Fig 2: NAS1 promotes EMT but inhibits tumorigenicity of breast cancer cells.a Representative IF images of EMT markers in CA1h-P2 with NAS1 knockdown (left) or MCF10AT with NAS1 overexpression (right). Scale bar, 100 µm. Random microscopic fields (n > 3) showed similar results. b, c Flow cytometric analyses of CD24/CD44 expression in CA1h-P2 with NAS1 knockdown (b), or MCF10AT with NAS1 or NR2F1 overexpression (c). Red vertical lines indicate the same abscissa value in different panels. Numbers denote the CD24-CD44+ percentages. d, e Flow cytometric analyses of ALDH+ BCSCs in CA1h-P2 with NAS1 knockdown (d) or in MCF10AT with NAS1 or NR2F1 overexpression (e). Numbers show the ALDH+ percentages. DEAB, the ALDH inhibitor. f Tumorsphere formation of CA1h-P2 with NAS1 knockdown (n = 3 culturing experiments). g Tumor incidence in NOD/SCID mice orthotopically injected with various numbers of CA1h-P2 cells after NAS1 knockdown. h Tumor sizes of the mouse groups injected with 20,000 cells in (g) at week 7 (n = 8, 10, and 10 tumors, respectively). i Tumorsphere formation of MCF10AT with NAS1 overexpression (n = 3 culturing experiments). j Tumor incidence in NOD/SCID mice orthotopically injected with various numbers of NAS1-overexpressing MCF10CA1a cells. k Tumor recurrence rate 2 weeks after surgically removing tumors at the size of about 1 cm3 in the mice injected with 20,000 cells in (j). Data represent mean ± SD (f, h, i). Statistical significance was determined by a two-tailed unpaired t test (f, h, i), two-sided chi-square test (k, g, j).

Fig 3: Optic nerve myelination is affected by reduced Nr2f1 dosage A–CNF1A (red, astrocytes) and Sox10 (green, oligodendrocytes) IF on cross-sections of WT and KO P8 ONs showing altered ratio of astrocytes and oligodendrocytes in mutants, particularly in KOs, as quantified in (C).D–JMBP (yellow, fully differentiated oligodendrocytes) IF on cross-sections of WT, HET, and KO P7 optic nerves (ONs) at the chiasm (D–F) and between the chiasm and the eye (G–I). Note the dramatic decrease of MBP+ oligodendrocytes at the chiasm in HET and KO, quantified in (J). The ratio of MBP+ cells is calculated over the total DAPI+ cell number per chiasmal/ON section.K–NEM thin sections of P8 WT, HET, and KO ONs depicting myelin as a dense, dark staining around axonal fibers illustrate a significantly lower number of myelinated fibers in HET and KO, as quantified in (N). Higher magnification images (insets) showing different degrees of myelination in HET and KO ONs. Arrowheads in insets point to myelinated fibers.O–QTEM images of P28 WT and HET confirming persistent hypomyelination in mutants at P28, as quantified in (Q).R, SHigh-magnification EM images of P28 WT and HET ONs showing the difference in myelin compaction between WT (arrowheads in R) and HET (arrowheads in S).Data information: In (C, J, N, Q), data are represented as means ± SEM. N = 4–5 for (C, J); N = 3 for (N, Q). Statistical significance was obtained by Student's t-test or by two-way ANOVA when comparing 2 or multiple conditions, respectively (*P < 0.05; **P < 0.01; ***P < 0.001). Nuclei (blue) were stained with DAPI. Scale bars: 50 µm, except (K–P) (4 µm) and (R, S) (1 µm).

