ALDOB Promotes Neuroblastoma Metastasis by Affecting EMT Pathway and is a Potential Prognostic Factor

Background(cid:0) Aldolase B (ALDOB) is a member of the aldolase family, which is the fourth enzyme in glycolysis process. In recent years, the non-enzymatic effects of some glycolytic enzymes have been reported to promote the formation of several human tumors, but the non-enzymatic action of ALDOB in neuroblastoma(NB) remains unclear. This study aims to explore the non-enzymatic effect of ALDOB in neuroblastoma. Methods(cid:0) We used immunohistochemistry to examine 63 patients tissue microarray samples and 3 pairs of lymph node metastases and the primary tissue samples, and evaluated the relationship between ALDOB expression level and clinical characteristics. We then analyzed the public datasets of NB based on microarray to verify the immunohistochemistry results. In addition, we conducted in vitro experiments on SK-N-BE(2) and SH-SY5Y cell lines to explore the molecular mechanism. Results(cid:0) Immunohistochemistry indicated ALDOB is signicantly associated with INSS stage and tumor metastasis in NB, public dataset analysis showed ALDOB is related to NB patient survival remarkably. In vitro experiments displayed silencing ALDOB may inhibit the cell migration by epithelial-mesenchymal transition (EMT) pathway. Conclusions(cid:0) Our nding demonstrated that ALDOB can affect the metastasis of NB by EMT pathway and may be a potential target for neuroblastoma therapy in the future. The prognostic value of ALDOB in four NB databases. a Kaplan-Meier analysis of OS of the Oberthuer database based on ALDOB expression, and the log-rank test P value was shown (n 251). b Kaplan-Meier analysis of OS of the Maris Database based on ALDOB expression, which shows the log-rank test P value (n 97). of OS the NRC database based on ALDOB expression, and P value of log-rank was database on in stage tumors of mean and signicance


Background
Neuroblastoma (NB) is the most frequently seen childhood malignancy, with 25-50 cases per million population [1], accounting for more than 7% of childhood malignancies and a mortality rate of 15% [2].
Neuroblastoma is a neuroendocrine tumor, may originate from any neural crest part of the sympathetic nervous system, the commonest location is the adrenal glands, but they can also occur in the neck, chest, abdomen and pelvic nerve tissue [3]. Many patients have tumor metastasis at the time of diagnosis [4], about half of the children who had metastasized developed distal metastases to bone marrow, cortical bone, non-adjacent lymph nodes, and liver [5], mortality in patients with bone metastases occurred in more than 93% of patients [6]. For more than a decade, the clinical staging and risk classi cation for neuroblastoma has dramatically improved [7], which is bene cial for guiding treatment. However, the prognosis with metastasis is poor and the chance of recurrence with neuroblastoma is still high [8].
Therefore, understanding the molecular mechanism of NB invasion and metastasis and nding biomarkers that lead to NB metastasis and relapse are of great signi cance for monitoring NB progression and guiding diagnosis and therapy.
ALDOB is a member of the aldolase family, which is the fourth enzyme in glycolysis process, including ALDOA, ALDOB and ALDOC three subtypes. The association between aerobic glycolysis (oxyglycolysis) and tumorigenesis has been known for decades after the "Warburg hypothesis" (also termed the "Warburg effect") was proposed by Otto Warburg, a German scientist [9]. However, in recent years, lots of studies have suggested that some glycolytic enzymes not only function in the glycolytic pathway [10], but also have extra non-glycolytic enzyme functions in many ways such as; the aldolase family has been reported to activate WNT signaling in a GSK-3-dependent mechanism that in uences the course of cancer development [11]. The non-enzymatic action of ALDOA, one of the isoenzymes of ALDOB, has been demonstrated in a variety of tumorigenesis [12][13][14]. ALDOA has been suggested to promote lung cancer metastasis through prolyl hydroxylases (PHD)-mediated stabilization of HIF-1 and subsequent activation of MMP9 [15]. It has also been reported that the expression of ALDOA can be used to predict poor prognosis of pancreatic cancer, partly due to its role in regulating E-cadherin expression [12]. ALDOB is mainly overexpressed in liver [16] and large intestine [17] tissue samples. ALDOB has been shown to promote tumor cell proliferation in liver metastasis of colon cancer [16] and It is a key molecule that promotes the metastasis of colorectal cancer (CRC) by regulating epithelial-mesenchymal transition (EMT) pathway [17]. However, the non-enzymatic role of ALDOB in neuroblastoma has not been studied.
Therefore, a series of molecular biological studies were conducted on the relationship between ALDOB expression and the development of neuroblastoma. We predict based on tissue microarray immunohistochemistry and public databases that ALDOB will promote the metastasis of neuroblastoma, leading to poor prognosis, and through in vitro experiments such as wound healing assay, transwell experiment, western blot, etc., it was veri ed that silencing ALDOB can inhibit the proliferation and metastasis of NB cells by inhibiting EMT pathway.

