Leukemia, including ALL and acute myeloid leukemia (AML), was diagnosed according to standard clinical and laboratory criteria (30). The present study included newly diagnosed patients with pediatric leukemia and healthy pediatric donors who presented no history of leukemia. The samples of the healthy group and the patient group in the experiment were collected from January 2018 to December 2018 in Kunming Children's Hospital (Kunming, China). The age range of the healthy group was 1–12 years old, including 15 males and 13 females; the age range of the patient group was 10 months to 13 years old, including 34 males and 27 females. The details of each patient group are presented in Table SI. Peripheral blood was collected from 18 patients with pediatric leukemia and 6 healthy children, and peripheral blood mononuclear cells (PBMCs) were isolated by density centrifugation at 2,000 × g for 10 min at 4°C using Ficoll solution, followed by washing with PBS. Plasma was collected from 55 patients with pediatric leukemia and 22 healthy children for ELISA analysis. Of the 18 patients with pediatric leukemia whose PBMCs were isolated, the plasma samples of only 12 patients were included in the ELISA analysis, as the amount of plasma obtained from the remaining 6 patients was insufficient.

Plasma was collected and stored at −80°C after total blood was centrifuged at 2,000 × g for 10 min at 4°C. Soluble Sema4D plasma levels were measured using Sema4D ELISA kit (cat. no. MBS705483; MyBioSource, Inc.) according to the manufacturer's instructions. In brief, a microtitration plate coated with anti-Sema4D antibodies was incubated with plasma samples or Sema4D protein standard for 2 h at 37°C. It was then incubated with biotin-conjugated anti-Sema4D antibodies for 2 h at 37°C. After being washed with the washing solution provided in the kit, it was incubated with HRP-coupled streptavidin for 20 min at 37°C. and then incubated with HRP substrate for 15 min at 37°C. Following termination of the reaction, the absorbance was measured at 450 nm.

Jurkat and BALL-1 cell lines (Shanghai Yihe Applied Biotechnology Co., Ltd. China) were both cultured in RPMI-1640 supplemented with 10% FBS and 1% penicillin and streptomycin (all from Gibco; Thermo Fisher Scientific, Inc.) at 37°C with 5% CO₂. The cell lines had been confirmed to be mycoplasma free and authenticated by STR profiling. Sema4D cDNA (Sema4D) and short hairpin (sh)RNA were subcloned into the lentiviral vector PWPI-GFP (kindly provided by Dr John Basile, University of Maryland) to obtain lentiviral Sema4D overexpression (Sema4D) and shRNA (S4DshRNA) constructs, and the empty lentiviral vector PWPI-GFP was used as negative control in further analysis. 293T cells (Institute of Medical Biology, Chinese Academy of Medical Sciences and Peking Union Medical College) were cultured in high glucose DMEM (Gibco; Thermo Fisher Scientific, Inc.) with 10% FBS and 1% penicillin and streptomycin at 37°C with 5% CO₂. Lentiviruses were produced by transfecting 293T cells (5×10⁵ cells in six-wells plates incubated overnight), with 1.5 µg lentiviral plasmid, 0.5 µg PVSVG plasmid and 1 µg PASPAX plasmid (kindly provided by Dr John Basile, University of Maryland) with Lipofectamine^® 2000 (Thermo Fisher Scientific, Inc.). Viral supernatants were collected 72 h after transfection. Jurkat and BALL-1cells were infected with lentiviruses at a MOI of 30 with 4 µg/ml polybrene (Beijing Solarbio Science & Technology Co., Ltd.) for 72 h before collection for subsequent analysis.

Jurkat and BALL-1 cells were lysed with RIPA buffer (Beijing Solarbio Science & Technology Co., Ltd.) containing protease inhibitors. Protein concentration was determined using the BCA method, and 50 µg protein/sample were separated by 10% SDS-PAGE and transferred to PVDF membranes, which were blocked at room temperature for 40 min using 5% BSA solution (Beijing Solarbio Science & Technology Co., Ltd.). The membranes were incubated at 4°C overnight with the following primary antibodies: Mouse anti-Sema4D (1:1,000; cat. no. 610671; BD Biosciences), rabbit anti-ERK (1:2,000; cat. no. 9126S), rabbit anti-phosphorylated (p)-ERK (1:2,000; cat. no. 4376S), rabbit anti-AKT (1:2,000; cat. no. 9272S), rabbit anti-p-AKT (1:2,000; cat. no. 9611S), rabbit anti-PI3K (1:2,000; cat. no. 4257S), rabbit anti-p-PI3K (1:2,000; cat. no. 4228S; all from Cell Signaling Technology, Inc.) and rabbit anti-β-actin (1:100,000; cat. no. AC026; ABclonal Biotech Co., Ltd.). Following primary antibody incubation, the membranes were incubated at room temperature for 1 h with goat anti-mouse lgG (H + L) (1:5,000; cat. no. 074-1806) and goat anti-rabbit lgG (H + L) antibodies (1:5,000; cat. no. 074-1506; both from KPL). The protein signal was detected using SuperSignal reagent (MilliporeSigma) and quantified using ImageJ software v1.8.0 (National Institutes of Health).

