap.[41]
2.6. The mRNA profiles of patients following immunotherapy
We tried to collect some cohorts of oncology patients receiv- ing immunotherapy with transcriptomic data and sur- vival outcomes in public studies and found that the cohorts including phs000452, PRJEB23709_PD1 for SKCM;[50,51] and PRJNA482620 for GBM,[52] were available. The cohort phs000452 was downloaded from Melanoma Genome Sequencing Project at https://www.ncbi.nlm.nih.gov/projects/ gap/cgi-bin/study.cgi?study_id=phs000452.v3.p1. The cohort PRJEB23709 was downloaded from the link https://www. ncbi.nlm.nih.gov/bioproject/?term=PRJEB23709. The cohort PRJNA482620 was downloaded from the link https://www. ncbi.nlm.nih.gov/bioproject/482620. To visualize overall sur- vival, forest plots and KM plots were generated using the R packages forestplot and survminer, respectively.[33,34]
2.7. Genomic variant profiles in pan-cancer
There were 10,288 cancer patients with genome sequencing data in TCGA, including 92 patients with ACC, 411 patients with BLCA, 1026 patients with BRCA, 291 patients with CESC, 36 patients with CHOL, 408 patients with COAD, 37 patients with diffuse large B-cell lymphomas, 185 patients with ESCA, 403 patients with GBM, 509 patients with HNSC, 66 patients with KICH, 370 patients with KIRC, 282 patients with KIRP, 141 patients with acute myeloid leukemia, 526 patients with LGG, 365 patients with LIHC, 569 patients with LUAD, 485 patients with LUSC, 82 patients with MESO, 412 patients with OV, 178 patients with PAAD, 184 patients with PCPG, 498 patients with prostate adenocarcinomas, 151 patients with READ, 239 patients with SARC, 468 patients with SKCM, 439 patients with STAD, 151 patients with TGCT, 500 patients with THCA, 123 patients with THYM, 531 patients with UCEC, 50 patients with UCS and 80 patients with UVM, respectively. The genetic variants data of mutations, including insertion (INS), deletion (DEL), single nucleotide polymorphism, was downloaded in the file named mc3.v0.2.8.PUBLIC.maf.gz from the official web- site of the TCGA Pan-Cancer Project at https://gdc.cancer.gov/ about-data/publications/pancanatlas. The genetic variants data of copy number variations was downloaded in the file named gene-level copy number (gistic2_thresholded) from the website at https://xenabrowser.net/datapages/?cohort=TCGA%20Pan- Cancer%20(PANCAN. The threshold values to -2 and 2 esti- mated by GISTIC2 represented copy number deep deletions and amplifications respectively.[53] The boxplots and heatmaps were performed in ggpubr and pheatmap, respectively.[40,41]
2.8. HEK293T cell cultures treated with the proposed ligand TX14A
We initiated our experiments by constructing chicken GPR37 and GPR37L1 pcDNA3.1(+) expression plasmids, using empty pcDNA3.1(+) plasmid as the negative control based on the methods described in our previous studies.[54-57] HEK293T cells, preserved in liquid nitrogen, were thawed and cultured in 10-cm cell culture dishes. Upon reaching the second generation, the
cells were transferred to 6-well plates and cultured under stan- dard conditions of 37℃ and 5% CO,. When the cell density reached approximately 80%, we transfected the HEK293T cells in 6-well plates with either the GPR37 or GPR37L1 receptor expression plasmid, or the pcDNA3.1(+) control plasmid. The transfection process utilized 800ng of either the pGL3-CRE- Luciferase reporter plasmid or the pGL3-serum response ele- ment (SRE)-Luciferase reporter plasmid, 2 uL of jetPRIME transfection reagent, and 100 µL of transfection buffer. Post- transfection, the cells were re-suspended and seeded in 96-well plates at a density of approximately 2 × 104 cells per well in 180 uL of culture medium. After 20 hours of incubation, the medium was aspirated and replaced with 60 uL of serum-free medium containing varying concentrations of TX14A for a 6-hour incu- bation period. TX14A, the proposed ligand, was sourced from TargetMol Chemicals Inc. Following the incubation, the medium was removed, and 50 µL of 1 x Passive Lysis Buffer was added to each well for cell lysis. From each well, 15 uL of the cell lysate was combined with 40 µL of luciferase substrate, and luciferase activity was measured using a luminometer. The HEK293T cell line used in vitro was purchased from the Bio-sample Bank of the State Key Laboratory of Sichuan University. Its acquisition and use adhered to national and institutional ethical guidelines for cell research, with no ethical conflicts.
2.9. Statistical analysis
All statistical analyses were conducted using R software (ver- sion 4.1.3, https://www.r-project.org/) and GraphPad Prism 7 (GraphPad Software) according to the methods described in our previous studies. [58-60] Differences between 2 groups were evaluated using the Wilcoxon rank sum test, while 1-way analysis of variance followed by Dunnett’s test was used for comparisons among multiple groups. Survival differences between groups were assessed via KM analysis with a log- rank test. Patients were divided into high-expression and low- expression groups based on the median mRNA expression level of GPR37/GPR37L1 and we used the log-rank test to compare survival curves between the 2 groups. This categorical group- ing method helps to clarify the correlation between receptor expression and prognosis. For correlation analyses, Spearman correlation analysis was applied to examine linear relationships between gene expression, immune checkpoint expression, and drug sensitivity. This method was selected because the contin- uous data in our study conformed to a normal distribution. mRNA levels of target genes were normalized to ß-actin and expressed as fold changes relative to selected tissues. Luciferase activities for each peptide treatment group were expressed as relative fold changes compared to the control group (without peptide treatment). Unless otherwise specified, a P-value < . 05 was considered statistically significant (*P <. 05; ** P <. 01; *** P < . 001). All experiments were independently repeated 2 to 3 times to ensure validity.
3. Results
3.1. Genetic variants of GPR37 and GPR37L1 in pan-cancer
We conducted a comprehensive pan-cancer analysis of genetic variants in GPR37 and GPR37L1, encompassing gene muta- tions, copy number amplifications and deep deletions. The cancers with the highest number of GPR37 variants were, in order, OV, UCEC, and SKCM, with 42, 40, and 40, respectively (Fig. 1A). To account for differences in sequencing sample sizes across tumor types, we performed the variance rate statistics for GPR37. Ovarian cancer was the tumor with the highest frequency of GPR37 variants, observed at the rate of 10.2% (42/412), and followed by SKCM at 8.5% (40/468) and UCEC
at 7.5% (40/531; Fig. 1B). For GPR37L1 variants, the highest patient populations were BRCA, UCEC, OV, SKCM and LIHC (Fig. 1C). The GPR37L1 gene variant was most frequently found in BRCA with a frequency of 13.0% (Fig. 1D). Like GPR37, the frequency of GPR37L1 variants was also relatively high in OV, SKCM, and UCEC, at 10.7%, 8.8%, and 9.2%, respectively (Fig. 1D).
3.2. Expression of GPR37 and GPR37L1 in pan-cancer
According to the most recent version of the human genome GRCh38.p14, GPR37 is located on chromosome 7, spanning positions 12,47,43,885 to 12,47,65,792, with a total length of 21,908 bases (Fig. 2A). GPR37 comprises of 2 exons, encoding a total of 613 amino acids, and contains a classic 7-transmembrane structure (Fig. 2B). GPR37 is relatively highly expressed in most cancers, although it exhibits low specificity (Fig. 2C).
In contrast, GPR37L1 is located on chromosome 1, span- ning positions 20,21,22,886 to 20,21,33,592, with a total length of 10,707 bases (Fig. 2D). GPR37L1 consists of 2 exons encoding a total of 481 amino acids and also has a classical 7- transmembrane structure (Fig. 2E). GPR37L1 is specifically highly expressed in GBM and LGG (Fig. 2F).
At the single-cell level of gliomas in dataset GSE102130 (Fig. 2G), GPR37 was predominantly expressed at high levels in oligodendrocytes, with lower expression in astrocyte-like (AC-like) malignant cells (Fig. 2H). In contrast, GPR37L1 showed predominant high expression in oligodendrocyte pre- cursor cell-like (OPC-like) and AC-like malignant cells, with sig- nificantly reduced expression in oligodendrocytes (Fig. 2I).
