Philippe Collas’s research while affiliated with University of Oslo and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (353)


Multimodal epigenetic and enhancer network remodeling shape the transcriptional landscape of beige adipocytes
  • Preprint

April 2025

·

19 Reads

Sarah Hazell Pickering

·

·

·

[...]

·

Nolwenn BRIAND

Epigenetic regulation is a key determinant of adipocyte fate and function, conferring phenotypic plasticity to adipose tissue in response to metabolic and thermal challenges. To understand the spatiotemporal regulation of chromatin during the establishment of a beige thermogenic adipocyte phenotype, we analyzed the transcriptomic, epigenetic, and enhancer connectome dynamics during white and beige adipogenesis. Using a machine learning approach, we find that the white-specific transcriptional program is associated with promoter modulations of H3K2ac levels and chromatin accessibility. In contrast, beige-specific mitochondrial gene expression correlates with promoter changes in H3K4me3 levels. Adipocyte beiging is also mediated by a remodeling of the 3D genome involving the recruitment of short range enhancers targeting fatty acid oxidation and thermogenic genes. These increased promoter-enhancer contacts correlate with increased chromatin opening at sites enriched for C/EBP transcription factor motifs. We notably identify the C/EBP transcription factor NFIL3 as differentially bound between white and beige adipocytes at enhancers regulating PDK4, a key metabolic switch promoting fatty acid oxidation. Our results highlight a multimodal, pathway-specific regulation of the transcriptional program underlying the beige adipocyte phenotype.


Repeated cycles of in vitro fasting and refeeding elicit changes in the expression of adipogenic genes and increase the storage of neutral lipids in ASCs. (A) Experimental design: Human adipose stem cells (ASCs) were fasted for 24 h in low glucose (LG; 5 mM glucose) or low glucose/high fatty acid medium (LGHF100/200; 5 mM glucose + 100 or 200 μM oleic acid) and then refed for 24 h with high glucose/low fatty acid medium (HGLF; 17.5 mM glucose). This was performed four times. For the non-fasted protocol (NF), the cells were cultured in HGLF medium for 8 days, with daily medium changes. (B) Relative expression of adipogenic marker genes in non-proliferating ASCs being subjected to different fasting and refeeding regimens. ASCs that had gone through four fasting and refeeding cycles were fixed with paraformaldehyde and neutral lipids stained with (C) Oil Red O, scale bar: 1 cm, or (D) BODIPY, scale bar: 40 µM. The pictures shown are representative of three biological replicates. (E) Relative expression of esterification genes. The gene expression data are presented as mean fold-change ± SEM of 3–4 biological replicates. Fasted (F) vs. Refed (R): * p < 0.05, ** p < 0.01, *** p < 0.001 and Basal (B) vs. F or R: # p < 0.05, ## p < 0.01, ### p < 0.001, determined by one-way ANOVA with Tukey’s multiple comparison test. NF: non-fasted; LG: low glucose; LGHF100/200: low glucose/high fatty acid (100/200 μM oleic acid).
F/R cycled ASCs display a lipolytic/re-esterifying gene profile during adipogenic differentiation. (A) Expression of selected genes in differentiating ASCs subjected to prior fasting and refeeding regimens relative to NF cells. (B) Table view of all genes analyzed by qPCR being significantly upregulated (+), downregulated (−) or unaffected (n.s.) in F/R cycled adipocytes on day 9 or day 15 of differentiation relative to NF cells. The corresponding graphs can be found in Figure S1. The data are presented as mean fold-change ± SEM of 3–4 biological replicates. * p < 0.05, ** p < 0.01, *** p < 0.001 as determined by one-way ANOVA with Dunnett’s multiple comparison test. NF: non-fasted; LG: low glucose; LGHF100/200: low glucose/high fatty acid (100/200 μM oleic acid); n.s: non-significant.
Lipid droplet dynamics during differentiation of fasted and refed ASCs. Differentiating ASCs subjected to prior fasting and refeeding regimens were fixed with paraformaldehyde on day 0, 9,and 15 after start of differentiation, and neutral lipids stained with (A) Oil Red O, scale bar: 1 cm, or (B) BODIPY, scale bar: 40 µM. Lipid droplets (LDs) were quantified using Fiji and presented as (C) violin plots (median + quartiles) and (D) frequency plots (bin width day 0: 0.2 μm²; day 9: 10 μm²; day 15: 40 μm²). The plots are calculated based on three biological replicates. The Y-axes are plotted with log2 scale, and the Day 9 X-axis cropped at 180 μm² to highlight the differences. Significance levels were determined by unpaired, two-tailed t test. NF: non-fasted; LG: low glucose; LGHF100/200: low glucose/high fatty acid (100/200 μM oleic acid).
Insulin sensitivity is reduced in adipocytes during adipogenic differentiation, independent of prior fasting and refeeding. Differentiating ASCs subjected or not to prior fasting and refeeding regimens were serum-deprived overnight and stimulated with 10 nM insulin for 10 min on day 0, 9, and 15 after start of differentiation. The relative level of phosphorylated Akt was determined by immunoblotting with phoshpo-Ser473 and total Akt antibodies and densitometric quantification of the resulting blots. The data are presented as mean p-Akt/Akt ratio ± SEM of 2–3 biological replicates. One representative immunoblot is shown per time point. Insulin-stimulated vs. non-stimulated: * p < 0.05, *** p < 0.001, and insulin-stimulated vs. insulin-stimulated between different F/R regimens: # p < 0.05, as determined by two-way ANOVA with Tukey’s multiple comparisons test. Only significant differences are indicated in the figure. NF: non-fasted; LG: low glucose; LGHF100/200: low glucose/high fatty acid (100/200 μM oleic acid).
F/R cycling-dependent activation of functional pathways are lost by day 15 of differentiation. Total RNA sequencing was performed on ASCs at baseline (Day 0; n = 5, 3 donors), following either four F/R cycles or 8 days in standard culture medium, and after consecutive induction of differentiation on Day 9 and Day 15 under the same prior conditions (n = 3, 1 donor). (A) Differentially expressed genes (DEGs) between time points and pre-treatments in differentiating ASCs (FDR-adjusted p-value < 0.05). (B) Venn diagram of DEGs in LG- and LGHF100-cycled compared to NF cells on day 0. (C) The intersecting DEGs from (B) were tested for overrepresentation against the Reactome gene sets. (D) Gene set enrichment analysis (GSEA) with the MSigDB Hallmark gene set. NF: non-fasted; LG: low glucose; LGHF100: low glucose/high fatty acid (100 μM oleic acid).

