lncRNA-Associated ceRNA Networks in Spleen of Nocth1-correlated T-ALL Leukemia Mice

Background Acute T-lymphocytic leukemia (T-ALL) is a highly aggressive malignant tumor in leukemia. Nocth1 is considered as an major oncogene in the development of T-ALL. Increasing evidences have revealed that the occurrence and progression of T-ALL referred to abnormal gene expression, pathway activation and the regulation between these genes. However, the potential lncRNA-associated competing endogenous RNA (ceRNA) network involved in spleen of Nocth1-correlated T-ALL leukemia mice remains unclear. Methods Overexpression of Notch intracellular domain (ICN) of Notch1 by retroviral infection was used to set up mouse T-ALL model. Deep RNA-sequencing analysis was performed the expression of lncRNAs and mRNA in spleen of T-ALL mice and C57BL/6 mice. Results The deep RNA-sequencing analysis shown that 1833 lncRNAs and 4626 mRNAs were deregulated according to the P-value (p<0.05) and fold change (>2-fold) in spleen of T-ALL leukemia mice compared with that of C57BL/6 mice. Gene Ontology(GO) and KEGG pathway analysis were performed to reveal the potential roles of differentiated expressed lncRNAs. Co-expression Network was performed to reveal the regulation relationship between the differentiated expressed lncRNAs and mRNAs. CeRNA prediction constructed the lncRNA-miRNA-mRNA model to find the core ceRNA based on regression model analysis and seed sequence matching methods. Conclusion This study provided a systematic overview of the altered lncRNAs and mRNA expression, pathway and ceRNA regulation network in the pathogenesis of Nocth1-correlated T-ALL.


3
ALL accounts for 20-25% of the total incidence of acute lymphoblastic leukemia in adults, for 10%-15% of childhood acute lymphoblastic leukemia [2]. The abnormal proliferation T cell can infiltration and damage to various of tissues and affect organ function. Despite the combined chemotherapy and allogeneic hematopoietic stem cell transplantation have applied to clinical treatment of T-ALL, its event-free survival rate is only 30%-50% [3].
Explore the mechanism and pathogenesis of T-ALL is of great significance in improving survival rate.
The formation of T-ALL is a multi-step process which includes the activation of oncogenes and inactivation of tumor suppressor genes [4]. Notch1 is a subtype of Notch receptor in Notch signaling pathway which participated in multiple pathological and physiological processes, including cancer [5]. Approximately 55% of T-ALL patients are related with acquired functional mutations in Notch1 [6]. Block Notch signaling pathway causes cell cycle arrest and apoptosis in T-ALL cell lines [7]. Notch1 is a class I transmembrane protein which transduces information from extracellular signals into nucleus directly [8]. The Notch intracellular domain (ICN) of the Notch1 is its active component, which can activate expression of target genes in nucleus. Overexpressed of ICN1 by retroviral infection in hematopoietic progenitor cells or thymocytes promote T ALL tumorigenesis, which was made to set up mouse T-ALL model [9].
lncRNAs are a special class of non-coding RNAs (ncRNAs), participated with varies of physiological cellular processes as well as cancer pathological process [10]. Increasing evidences have revealed that lncRNAs play a critical role in many types of cancers, including hepatocellular carcinoma, renal cell carcinoma and leukemia [11]. Investigating lncRNAs specifcally transcriptional profiles in T-ALL is of great significance to understand the gobal altered expression of lncRNAs. lncRNAs have diverse mechanism to regulate gene expression. Some studies show that lncRNAs act as competing endogenous RNAs 4 (ceRNA) through a lncRNA-miRNA-mRNA model to regulate gene expression [12]. Based on the lncRNAs and mRNAs specifcally transcriptional profiles, construction the ceRNA network can enrich the raw data in studying the potential mechanism in the formation of T-ALL.
In the present study, we performed the deep RNA-sequence to analyze expression profile in spleen of T-ALL leukemia mice with the aim of explore the lncRNAs catalogue. We showed the differentially expressed lncRNAs and mRNAs. Gene Ontology(GO) and KEGG pathway analysis were performed to reveal the potential roles of these mRNAs. Coexpression Network was performed to reveal the regulation relationship between the differentially expressed lncRNAs and mRNAs. Meanwhile, we performed ceRNA prediction to construct the lncRNA-miRNA-mRNA model. Thus, this study provided a systematic overview of the altered RNAs expression, pathway and ceRNA network in the pathogenesis of Nocth1-correlated T-ALL.