Clinical specimens
The clinical samples used in this research were primary pediatric NB patients histologically diagnosed in Xinhua Hospital a liated to Shanghai Jiao Tong University School of Medicine from September 2012 to February 2015. Tumor samples from all the patients in the study were surgically removed. All specimens were rst quick-frozen in liquid nitrogen and then stored at -80°C for tissue microarray analysis. The research and consent procedure have been validated by the Clinical Committee of Xinhua Hospital, Shanghai Jiao Tong University School of Medicine.
Tissue microarray (TMA) preparation and immunohistochemistry (IHC) Separate a small part of the tissue specimen and shape it in a customized mold for chip production. Then x overnight in 4% paraformaldehyde, after that tissue blocks were embedded in para n in a prepared array. Then the sample was sliced (5 μm) and adhered to a poly-L-lysine coated glass slide for immunohistochemical staining, which was performed as previously described [18,19], using speci c antibody against ALDOB (1:100 dilutions, Cat.18065-1-AP, Proteintech Group, Chicago, USA). Without knowing the clinicopathological characteristics of the tumor, the immunoreactivity in tissue sections was observed under three random microscopes, and then evaluated by three pathologists. Differences in scoring were discussed until consensus was reached. The tissue sections were scored under an optical microscope according to the degree of staining (0~3 points were negative staining, light yellow, light brown, dark brown) and the positive range (1~4 points were 0~25%, 26~50%, 51~75%, 76~100%). The two scores are then multiplied to get the nal score for comparison. Finally, the intensity of staining was divided into four levels according to the nal score: "0" means negative expression (the nal score was 0), "1" means weak positive expression (the nal scores were 1~4 points), "2" means moderate expression (the nal scores were 5~8 points), and "3" means strong expression (the nal scores were 6~12 points).

Public database analysis
Four publicly available data sets were selected from the R2: Microarray analysis and Visualization platform (http://r2.amc.NL) : Oberthuer-251-custom-ampexp255, Maris-101-custom-u95a, NRC-283-rma_sketch(bc)-huex10t and Asgharzadeh-249-Custom-huex10Tx, which contained information on the clinical characteristics and prognosis of patients with neuroblastoma, the target genes were obtained and selected for analysis. All Kaplan-Meier analyses were performed online, and the optimal P value and corresponding cutoff value for separating the high-expression group from the low-expression group were selected by the median.

Cell Culture
Human neuroblastoma cell lines SK-N-BE (2) and SH-SY5Y were acquired from American Type Culture Collection (Manassas, USA) and maintained in a mixture of Eagle basic medium and Ham nutrient mixture F12 1:1, 10% fetal bovine serum (FBS) , both from Gibco, USA, and cultivated in a 5% CO2 moist incubator at 37°C.

Production and infection of lentivirus
Lentivirus particles expressing shRNA sequence targeting human ALDOB gene (5'-CCTATTGTTGAACCAGAGGTATA-3') were designed and constructed, and shRNA particles expressing nonsense sequences were used as negative control (shNC). Lentivirus particles expressing shALDOB or shNC were added to 6-well plates with SH-SY5Y and SK-N-BE(2) growing to 30%-60% con uency, and the infection complex (MOI) was 20. After 48 hours of transfection, uorescent microscope (Olympus) was used to detect the transfection e ciency of stably expressed green uorescent protein (GFP) cells, and 1 g/ml puromycin (Cat. #ST551, Beyotime, China) was used to screen the transfected cells, and stable transfected cell lines were obtained after 48 hours of continuous treatment. After screening, some cell RNA was collected to con rm the ALDOB mRNA level by qPCR, and the extracted protein was detected by Western blot.