Cell viability was analyzed by CCK-8 assay (Tongren Institute of Chemistry). Jurkat and BALL-1 cells (5×10⁶) were lentivirally transduced for 72 h, and 5×10⁴ cells/ml were seeded in 96-well plates. At 1, 2, 3, 4 and 5 days, cells were incubated with 10 µl CCK-8 solution for 2 h at 37°C with 5% CO₂ and the absorbance at 450 nm was measured.

Cell migratory capacity was detected by Transwell assay. The lentivirally transduced Jurkat and BALL-1 cells were cultured to logarithmic growth phase, adjusted to 1×10⁴ cells/ml with RPMI-1640 basal medium containing 0.1% BSA (Beijing Solarbio Science & Technology Co., Ltd.), and placed on top of each compartment of 0.8 µm, 24-well Transwell chamber (Corning, New York, USA), and RPMI-1640 medium containing 10% FBS was added to lower chamber of the plates. After the cells were cultured for 24 h at 37°C, the cell number in the lower chamber was counted using a hemocytometer. Cell invasive capacity was also determined by Transwell cell assay. A total of 100 µl 300 µg/ml Matrigel (BD Biosciences) were thawed at 4°C overnight before being placed into the upper chamber and incubated at 37°C for 1 h. The lentivirally transduced Jurkat and BALL-1 cells were cultured to logarithmic growth phase, adjusted to 1×10⁴ cells/ml with RPMI-1640 basal medium containing 0.1% BSA and placed on top of each compartment. RPMI-1640 medium with 10% FBS was then added to the lower chambers. After the cells were cultured for 24 h at 37°C, the cell number in the lower chamber was counted using a hemocytometer.

Apoptosis was detected using PE Annexin V Apoptosis Detection Kit I (BD Biosciences). After lentiviral transduction for 72 h, Jurkat and BALL-1 cells were washed with pre-chilled PBS and were adjusted to 1×10⁶ cells/ml with 1× Binding Buffer. The cells were incubated at room temperature with 5 µl PE Annexin V and 5 µl 7-AAD for 15 min before being analyzed by Attune NxT flow cytometry (Thermo Fisher Scientific, Inc.), and the data were analyzed using Treestar FlowJo version 10 software (Becton, Dickinson and Company).

After lentiviral transduction for 72 h, Jurkat and BALL-1 cells were washed with pre-chilled PBS, before being suspended in pre-chilled PBS and fixed with ice-cold absolute ethanol at −20°C overnight. After being washed with ice-cold PBS, the cells were incubated with 50 µg/ml RNase A and 65 µg/ml propidium iodide (BD Biosciences) at 4°C for 30 min before analysis by Attune NxT flow cytometry (Thermo Fisher Scientific, Inc.), and the data were analyzed using Treestar FlowJo version 10 software (Becton, Dickinson and Company).

The analysis was performed on the GEPIA website (http://gepia.cancer-pku.cn) using the median expression of Sema4D as cut-off to differentiate low and high Sema4D expression groups and 95% confidence interval.

The statistical analysis was performed by SPSS software version 21.0 (IBM Corp.). Experimental values are presented as the mean ± SD. Statistical analysis in Fig. 1 and Figs. 3–6 was performed using one-way ANOVA followed by Bonferroni's post hoc test. Statistical analysis in Fig. 2 was performed using two-way ANOVA followed by Bonferroni's post hoc test. Statistical analysis in Fig. 7B was performed using unpaired two-tailed Student's t-test. The correlation between the level of p-PI3K, p-ERK or p-AKT and Sema4D in Fig. 7C was evaluated using Pearson's correlation analysis (two-sided). P<0.05 was considered to indicate a statistically significant difference.

