Fig. 4

a DANCR and ROCK1 shared two similar MREs for miR-335-5p and miR-1972 as predicted by DIANA-LncBase (http://carolina.imis.athena-innovation.gr) and TargetScan (http://www.targetscan.org/vert_71), individually. b miR-335-5p and miR-1972 in tissue level were detected by using ISH, **, P < 0.01, normalized to para-tumor tissue group. c The expression of miR-335-5p and miR-1972 in osteosarcoma tissue and cell lines were assessed by qRT-PCR, ##, P < 0.01 and **, P < 0.01 as normalized to para-tumor tissue group and hFOB 1.19 group, separately. d-e Correlations between DNACR and miR-335-5p (d) as well as DANCR and miR-1972 (e) were qualified by a Pearson correlation analysis, P < 0.0001 for miR-335-5p and P = 0.0009 for miR-1972, respectively. f-g A qRT-PCR assay verified that increased or reduced DNACR negatively regulated miR-335-5p (f) and miR-1972 (g) expression, **, P < 0.01 and ##, P < 0.01 as normalized to pcDNA group and NC shRNA group. h-i An elevation and a depression of miR-335-5p (h) and miR-1972 (i) also inversely regulated DANCR expression, separately. Determined by qRT-PCR, **, P < 0.01 and ##, P < 0.01 as normalized to mimic control group and inhibitor control group. j-m Up-regulation or down-regulation of miR-335-5p (j and l) and miR-1972 (k and m) functioned on ROCK1 expression inversely as measured by qRT-PCR and western blot. **, P < 0.01 and ##, P < 0.01 as normalized to mimic control group and inhibitor control group. p-s Elevation of miR-335-5p and miR-1972 significantly attenuated the facilitative effect of DANCR on ROCK1 expression as determined by qRT-PCR (p and q) and western blot (r and s) in MNNG/HOS and 143B cells, respectively. **, P < 0.01, normalized to pcDNA-DANCR group. Data were shown as mean ± SD from three independent experiments