3.3. The prognosis values of GPR37 and GPR37L1 in pan-cancer
Our survival analysis revealed significant associations between the expression levels of GPR37 or GPR37L1 and clinical outcomes across various cancer types. In the forest plot for GPR37 (Fig. 3A), a higher GPR37 expression was associated with a poorer prognosis in ACC, BLCA, LGG, LUAD, LUSC, STAD, and THYM, but indicated a better prognosis in BRCA, CESC, OV, READ, UCEC, and UVM. Conversely, in the for- est plot for GPR37L1 (Fig. 3B), a higher GPR37L1 expres- sion was linked to a better prognosis in ACC, BRCA, KICH, LUSC, MESO, SARC, SKCM, and STAD, but a poorer progno- sis in CESC, KIRC, KIRP, LGG, LIHC, LUAD, PAAD, THCA, UCEC, and UVM. Interestingly, GPR37 and GPR37L1 may exert similar effects on the tumor environment in BRCA, LGG and LUAD (Fig. 3C-H). For example, in BRCA, both GPR37 and GPR37L1 could serve as risk factors, with median over- all survival (mOS) of 131.5 vs 102.1 months and 132.0 vs 87.9 months, respectively (Fig. 3C, F). And both GPR37 and GPR37L1 could be prognostic factors for favorable outcome in LGG (Fig. 3D, G, mOS: 44.6 vs 106.7 months and 62.9 vs 95.8 months) and LUAD (Fig. 3E, H, mOS: 38.2 vs 67.6 months and 31.7 vs 54.4 months). Additionally, GPR37, but not GPR37L1, can be used as a prognostic biomarker in BLCA, OV, READ, and THYM (Fig. 3A, B). Conversely, GPR37L1, but not GPR37, has the potential to predict prognosis in patients with KICH, KIRC, KIRP, LIHC, MESO, PAAD, SARC, SKCM, and THCA (Fig. 3A, B).
Although we found that both genes had consistent effects on BRCA, LGG, and LUAD during survival analysis (Fig. 3C-H), their impacts differed significantly in many other cancer types (Fig. 4). In ACC, the increased GPR37 expression was associ- ated with shorter overall survival (P = . 018, mOS: 39.9 months vs not reached), whereas increased GPR37L1 expression was linked to longer survival (P = . 017, mOS: not reached vs 40.1 months; Fig. 4A). In CESC, elevated GPR37 expression was associated with a better prognosis (P = . 020, mOS: 136.2 vs
A
GPR37_Variants
0
1-10
11-20
21-40
42
Numbers of patients with GPR37 Variants
40
40
40
30
30
24
24
23
20
19
17
13
14
10
10
11
7
7
7
9
9
8
2
3
0
1
1
0
0
0
1
0
1
1
3
0
0
0
ACC
BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LAML
LGG
LIHC
LUAD
LUSC
MESO
MISC
OV
PAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THCA
THYM
UCEC
UCS
UVM
B
10
Mutation
1.1
2.2
0.4
1.4
4.4
2.2
1.0
2.4
1.1
0.6
0.8
3.3
3.3
1.2
0.2
5.3
2.1
4.9
3.9
0.7
0.2
7.0
6.0
Amplication
0.7
8
0.9
0.3
1.5
0.4
1.5
0.8
1.4
4.0
1.6
1.4
1.2
8.7
0.6
0.6
2.1
4.1
0.7
0.2
1.1
0.2
0.1
0.7
0.2
2.7
1.6
0.6
0.5
0.4
0.2
1.0
0.8
0.7
0.4
6
Deep_deletion
All_variation
2.2
3.2
1.4
2.4
4.7
2.7
3.8
2.5
3.3
1.5
0.8
2.5
4.6
2.5
5.3
4.9
10.2
0.6
1.8
5.3
4.6
8.5
5.2
0.7
0.2
7.5
6.0
4
GPR37
ACC
BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LAML
LGG
LIHC
LUAD
LUSC
MESO
MISC
OV
PAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THCA
THYM
UCEC
UCS
UVM
2
C
140-
GPR37L1_Variants
0
1-10
11-20
21-40
133
Numbers of patients with GPR37L1 Variants
130
120
50
49
41
44
40
37
41
30
20
17
20
12
10
11
10
4
4
3
5
5
2
2
4
3
3
3
4
3
5
5
1
3
0
1
O
0
0
0
1
ACC
BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LAML
LGG
LIHC
LUAD
LUSC
MESO
MISC
OV
PAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THCA
THYM
UCEC
UCS
UVM
D
Mutation
1.7
0.4
0.7
3.7
0.5
0.7
1.4
0.3
0.7
0.6
0.3
0.4
0.6
0.5
0.6
0.7
0.4
4.9
2.5
0.2
4.0
12
Amplication
1.1
0.5
12.6
0.7
11.1
0.5
8.1
2.2
0.5
0.8
0.3
0.2
11.0
6.0
1.4
3.7
10.0
1.7
1.6
1.3
1.7
4.1
1.8
0.8
0.8
5.3
6.0
1.3
10
8
Deep_deletion
0.7
0.2
0.2
0.2
0.2
6
All_variation
1.1
2.9
13.0
1.4
11.1
4.2
8.1
2.7
1.2
2.2
0.5
0.7
0.8
11.2
6.5
2.1
3.7
10.7
1.7
1.6
0.8
2.0
2.1
8.8
4.6
1.0
0.8
9.2
6.0
1.3
4
GPR37L1
ACC
BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LAML
LGG
LIHC
LUAD
LUSC
MESO
MISC
OV
PAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THCA
THYM
UCEC
UCS
UVM
2
41.5 months), but elevated GPR37L1 expression was associated with a worse prognosis (P = . 024, mOS: 101.5 vs not reached; Fig. 4B). In LUSC and STAD, increased GPR37 expression was associated with shorter overall survival, whereas increased GPR37L1 expression was linked to longer survival (Fig. 4C, D). In UCEC and UVM, elevated GPR37 expression was associated with a better prognosis, but elevated GPR37L1 expression was associated with a worse prognosis (Fig. 4E, F).
3.4. The expression of GPR37 and GPR37L1 in association with chemotherapy
Chemotherapy is one of the most crucial methods of can- cer treatment, and we assessed the association between the expression of GPR37 and GPR37L1 and various che- motherapeutic drugs. The correlation between GPR37
expression and IC50 data for 21 common chemotherapy drugs was evaluated in the 9 cancers where GPR37 could serve as a potential prognostic marker (Fig. 5A). Increased expression of GPR37 was found to be associated with increased drug resistance in CESC, LUSC, BLCA, and LGG, while it was associated with enhanced drug sensitivity in UCEC, BRCA, and LUAD (Fig. 5A). Specifically, in UCEC, GPR37 expression was significantly negatively correlated with sensitivity to paclitaxel, suggesting increased sensi- tivity to paclitaxel (Fig. 5B). Similarly, in BRCA, GPR37 expression was significantly negatively associated with sensitivity to gemcitabine and vinorelbine (Fig. 5C, D). In LUAD, GPR37 expression showed a significant negative correlation with sensitivity to cisplatin and gemcitabine (Fig. 5E, F). For GPR37L1, the correlation with IC50 data for 21 chemotherapy drugs was assessed in thirteen
Zhu et al. · Medicine (2026) 105:10
A
B
p22.3
GPR37 log2(TPM+1)
p22.1
10
Cytoplasmic
Transmembrane
Extracellular
Topology
GPR37
O
Exon
D
p21.3
01
E
0
p21.2
0
11
GPR37L1 log2(TPM+1)
Extracellular
Topology
GPR37L1
ACC
€
p21.1
Exon
BLCA
p15.3
27
G
10
Cytoplasmic
p36.13
BRCA
5
M1
p36.11
p14.3
0
CESC
= G protein-coupled receptor 37 like 1.
