Fig 2: Tumor selectivity and the involvement of CD44. A5 = MLO-A5 osteocytes, ß-cat = ß-catenin plasmids, BML = BML284, Eno1 = Enolase 1, Ubc = ubiquitin C. The signal and double asterisk indicates p < 0.01 and p < 0.05, respectively. (A) Comparison of MTT-based viability of four non-tumor cells (KTB34, KTB6, MC3T3, MSC) and four tumor cells (EO771, TRAMP, 4T1.2, PANC-1) in response to ß-catenin-iTS CM and BML-iTS CM in 2 days. (B) Tumor selectivity of EO771 CM, 4T1.2 CM, TRAMP CM, and PANC-1 CM. MTT-based tumor selectivity is defined as the ratio of (reduction in tumor cells) to (reduction in non-tumor cells). The selectivity value above 1 indicates that MTT-based inhibition is more selective to tumor cells than non-tumor cells. Of note, N.D. = not defined since the viability of non-tumor cells is stimulated. (C&D) Tumor selectivity of Enolase 1 and ubiquitin C. (E) CD44 was co-immunoprecipitated with Eno1. The protein extracts of EO771 cells were incubated with anti-Eno1 antibody using the protein A/G beads. Immunoprecipitates and total cell lysates were analyzed by Western blotting with anti-CD44 and anti-Eno1 antibodies as indicated. (F&G) siRNA knockdown of CD44 suppressed Eno1-mediated inhibition of the proliferation of EO771 cells. (H) siRNA knockdown of CD44 suppressed Eno1-mediated downregulation of MMP9, Runx2, and Snail in EO771 cells.

Fig 3: ENO1 promotes proliferation and suppresses expression of contractile proteins in PASMC. a–c Normal PASMCs were transfected with lentivirus containing shRNA targeting ENO1 to silence ENO1 (shENO1). The shENO1 and control (shCTRL) PASMC were treated with hypoxia (1% O2) for 24 h. We measured the expression level of ENO1 and PCNA by western blotting in cell lysates of shCTRL and shENO1–PASMC under normoxic and hypoxic conditions a. b shows the normalized quantification of ENO1 levels (n = 3 per group, **P < 0.01). c shows the normalized quantification of PCNA levels (n = 3 per group, *P < 0.05, **P < 0.01). d BrdU assay was used to measure the cell proliferation in shENO1–PASMC under normoxic and hypoxic conditions (n = 6, **P < 0.001). e Western blotting images of contractile proteins (Myocardin, SMA, MHC, and Calponin) in shCTRL and shENO1–PASMC under normoxic and hypoxic conditions. f PASMC were treated with 10 µM ENOblock and hypoxia (1% O2) for 24 h, and the PCNA levels were measured by western blotting. g Normalized quantification of PCNA level in ENOblock-treated PASMCs under normoxic and hypoxic conditions (n = 3–4 per group, *P < 0.05, **P < 0.01). h PASMCs were treated with ENOblock (in gradient concentrations) and hypoxia (1% O2) for 24 h, afterwards the cell proliferation (using BrdU assay) and i cell viability were measured (n = 4 per group, *P < 0.05, **P < 0.01). j PASMC were treated with 10 µM ENOblock and hypoxia (1% O2) for 24 h, the levels of SMC contractile proteins (Myocardin, Calponin, MHC, and SMA) were measured by western blotting. k PASMCs were transfected with pCMV3-ENO1-GFP. The transfection efficiency was determined by the detection of GFP and GFP-tagged ENO1 by GFP antibody and the overexpression of ENO1 was validated by ENO1 antibody (ENO1 and ENO1-GFP). The PCNA levels were also measured after overexpression of ENO1. l Normalized quantification of PCNA level in PASMC transfected with pCMV3-ENO1-GFP (n = 4 per group, *P = 0.0202). m BrdU assay of ENO1-overexpressed PASMC (n = 6 per group, **P < 0.0001). n Western blotting images and o normalized quantification of SMC contractile proteins (Myocardin, SMA, MHC, and Calponin) in ENO1-overexpressed PASMCs (n = 3 per group, *P < 0.05). Data represent the mean ± SEM. Student t test and one-way ANOVA were used to compare two and multiple groups. Bonferroni post-tests were carried out after ANOVA

Fig 4: Immunodetection of glycolytic enzymes within bat and human cells. a-d Immunodetection of Eno1, Tpi1, Pgam1 and ß2-Tub (load control) in Poly I:C transfected and Control cells at 4, 8 and 20 hpt (a: PaKiT03, b: PaLuT02, c: HEK293T, d: HeLa). e-h Densitometry analysis following immunodetection of Eno1, Tpi1, Pgam1. Average ratios were calculated as Poly I:C/Control normalised to ß-Tub. Error bars represent the standard error of the mean from two technical replicates (e: PaKiT03, f: PaLuT02, g: HEK293T, h: HeLa)

Fig 5: FAT relies on glycolytic enzymes.(a) Silencing of the glycolytic enzymes from the pay-off phase reduces the velocity of BDNF-mCherry vesicles (left panel). Mean and s.e.m. for anterograde and retrograde velocities of control and si-GAPDH (anterograde velocity, t=6.355, P=4.3 × 10-10, control: n=321, si-GAPDH: n=235; retrograde velocity, t=5.4, P=9.8 × 10-8, control: n=254, si-GAPDH: n=239), control and si-PGK (anterograde velocity, t=11.88, P=1.9 × 10-31, control: n=973, si-PGK: n=915; retrograde velocity, t=9.6, 1.8 × 10-21, control: n=818, si-PGK: n=829); control and si-PGM (anterograde velocity, t=6.289, P=4.2 × 10-10, control: n=890, si-PGM: n=1,126; retrograde velocity, t=5.940, P=3.6 × 10-9, control: n=802, si-PGM: n=981); control and si-ENO1+2 (anterograde velocity, t=6.852, P=1.5 × 10-11, control: n=455, si-ENO1+2: n=325; retrograde velocity, t=6.498, P=1.6 × 10-10, control n=446, si- ENO1+2=335), and control and si-PK1 (anterograde velocity, t=9.974, P=2.9 × 10-22, control: n=441, si-PK1: n=437; retrograde velocity, control: t=11.58, P=7.3 × 10-24, n=576, si-PK1: n=618). Representative kymographs showing the trajectories of BDNF-mCherry vesicles of different conditions and analysed trajectories with colour-code red for retrograde, green for anterograde and blue for static vesicles (right panel). Scale bar, 10 µm and 10 s. (b) The preparatory phase of glycolysis is required for FAT. Inhibition of HK by 2-DG reduces FAT that is rescued by addition of glucose. Mean and s.e.m. for anterograde and retrograde velocities of control, 2-DG and glucose (anterograde velocity, F(2,784)=27.96, P=1.9 × 10-12, control: n=248, 2-DG: n=151, glucose: n=388; retrograde velocity, F(2,783)=35.42, P=1.9 × 10-15, control: n=242, 2-DG: n=165, glucose: n=379); (c) PK activation by PEP rescues transport defect induced by GAPDH silencing. Anterograde and retrograde velocities are represented as mean+s.e.m. (anterograde velocity, F(2,293)=40.02, P=4.3 × 10-16, control: n=115, si-GAPDH: n=93, si-GAPDH+PEP: n=87; retrograde velocity, F(2, 262)=34.64, P=4.5 × 10-14, control: n=112, si-GAPDH: n=97, si-GAPDH+PEP: n=85). ***P<0.001 and ****P<0.0001.