Fig 1: (A) GSEA showing that the pathways of mitochondrial gene expression, organization and metabolism include genes mostly down-regulated in cells treated with the combination of AA7.1 and SP-2577 compared to vehicle cells. (B) Cytoscape representation of the network of down-regulated biological functions when the combination of AA7.1 and SP-2577 is used. (C) Schematic representation of our model. The overexpression of LSD1 in Gr3 MB positively modulates PRUNE-1 expression via binding to its promoter region leading to activation of TGF-β pathway, through SMAD2/3 phosphorylation and nuclear translocation and PTEN inhibition. AA7.1 is responsible for the proteasomal-dependent degradation of Prune-1 protein [15]. When LSD1/KDM1A inhibitor SP-2577 is used in combination with PRUNE-1 inhibitor AA7.1, LSD1 is not able to enhance PRUNE-1 transcription. Then the activation of TGF-β pathway is lower and PTEN is up-regulated. This determines a deregulation of genes involved in neuronal commitment, antigen presenting cells, a cytotoxic mediated T cell response and a significant impairment of mitochondrial metabolism and OXPHOS. Created with Biorender.
Fig 2: (A) Expression of LSD1/KDM1A derived from multiple datasets of MB (Kool, n = 62; Delattre, n = 57; Pfister, n = 223; Gilbertson, n = 76) compared to that in normal cerebellum (Roth, n = 9) (p < 0.001). LSD1/KDM1A was highly expressed in MB. (B) RNA log2 expression of LSD1/KDM1A derived from publicly available dataset of MB (Cavalli; n = 763; p < 0.001), grouped according to the molecular group disease variants. LSD1/KDM1A was highly expressed in Gr3 MB. (C) Kaplan-Meyer analysis for event-free survival (EFS) of MB and its groups patients according to LSD1/KDM1A expression levels. Patients who showed higher levels of LSD1/KDM1A expression (n = 305) showed shorter event-free survival compared to those with lower levels of LSD1/KDM1A expression (n = 307) (p = ns). Red = Low; Blue = High. (D) Gr3 MB patients who showed higher levels of LSD1/KDM1A expression (n = 56) showed significantly shorter event-free survival compared to those with lower levels of LSD1/KDM1A expression (n = 57) (p < 0.001). (E) Correlation between the expression of LSD1/KDM1A and PRUNE1 in MB (Cavalli; n = 763; R = 0.030; p = ns). (F) Correlation between the expression of LSD1/KDM1A and PRUNE1 in MB (Cavalli; n = 144; R = 0.168; p < 0.001). (G) Correlation between the expression of LSD1/KDM1A and PRUNE1 in MB and (Pfister; n = 223; R = 0.358; p < 0.001). (H) Correlation between the expression of LSD1/KDM1A and PRUNE1 in Gr3 MB (Pfister; n = 56; R = 0.416; p < 0.001). (I) Survival curves of patients affected by Gr3 MB showing PRUNE1 high (n = 73) or LSD1/KDM1A high (n = 64; p = ns). Green = PRUNE1 High; Orange. (J) Combined survival data show that a high expression of both PRUNE1 and LSD1/KDM1A is correlated with a lower percentage of survival in MB (n = 606; p = ns). Blue = PRUNE1 Low/KDM1A Low; Red = PRUNE1 High/KDM1A Low; Green = PRUNE1 Low/KDM1A High; Violet = PRUNE1 High/KDM1A High. (K) Combined survival data show that a high expression of both PRUNE1 and LSD1/KDM1A is correlated with a lower percentage of survival Gr3 MB (low/low = 21; high/high = 34) (p < 0.05). Blue = PRUNE1 Low/KDM1A Low; Violet = PRUNE1 High/KDM1A High.
Fig 3: Generation of the Prune1 conditional allele by CRISPR/Cas9-based gene-targeting. (a) Genomic structure of the mouse Prune1 locus. The regions used for inserting two loxP sites that flank exon 6 (E6) are shown. (b) Schematic representation of a targeting procedure that allowed to efficiently generate the Prune1 conditional alleles within five months. It consists of three steps: (1) In vitro analysis to test targeting efficiency, (2) the first round of targeting to generate mice homozygous for 3′ loxP, and (3) the second round of targeting of the 5′ loxP in zygotes made from the sperm of mice homozygous for 3′ loxP. (c) PCR genotyping of mouse offspring generated from the second round of targeting. Among seven offspring, three contained homozygous 5′ loxP and heterozygous 3′ loxP (1, 6, and 7 indicated by arrows), which was further confirmed by DNA sequencing. The PCR products of mutant (mut) and wt, as well as the formation of heteroduplexes (which reflects heterozygosity), are indicated. Controls used in this genotyping were genomic DNA prepared from: (1) wt C57BL/6J; (2) and (3): Blastocysts carrying homozygous 3′ loxP, and (4) blastocyst carrying homozygous 5′ loxP. The base pair (bp) of DNA marker (M) is indicated.
Fig 4: Characterization of the Prune1Δexon6/Δexon6 as a null allele. (a) Western blot analysis, demonstrating the complete absence of PRUNE1 protein in mouse embryonic fibroblast (MEF) cells derived from E10.5 Prune1Δexon6/Δexon6 embryo. Lane 1, Protein molecular weight marker; Lane 2, wt MEF; Lane 3, the Prune1Δexon6/Δexon6 MEF. The arrow indicates PRUNE1 protein (~58 kDa) which was only detected in wt MEF cells. Ponceau S staining is shown as a normalization control. (b,c) Whole mount imaging of yolk sac and embryos dissected from Prune1Δexon6 intercross at E10.5. The Prune1Δexon6/Δexon6 displayed an avascular yolk sac with primitive vascular plexus and pericardial effusion (arrowhead) compared to the control littermate. (d–g) Hematoxylin-eosin (H&E) staining of sections prepared from E10.5 yolk sac and embryos showing distended capillary with defective hematopoiesis (arrow in (e)) and a thin heart wall with severe reduction of cardiomyocytes and very little trabeculae (arrow in (g)) in the Prune1Δexon6/Δexon6 mutants compared to the control. (h,i) Immunostaining of transverse sections of E10.5 embryos with anti-Endomucin antibody (a marker of endothelial cells), demonstrating lack of invaded vascular plexus in the neural tube of Prune1Δexon6/Δexon6 compared to the control. Arrow indicates invaded blood vessel in the control neural tube.
Fig 5: Family pedigrees, genotype and PRUNE mutation. Mutations in PRUNE1 detected in Omani (A); Iranian (B); Italian (C) and Indian (D) families. (E) Alignment of PRUNE amino acid sequence showing stringent conservation of the Asp30; Pro54; Asp106 and Arg297 residues. (F) 3D model of PRUNE showing the location and close proximity of the Asp30 and Arg297 amino acid residues.
Supplier Page from Abcam for Anti-PRUNE antibody