populations in the GSE102130. (I) GPR37L1 expression across different cell populations in the GSE102130. GPR37 = G protein-coupled receptor 37, GPR37L1
of GPR37L1 in pan-cancer. (G) The distributions of various cell types in single cell dataset GSE102130 of gliomas. (H) GPR37 expression across different cell the GPR37L1 gene in the human genome. (E) Information on the extracellular, 7 transmembrane, and cytoplasmic structures of GPR37L1. (F) Expression levels
Figure 2. Profiling GPR37 and GPR37L1 expression in cancers. (A) The chromosomal localization of the GPR37 gene in the human genome. (B) Information on the extracellular, 7 transmembrane, and cytoplasmic structures of GPR37. (C) Expression levels of GPR37 in pan-cancer. (D) The chromosomal localization of
Human (GRCh38.p14)
Oligodendrocyte
· OPC-like Malignant OC-like Malignant
AC-like Malignant
Transmembrane
ACC
Y
0
Human (GRCh38.p14)
CHOL
BLCA
p34.3
100
p14.1
26
p34.2
COAD
BRCA
p33
DLBC
p12.3
CESC
p32.3
ESCA
p12.1
Oligodendrocyte
Glioma_GSE102130
CHOL
GBM
p11.2
Chr 7: 124,743,885-124,765,792
COAD
p31.3
1
HNSC
ـعجيز
DLBC
IS
100
p31.1
KICH
200
q11.21
AC-like Malignant
M1
1
q11.22
ESCA
-
OC-like Malignant
Chr 1: 202,122,886-202,133,592
KIRC
OPC-like Malignant
GBM
KIRP
q11.23
HNSC
134
p21.3
LAML
q21.11
KICH
p21.1
LGG
265
155
KIRC
p13.3
168
p13.2
LIHC
286
300
q21.13
H
KIRP
189
300
321
210
200
LUAD
LAML
341
q21.3
205
LUSC
H
335
q22.1
LGG
q12
MESO
226 252
356
q22.3
GPR37
Size: 21,908 bases
LIHC
q21.1
OV
LUAD
380
401
400
q31.1
PAAD
LUSC
273
q21.3
Size: 10,707 bases
PCPG
5
q31.2
q31.31
MESO
q23.3
PRAD
300
443
q31.32
READ
q31.33
OV
PAAD
310
q25.2
SARC
464 494
2
a
PCPG
q25.3
500
q33
331
q31.1
SKCM
515
PRAD
2
q34
-2.5
q31.3
STAD
-1.5
2.0
READ
362
TGCT
531
q35
-0.5
-1.0
SARC
383
q32.1
-0.0
q32.2
THCA
553
q36.1
SKCM
400
Medicine
398
THYM
613aa
q36.3
STAD
q41
UCEC
GPR37L1
TGCT
419
q42.13
q42.2
UCS
THCA
q43
UVM
THYM
481aa
q44
UCEC
UCS
A
UVM
1
2.5
-3.0
3.5
- 1.0
1.5
2.0
-0.0
0.5
A
GPR37 expression associated with overall survival
B
GPR37L1 expression associated with overall survival
Tumor type
log-rank p value
HR (95% CI)
Tumor type
log-rank p value
HR (95% CI)
ACC
0.018
2.44 (1.04 - 5.73)
ACC
0.017
0.42 (0.18 - 0.97)
BLCA
< 0.001
1.76 (1.35 - 2.28)
BLCA
0.092
1.30 (0.95 - 1.78)
BRCA
0.041
0.61 (0.36 - 1.00)
BRCA
0.002
0.52 (0.30 - 0.89)
CESC
0.020
0.46 (0.20 - 0.99)
CESC
0.024
3.48 (1.76 - 6.90)
KICH
0.196
0.37 (0.12 - 1.20)
KICH
0.049
0.27 (0.04 - 0.94)
KIRC
0.107
1.40 (1.01 - 1.94)
KIRC
0.004
1.80 (1.08 - 3.01)
KIRP
0.157
1.56 (0.85 - 2.87)
KIRP
0.049
1.87 (1.03 - 3.38)
LGG
< 0.001
2.31 (1.58 - 3.38)
LGG
<0.001
1.95 (1.24 - 3.07)
LIHC
0.289
0.81 (0.58 - 1.13)
LIHC
0.002
1.72 (1.20 - 2.47)
LUAD
< 0.001
1.76 (1.35 - 2.30)
LUAD
< 0.001
1.69 (1.19 - 2.38)
LUSC
0.008
1.46 (1.14 - 1.85)
LUSC
0.021
0.73 (0.56 - 0.95)
MESO
0.111
0.64 (0.41 - 1.01)
MESO
< 0.001
0.41 (0.25 - 0.65)
OV
< 0.001
0.63 (0.50 - 0.80)
OV
0.213
0.84 (0.64 - 1.10)
PAAD
0.085
0.58 (0.37 - 0.93)
PAAD
0.042
1.68 (1.08 - 2.64)
PRAD
0.214
Inf ( Inf - Inf)
PRAD
0.003
0.19 (0.03 - 0.98)
READ
0.024
0.40 (0.16 - 0.98)
READ
0.095
0.37 (0.15 - 0.92)
SARC
0.089
1.43 (1.00 - 2.03)
SARC
0.003
0.42 (0.26 - 0.66)
SKCM
0.187
0.79 (0.59 - 1.06)
SKCM
0.011
0.71 (0.53 - 0.94)
STAD
0.003
1.88 (1.36 - 2.59)
STAD
0.023
0.58 (0.32 - 0.96)
THCA
0.149
0.47 (0.19 - 1.12)
THCA
< 0.001
7.31 (1.39 -38.53)
THYM
0.041
3.79 (1.11 -12.91)
THYM
0.304
2.02 (0.55 - 7.46)
UCEC
0.013
0.26 (0.15 - 0.46)
UCEC
< 0.001
2.59 (1.25 - 5.36)
UVM
0.005
0.30 (0.14 - 0.62)
UVM
0.016
3.43 (1.50 - 7.86)
1/8
1/4
1/2
1
2
4
8
16
1/8
1/4
1/2
1
2
4
8
16
C
GPR37-Low + GPR37-High
D
Strata + GPR37-Low + GPR37-High
E
Strata
Strata + GPR37-Low + GPR37-High
1.00
1.00
1.00
BRCA
LGG
LUAD
Survival probability
0.75
Survival probability
0.75
Survival probability
0.75
p = 0.041
p < 0.001
p < 0.001
0.50
0.50
0.50
0.25
mOS: 131.5 vs 102.1
0.25
mOS: 44.6 vs 106.7
0.25
mOS: 38.2 vs 67.6
GPR37
GPR37
GPR37
0.00
0.00
0.00
0
24
48
72
96
120
0
24
48
72
96
120
0
24
48
72
96
120
months
months
months
Note the risk set sizes
Note the risk set sizes
Note the risk set sizes
Strata
110
51
24
16
7
4
Strata
363
187
73
42
21
14
Strata
282
132
54
29
12
7
984
559
311
186
100
40
151
58
19
10
7
2
224
93
26
11
5
3
0
24
48
72
96
120
0
24
48
72
96
120
0
24
48
72
96
120
months
months
months
F
Strata + GPR37L1-Low - GPR37L1-High
G
Strata + GPR37L1-Low + GPR37L1-High
H
Strata + GPR37L1-Low - GPR3TL1-High
1.00
1.00
1.00
BRCA
LGG
LUAD
Survival probability
0.75
Survival probability
0.75
Survival probability
0.75
p = 0.002
p < 0.001
0.50
p < 0.001
0.50
0.50
0.25
mOS: 132.0 vs 87.9
mOS: 31.7 vs 54.4
0.25
mOS: 62.9 vs 95.8
0.25
GPR37L1
GPR37L1
GPR37L1
0.00
0.00
0.00
0
24
48
72
96
120
0
24
48
72
96
120
0
24
48
72
96
120
months
months
months
Note the risk set sizes
Note the risk set sizes
Note the risk set sizes
Strata
112
61
29
18
7
4
Strata
401
199
77
46
24
13
Strata
374
169
68
34
15
8
982
549
306
184
100
40
113
46
15
6
4
3
132
56
12
6
2
2
0
24
48
72
96
120
0
24
48
72
96
120
0
24
48
72
96
120
months
months
months
cancers where GPR37L1 could act as a prognostic marker (Fig. 5G). Increased expression of GPR37L1 was asso- ciated with enhanced drug resistance in BRCA, UCEC, LGG, SKCM, PAAD, and STAD, but with enhanced drug sensitivity in LUAD, CESC, and LUSC (Fig. 5G). In LUAD, GPR37L1 expression was significantly negatively associ- ated with sensitivity to cisplatin, docetaxel, gemcitabine, and paclitaxel (Fig. 5H-K). Similarly, in LUSC, GPR37L1 expression was significantly negatively correlated with sen- sitivity to docetaxel (Fig. 5L).