Colony formation test
The stably transfected SH-SY5Y and SK-N-BE(2) cells were inoculated into a 6-well plate at a density of 1000 cells per well. The cells were cultured in a cell incubator for 14 days, and the liquid was changed every three days, the cell state and the size of the clones was observed under the microscope. When each clone was larger than 20 cells, the culture medium was removed, paraformaldehyde xed the cell, then crystal violet was used to stain cells. Next the cells were washed with PBS, and photographs were taken.
Determination of wound healing SH-SY5Y and SK-N-BE(2) cells were cultured in a six-well plate and grown to con uence 90%. Arti cial wound marks were created by scraping a single layer of converging cells with a pipette suction head and incubating the cells with a serums free medium for 48 hours. After PBS rinsing, photographed transplanted cells at 0 and 48 hours, and calculated the migration distance during wound healing was.

Cell invasion
Cell invasion assay was performed in transwell chamber using Matrigel (BD, Biosciences, CA). 6×10 cells in the matrix medium with no fetal bovine serum were incubated in Transwell upper chamber with matrix gel. In the lower chamber, there was the cell culture medium with 10% FBS. Then cultivated at 37°C for 48 hours. After 48 hours, 4% paraformaldehyde xed the cells on the submembrane surface of the compartment and then stained with crystal violet.

Western blot analysis
Lyse cells on ice with RIPA buffer which contains PMSF (Cat. # ST506, Beyotime, China) (Cat. # P0013B, Beyotime, China) for 30 minutes, centrifugation at 20,000 g for 15 minutes, collected the supernatant, and used the BCA kit (Thermo Scienti c, Rockford) to test the protein concentration. Gel electrophoresis used 10% SDS-PAGE gel and contains the same amount of protein per lane. Protein transfer used PVDF membranes (Millipore, Sigma Aldrich, USA). Then took the 5% BSA TBST (Tris buffered saline, pH 7.6, 0.1% Tween®20) to seal the PVDF membranes at room temperature for 1 hour, followed by primary incubation at 4°C overnight. TBST washed the membrane 3 times for 5 minutes each time, then incubated with TBST and an appropriate secondary antibody at room temperature for 30 minutes. Next, TBST wash the membrane 3 times for 5 minutes each time. Proteins are visualized by electrochemical luminescence (ECL). The primary antibodies of ALDOB (catalog number AB75751) and β-actin (catalog number AB8227) were purchased from Abcam. E-cadherin, Vimentin, N-cadherin, Claudin-1, Slug, and ZEB1 were purchased from Cell Singling Technology (CST). Image J (X64, v. 2.1.4) software for quantitative analysis of imprinting.

Statistical analysis
The experimental data were showed as mean ± standard deviation (SD). Statistical analysis was performed using GraphPad Prism8 software (GraphPad Software, Inc. La Jolla, USA). Between the two groups, we used Student's T-test to estimate the differences, and between the three groups, we used and one-way ANOVA. Kaplan-Meier analysis was used for survival assessment, and survival differences was analyzed by log-rank test. P <0.05 was considered as a statistically signi cant difference. Signi cance was displayed as: * P <0.05, ** P <0.01, and *** P <0.001 Results ALDOB is signi cantly associated with INSS stage and tumor metastasis in NB Firstly, we stained the tissue chips from patients with 63 primary neuroblastomas using immunohistochemical (IHC) staining. We observed that ALDOB was predominantly located in the cytoskeleton of neuroblastoma cells, with no staining in adjacent non-tumor and normal tissues (Fig. 1a,   b). We next analyzed the expression of ALDOB and found that ALDOB was expressed in 56 out of 63 (88.9%) primary neuroblastomas. Among them, 24 patients (38.1%) showed weak expression of ALDOB, 27 patients (42.9%) showed moderate expression, and 5 patients (7.9%) showed strong expression (Fig. 1c). Subsequently, we analyzed whether the expression level of ALDOB in tissue samples was signi cantly correlated with clinical characteristics, and one-way ANOVA showed that the expression level of ALDOB was signi cantly correlated with tumor stage (P = 0.0223). ALDOB expressed in International Neuroblastoma Staging System(INSS)stage 3, 4 neuroblastoma tissues were signi cantly higher than those in the INSS stage 1, 2 neuroblastoma tissues (Fig. 1d). The expression level in patients with both lymph node metastasis and bone marrow metastasis was obviously higher than that of patients without metastasis (P = 0.0202) (Fig. 1e), but it was not related to gender, histopathology diagnosis or differentiation degree (Table 1).
We then stained paired samples of both the lymph node tissue and primary tumor tissue of three patients with lymph node metastasis (Fig. 1f). We found that ALDOB expression in lymph node tissues was higher than that in primary tumor tissues signi cantly (P = 0.0232) (Fig. 1g). Together, the above mentioned results indicate that overexpression of ALDOB in neuroblastoma patients may be associated with the metastasis of neuroblastoma.
ALDOB is signi cantly related to NB patient survival and is a potential prognostic factor for NB To further verify the role of ALDOB in neuroblastoma, we analyzed the expression of ALDOB in four different public datasets of tumor neuroblastoma(Oberthuer-251-custom-AMPexp255, Maris-101custom-U95a,CNC-283-RMA-Sketch(BC)-Huex10T and Asgharzadeh-249-custom-huex10Tx) online on R2 platform. The KM curves re ecting the relationship between ALDOB expression and overall survival rate (OS) were drawn. Patients with high ALDOB mRNA level showed poor OS with p values of 4.2E-05, 0.034, 0.029 and 0.014 (Fig. 2a, b, c, d), respectively. From the above, our data suggested that a high level of ALDOB expression correlates with a poor prognosis in neuroblastoma.

