LUSC is a malignant tumor with a high mortality rate but very few therapeutic options. Thus, it is critical to explore more reliable and applicable prognostic factors by based on multiple LUSC datasets. In the current bioinformatics analysis of three LUSC datasets, ALDH7A1, PAEP, and VAV2 were screened out for predicting reduced OS with the ability to improve the predictive accuracy of traditional CPPs. We also determined that the immune microenvironment and EMT might be important mechanisms underlying the key roles of ALDH7A1, PAEP, and VAV2 in the regulation of LUSC prognosis.
ALDH7A1 is member A1 of subfamily 7 of the aldehyde dehydrogenase superfamily (ALDHs), all of which are considered potential biomarkers of cancer stem/initiating cells and involved in detoxification, drug resistance, cell proliferation, differentiation, and oxidative-stress response via catalyzing the oxidation of endobiotic and exobiotic aldehyde compounds generated by a variety of factors including alcohol metabolism, lipid peroxidation, lysine catabolism, chemotherapies, etc(Stoll et al., 2019). There have been limited but deep explorations of the roles of ALDHs, including ALDH7A1, in the oncogenesis and development of cancers.
Accordingly, ALDH7A1 has been found to express in stem/precancerous/particular cancer cells and is supposed to be related to the oncogenesis of multiple cancers, such as prostate(van den Hoogen et al., 2011), colon(M. Y. Jiang et al., 2016), cervical(Prokopczyk, Sinha, Trushin, Freeman, & El-Bayoumy, 2009), ovarian(Saw et al., 2012) and esophageal squamous cell cancer(Wang et al., 2014), and superficial spreading melanoma(Rose et al., 2011). As for the roles of ALDH7A1 in cancer development and patients’ prognosis, a variety of results has been published. For example, the depletion of ALDH7A1 predicted poor clinical outcomes for patients with hepatocellular and renal clear cell carcinomas, with reduced cellular metabolism and decreased metabolism regulator PPARs observed(Andrejeva et al., 2018). In non-small cell lung cancer (NSCLC), high ALDH7A1 expression is associated with cisplatin resistance in vitro(Stoll et al., 2019), and predicted increased cancer recurrence as well as decreased recurrence-free and OS in stage I surgically treated American patients(Giacalone et al., 2013) but showed no effect on the progression-free survival or OS of European patients (CHEMORES cohort, www.chemores.ki.se)(Stoll et al., 2019). Here, we revealed ALDH7A1 as one of three genes, significantly predicting reduced OS in LUSC with the predictive accuracy and independence being well-validated by using multiple datasets and comparing results with traditional CPPs. Our result is consistent with but might be more applicable than that from the only other published study identifying ALDH7A1 as 1 of 14 prognostic genes in LUSC using bioinformatic analysis of only the TCGA dataset(J. Li, Wang, Chen, Yang, & Chen, 2017).
To elucidate the underlying mechanism(s) of ALDH7A1-mediated poor prognosis in LUSC, we further performed GSEA and correlation analysis. Interestingly, we found that a high level of ALDH7A1 was significantly enriched in the immune microenvironment and EMT pathways, and correlated with the level of mesenchymal biomarker vimentin but not with that of the epithelial indicator E-cadherin. Our results are well supported by previous reports that ALDH7A1 depletion leads to a paucity of immune effectors within the tumor bed in an NSCLC bioinformatic analysis(Stoll et al., 2019), and decreased the colonization, migration, and bone metastasis of human prostate cancer cells in both in vitro and in vivo studies(van den Hoogen et al., 2011). We also present novel evidence for the influence of ALDH7A1 on cellular EMT and correlation of ALDH7A1 with vimentin in LUSC, which supplements previous observations that ALDH7A1 knock-down reduces the expression of EMT-related factors (including snail, snail2, twist, N-cadherin, and osteopontin) in human prostate cancer cells(van den Hoogen et al., 2011). Further studies are warranted to explore whether ALDH7A1 could be used to predict the efficacy of immunotherapy and applied as a treatment target to prevent metastasis in LUSC.
VAV2 is a ubiquitously expressed member of the VAV proteins, which control cytoskeleton organization and are involved in multiple cellular processes, including the proliferation and migration of various cells, the stability of intercellular contacts and microenvironment, the formation of the immunological synapse, phagocytosis, and the aggregation and distribution of platelets(Hornstein, Alcover, & Katzav, 2004). Many clinical studies are showing the critical roles of VAV2 in mediating the invasive phenotype and poorer clinical outcomes in a variety of cancers, including gastric cancer(Tan et al., 2017), colorectal cancer(Sho, Court, Winograd, Russell, & Tomlinson, 2017), breast cancer(Y. Jiang et al., 2014), and adrenocortical carcinoma(Sbiera et al., 2017). To the best of our knowledge, we are the first to show the prognostic, predictive value of VAV2 in lung cancer, especially LUSC.