3.5. The expression of GPR37 and GPR37L1 in association with immune infiltration and chemokines
Infiltration of macrophage M1 was significantly positively cor- related with GPR37 expression in BLCA and LGG but nega- tively correlated in STAD and UVM (Fig. 6A). Plasma B cell infiltration was significantly negatively correlated with GPR37 expression in BLCA and LUAD (Fig. 6A). Endothelial cell infil- tration was positively correlated with GPR37 expression in COAD, KIRP, PCPG, and TGCT, but negatively correlated in
A
B
Strata + OPRIT-Lo - OPŁAT-High
Strata + GPRITL1-Low + GPR37L1-High
Strata + OPR37-Low + OPRI7-High
Strata + OPR37L1-Low + GP/037L1-High
1.00
1.00
1.00-
1.00-
ACC
ACC
CESC
Survival probability
0.75
Survival probabilay
0,75
Survival probability
0,75
Survival probabilay
9.75
p = 0.018
p =0.017
p =0.020
0.50
p =0.024
0.50
0.50
0.50
GPR37L1
mOS: 136.2 vs 41.5
CESC
0.25
mOS: 39.9 vs NA
GPR37
0.25
MOS: NA vs 40.1
0.25
GPR37
0.25
MOS: 101.5 vs NA
GPR37L1
0.00
0 00
6.50
0.00
Q
24
48
120
·
24
48
72
06
120
months
12
9
9
24
40
months
12
96
120
moutha
0
24
48
months
12
96
120
Note the risk set sizes
Note the risk set sizes
Note the risk set sizes
Note the risk set sizes
Strata
60
51
26
13
7
3
Strata
25
15
5
3
1
0
Strata
31
10
4
3
2
1
Strata
32
18
9
6
4
3
19
8
4
3
2
2
54
44
25
13
8
5
273
134
60
32
23
16
272
126
55
29
21
14
0
24
48
72
90
120
0
24
48
72
96
120
0
24
48
72
96
120
months
months
moutha
0
24
48
72
06
120
months
C
Strata - GPR37-Low + GPR37-High
Strata - GPR37LT-Low *+ GPRI7LT-High
D
Strata
1.00
1.004
LUSC
LUSC
1.00
1.00
STAD
STAD
Survival probability
6.75
Survival probability
0.75
Survival probability
0.75
Survival probability
8.75
p = 0.023
p = 0.008
p = 0.021
0.50
0.50
0.50
p = 0.003
GPR37
mOS: 36.5 vs 19.6
0.50
025
mOS: 44.6 vs 66.1
mOS: 66.1 vs 43.8
0.25
mOS: 26.7 va $6.2
0 25
GPR37
GPR37L1 :
GPR37L1
4.00
4.00
9.00
0.00
0
24
48
months
12
96
120
0
24
48
12
months
96
120
0
24
48
12
90
120
0
24
48
72
months
months
90
120
Note the risk set sizes
Note the risk set sizes
Note the risk set siren
Note the risk set siren
Strata
204
102
50
24
15
10
Strata
263
112
49
25
14
6
Strata
99
33
7
1
0
0
Strata
40
5
1
0
0
0
291
120
60
32
14
7
232
118
61
31
15
11
273
72
16
7
4
1
369
105
23
8
4
1
0
24
48
72
50
120
0
months
24
48
72
06
120
months
.
24
48
12
months
96
120
0
24
48
12
months
9%
120
E
Strata + OPR57-Low ++ GPREST-High
Strata + OPRITLI-Low + GPRSTL1-High
F
Strata
-OPR37-Low + GPRS7-High
Strata
GPRSFL1-Low -+ GPR37L1-High
.00
1.00
1.00
UCEC
T
1.00-
UVM
UVM
Survival probability
0 75
Survival probability
a.Ts
Survival probability
975
Survival probability
8.7%
p = 0.013
UCEC
0 50
p < 0.001
p = 0.005
p =0.016
0.00
8.50
0.50
GPR37L1
0 25
MOS: NA vs 114.1
GPR37
0.25
MOS: NA vs NA
GPR37L1
0,25
MOS: NA vs 42.3
GPR37
025
mOS: 43.8 vs 52.7
1
0.00
0 00
0.00
₼ đô
0
24
48
months
12
96
120
6
24
48
72
monthe
06
120
6
24
48
72
06
120
6
24
45
12
96
120
months
months
Note the risk set sizes
Note the risk set sizes
Note the risk set sizes
Note the risk set sizes
Strata
471
282
137
68
21
5
Strata
477
286
142
73
22
6
Strata
43
22
1
0
0
0
Strata
-
26
21
4
1
0
0
60
35
17
12
4
2
54
31
12
7
3
1
37
23
6
2
0
0
54
24
3
0
0
0
24
48
72
96
120
months
0
24
48
months
12
56
120
A
24
48
months
72
96
120
0
24
48
months
72
95
120
THCA (Fig. 6A). CD8+ T cell infiltration was positively cor- related with GPR37 expression in BLCA, LGG, and READ, while negatively correlated in GBM, KIRC, LIHC, SKCM, TGCT, THYM, and UVM (Fig. 6A). On the other hand, Macrophage M1 infiltration was significantly positively cor- related with GPR37L1 expression in THYM, while showing a significant negative correlation in HNSC and STAD (Fig. 6B). Plasma B cell infiltration was significantly positively correlated with GPR37L1 expression in THYM (Fig. 6B). Endothelial cell infiltration was significantly positively correlated with GPR37L1 expression in TGCT and THCA but negatively cor- related in BRCA, HNSC, KIRC, KIRP, LIHC, LUAD, MESO, PAAD, SARC, and UCEC (Fig. 6B). CD8+ T cell infiltration exhibited significant negative correlations with GPR37L1 expression in BLCA, BRCA, CHOL, COAD, ESCA, GBM, HNSC, KIRP, LIHC, LUAD, LUSC, PAAD, PCPG, STAD, TGCT, THYM, and UCEC (Fig. 6B).
Significant negative correlations were found between GPR37 expression and CCL3, CCL4, CCL5, CCL7, CCL8, CCL14, CCL18, CCL22, CCL23, and CXCL12 in SKCM (Fig. 6C). Additionally, GPR37L1 expression was significantly nega- tively correlated with CCL2, CCL3, CCL4, CCL8, CCL11, CCL13, CCL14, CCL16, CCL20, CCL24, CCL26, CXCL1, CXCL2, CXCL3, CXCL5, CXCL6,CXCL12, XCL1, and XCL2 in GBM (Fig. 6D). Moreover, significant positive correlations were observed between GPR37 and CXCL1 in BRCA, LUSC, and THCA, and between GPR37 and CXCL5 in BRCA, ESCA, LGG, and THCA. GPR37L1 also exhibited significant positive
correlations with CXCL1 in CHOL, KIRC, KIRP, LIHC, and UCS, and with CXCL5 in KIRP, LIHC, UCS, and UVM (Figs. 6C, D).
3.6. The expression of GPR37 and GPR37L1 in association with immunotherapy
We collected 3 cohorts of cancer patients who underwent immu- notherapy, including their survival information and GPR37 and GPR37L1 expression levels. These cohorts are phs000452 and PRJEB23709_PD1 for SKCM,[50,51] and PRJNA482620 for GBM.[52]
In the phs000452 cohort, GPR37 expression was significantly higher in patients with progressive disease (PD)/stable disease (SD; progressive/stable disease) compared to those with PR/ CR (partial/complete response; P = . 02; Fig. 7A). Additionally, SKCM patients with low GPR37 expression showed a trend toward longer overall survival following immunotherapy, though this difference did not reach statistical significance (P =. 061; hazard ratio [HR] = 1.77, 95% CI 0.92-3.02; Fig. 7A). In the PRJEB23709_PD1 cohort, the median transcript per million of GPR37 was likely higher in patients with PD/SD compared to those with PR/CR (Fig. 7B), and SKCM patients with low GPR37 had significantly longer overall survival (P =. 023; HR = 2.53, 95% CI 0.92-6.99; Fig. 7B). In the PRJNA482620 cohort, the median transcript per million of GPR37 was likely lower in patients with PD/SD compared to those with PR/CR, and there
A
CESC
0.3
B
LUSC
0.2
200-
R =- 0.146, p <0.001
BLCA
…
LGG
0.1
150
UCEC
OV
0
Paclitaxel
100
UCEC
STAD
-0.1
BRCA
50
Carmustine
Cisplatin
Cyclophosphamide
Cytarabine
Dactinomycin
Docetaxel
Epirubicin
Fludarabine Gemcitabine
LUAD
-0.2
Irinotecan
Mitoxantrone
Nelarabine
Oxaliplatin
Paclitaxel
Temozolomide
Teniposide
Topotecan
Vinblastine
Vincristine
Vinorelbine
Vorinostat
5.Fluorouracil
0
-0.3
0.0
2.5
5.0
7.5
10.0
GPR37
GPR37
C
D
E
F
6000
R =- 0.068, p = 0.024
0.6
R =- 0.106, p <0.001
R =- 0.169, p < 0.001
15-
R =- 0.193, p < 0.001
Gemcitabine
4000
BRCA
BRCA
2000
LUAD
Vinorelbine
LUAD
0.4
10
Cisplatin
Gemcitabine
2000
0.2
1000
5
0
0.0
0
0
0.0
2.5
5.0
7.5
10.0
0.0
2.5
5.0
7.5
10.0
3
6
9
3
6
9
GPR37
GPR37
GPR37
GPR37
G
BRCA
0.3
UCEC
0.2
H
LUAD
…
KIRC
0.1
R =- 0.263, p < 0.001
THCA
0
CESC
2000
…
LUSC
-0.1
Cisplatin
LUAD
MESO
-0.2
1000
LIHC
LGG
-0.3
SKCM
0
PAAD
0
2
4
6
8
Carmustine
Cisplatin
Cyclophosphamide
Cytarabine
Dactinomycin
Docetaxel
Epirubicin
Fludarabine
Gemcitabine
Irinotecan Mitoxantrone
STAD
GPR37L1
Nelarabine
Oxaliplatin
Paclitaxel
Temozolomide
Teniposide
Topotecan
Vinblastine
Vincristine
Vinorelbine
Vorinostat
5.Fluorouracil
GPR37L1
V
K
L
200
R =- Q113, p=0.012
15-
R =- 0.256, p<0.001
800
R =- 0.125, p=0.005
0.08
R =- 0.134, p=0.003
150
LUAD
Gemcitabine
10
LUAD
600
LUAD
0.06
LUSC
Docetaxel
Paclitaxel
Docetaxel
100
400
0.04
5
50
200-
0.02
0
0
0
0.00
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0.0
2.5
5.0
7.5
GPR37L1
GPR37L1
GPR37L1
GPR37L1
was no significant difference in overall survival between GPR37- high and GPR37-low GBM patients (Fig. 7C).