Fig 4: Axonal conduction velocity and visually-dependent operant conditioning are impaired in Nr2f1 HET mice AAnimals were implanted with recording electrodes in lateral geniculate nucleus (LGN), superior colliculus (SC), and visual cortex (V1) and stimulated with a flashing stroboscope.BNissl-stained sections illustrating the location of the recording electrodes (arrows). dhc, dorsal hippocampal commissure; PAG, periaqueductal gray; Th, thalamus.CExamples of field potentials (averaged = 20 times) evoked by flash stimulation in the recording sites of WT and HET mice. Dotted lines indicate the starting point of evoked field potentials.DLatency of evoked field potentials for WT (N = 13) and HET (N = 18) mice.EIn a first series of experiments, animals [WT (N = 13) and HET (N = 18)] were trained to press a lever to obtain a pellet of food in an illuminated Skinner box with a fixed-ratio (1:1) schedule. After 10 days of training, they were transferred to a light on/light off protocol where lever presses were only rewarded when a small light bulb located over the lever was switched on. Lever presses while the bulb was off were punished with a time penalty of = 10 s during which the bulb would not turn on.F, GMean days to criterion (F) and mean number of lever presses during the last two training sessions (G) for the fixed-ratio (1:1) paradigm. No significant differences were observed.HPerformance of WT (N = 13) and HET (N = 18) mice in the light on/light off test. Criterion (dotted line) was that the animal had to press the lever more times during the small bulb light on period than during the light off one. WT mice performed better than HET animals and reached the criterion by the 6th session. The HET failed to reach the criterion for the duration of the test.IThe percentage of lever presses during the light on and light off periods for WT and HET mice was also different, because the WT group pressed the lever preferentially during the light on period. Data were averaged from the 9th and 10th sessions. For statistical significance, see Materials and Methods (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig 5: NAS1 and PTBP1 cooperatively enhance the IRES function of NR2F1-5'UTR.a PTBP1 pulldown by NAS1 RNA in MCF10CA1h lysates. A schematic of the RNA pulldown assay was shown on the left. b NAS1 RNA level in RIP assays for PTBP1-interacting RNAs in MCF10CA1h expressing PTBP1-HA. c NR2F1 protein level after PTBP1 knockdown in NAS1-overexpressing MCF10AT (left) or after PTBP1 overexpression in CA1h-P2 with NAS1 knockdown (right). d NR2F1 mRNA level in RIP assay for PTBP1-interacting RNAs in MCF10CA1h expressing PTBP1-HA. e RIP assays with the MS2–GFP system to analyze the RNA–RNA interaction between NAS1 and NR2F1. A schematic of the assay was shown on the left. f RIP analyses of the binding of PTBP1 to NR2F1 mRNA in CA1h-P2 with NAS1 knockdown (n = 3 replicates from one experiment). g RIP analyses of RNA–RNA interaction between NAS1 and NR2F1 in MCF10CA1h with PTBP1 knockdown. The ratios of NR2F1 to NAS1 were shown (n = 3 replicates from one experiment). h Schematic of the pGF plasmid and the NR2F1-5'UTR segments for IRES activity analyses. Numbers in parentheses indicate the start and end base-pair positions of the segments. GFP and firefly luciferase are transcribed by the same CMV promoter, while luciferase translation is dependent on the IRES activity of the insert sequence between the two genes. i IRES activity of NR2F1-5'UTR was analyzed by a dual-luciferase reporter assay of the pGF system (n = 3 wells from one experiment). EMCV IRES was used as a positive control. NC, negative control. j PTBP1 binding sites on NR2F1-5'UTR predicted by catRAPID (top), and NAS1 binding sites on NR2F1-5'UTR predicted by IntaRNA (bottom). k RIP analyses of the binding of NAS1 to different regions of NR2F1 mRNA. l RNA pulldown assays for interaction between PTBP1 and regions of NR2F1 mRNA. m Effect of PTBP1 on IRES activity of NR2F1-5'UTR-PT (n = 3 wells from one experiment). n Effect of NAS1 overexpression on IRES activity of NR2F1-5'UTR (n = 3 wells from one experiment). o Schematic of the mechanism by which PTBP1 and NAS1 regulate the IRES activity of NR2F1-5'UTR. PTBP1-mediated NR2F1 translation was halted by the GC-rich segment (i) and could be activated by the binding of NAS1 (ii). TSS transcription start site. Data represent mean ± SD. Statistical significance was determined by a two-tailed unpaired t test. Experiments were repeated at least three times independently with similar results; data from one representative experiment are shown.