+4

Transcriptional and Metabolic Changes Following Repeated Fasting and Refeeding of Adipose Stem Cells Highlight Adipose Tissue Resilience
  • Article
  • Full-text available

December 2024

·

35 Reads

Background: Obesity and related metabolic disorders have reached epidemic levels, calling for diverse therapeutic strategies. Altering nutrient intake, timing and quantity by intermittent fasting seems to elicit beneficial health effects by modulating endocrine and cell signaling networks. This study explores the impact of cyclic nutrient availability in the form of every-other-day fasting (EODF) on human adipose stem cells (ASCs). Methods: We subjected ASCs to repeated fasting/refeeding (F/R) cycles, mimicking low glucose/high fatty acid (LGHF) conditions, and assessed phenotypic and transcriptomic changes, lipid storage capacity, insulin sensitivity, and differentiation potential. Results: Four consecutive F/R cycles induced significant changes in adipogenic gene expression, with upregulation of FABP4 and PLIN1 during fasting, and increased lipid storage in the ASCs. Upon differentiation, ASCs exposed to LGHF conditions retained a transient increase in lipid droplet size and altered fatty acid metabolism gene expression until day 9. However, these changes dissipated by day 15 of differentiation, suggesting a limited duration of fasting-induced transcriptional and adipogenic memory. Despite initial effects, ASCs showed resilience, returning to a physiological trajectory during differentiation, with respect to gene expression and lipid metabolism. Conclusions: These findings suggest that the long-term effects of EODF on the ASC niche may be transient, emphasizing the ability of the adipose tissue to adapt and restore homeostasis.