In the phs000452 cohort, SKCM patients with high GPR37L1 expression demonstrated shorter overall survival compared to those with low GPR37L1 expression (P =. 033; HR = 1.85, 95% CI 1.02-3.75), despite no significant difference in GPR37L1 expres- sion between patients with PD/SD and those with PR/CR (Fig. 7D).
Similarly, in the PRJEB23709_PD1 cohort, SKCM patients with high GPR37L1 expression had also shorter overall survival compared to those with low GPR37L1 expression (P = . 007; HR = 4.03,95% CI 1.60-10.18; Fig. 7E). Notably, in the PRJNA482620 cohort, GBM patients with high GPR37L1 expression had a shorter median over- all survival (31.7 vs 54.2 months), but this difference did not reach statistical significance (Fig. 7F).
A
Macrophage M1 by CIBERSORT
0.4
…
Macrophage M1 by XCELL
B cell plasma by XCELL
0.2
B cell plasma by CIBERSORT
Endothelial cell by MCPCOUNTER
0
Endothelial cell by XCELL
T cell CD8 by TIMER
-0.2
T cell CD8 by CIBERSORT
T cell CD8 by MCPCOUNTER
-0.4
T cell CD8 by XCELL
ACC
BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LGG
LIHC
LUAD
LUSC
MESO
OV
PAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THCA
THYM
UCEC
UCS
UVM
GPR37
B
Macrophage M1 by CIBERSORT
0.6
Macrophage M1 by XCELL
0.4
B cell plasma by XCELL
B cell plasma by CIBERSORT
0.2
Endothelial cell by MCPCOUNTER
0
Endothelial cell by XCELL
-0.2
T cell CD8 by TIMER
-0.4
T cell CD8 by CIBERSORT
T cell CD8 by MCPCOUNTER
-0.6
ACC BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LGG
LIHC
LUAD
LUSC
MESO
OV
PAAD
PCPG PRAD
READ
SARC
SKCM
STAD TGCT
THCA THYM
T cell CD8 by XCELL
UCEC
UCS
UVM
GPR37L1
C
GPR37
D GPR37LI
es
CCL1
CCL1
CCL2
.
CCL3
0.5
CCL2
CCL3
0.5
CCL4
CCL4
CCL5
CCL5
0
CCL7
CCL7
CCL8
-05
.
CCL8
-0.5
CCL11
CCL11
CCL13
CCL13
CCL14
CCL14
·
CCL15
9
CCL15
CCL16
CCL 16
CCL17
CCL17
CCL18
CCL18
CCL19
CCL19
.
CCL20
·
9
CCL20
CCL21
CCL21
U
CCL22
9
CCL22
CCL23
CCL23
CCL24
CCL24
·
CCL25
9
CCL25
CCL26
CCL26
CCL27
CCL27
.
CCL28
·
CCL28
CXCL1
6
CXCL1
Q
CXCL2
CXCL2
CXCL3
CXCL3
CXCL5
P
CXCL5
0
CXCL6
9
CXCL6
CXCL9
CXCL9
0
O.
0
CXCL 10
.
0
CXCL10
9
CXCL 11
·
·
CXCL11
CXCL 12
CXCL12
CXCL13
CXCL13
a
CXCL 14
0 CXCL14
.
·
.
. CXCL 16
·
.
CXCL16
CXCL 17
9
CXCL17
·
XCL1
XCL1
9
4
XCL2
XCL2
0
®
2
D
0
0
0
0
0
9
CX3CL1
0
0
9
0
0
0
·
·
9
CX3CL1
ACC
GLCA
ARC
CESC
CHOL
COND
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LGG
LINC
LUAD
MESO
2
PAAD
POPO
PRAD
READ
GARC
SKCM
STAD
TGCT
INGA
THYM
UCEC
UVM
ACC
BECA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
NICH
KIRC
KIRD
LAML
LGG
LIHC
LUAD
LUSC
MESO
O
DAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THICA
THYM
UGEC
UVM
330
3.7. Experiments for the identification of TX14A as a ligand
To address the issue of conflicting reports from different research groups, we examined the activation potential of PSAP and TX14A on non-human GPR37 and GPR37L1. Chicken GPR37 and GPR37L1 were cloned, and their downstream signaling was assessed using luciferase reporter assays. The pGL3-CRE-Luciferase and pGL3-SRE-Luciferase systems were employed to measure cAMP production and ERK/mitogen- activated protein kinase pathway activity, respectively, in HEK293T cells,[61,62] with an empty pcDNA3.1(+) plasmid serv- ing as a negative control. Our results demonstrated that TX14A did not activate CRE or SRE signaling through chicken GPR37 (Fig. 8A, B), nor did it activate these pathways through chicken GPR37L1 (Fig. 8C, D). These observations suggest that PSAP
and its synthetic peptide TX14A are unlikely to induce down- stream signaling of GPR37 and GPR37L1 in HEK293T cells under the present experimental conditions.
4. Discussion
GPR37 was identified to be potentially influencing patient prog- nosis in our prior avian ovarian cancer model.[10] The current study extends this by revealing high-frequency genetic alter- ations in both GPR37 and its homologous ligand, GPR37L1 across OV, SKCM and UCEC (Fig. 1), which may suggest their pan-cancer significance. Besides confirming the estab- lished prognostic roles of GPR37/GPR37L1 in LUAD and LGG (Fig. 3),[4,6,11,63,64] we reveal significant associations in 7
A
B
C
response_NR 3 POSO F PR/CR
response_NR E=] PO/SO ET PR/CR
1
response_NR @ PO/SDE PRIOR
p=0.02
SKCM phs000452
V
a
p= 0.31
SKCM PRJEB23709_PD1
100
P = 0.93
GPR37
GBM
PRJNA482620
The TPM of GPR37
Survival probability
20
P = 0.061
GPR37
The TPM of GPR37
Survival probability
Survival probability
S
P = 0.023
GPR37
The TPM of GPR37
75
p=0.83
MOS: 20.8 va NA
HỆ T.TT: ĐƠN CHỐNG-0 82
HR.2.33: 93% C1 0.82-6.99
50
MOS: 44 2 vs 54.2
10
HR, 1.11; 85% CI 0.36-3.63
-
2
J
.
60
4
Q
-
0
-
25
4
-
-
Ne
0
Strate
34
23
4
2
·
0
30
21
4
Hoàn dha tích cơ sữaes
POISD
PRICR
53
23
5
2
·
PD/SO
PRICR
Strada
0
+
3
0
0
Strata
3
3
2
.
2
®
Đ
.
1
34
4
-
PD/SO
PRICR
31
-
12
?
·
®
-
-
4
-
-
12
D
E
F
response_NR E;3 POSO E PR/CR
response_NR 3 POISD E;] PRUCR
response_NR 3 PO/SO EI PR/CR
IL
-
1.00-
p= 0.93
1
SKCM phs000452
0.20
p =0.56
SKCM PRJEB23709-PD1
p= 0.68
40
The TPM of GPR37L 1
Survival probability
The TPM of GPR37L1
Survival probability
-
GBM
PRJNA482620
0.75
D = 0,033
0.15
P = 007
GPR37L1
The TPM of GPR37L1
GPR37L1
Survival probability
p = 0.066
GPR37L1
30
0,50
-
MOS: 9 868 22 3
HRU 1.AN: ĐỘNG CƠ 1.02-3,75
0.10
HR, 403, ĐỒNG CI 1.60-10 TH
2
MOS: 31.7 v9 54.2
0.25
-
54
0.05
10
:
12
38
-
-
TF
0.00
-
66
36
.
p
0.00
14
,
·
8
0
POISD
PR/CR
22
5
POISD
PRICR
Strate
5
5
.