Animals and tissues
We used T-ALL leukemia mice overexpressing the Notch l intracellular domain (NICD) as the research model. T-ALL leukemia cells presented by the Institute of Hematology, Peking Union Medical College, Chinese Academy of Medical Sciences. We cotransfected with retroviral plasmid MSCV-ICN1-IRES-GFP (ICN1-GFP), the reverse transcription packaging protein CMV-VSVG and and Kat into 293T. C57BL/6 mice were divided into two groups randomly, T-ALL mice and C57BL/6 mice, and each group has three mice. T-ALL mice group were treated as follows, collected the viral supernatant to infect C57BL/6 mouse bone marrow cells (Lin-Scal + ), and then transplanted into C57BL/6 mice (10 6 per mouse)after semi-lethal dose irradiation through tail vein injection. C57BL/6 mice group 5 were transplanted culture medium in the same way. Mice were numbered, divided into three cages randomly and reared routinely. C57BL/6 mice, 6~8 weeks female, were purchased from Charles River Experimental Animal Technology Co., Ltd. (Beijing, China). 2 weeks after, the mice were sacrificed by cervical dislocation and removed the spleen by laparotomy (Fig.1A). Mice quarantined in a 12 h light and 12 h dark photoperiod pathogen free environment, received water and food in clean class animal room of Hebei Medical University. All animals were housed and cared in accordance with the Declaration of Helsinki and the guidelines and regulations of the Institutional Animal Care and Use Committee of the Second Hospital of Hebei Medical University.

RNA-Seq
RNeasy mini kit (Qiagen, Germany) was used to isolate the total RNA. TruSeq™

qRT-PCR
Differently expressed lncRNAs were selected for validation by qRT-PCR. Total RNA was 6 reverse-transcribed using SureScript™ First-Strand cDNA Synthesis Kit (Genecopoeia, USA) according to the manufacturer's guide. qPCR reactions was performed using SYBR-Green (Invitrogen) according to the manufacturer's guide. BioRad iQ5 Real-Time thermocycler was used to perform qPCR reactions. The Cycling conditions was denaturing at 95°C, followed by 39 cycles of 95℃ (10 s) and 55℃ annealing (30 s). Specific primers of each lncRNA were listed in Table 2.

Statistical analysis
The differentially expression levels of lncRNAs and mRNAs in T-ALL mice or C57BL/6 mice spleen was analyzed by Bioconductor package (limma version 3.26.1) and R (version 3.2.2) software. Spearman correlation test was used to analyze co-expression relationships between the lncRNAs. CeRNA prediction was analyzed by Pearson's correlation coefficients. qRT-PCR data was shown as the means±S.E.M. Differences between two groups were analyzed by Student's t test. P<0.05 was considered as significantly.

1.
Deep RNA-sequence lncRNAs and mRNA expression profiles in spleen of T-ALL mice.
Before injection the mice were healthy and two weeks later, deep RNA-sequence was  (Fig.1DE).

mRNA GO and KEGG pathway analysis
In order to clarify the biological processes, cellular components and molecular functions of differentially expressed mRNAs, we performed GO terms enrichment and KEGG pathway analysis. The GO terms enrichment for differentially expressed mRNAs were related to melanocyte differentiation, myosin complex and translation repressor activity in biological processes( Fig.2A), cellular component( Fig.2B) or molecular function (Fig.2C), respectively.
KEGG pathway analysis showed that showed that 303 pathways were significantly enriched among the transcripts (Fig.2D). Acute myeloid leukemia, protein export and glycosaminoglycan biosynthesis -keratan sulfate were the 3 significantly enriched pathways.

Coding/non-coding co-expression analysis
To predict the functions of lncRNAs, we constructed lncRNA-mRNA co-expression network (Fig.3).

Construction of a ceRNA network.
lncRNAs could function as ceRNA to competing the binding between miRNA and mRNA.
Analysis of ceRNA network helped to understand the characterization of lncRNAs. As shown in Fig.4, 71 differentially expressed lncRNAs and 123 expressed mRNAs were selected, which predicted 11 miRNAs sharing binding sites with the differentially expressed lncRNAs and mRNAs.

Validation of differentially lncRNAs
Nice differentially expressed lncRNAs from ceRNA network prediction were selected to confirm RNA sequence results by qRT-PCR (Table 1). Ten pairs of mice spleen which contain both T-ALL leukemia mice or C57BL/6 mice were selected. 10 downregulated lncRNAs(p<0.05; FC>5-fold) were selected. All these lnRNAs expression were downregulated and consistent with RNA sequence results(P<0.05). Moreover, NONMMUT026003.2 was maximal changed lncRNAs in these lncRNAs.