To investigate the expression level of Sema4D in pediatric leukemia cells, the mononuclear cell lysates of 18 patients with pediatric leukemia and 6 healthy children were subjected to western blot analysis. The clinical details of the 18 patients with pediatric leukemia are presented in Table SII. The results revealed that Sema4D was highly expressed in patients with pediatric leukemia compared with healthy participants (Fig. 1A). As the majority of patients were B cell-ALL patients, it was subsequently determined whether there were any differences in Sema4D expression level between B cell-ALL and other types of leukemia. Therefore, patients were divided into either B cell-ALL (B-ALL) or non-B cell-ALL groups (Non-B-ALL). The analysis indicated that Sema4D expression level in both groups was significantly higher compared with that in the control group (P<0.01), and there was no difference in Sema4D expression level between the B-ALL and Non-B-ALL groups (Fig. 1B).

As Sema4D can be cleaved to functional soluble Sema4D, which is secreted from cells (23), the level of soluble Sema4D in the plasma of patients with leukemia was examined. The plasma of 55 pediatric patients with leukemia and 22 healthy children was collected and analyzed by ELISA. The clinical details of the 55 pediatric patients with leukemia are listed in Table SII. The results revealed that the level of soluble Sema4D in the B-ALL and Non-B-ALL groups was higher compared with that in healthy children (P<0.01 for B-ALL; P<0.05 for Non-B-ALL), and there was no difference in Sema4D level between the B-ALL and Non-B-ALL groups (Fig. 1C).

Taken together, the results indicated that Sema4D was highly expressed in patients with pediatric leukemia, and soluble Sema4D was released into the circulation in these patients. However, there was no difference in Sema4D levels between B-ALL and other types of leukemia in both leukemia cells and plasma.

As the results indicated that Sema4D was highly expressed in pediatric leukemia, the potential role of Sema4D in leukemia development was subsequently investigated. To examine the function of Sema4D, the Jurkat cell line, which originates from T cell ALL, and the BALL-1 cell line, which originates from B cell ALL, were selected for the study. Sema4D overexpressing and S4DshRNA lentiviruses were transduced into Jurkat and BALL-1 cells. Sema4D expression level in Jurkat and BALL-1 cells was detected by western blotting after transduction for 72 h. The results indicated that Sema4D expression in both cell lines was reduced after transduction with S4DshRNA, and increased after transduction with the Sema4D overexpressing lentivirus compared with uninfected cells or cells infected with the empty vector (GFP) (Fig. 2A). The viability of Jurkat and BALL-1 cells was assessed by CCK-8 assay after transduction for 72 h. The results demonstrated that the viability of both Jurkat and BALL-1 cells was significantly decreased after transduction with S4DshRNA (P<0.001 in Jurkat and P<0.01 in BALL-1), while it was significantly increased after transduction with Sema4D overexpression construct (both P<0.001) (Fig. 2B). The results suggested that Sema4D promoted the proliferation of Jurkat and BALL-1 cells.

Subsequently, the effect of Sema4D on cell cycle was examined in Jurkat and BALL-1 cells. The results revealed that the percentage of cells in the G₀/G₁ phase significantly increased after transduction with S4DshRNA (P<0.01), and significantly decreased after transduction with the Sema4D overexpression construct (P<0.05) in BALL-1 cells (Fig. 3A and B). The cell percentage in the G₀/G₁ phase also significantly decreased after transduction with the Sema4D overexpression construct in Jurkat cells (P<0.01), but was not altered after transduction with S4DshRNA, as the percentage of Jurkat cells in the G₀/G₁ phase was already high (Fig. 3A and B). As a consequence of the decreased abundance of cells in the G₀/G₁ phase caused by Sema4D overexpression, the percentage of Jurkat cells in the S phase or BALL-1 cells in the G₂/M phase increased significantly compared with the control (P<0.01 for Jurkat; P<0.05 for BALL-1 cells). The results suggested that downregulation of Sema4D induced cell cycle arrest at the G₀/G₁ phase, and Sema4D modulated the G₀/G₁ phase of the cell cycle.

The effect of Sema4D on apoptosis was also investigated. In Jurkat cells, the number of apoptotic cells was significantly higher in the S4DshRNA group (P<0.001; Fig. 4A and B). Similarly, in BALL-1 cells, the number of apoptotic cells was significantly higher in the S4DshRNA group (P<0.05) and was significantly lower in Sema4D overexpression group compared with the GFP control group (P<0.05) (Fig. 4A and B). These results indicated that Sema4D protected leukemia cells from apoptosis.