Strata
21
5
16
·
·
1
-
-
na
POISO
PRICR
t
12
·
0
0
·
N
-
1
Fold change in luciferase > activity (/Control)
Fold change in luciferase w activity (/Control)
4
pcDNA3.1
GPR37
4
pcDNA3.1
GPR37
3
GPR37
CRE
3
GPR37
SRE
2
2
₹
1
1
0
0
-13
-12
-11
-10
-9
-8
-7
-6
-5
-13
-12
-11
-10
-9
-8
-7
-6
-5
Fold change in luciferase activity (/Control)
TX14A (log M, 6 h)
TX14A (log M, 6 h)
Fold change in luciferase activity (/Control)
4.
pcDNA3.1
GPR37L1
4
pcDNA3.1
GPR37L1
3
GPR37L1
CRE
3
GPR37L1
SRE
2
2
₹
1
1
0
0
-13
-12
-11
-10
-9
-8
-7
-6
-5
-13
-12
-11
-10
-9
-8
-7
-6
-5
TX14A (log M, 6 h)
TX14A (log M, 6 h)
additional malignancies, including breast (BRCA), adrenocorti- cal (ACC), cervical (CESC), lung squamous (LUSC), stomach (STAD), endometrial (UCEC), and UVM (Figs. 3 and 4). It has been reported that the upregulation of GPR37 enhances the proliferation, migration, and invasion of LUAD or LGG cells in vitro by inducing cell cycle arrest at the G1 phase,[2,3,5] whereas its knockdown suppresses these malignant phenotypes.[64] Similarly, elevated GPR37L1 expression in LGG corroborates our observed pro-proliferative effects of both receptors (Fig. 3D, G).[65] The REG4-GPR37 complex drives gastric cancer peri- toneal metastasis, aligning with its adverse prognostic role in STAD (Fig. 4D).[8] Notably, our study also identified GPR37L1 as being associated with favorable outcomes in STAD (Fig. 4D). Besides STAD, we observed different impacts of GPR37 and GPR37L1 on patient prognosis in ACC, CESC, LUSC, UCEC, and UVM (Fig. 4), which could reflect differences in molecular mechanisms but may also result from tumor heterogeneity or co-expression with other GPCRs rather than direct causal roles.
Chemotherapy remains important to cancer treatment despite therapeutic advances.[66,67] While GPR37 was previously associated with cisplatin response,[11] GPR37L1-chemotherapy relationships were unreported. We validated the cisplatin resis- tance phenotype from upregulated GPR37 consistent with previous findings (Fig. 5E).[11,63] Systematic analysis of 21 com- mon chemotherapeutic agents (Fig. 5A, G) revealed clinically significant association with GPR37 and GPR37L1 that war- rants further investigation (Figs. 5A, G). Notably, cisplatin and gemcitabine, both first-line treatments for lung cancer, showed negative correlations with GPR37 and GPR37L1 expres- sion (Fig. 5F, J, and H).[68,69] Similar negative correlations were observed for docetaxel and paclitaxel, which are widely used in LUAD and LUSC treatment (Fig. 5I, K, and L).[70,71] These asso- ciations extended beyond lung cancer, encompassing GPR37- paclitaxel correlations in UCEC patients (Fig. 5B),[72] as well as GPR37-gemcitabine/vinorelbine relationships in BRCA patients (Fig. 5C, D).[73,74] These findings suggest the possibility that both receptors are involved in pan-cancer chemotherapy resistance, and highlight the need for further functional validation to sup- port their potential as predictive biomarkers and therapeutic targets.
Immunotherapy has emerged as a cornerstone in oncology, yet most patients derive limited clinical benefit.[75,76] Predictive biomarkers are therefore critical for identifying immunotherapy responders.[77-79] Chemokines modulate both tumor cell dynam- ics and immune cell trafficking, directly influencing therapeutic outcomes.[80,81] It has been reported that an elevated expres- sion of GPR37 significantly increased the levels of CXCL1 and CXCL5 in the SW480 and DLD-1 cell lines.[7] Our pan-cancer analysis also reveals significant inverse correlations between GPR37 and CXCL1 in BRCA, LUSC, and THCA (Fig. 6C), and between GPR37 and CXCL5 in BRCA, ESCA, LGG, and THCA (Fig. 6C). Moreover, GPR37L1 exhibited a significant negative correlation with CXCL1 in CHOL, KIRC, KIRP, LIHC, and UCS, and with CXCL5 in KIRP, LIHC, UCS, and UVM (Fig. 6C, D). Besides chemokine regulation, GPR37 promotes macrophage phagocytosis and macroautophagy.[13-15,82] Our immune deconvolution analysis (CIBERSORT and XCELL) shows GPR37 expression positively correlates with M1 mac- rophages in BLCA/LGG, while GPR37L1 associates with M1 infiltration in THYM (Fig. 6A, B). Conversely, both receptors demonstrate significant negative correlations with CD8 + T cell infiltration across most cancers including SKCM and GBM (Fig. 6A, B), aligning with increased CD8 + T cells in GPR37- knockout models.[7] Critically, immunotherapy-treated mela- noma cohorts (phs000452, PRJEB23709_PD1) demonstrated that responders had lower expression levels of GPR37 and GPR37L1 compared to non-responders (Fig. 7A, B, D, and E). Additionally, KM survival curve analysis also demonstrated a better prognosis for SKCM patients with either low GPR37 or GPR37L1 expression (Fig. 7A, B, D, and E). These results, from
retrospective analyses, provide initial clinical evidence suggest- ing the potential of both receptors as immunotherapy response predictors, though functional validation in human systems is needed to confirm this role.
In the in vitro experiments of this study, the pGL3-SRE- Luciferase and pGL3-CRE-Luciferase reporter systems in HEK293T cells were employed to investigate whether GPR37 and GPR37L1 can be activated by TX14A. The luciferase reporter assay system with 2 response elements, CRE and SRE, was used to measure cAMP production and ERK/mitogen- activated protein kinase activity, respectively.[61,62,83] However, HEK293T cells with GPR37 or GPR37L1 overexpression did not exhibit a significant increase in CRE or SRE activity follow- ing TX14A treatment (Fig. 8A-D). The observation is in con- trast with several studies that reported PSAP and its synthetic analog TX14A can bind to GPR37 and GPR37L1, inducing ERK1/2 phosphorylation, calcium signaling, and inhibition of forskolin-stimulated cAMP levels in HEK293T cells.[19,23,84,85] Till present, artesunate (ARU), neuroprotectin D1 (NPD1), and osteocalcin (OCL) have been found to interact functionally with GPR37 to activate Gai-coupled receptor signaling.[13,14,86] In addition, recently, the pro-resolving lipid mediator maresin 1 (MaR1) was identified as a novel ligand for GPR37L1 through lipid overlay and protein pulldown experiments conducted by the team, who had previously identified NPD1 as a ligand for GPR37.[14,87] Since their computer simulations revealed that GPR37L1 could physically interact with MaR1 in the forms of hydrogen bonds rather than NPD1, indicating that GPR37L1 exhibits different molecular pharmacology compared to GPR37.[87] The potential ligands for the GPR37 and GPR37L1 thus need to be further investigated.
Although many novel findings concerning GPR37 and GPR37L1 were revealed, there are several limitations in this study that should be acknowledged. First, the prognostic assess- ment of GPR37 and GPR37L1 relied on retrospective data from public databases. This retrospective design cannot control for unmeasured confounders or establish causal relationships between receptor expression and clinical outcomes, and these analyses may be influenced by biases inherent in data collec- tion and processing. Second, our analysis of immunotherapy responses only included 3 public cohorts. The limited immu- notherapy datasets restrict the generalizability of results across a wider scope of cancer types and regimens. Third, past studies on the ligands of GPR37 and GPR37L1 has yielded conflict- ing findings. To provide new insight on cross-species ligand- receptoractivation, our in vitro assays used chicken instead of human GPR37 and GPR37L1. Unfortunately, the most prom- ising ligand PSAP/TX14A still failed to activate these chicken receptors, hinting the search for their ligands goes on. Fourth, further mechanistic validation experiments will be necessary to clarify how GPR37 and GPR37L1 regulate prognosis, chemo- therapy resistance, and immune infiltration. Future prospective studies, expanded immunotherapy cohorts, ligand identifica- tion experiments, and mechanistic experiments help strengthen the understanding of the clinical significance of GPR37 and GPR37L1.
5. Conclusions
Overall, we conducted a pan-cancer biomarker investigation of GPR37 and GPR37L1 leveraging retrospective cohorts with real-world clinico-genomic data and in vitro experiments. Our results suggest that both GPR37 and GPR37L1 warrant fur- ther validation as candidate biomarkers for prognosis and may serve as valuable references for guiding chemotherapy and immunotherapy across multiple cancers. Our study provides insights into the molecular basis underlying the role of GPR37 and GPR37L1 in tumor progression, providing preliminary evidence to support their potential as novel targets for cancer treatment.
Acknowledgments
The authors acknowledge the invaluable support from public databases, websites, and software used in this paper.
Author contributions
Formal analysis: Guoqiang Zhu, Juan Li.
Methodology: Guoqiang Zhu.
Software: Guoqiang Zhu, Jiliang He, Ningkun Shi, Susanna Chau Yi Wang.
Supervision: Mao Zhang, Yajun Wang.
Validation: Zhongyao Cai, Jiannan Zhang.