Discussion
Notch1 is one of the major driving oncogene in T-ALL. About 55% of T-ALL patients occurred Notch1 mutation in the trans-membrane region and the intracellular PEST domain, which resulted abnormal activation of the Notch signaling pathway.

Overexpression of ICN1 by retroviral infection in hematopoietic progenitor cells or
thymocytes promote T ALL tumorigenesis, which was made to set up mouse T-ALL model.
Over the past decades, the molecular mechanism of Notch1-correlated T-ALL has been extensively investigated. However, the precise pathogenesis of Notch1-correlated T-ALL is still unknown. Recent years, ncRNAs, including lncRNAs, have been found to be related with humorous number of biological regulatory functions. lncRNA NALT activating Notch signaling pathway promoted cell proliferation in T-ALL [13]. lncRNA-IUR acted as a tumor suppressors by suppressing the STAT5-CD71 pathway in Bcr-Abl-mediated tumorigenesis of T-ALL [14]. To further confirm the significant differences in the expression of lncRNAs and mRNAs, we removed the spleen from T-ALL leukemia mice and C57BL/6 mice and performed deep RNA sequence to study the different expression of lncRNAs and mRNAs.
We found that lncRNAs' expression altered significantly in the spleen of Notch1-correlated T-ALL leukemia mice compared with that of C57BL/6 mice which was the first time to report the altered expression of lncRNAs. 1873 lncRNAs and 5626 mRNAs were differentially expressed in the spleen from T-ALL leukemia mice compared with that of C57BL/6 mice.

9
we performed GO terms enrichment to study the biological functions of differently expressed mRNAs. The most enriched GO terms for differentially expressed mRNAs were related to melanocyte differentiation, myosin complex and translation repressor activity in biological processes, cellular component or molecular function. To understand the function of differentially expressed mRNAs further, we performed KEGG pathway analysis and found that 303 pathways were significantly enriched among the altered transcripts. Acute myeloid leukemia, Protein export ad Glycosaminoglycan biosynthesis-keratan sulfate were the 3 significantly enriched pathways. Apoptosis, Notch signaling pathway and PI3K-Akt signaling pathway which was closely related with the pathology process of T-ALL was involved in the enriched pathways [15]. Furthermore, we constructed co-expression network to investigate the relation between lncRNAs and the coding genes.
Competing endogenous RNAs (ceRNAs) was raised that ceRNA molecules could sponge miRNA through miRNA response elements (MREs) and regulate gene expression. Numbers of molecules could act as ceRNAs, including lncRNAs, circRNAs or pseudogene. Wang et al found that lncRNA CHRF regulates cardiac hypertrophy by targeting miR-489 [16]. circRNA MTO1 sponges miR-9 to suppress hepatocellular carcinoma progression [17]. Chan et al found that A FTH1 gene:pseudogene, can sponge miRNAs to regulates tumorigenesis in prostate cancer [18]. In this study, a lncRNA-associated ceRNA network analysis was perform and showed that 71 differentially expressed lncRNAs and 123 expressed mRNAs were selected, which predicted 11 miRNAs sharing binding sites with the differentially expressed lncRNAs and mRNAs.
Nice differentially expressed lncRNAs from ceRNA network prediction were selected to confirm deep RNA sequence results by qRT-PCR. We selected ten pairs of mice spleen which contain both T-ALL leukemia mice or C57BL/6 mice to further validation. All of the nice downregulated lncRNAs expression were consistent with RNA sequence results.
NONMMUT117521.1 was located in intron of ECE-1 which has endopeptidase activity and membrane-bound metalloprotease. Bao et al found that ECE-1 promoted Ischemia/Reperfusion-Induced Injury [19]. ENSMUST00000195494 was located in Pfkfb3 which participated in the glucose metabolism and promoted cell proliferation, apoptosis and autophage in many types of cancer [20]. NONMMUT008951.2 was located in Bcl11a which inhibited proliferation and promoted apoptosis in B lymphoma cell lines [21].
NONMMUT026003.2, in the ceRNA network of differentially expressed lncRNAs, was located Bcl6 which participated humorous of processes, including inflammatory response, growth and differentiation [22].
This study provides the first Notch1-correlated ceRNA network prediction of lncRNAs and mRNAs which needed some studies to explore the roles of these differentially expressed lncRNAs. Because of the limitation of mice model, patients' samples addition is valuable.
Abbreviations   Validation of deep RNA sequence results by qRT-PCR from spleen of T-ALL mice or C57BL/6 mice.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.
additional data.xls ARRIVE Guidelines Checklist.pdf