As for the mechanisms, our current bioinformatics analysis using LUSC high-throughput data uncovered a significant correlation between VAV2 with the immune microenvironment and EMT and further revealed a consistent level of VAV2 with the EMT-inducer vimentin. The VAV2-mediated immune response is mostly reported for various immune effectors such as T cells(Fray, Charpentier, Sylvain, Seminario, & Bunnell, 2020), B cells(Doody et al., 2001), and macrophages(Y. Li et al., 2012), but rarely reported in cancers other than in a study showing VAV2 recruitment of pro-inflammatory cells to the microenvironment to promote oncogenesis in skin cancers. In both in vitro and in vivo models, VAV2 has been identified as a pro-metastasis factor in cancers such as gastric(Tan et al., 2017), lung(Brantley-Sieders et al., 2009), head and neck squamous carcinomas(Patel et al., 2007), sarcomas(Tie et al., 2016), and melanomas(Brantley-Sieders et al., 2009). However, studies of the relationship of VAV2 with EMT and its inducer vimentin in cancers are limited. VAV2 was declared as promoting EMT in NSCLC cells, but mediating EMT's bidirectional regulation in breast cancer. Additionally, vimentin has been shown to be necessary for the phosphorylation and activity of VAV2 during the promotion of migration, invasion and metastasis of lung cancer cells. Our current study not only presents further evidence for a VAV2-regulated immune response and cancer invasion, but also indicates future directions for the study of VAV2 in LUSC.
PAEP encodes a secreted immunosuppressive glycoprotein, glycodelin, which has been reported in several types of cancers. Based on these studies, PAEP/glycodelin is detectable in serum, tissue, and cystic fluid(Jeschke et al., 2009) and has been demonstrated by most studies to promote oncogenesis, proliferation, differentiation, invasion, angiogenesis, and immune surveillance of cancer cells. However, it exerts a contradictory influence on the metastasis/recurrence of disease and patients' outcome. Moreover, the expression of PAEP might be influenced by multiple inducers such as phorbol 12myristate 13 acetate (PMA), lysophosphatidic acid (LPA), mitogenactivated protein kinase (MEK)1/2, protein kinase A, GATAbinding protein 3 (GATA3), the microphthalmiaassociated transcription factor (MITF) and the specific protein1 (Sp1), and its suppressor Krüppellike transcription factor11 (KLF11). (Weber et al., 2019)
In NSCLC, PAEP/glycodelin was originally demonstrated as a serum and tissue biomarker that predicted cancer recurrence/metastasis and reduced OS, but was further observed to downregulate immune system modulators such as PDL1 and proliferation stimulators such as HBEGF and to enhance the cellular migration by unclarified mechanisms. In LUSC, a single study showed that the cellular PAEP/glycodelin expression could be stimulated by the canonical transforming growth factor (TGF)β pathway, the MEK/extracellular-signal-regulated kinase (ERK) pathway, and the protein kinase C signaling cascade, all of which are the targets of clinically approved or investigated drugs for NSCLC(Weber et al., 2019).
Currently, PAEP is known to independently predict reduced OS of patients and increase the prognosis prediction efficacy of CPPs in LUSC, and that the involvement of PAEP in the immune microenvironment might be important for LUSC development. For the first time, a significant correlation between PAEP and EMT has been found using LUSC clinical specimens, and the PAEP–EMT signaling should be further studied to explore the detailed effects on LUSC progression, as well as the potential clinical value in other types of cancers.
Based on the current results and previous publications, the following issues are notable and deserve further clarification. First, our findings are based on high-throughput bioinformatic analysis and on-line datasets and need further experimental and clinical validation. Second, more studies should be done to compare the expression levels of ALDH7A1, PAEP, and VAV2 in cancer and non-cancer tissues and illustrate the leading-edge genes and the co-expression network related to ALDH7A1, PAEP, and VAV2. Third, previous studies have shown that ALDH7A1, PAEP, and VAV2, respectively, belong to different families with different molecular functions in cells. In contrast, we proved consistent relationships between the three factors with the immune microenvironment and EMT, raising an interesting question about whether ALDH7A1, PAEP, and VAV2 have interactions/cooperation/negotiations in the processes of immune microenvironment/EMT-induced progression of cancers, including LUSC. Fourth, immunotherapy has led to significant clinical benefit in LUSC, whereas the corresponding immunotherapy targets, such as PD-L1, have failed to predict LUSC patients' prognosis in the published clinical studies.
Furthermore, TMB has been demonstrated by some researchers to play a predictive role in response to immunotherapy(Doroshow et al., 2019; Paik, Pillai, Lathan, Velasco, & Papadimitrakopoulou, 2019). Here, we revealed that ALDH7A1, PAEP, and VAV2, respectively, indicated the OS of patients and were significantly related with immune microenvironment signals in LUSC, but had no relationship with either PD-L1 (data not shown) or TMB. It is supposed that ALDH7A1, PAEP, and VAV2 might be used as biomarkers to identify patients most likely to benefit from cancer treatment, including immunotherapy, and might help clarify primary and acquired resistance to LUSC.