Writing - original draft: Guoqiang Zhu, Mao Zhang, Yajun Wang.
Writing - review & editing: Guoqiang Zhu, Mao Zhang, Yajun Wang.
References
[1] Liu F, Zhu C, Huang X, et al. A low level of GPR37 is associated with human hepatocellular carcinoma progression and poor patient sur- vival. Pathol Res Pract. 2014;210:885-92.
[2] Xie X, Cai X, Zhou F, et al. GPR37 promotes cancer growth by binding to CDK6 and represents a new theranostic target in lung adenocarci- noma. Pharmacol Res. 2022;183:106389.
[3] Wang J, Xu M, Li DD, Abudukelimu W, Zhou XH. GPR37 promotes the malignancy of lung adenocarcinoma via TGF-beta/Smad pathway. Open Med. 2021;16:24-32.
[4] Chen C, Wan M, Peng X, Zhang Q, Liu Y. GPR37-centered ceRNA net- work contributes to metastatic potential in lung adenocarcinoma: evi- dence from high-throughput sequencing. Transl Oncol. 2024;39:101819.
[5] Zhang Y, Wang L. Up-regulation of GPR37 promotes the proliferation of human glioma U251 cells. Chin J Cell Mol Immunol. 2018;34:341-5.
[6] Liang K, Guo Z, Zhang S, et al. GPR37 expression as a prognostic marker in gliomas: a bioinformatics-based analysis. Aging (Milano). 2023;15:10146-67.
[7] Zhou J, Xu W, Wu Y, et al. GPR37 promotes colorectal cancer liver metastases by enhancing the glycolysis and histone lactylation via Hippo pathway. Oncogene. 2023;42:3319-30.
[8] Wang H, Hu L, Zang M, et al. REG4 promotes peritoneal metastasis of gastric cancer through GPR37. Oncotarget. 2016;7:27874-88.
[9] Huang X, Wang Y, Nan X, et al. The role of the orphan G protein- coupled receptor 37 (GPR37) in multiple myeloma cells. Leuk Res. 2014;38:225-35.
[10] Zhu G, Wang X, Wang Y, et al. Comparative transcriptomic study on the ovarian cancer between chicken and human. Poult Sci. 2024;103:104021.
[11] Liu H, Zhu Y, Niu H, et al. Activation of PI3K/Akt pathway by G protein-coupled receptor 37 promotes resistance to cispla- tin-induced apoptosis in non-small cell lung cancer. Cancer Med. 2023;12:19777-93.
[12] Xu J, Bang S, Chen O, et al. Neuroprotectin D1 and GPR37 protect against chemotherapy-induced peripheral neuropathy and the transition from acute to chronic pain. Pharmacol Res. 2025;216:107746.
[13] Bang S, Donnelly CR, Luo X, et al. Activation of GPR37 in macro- phages confers protection against infection-induced sepsis and pain- like behaviour in mice. Nat Commun. 2021;12:1704.
[14] Bang S, Xie Y-K, Zhang Z-J, Wang Z, Xu Z-Z, Ji R-R. GPR37 regulates macrophage phagocytosis and resolution of inflammatory pain. J Clin Invest. 2018;128:3568-82.
[15] Bhattacharya P, Dhawan UK, Hussain MT, et al. Efferocytes release extracellular vesicles to resolve inflammation and tissue injury via prosaposin-GPR37 signaling. Cell Rep. 2023;42:112808.
[16] Mouhi S, Martin B, Owino S. Emerging roles for the orphan GPCRs, GPR37 and GPR37 L1, in stroke pathophysiology. Int J Mol Sci. 2022;23:4028.
[17] Bolinger AA, Frazier A, La J-H, Allen JA, Zhou J. Orphan G protein-coupled receptor GPR37 as an emerging therapeutic target. ACS Chem Neurosci. 2023;14:3318-34.
[18] Smith NJ. Drug discovery opportunities at the endothelin B receptor- related orphan G protein-coupled receptors, GPR37 and GPR37L1. Front Pharmacol. 2015;6:275.
[19] Meyer RC, Giddens MM, Schaefer SA, Hall RA. GPR37 and GPR37L1 are receptors for the neuroprotective and glioprotective factors prosap- tide and prosaposin. Proc Natl Acad Sci USA. 2013;110:9529-34.
[20] Costain WJ, Haqqani AS, Rasquinha I, et al. Proteomic analysis of synaptosomal protein expression reveals that cerebral ischemia alters lysosomal Psap processing. Proteomics. 2010;10:3272-91.
[21] Hiraiwa M, Liu J, Lu A-G, et al. Regulation of gene expression in response to brain injury: enhanced expression and alternative splicing of rat prosaposin (SGP-1) mRNA in injured brain. J Neurotrauma. 2003;20:755-65.
[22] Sano A, Matsuda S, Wen T-C, et al. Protection by prosaposin against ischemia-induced learning disability and neuronal loss. Biochem Biophys Res Commun. 1994;204:994-1000.
[23] Liu B, Mosienko V, Cardoso BV, et al. Glio-and neuro-protection by prosaposin is mediated by orphan G-protein coupled receptors GPR37L1 and GPR37. Glia. 2018;66:2414-26.
[24] Ngo T, Wilkins BP, So SS, et al. Orphan receptor GPR37L1 remains unliganded. Nat Chem Biol. 2021;17:383-6.
[25] Southern C, Cook JM, Neetoo-Isseljee Z, et al. Screening ß- arrestin recruitment for the identification of natural ligands for orphan G-protein-coupled receptors. J Biomol Screen. 2013;18:599-609.
[26] Coleman JLJ, Ngo T, Smythe RE, et al. The N-terminus of GPR37L1 is proteolytically processed by matrix metalloproteases. Sci Rep. 2020;10:19995.
[27] Giddens MM, Wong JC, Schroeder JP, et al. GPR37L1 modulates seizure susceptibility: evidence from mouse studies and analyses of a human GPR37L1 variant. Neurobiol Dis. 2017;106:181-90.
[28] Wheeler DL, Chappey C, Lash AE, et al. Database resources of the national center for biotechnology information. Nucleic Acids Res. 2000;28:10-4.
[29] Garcia P, Wang Y, Viallet J, Jilkova ZM. The chicken embryo model: a novel and relevant model for immune-based studies. Front Immunol. 2021;12:791081.
[30] Johnson PA, Giles JR. The hen as a model of ovarian cancer. Nat Rev Cancer. 2013;13:432-6.
[31] Jefferies B, Lenze F, Sathe A, et al. Non-invasive imaging of engineered human tumors in the living chicken embryo. Sci Rep. 2017;7:4991.
[32] Dodgson JB, Romanov MN. Use of chicken models for the analysis of human disease. Curr Protocols Human Genetics. 2004;40:15.5.1-15.5.12.
[33] Therneau TM, Lumley T. Package “survival.”. R Top Doc. 2015;128:28-33.
[34] Gordon M, Lumley T, Gordon MM. Package “forestplot.” Advanced forest plot using ‘grid’graphics. The Comprehensive R Archive Network, Vienna. 2019.
[35] Filbin MG, Tirosh I, Hovestadt V, et al. Developmental and onco- genic programs in H3K27M gliomas dissected by single-cell RNA-seq. Science. 2018;360:331-5.
[36] Han Y, Wang Y, Dong X, et al. TISCH2: expanded datasets and new tools for single-cell transcriptome analyses of the tumor microenviron- ment. Nucleic Acids Res. 2023;51:D1425-31.
[37] Yang W, Soares J, Greninger P, et al. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic biomarker discovery in can- cer cells. Nucleic Acids Res. 2012;41:D955-61.
[38] Bukowski K, Kciuk M, Kontek R. Mechanisms of multidrug resistance in cancer chemotherapy. Int J Mol Sci. 2020;21:3233.
[39] Maeser D, Gruener RF, Huang RS. oncoPredict: an R package for pre- dicting in vivo or cancer patient drug response and biomarkers from cell line screening data. Brief Bioinform. 2021;22:bbab260.
[40] Kassambara A. ggpubr:“ggplot2”based publication ready plots. R package version. 2018:2.
[41] Kolde R, Kolde MR. Package “pheatmap.” R package. 2015;1:790.
[42] Huinen ZR, Huijbers EJ, van Beijnum JR, Nowak-Sliwinska P, Griffioen AW. Anti-angiogenic agents - overcoming tumour endothe- lial cell anergy and improving immunotherapy outcomes. Nat Rev Clin Oncol. 2021;18:527-40.
[43] Sharonov GV, Serebrovskaya EO, Yuzhakova DV, Britanova OV, Chudakov DM. B cells, plasma cells and antibody repertoires in the tumour microenvironment. Nat Rev Immunol. 2020;20:294-307.
[44] Mills CD, Lenz LL, Harris RA. A breakthrough: macrophage-directed cancer immunotherapy. Cancer Res. 2016;76:513-6.
