Bioinformatics and survival analysis of glia maturation factor-γ in pan-cancers

Glia maturation factor-γ (GMFG) is a 17 kDa small protein, and its gene sequence is conserved from yeast to mammalian. GMFG was initially identified as a glia maturation factor that can induce brain cell differentiation [1, 2]. Later reports found that GMFG can regulate the actin cytoskeleton because it actually belongs to the actin-depolymerizing factor homology (ADF-H) family [3, 4]. All ADF-H proteins can remodel actin cytoskeleton through binding to either actin and/or actin-related proteins (Arps) [5], and the reorganization of actin cytoskeleton is crucial for immune system function, angiogenesis and cell motility [6]. As a novel regulator of the Arp2/3 complex, GMFG is reported to enhance the angiogenic sprouting of zebrafish [7] and to regulate the migration of airway smooth muscle cell [8‐10], as well as to promote the chemotaxis of neutrophils and T lymphocytes [11, 12].

Considering the main function of GMFG in actin remodeling vital for cancer immunity, angiogenesis, cell division and motility [13, 14], GMFG is supposed to have important roles in tumor development, while up to now, only two studies described the role of GMFG in cancers [15, 16]. These two studies demonstrated that GMFG could promote the migration and proliferation of ovarian and colorectal cancer cells, and the high expression of GMFG was related to poor survival outcome for ovarian cancer patients. To better understand the potential function of GMFG in pan-cancers and to draw more attention to GMFG for the future cancer research, in this study, we investigated the clinical values of GMFG using The Cancer Genome Atlas (TCGA) data, and analyzed the underlying functional mechanisms of GMFG through conducting Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichments. Furthermore, we also evaluated the impact of GMFG on cancer immunity by analysis of the association between GMFG and immunomodulators.

Our study found that, compared with in normal samples, GMFG was increased in six cancers (GBM, KIRC, LGG, LAML, PAAD, and TGCT), but was decreased in three cancers (KICH, LUAD, and LUSC). In addition, high expression of GMFG was associated with early pathological stage in four cancers (BLCA, LUAD, SKCM, and THCA), but was correlated with advanced pathological stage in STAD. Moreover, the associations of GMFG expression with OS and DFS were also investigated, and high expression of GMFG predicted worse OS in four cancers (GBM, LGG, LUSC, and UVM), and worse DFS for LGG and PRAD, but was associated with better OS in SKCM and THYM, better DFS in CHOL. These findings demonstrated significant differences in the expression patterns and prognostic values of GMFG in different cancers, indicating that GMFG may be a valuable biomarker for the diagnosis and prognosis of some cancers.

High expression of GMFG predicted different prognoses for different cancers. A possible explanation for this result is that GMFG functions differently in different cancers. First of all, combining the previous reports, we conclude that GMFG may enhance the motility of both cancer cells and immunocytes, because GMFG is mainly involved in protein binding in most cancers according to our study, which is consistent with the fact that GMFG binds to Arp2/3 complex to depolymerize the actin cytoskeleton [13]. In the literature, GMFG promotes the process of angiogenesis [7], and the proliferation and motility of cancer cell through altering the actin cytoskeleton [15, 16]. Thus, our research data suggest that GMFG plays an important role in tumor development largely through the protein-protein interaction. On the other hand, GMFG is also reported to enhance the chemotaxis of neutrophils and T lymphocytes through remodeling the actin cytoskeleton [11, 12]. In our study, GMFG is involved in immune response including T cell receptor signaling pathway, chemotaxis, T cell co-stimulation, leukocyte migration and T cell proliferation regulation, which may also be associated with the function of actin remodeling. Additionally, the biological process of immune response was also found in the KEGG pathways, such as Natural killer cell mediated cytotoxicity and B cell receptor signaling pathway. Taken together, that GMFG enhances both the cancer progression and the immune defense, and patients’ survival outcome may depend on the balance between the speed of cancer development and the ability of immune system against cancer tissues.

Another mechanism causing the different survival outcomes may be that GMFG appears strongly correlated with different immunomodulators including co-stimulators, co-inhibitors, ligands, receptors, cell adhesion and antigen presentation molecules in different cancers. Meanwhile, the strongest correlations between GMFG and immunomodulators were found in TGCT, SKCM, UVM, LUSC and COADREAD. However, GMFG was negatively correlated with most of the immunomodulators in THYM, and was weakly correlated with the immunomodulators of DLBC, LAML, and LIHC. These findings indicate that GMFG may participate in the regulation of cancer immunology. Recent studies reported that GMFG regulates monocyte migration by modulating ITGB1 [22], and functions as a negative regulator of Toll-like receptor 4 (TLR4) signaling through facilitating TLR4 endocytic trafficking in macrophages [23]. Interestingly, our study also found a strong correlation between GMFG and TLR4 as well as ITGB2. The significant associations between GMFG and immunomodulators show that GMFG modulates the migration, adhesion and activation of immunocytes through regulating immunomodulators, thus functions in the immune response of various cancers. Besides, GMFG is a cytokine-responsive protein mediating the pluripotential and lineage commitment of human hematopoietic stem cells [24], and is downregulated in the process of erythroid maturation and the response to LPS [25]. Apart from immune response, we show that GMFG also participates in cytokine-cytokine receptor interaction and hematopoietic cell lineage pathways, which is consistent with the previous studies. Furthermore, a study [26] reported that GMFG modulates the iron metabolism and M2 polarization of macrophage via inducing mitochondrial ROS and serves as a novel therapeutic target in immune and metabolic disorders. Therefore, GMFG may also act as a regulator of immunomodulators and a cytokine-responsive protein in modulating tumor immunity and hematopoiesis in addition to serving as an ADF-H protein in remodeling the actin cytoskeleton. GMFG may involve in the cancer immunity and hematopoiesis through interacting with different immunomodulators and cytokines, leading to different immune response status and tumor progression in different types of cancer.