Aldob Knockdown Reduces Invasion And Migration Of Nb Cells
SH-SY5Y cells with high ALDOB expression (Fig. S1 shows this in more detail, see additional le 1) were selected for transfection with shALDOB and non-silenced shNC lentivirus to clarify the biological function of ALDOB in NB; SK-N-BE(2) cells with stable knockdown were also constructed for veri cation. Through uorescence microscope observation, it was found that more than 80% of the cells expressed GFP, showing a high e ciency of infection (Fig. 3a). As shown in Fig. 3b and Fig. 3c, mRNA and protein expression levels of ALDOB in SH-SY5Y cells infected with shALDOB were signi cantly lower than those of cells infected with shNC. Cell viability of SH-SY5Y and SK-N-BE (2) were detected by CCK-8 and colony formation tests. The results showed that cell proliferation of shALDOB group was lower than that of shNC group (Fig. 3d, e), which proved the effect of ALDOB knockdown on cell proliferation in vitro. Then, in order to explore the effect of ALDOB knockdown on migration and invasion of SH-SY5Y and SK-N-BE (2) cells, wound healing test and Transwell Matrigel assay were performed. The results were shown in Fig. 4a and Fig. 4b. The scratch closure rate of SH-SY5Y cells transfected with shALDOB was lower than that of the control group and the decrease of ALDOB expression signi cantly inhibited the invasion of SH-SY5Y cells (P < 0.001). These ndings suggested that ALDOB may promote NB cells migration and invasion.

Aldob Knockdown Inhibits Emt Pathway
To further understand the molecular mechanism of ALDOB knockdown on NB cell migration and invasion, we performed western blot experiments on SH-SY5Y cells. As the Fig. 5 shows, down-regulation of ALDOB enhances the expression of E-cadherin, claudin-1, while the expression of Vimentin, N-cadherin, Slug, and ZEB1 decreased. These ndings suggest that ALDOB may promote the metastasis of neuroblastoma by regulating the EMT pathway and affect the prognosis.

