Previous studies regarding the transcriptic results of B-ALL patients have identified a number of lncRNAs and vital protein–coding genes that are associated with the etiology or clinical outcome of B-ALL [17–19]. To the best of the authors’ knowledge, although some studies have performed transcriptome profile analysis to investigate the role of lncRNA in the pathogenesis of B-ALL, the small sample sizes of parallel their controls and the restricted numbers of isolated cell types limited the general conclusions of these studies [20–22]. Moreover, microarray data prevented the identification of novel transcripts. In the present study, a comprehensive analysis of lncRNA and mRNA profiling data was conducted between B-ALL patients and CBD controls by NGS. Several differently expressed lncRNAs and mRNAs were identified, in which some were validated by qRT-PCR; functional annotations were also carried out. Furthermore, lncRNA-mRNA network analyses were preformed to extract candidate lncRNAs interacting with the aberrantly expressed mRNAs, revealing that five key lncRNAs were co–expressed with at least seven mRNAs in the network, bases on a very high threshold of 0.99 for the Pearson’s correlation. These findings could promote the understanding of the functional mechanism through which lncRNA interacts with mRNA, thus playing an important role in B–ALL pathogenesis.
The biological functions and potential pathways of mRNAs associated with differentially expressed lncRNAs were preliminarily predicted by GO and KEGG pathway analyses. Extraordinarily, significantly enriched pathways were closely associated with signal transduction, metabolic process, infections, and the immune system, which all play important roles in B-ALL development and disease treatment. The KEGG and co-expression analysis results showed that the Rap1 signaling pathway was exclusively highly enriched in B-ALL, when compared to controls. It was demonstrated that the Rap1 signaling pathway is associated with leukemia cell adhesion and migration, suggesting that Rap1 could participate the process of notch activation and leukemogenicity of T-ALL. This suggests that the Rap1 signal may function in the notch-dependent B-ALL development and its progression[14, 23, 24]. Infante et al. reported that Rap1b is a subtype of Rap1, and that its depletion could be used to reduce tissue invasion in T-ALL. Furthermore, another integrative network analysis of pediatric acute lymphoblastic leukemia revealed that gene expression and methylation consistently targeted the Rap1 signaling pathway [25]. In addition, the direct activation of Rap1 can lessen the cellular actions of leptin and correct glucose imbalance in obese mouse, which suggests that Rap1 is associated with energy metabolism [26]. VEGFA and FLT1, two differentially expressed genes, presented the highest correlation score of 0.99 by STRING in the Rap1 signaling pathway. Previous studies have shown that increased levels of Helios Treg cells promoted angiogenesis in the bone marrow of ALL mice via the VEGFA/VEGF receptor 2 pathway [27]. Furthermore, FLT-1 is generally expressed in pediatric ALL and VEGFA/FLT-1 signaling enhances the migration and survival of leukemia [28, 29]. FLT-1 activation on ALL cells results in cell migration and proliferation in vitro. In addition, FLT-1 neutralization affects leukemia localization, increases leukemia apoptosis, and impedes the exit of ALL cells, thus prolonging the survival of inoculated mice. All of the above results indicate that the VEGFA/FLT-1 interaction could play important roles in regulating the development of B-ALL. PRKCZ is a calcium- and diacylglycerol-independent serine/threonine protein kinase that functions in the phosphatidylinositol 3-kinase (PI3K) pathway and the mitogen-activated protein (MAP) kinase cascade. It is also involved in NF-kappa-B activation, mitogenic signaling, cell proliferation, and inflammatory response. AKT, PI3K2CB, and some other genes also feature in the pathway upstream of PRKCZ. The PI3K/Akt and NF-κB signaling pathways have been extensively demonstrated to participate in the regulation of migration, and in the viability of leukemic cells, both in vivo and in vitro [30–32]. It has also been reported that the knockdown of PRKCZ leads to more rapid losses of DNA mismatch repair enzyme (MSH2) proteins, resulting in significant reductions in DNA mismatch repair (MMR) and increased resistance to thiopurines for ALL [33]. Arsenic trioxide (ATO) has been demonstrated to interrupt the function of the PI3K/Akt pathway in ALL; PRKCZ may be responsible for the provocation of resistance to ATO [34]. The most aberrant upregulated lncRNA, ENST00000473898, showed a high correlation score of 0.95 with PRKCZ, suggesting that an in-depth study should be conducted to elucidate its potential regulatory mechanism.
The lncRNA-mRNA network constructed in the present study revealed that some lncRNAs may exert significant effects in B-ALL by regulating downstream mRNAs. Affinito et al. specially constructed a lncRNAs-mRNAs co-expression network in childhood B-ALL, based on previous RNA-sequencing experiments [35]. In said experiments, gene expression profiling using RNA-seq of leukemic cells purified from three pediatric B-ALL patients (compared with that observed in mature B cells from the peripheral blood of three healthy donors) highlighted that 24 key lncRNAs and their co-expressed mRNAs may play important roles in B-ALL pathogenesis [36]. Unlike this study, however, here larger sample sizes of B-ALL patients were used alongside quantity-matched controls. Furthermore, the control group used the same cell type of mononuclear cells derived from bone marrow. In addition, trans-patterns were used to predict the co-expression network of lncRNAs with mRNAs, unlike the guilt–by–association approach of the aforementioned previous study. The present study also obtained more high thresholds of FDR (<0.01) and high Pearson correlation scores (0.99). Here, five novel lncRNAs were identified, which interacted with at least seven mRNAs, including FAT1, SOX11, and EYA4. The functional roles of these lncRNAs remain unknown. Nevertheless, according to the alternative mRNAs in coexpression networks, these lncRNAs may exert exclusive characters by regulating downstream mRNAs. Among them, FAT1 was highly expressed in a large proportion of cases of T-ALL and B-ALL was implicated in Wnt signaling and hippo signaling. FAT1 was also found to cooperate with NOTCH in driving T-ALL in vivo; this suggests that it might be a potential biomarker in carcinogenic roles [37, 38]. Moreover, previous studies have shown SOX11 to be a useful diagnostic marker for mantle cell lymphoma [39]. It has been reported that the high expression of SOX11 leads to alterations of gene expression that are typically associated with cell adhesion, migration, and differentiation. Furthermore, its expression marks a group of patients with good outcomes [40]. Furthermore, Huang et al. demonstrated that AML1-ETO-fused protein triggers the epigenetic silencing of the EYA4 gene, contributing to leukemogenesis in t(8;21) AML. This suggests that the EYA4 gene might be a novel therapeutic target for AML. In addition, although here the co–expression lncRNA–mRNA pairs were captured from RNA–Seq data, it is well established that co–expression correlation does not imply causation. Thus, gene perturbation experimental data would be necessary to gain insight into the possible regulatory relationships, or to infer causal gene co–expression patterns.