Moreover, we found that GMFG had a significantly positive association with PD-1 and PD-L2, but not with PD-L1 in all cancers except for THYM, DLBC, LAML, and LIHC. Interestingly, the expression of GMFG was also obviously higher in THYM, DLBC and LAML than that in other cancer types, and GMFG was mainly enriched in the plasma membrane in most cancers, but was strongly concentrated on the nucleoplasm or extracellular exosome in THYM, DLBC and LAML. A previous bioinformatics study also demonstrated a weak correlation between PD-1 and other immunomodulators in THYM, LAML, and DLBC, while a strong correlation among them was observed in other types of cancer [27]. These results imply that GMFG has a weak interaction with immunomodulators and a different role in THYM, LAML and DLBC, and these immune and hematopoiesis system-related tumors may have a different immune tumor microenvironment (TME) compared to other solid cancers, which has been proved in recent studies [28‐30]. PD-1 is a critical immune checkpoint in TME [31] which functions in adaptive resistance for cancer in immune escape [32]. PD-L1 and PD-L2 are important co-inhibitors for the immune inhibitory impact of PD-1 [33], PD-L1 is expressed in both tumor-infiltrating immune cells and tumor cells [34]. While it was initially suggested that PD-L2 expression is much more restricted than PD-L1 expression, and PD-L2 is expressed mainly in immune cells [35], but recent studies demonstrate that PD-L2 is also expressed in various types of tumor cells, depending on the different TME situations [36, 37]. What’s more, PD-L2 is also found to be significantly associated with the progression of some cancers [38, 39]. Thus, GMFG may interact with PD-1 and PD-L2 to promote the immune escape and progression of cancers.

Our study preliminary identified the potentially functional mechanisms of GMFG including protein binding, immune response and cytokine-cytokine receptor interaction. These findings may help to understand the real mechanisms of GMFG in regulating the tumor progression, immune response status and TME in different cancers. Therefore, in order to fully understand the role of GMFG, future cancer research on GMFG should not only focus on the function of GMFG in modulating of the actin cytoskeleton of cancer cell, cancer-associated vascular endothelial cells or immunocytes, but also pay attention to the cytokine-responsive function and immunomodulation impact of this factor, as well as its role in the tumor hematopoiesis and immune TME, especially its impacts on the motility and activation of the tumor-infiltrating immune cells, since the immune composition of the TME including the innate and adaptive immunocytes is a critical determinant of tumor development [40]. Meanwhile, the crosstalk between GMFG and immunomodulators should also be addressed in future studies. What’s more, our study also provides the overlapping co-expression genes in the cytokine-cytokine receptor interaction pathways, future studies may investigate the interaction between GMFG and these cytokines to better understand the role of GMFG in cancers. However, there are some limitations should be mentioned in this study. First, results of this study were mainly obtained by bioinformatics, the detailed mechanisms have not been verified by experiments. Thus, future experimental research should be performed to prove these pathways. Second, except for pathological stages, we failed to analyze the relationship between GMFG and other clinicopathological parameters of pan-cancers. In addition, this study mainly discussed GMFG-related KEGG pathways of cytokine-cytokine receptor interaction, the biological process of immune response, and the molecular function of protein binding, only limited attentions were paid to other GO/KEGG pathways such as cell adhesion molecules, signal transduction and receptor activity, which might also be critical. Finally, we mainly focused on mining and discussing the potential signaling of most cancers since they shared the same GO/KEGG pathways, we did not pay too much attention to the pathways of DLBC, LAML, and LIHC that were different from that of most cancer types.

Our study demonstrates that the expression of GMFG varies in different cancers, and high expression of GMFG predicts different prognoses for different cancers. GMFG is mainly involved in immune response, protein binding and cytokine-cytokine receptor interaction pathways, and is positively associated with immunomodulators in most cancers. Overall, our findings may help to understand the real mechanisms of GMFG in regulating the tumor progression, immune response status and tissue-specific TME of different cancers.