[45] Raskov H, Orhan A, Christensen JP, Gogenur I. Cytotoxic CD8(+) T cells in cancer and cancer immunotherapy. Br J Cancer. 2021;124:359-67.
[46] Becht E, Giraldo NA, Lacroix L, et al. Estimating the population abun- dance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol. 2016;17:1-20.
[47] Chen B, Khodadoust MS, Liu CL, Newman AM, Alizadeh AA. Profiling tumor infiltrating immune cells with CIBERSORT. Cancer Syst Biol. 2018;2:243-59.
[48] Li T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48:W509-14.
[49] Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol. 2017;18:220.
[50] Gide TN, Quek C, Menzies AM, et al. distinct immune cell popula- tions define response to Anti-PD-1 monotherapy and anti-PD-1/Anti- CTLA-4 combined therapy. Cancer Cell. 2019;35:238-55.e6.
[51] Van Allen EM, Miao D, Schilling B, et al. Genomic correlates of response to CTLA-4 blockade in metastatic melanoma. Science. 2015;350:207-11.
[52] Zhao J, Chen AX, Gartrell RD, et al. Immune and genomic correlates of response to anti-PD-1 immunotherapy in glioblastoma. Nat Med. 2019;25:462-9.
[53] Mermel CH, Schumacher SE, Hill B, Meyerson ML, Beroukhim R, Getz G. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. 2011;12:R41.
[54] Wan Y, Zhang Z, Lin D, et al. Characterization of CRH-Binding Protein (CRHBP) in chickens: molecular cloning, tissue distribution and inves- tigation of its role as a negative feedback regulator within the hypothal- amus-pituitary-adrenal axis. Genes. 2022;13:1680.
[55] Wan Y, Zhang J, Zhang X, et al. Characterization of the C5H11ORF96 gene in chickens: cloning, tissue distribution and investigation of its potential function in stress response regulation. Poult Sci. 2025;104:104500.
[56] Zhang X, Su J, Huang T, et al. Characterization of the chicken melano- cortin 5 receptor and its potential role in regulating hepatic glucolipid metabolism. Front Physiol. 2022;13:917712.
[57] Zhang X, Zhang J, Huang T, et al. SSTR2 mediates the inhibitory effect of SST/CST on lipolysis in chicken adipose tissue. Animals. 2024;14:1034.
[58] Wang X, Zhang J, Su J, et al. Genome-wide mapping of the binding sites of myocyte enhancer factor 2A in chicken primary myoblasts. Poult Sci. 2024;103:104097.
[59] Zhu G, Fang C, Mo C, Wang Y, Huang Y, Li J. Transcriptomic analysis of granulosa cell populations proximal and distal to the germinal disc of chicken preovulatory follicles. Sci Rep. 2021;11:4683.
[60] Zhu G, Fang C, Li J, Mo C, Wang Y, Li J. Transcriptomic diversification of granulosa cells during follicular development in chicken. Sci Rep. 2019;9:5462.
[61] Zhang X, Zhang J, Huang T, et al. SSTR2 Mediates the Inhibitory Effect of SST/CST on lipolysis in chicken adipose tissue. Animals (Basel). 2024;14:1034.
[62] Dijon NC, Nesheva DN, Holliday ND. Luciferase complementation approaches to measure GPCR signaling kinetics and bias. G Protein- Coupled Receptor Screening Assays: Methods and Protocols. Springer; 2021:249-74.
[63] Hu J, Meng F, Lv L, et al. GPR37-enhanced ubiquitination of ATP1A1 inhibits tumor progression and radiation resistance in esophageal squa- mous cell carcinoma. Cell Death Dis. 2024;15:933.
[64] Wang J, Xu M, Li D-D, Abudukelimu W, Zhou X-H. GPR37 promotes the malignancy of lung adenocarcinoma via TGF-B/Smad pathway. Open Med. 2020;16:024-32.
[65] Di Pietro C, La Sala G, Matteoni R, Marazziti D, Tocchini-Valentini GP. Genetic ablation of GPR37L1 delays tumor occurrence in Ptch1(±) mouse models of medulloblastoma. Exp Neurol. 2019;312:33-42.
[66] Tilsed CM, Fisher SA, Nowak AK, Lake RA, Lesterhuis WJ. Cancer chemotherapy: insights into cellular and tumor microenvironmental mechanisms of action. Front Oncol. 2022;12:960317.
[67] Behranvand N, Nasri F, Emameh RZ, et al. Chemotherapy: a double-edged sword in cancer treatment. Cancer Immunol Immun. 2022;71:507-26.
[68] Fennell DA, Summers Y, Cadranel J, et al. Cisplatin in the modern era: the backbone of first-line chemotherapy for non-small cell lung cancer. Cancer Treat Rev. 2016;44:42-50.
[69] Razak SAA, Gazzali AM, Fisol FA, et al. Advances in nanocarriers for effective delivery of docetaxel in the treatment of lung cancer: an over- view. Cancers (Basel). 2021;13:400.
[70] Li Z, Liu Z, Wu Y, et al. Efficacy and safety of apatinib alone or apa- tinib plus paclitaxel/docetaxel versus paclitaxel/docetaxel in the treat- ment of advanced non-small cell lung cancer: a meta-analysis. Thorac Cancer. 2021;12:2838-48.
[71] Sousa-Pimenta M, Estevinho LM, Szopa A, et al. Chemotherapeutic properties and side-effects associated with the clinical practice of ter- pene alkaloids: paclitaxel, docetaxel, and cabazitaxel. Front Pharmacol. 2023;14:1157306.
[72] Pignata S, Scambia G, Schettino C, et al. Carboplatin and pacli- taxel plus avelumab compared with carboplatin and paclitaxel in advanced or recurrent endometrial cancer (MITO END-3): a multi- centre, open-label, randomised, controlled, phase 2 trial. Lancet Oncol. 2023;24:286-96.
[73] Tolaney SM, Tarantino P, Graham N, et al. Adjuvant paclitaxel and trastuzumab for node-negative, HER2-positive breast cancer: final 10-year analysis of the open-label, single-arm, phase 2 APT trial. Lancet Oncol. 2023;24:273-85.
[74] Dan VM, Raveendran RS, Baby S. Resistance to intervention: paclitaxel in breast cancer. Mini Rev Med Chem. 2021;21:1237-68.
[75] Fukumura D, Kloepper J, Amoozgar Z, Duda DG, Jain RK. Enhancing cancer immunotherapy using antiangiogenics: opportunities and chal- lenges. Nat Rev Clin Oncol. 2018;15:325-40.
[76] Hammer C, Mellman I. Coming of age: human genomics and the can- cer-immune set point. Cancer Immunol Res. 2022;10:674-9.
[77] Gibney GT, Weiner LM, Atkins MB. Predictive biomarkers for check- point inhibitor-based immunotherapy. Lancet Oncol. 2016;17:e542-51.
[78] Sankar K, Ye JC, Li Z, Zheng L, Song W, Hu-Lieskovan S. The role of biomarkers in personalized immunotherapy. Biomark Res. 2022;10:32.
[79] Sharma P, Allison JP. The future of immune checkpoint therapy. Science. 2015;348:56-61.
[80] Nagarsheth N, Wicha MS, Zou W. Chemokines in the cancer micro- environment and their relevance in cancer immunotherapy. Nat Rev Immunol. 2017;17:559-72.
[81] Vilgelm AE, Richmond A. Chemokines modulate immune surveillance in tumorigenesis, metastasis, and response to immunotherapy. Front Immunol. 2019;10:333.
[82] Marazziti D, Di Pietro C, Golini E, Mandillo S, Matteoni R, Tocchini- Valentini GP. Induction of macroautophagy by overexpression of the Parkinson’s disease-associated GPR37 receptor. FASEB J. 2009;23:1978-87.
[83] Hauser AS, Attwood MM, Rask-Andersen M, Schioth HB, Gloriam DE. Trends in GPCR drug discovery: new agents, targets and indica- tions. Nat Rev Drug Discov. 2017;16:829-42.
[84] Lundius EG, Vukojevic V, Hertz E, et al. GPR37 protein trafficking to the plasma membrane regulated by prosaposin and GM1 gangliosides promotes cell viability. J Biol Chem. 2014;289:4660-73.
[85] Breitwieser GE, Cippitelli A, Wang Y, et al. Rare GPR37L1 variants reveal potential roles in anxiety and migraine disorders. bioRxiv. 2023:2023.07.05.547546.
[86] Qian Z, Liu C, Li H, et al. Osteocalcin alleviates lipopolysaccharide- induced acute inflammation via activation of GPR37 in macrophages. Biomedicines. 2022;10:1006.
[87] Bang S, Jiang C, Xu J, et al. Satellite glial GPR37L1 regulates maresin and potassium channel signaling for pain control. bioRxiv. 2023:2023.12.03.569787.