Discussion
NB is one of the most common pediatric neuroectodermal solid tumors in infancy and childhood. It can occur in the tissues of the sympathetic nervous system, usually in the paravertebral ganglia or adrenal medulla, and detectable masses can occur in the chest, neck, pelvis, and abdomen [1]. NB accounts for 7% of all pediatric cancers, and the 5-year survival rate of low-risk people is > 95%, while the 5-year survival rate of high-risk patients is only 40% [1]. About 65% of cases have metastasized at the time of diagnosis [20]. Surgery alone or in combination with minimal treatment regimens can improve survival in low-risk patients, but in most cases, surgery is not an option because the disease is clearly metastatic at the time of diagnosis. In many stage III and IV NB patients, recurrence occurs immediately after chemotherapy, and resistance to chemotherapy is also associated with organ metastasis. In neuroblastoma, the main cause of death is associated with metastasis. Preventing recurrence and reducing metastasis remains a major clinical challenge in treating the NB.
The role of glycolysis in tumorigenesis and development has long been recognized, glycolytic enzymes control the progress of glycolysis, and these metabolic changes regulate the biosynthesis and energy production of glucose, amino acids, and fatty acid-dependent metabolites [21,22]. However, in recent years, studies have found that some glycolytic enzymes have extra non-enzymatic action in tumorigenesis and development [10], which brings us new research directions. To name only a few, pyruvate kinase M2 (PKM2) promotes the migration and drug resistance of colon cancer cell by regulating STAT3 signaling [23,24]. In non-small cell lung cancer, enolase (ENO1) promotes tumor growth, invasion and migration through the pseudo-drug-mediated PI3K/AKT pathway [25]. In particular, aldolase interacts with a variety of proteins independent of glycolytic enzymes, including F-actin[26], ARNO [27] and Tubulin[28] and is essential for cancer cells to promote their proliferation [29]and metastasis [30] through non-enzymatic pathways. In this research, we have 63 patients of primary neuroblastoma tissue microarray and 3 pairs of samples of lymph node metastases and the primary tissue samples from different patients with lymph node metastasis with immunohistochemical staining were analyzed, we found that the expression level of glycolytic enzyme ALDOB in cancerous tissue is associated with tumor metastasis and stages. Further a public database through the analysis of the R2 platform for veri cation, found that glycolytic enzyme ALDOB will affect the survival of patients with neuroblastoma and the higher its expression, the more unfavorable to the survival of patients. Therefore, we suspect ALDOB can be used as a prognostic marker of NB. Our series of in vitro experiments also demonstrated that silencing ALDOB expression can inhibit the metastasis and invasion of neuroblastoma cells.
The EMT pathway is a key event that promotes the dissociation, invasion, and metastasis of cancer cells [31]. Loss of E-cadherin is a marker of EMT and is usually associated with high tumor stage and metastasis. Several transcription factors, such as Slug and ZEB1, have been involved in transcriptional inhibition of E-cadherin and induction of EMT [32]. Since ALDOB knockdown suppressed the NB cell proliferation, metastasis and invasion, EMT was one of the critical processes to enhance the invasion and metastasis of tumor, we speculated whether ALDOB played a role on neuroblastoma by inducing EMT. As we expected, the down-regulation of ALDOB expression resulted in an increase in epithelial marker proteins (E-cadherin, Claudin-1) and a decrease in mesenchymal marker proteins (N-cadherin, Vimentin, Slug, and ZEB1), displaying that silencing of ALDOB may cause the reversal of EMT progress in neuroblastoma. However, the non-enzymatic mechanism of glycolytic enzyme ALDOB in neuroblastoma has not been discovered and proven so far.
Although this study brings us initial revelation about the role and clinical signi cance of ALDOB in NB, there are still lots of limitations to be further discussed. First of all, the number of patients included in this research is not large enough, especially the lack of samples of stage 4S patients. Second, we'd better do an overexpression experiment to rule out the off-target effect. Thirdly, in vivo experiments are necessary to strengthen the evidence that ALDOB promotes the metastasis of neuroblastoma.
To sum up, this study con rmed that the high expression of ALDOB is an independent poor prognostic factor of OS in NB. It plays an important part in the progression and metastasis of neuroblastoma by inducing the EMT pathway. ALDOB is expected to become a new predictor for NB to guide novel clinical therapy. It may help clinicians to identify patients with a high risk of early recurrence. These patients should be recommended for close follow-up and appropriate adjuvant therapy to prolong survival.

Conclusions
There is still a need to nd potential biological targets for NB. Our study found for the rst time that high expression of ALDOB will promote the metastasis of NB by regulating the EMT pathway and affect the prognosis, indicating a critical tumor promoting role in NB progression.

Consent for publication
Not applicable Availability of data and materials The data used and/or analyzed during the current study are available from the corresponding author.  The  The values were mean ±SEM, and the statistical signi cance was *P< 0.05, **P<0.01, *** P< 0.001 Figure 4 The effects of ALDOB knockdown on NB cell migration and invasion in vitro. a Wound healing test of SH-SY5Y cells after transfection with shALDOB or shNC was performed in vitro to investigate the migration ability. b SH-SY5Y and SK-N-BE(2) were transfected with shALDOB or shNC to conduct Transwell test to assess the invasion ability, and the invasion was quanti ed by counting the cells in 3 random microscope elds. The values were mean±SEM, and the statistical signi cance was *P< 0.05, **P<0.01, *** P< 0.001