To evaluate the effects of Sema4D on cell invasion and migration, Jurkat and BALL-1 cells were transduced with Sema4D overexpression construct or S4DshRNA, and their migratory and invasive abilities were examined by Transwell assay. In Jurkat cells, no difference was observed in the invasive ability among the GFP control, S4DshRNA and Sema4D overexpression groups (Fig. 5A), while the migratory ability increased significantly in the Sema4D overexpression group compared with the GFP control (P<0.05; Fig. 5C). In BALL-1 cells, compared with the GFP group, the invasive ability increased significantly in the Sema4D overexpression group (P<0.001; Fig. 5B), while no difference was observed in the migratory ability among the GFP control, S4DshRNA and Sema4D overexpression groups (Fig. 5D). As Jurkat and BALL-1 cell lines originate from T-ALL and B-ALL, respectively, their different migratory and invasive capabilities may be due to their original characteristics. Overall, these data suggested that Sema4D moderately promoted the invasive and migratory ability of ALL cells.

As the results had indicated that Sema4D promoted proliferation and inhibited apoptosis, it was subsequently investigated how these effects were mediated. PI3K, AKT and ERK proteins, which are widely expressed in a variety of tumor tissues (31–35), promote cancer development by mediating activation of downstream effectors (31,36–39). The role of Sema4D in activating PI3K, AKT and ERK was examined in Jurkat and BALL-1 cells. In Sema4D-overexpressing BALL-1 cells, the phosphorylation level of PI3K was significantly increased compared with the GFP control cells (P<0.01; Fig. 6). However, potentially owing to the high endogenous expression of Sema4D in both Jurkat and BALL-1 cells, Sema4D overexpression did not consistently elevate the phosphorylation level of PI3K, ERK and AKT.

In Sema4D knocked-down BALL-1 cells, the phosphorylation level of PI3K and ERK significantly decreased compared with the GFP control cells (P<0.001 for PI3K and P<0.01 for ERK; Fig. 6A and B). In Sema4D knocked-down Jurkat cells, the phosphorylation level of PI3K was also significantly decreased (P<0.05; Fig. 6A and B). The alteration in the AKT phosphorylation level was not statistically significant in both Jurkat and BALL-1 cell lines. The Sema4D knockdown results indicated that Sema4D mediated the activation of PI3K and ERK in ALL cells.

In order to ascertain the clinical relevance of PI3K, ERK and AKT pathways, the expression and phosphorylation levels of PI3K, ERK and AKT were analyzed by western blotting in 18 pediatric leukemia samples, in which the expression of Sema4D had been examined (Fig. 1A and B). Since no difference was observed in the expression level of Sema4D between the B-ALL and Non-B-ALL groups, both groups were analyzed together regarding the expression and phosphorylation of PI3K, ERK and AKT. There was no significant difference in the expression level of PI3K, ERK and AKT between patients with leukemia and healthy control subjects (data not shown). The total phosphorylation level of PI3K and ERK in the leukemia group was significantly higher compared with the healthy control group (P<0.01), and the total phosphorylation level of AKT in the leukemia group was higher compared with the healthy control group (all normalized to β-actin), but with no statistical significance (Fig. 7A and B). The relative phosphorylation level normalized to total PI3K, ERK and AKT was not significantly different between the groups (data not shown). Therefore, the results indicated that the total phosphorylation level of PI3K, ERK and AKT was enhanced in pediatric leukemia. The correlation of Sema4D expression with the phosphorylation level of PI3K, ERK and AKT was further analyzed, and the results indicated that Sema4D expression was significantly correlated with p-PI3K and p-ERK (P<0.001 for p-PI3K; P<0.0001 for p-ERK) and moderately correlated with p-AKT (P=0.0042), and the expression level of Sema4D and the phosphorylation level of PI3K, ERK and AKT were higher in patients with leukemia compared with healthy subjects (Fig. 7C). The expression of Sema4D was not correlated with the expression of PI3K, ERK and AKT proteins (data not shown). The results depicted in Figs. 6 and 7 suggested that Sema4D activated the PI3K, ERK and AKT pathways in leukemia cells.

As there were no ALL data in the GEPIA website, overall survival was only analyzed in patients with AML, and it was revealed that high expression of Sema4D was associated with shorter overall survival in patients with AML (Fig. 8). Therefore, Sema4D may be a poor prognostic biomarker in leukemia.