Download

Model validation. (A) Bodipy staining of lipid droplets of day 15 differentiated white and beige adipocytes from 6 unrelated human subjects. Scale bar: 10µM. (B) Lipid droplet area per field quantified from (A) (ns non-significant; two-way ANOVA with Tukey’s multiple comparison; n>= 15 fields per condition from 3 independent experiments). (C) Western blot analysis of Perilipin1, CITED1, and UCP1 expression in day 15 differentiated white (W) and beige (B) adipocytes from 6 subjects. γTubulin is shown as a loading control. (D) Perilipin1 signals normalized to γTubulin, quantified from western blots (ns, non-significant; two-way ANOVA with Sidák’s multiple comparisons test; n=3). (E) Overrepresentation analysis of genes upregulated in beige vs white adipocytes from all donors (p < 0.05; RNA-seq) using GO and Hallmark gene sets from MSigDb (adjusted p-values < 1.2 x10⁻⁶). The gene ratio is the fraction of differentially expressed genes in each gene set. (F) Heatmap of relative gene expression (z-score transformed FPKM) for differentially expressed beige marker genes (p < 0.01 in any donor, eBayes method, limma package) from the BATLAS gene set (37). (G) Median gene expression of beige marker genes from the BATLAS gene set in day 15 white and beige adipocytes from 6 human subjects (*p < 0.05, **p < 0.01, ***p < 0.001; two-way ANOVA and t-tests with Holmberg adjustment).
Functional classification of beige and white alternative isoforms. (A) Volcano plot of changes in exon-exon junction percentage spliced-in (PSI) between white and beige adipocytes and adjusted p-values per cluster from LeafCutter. (B) Annotation of exon junctions to transcript databases. (C) Venn diagram showing the overlap between differentially spliced genes (DSGs) and differentially expressed genes (DEGs) (D) Mean TRIFID score for exon junctions with white and beige enrichment (| ΔPSI> 0.1 & p < 0.05) compared to junctions with low differential enrichment (| ΔPSI< 0.1) from significant clusters (p < 0.05; n subsampled to 280), and non-significant junctions (p > 0.05; n subsampled to 280; Kruskal-Wallis test and t-tests with Holmberg adjustment). (E) Change in normalized TRIFID scores (Δ TRIFID) between white and beige isoforms (averaged per exon junction) plotted against LeafCutter ΔPSI adjusted p-value. (F) Network representation of genes pertaining to “Mitochondrial Matrix” (GO:0005759) “Fatty acid metabolism” (GO:0006631) and “Adaptive thermogenesis” (GO:1990845) GO terms, colored by Δ TRIFID score.
Splice variants arise from differential TSS usage. (A) Proportion of differentially spliced junctions (DSJs) mapping to the first intron of a transcript, arranged by annotation database. (B) Enrichment of H3K4me3 ChIP-Seq (log ratio of input) at differential TSSs in white and beige adipocytes. TSSs belonging to non-significant DEGs and non-significant DSGs are used as a control (n=23,529). (C) Quantification of H3K4me3 enrichment around the TSS (-1 kb/+2 kb), scaled within each condition (ns, non-significant, ** p < 0.01: ***p < 0.001; two-way ANOVA and Wilcoxon test with Holmberg adjustment). (D) Top 10 differentially enriched transcription factor binding sites at beige vs white TSSs (-2 kb/+0.5 kb) identified by Unibind. (E) Proportion of promoters (TSSs -2 kb/+0.5 kb) that intersect PPARγ ChIP-Seq peaks (beige promoters n=170; white promoters n=174) and promoters from non-significant DSGs and DEGs (n=12 257). Overlapping white and beige promoter regions are shown separately (n=63). (F, H) PPARγ and MED1 ChIP profiles around white (left panel) and beige (right panel) TSSs. (G, I) Quantification of PPARγ (TSS ± 250 bp) and MED1 (TSS ± 500 bp) ChIP enrichment around white, beige, and non-significant TSSs (***p < 0.0001; two-way ANOVA and Wilcoxon test with Holmberg adjustment).
Beige specific isoform expression of PPARG. (A) Schematic representation of differential splicing pattern across the PPARG gene. (B) Integrative genomics viewer (IGV) tracks showing an overlay of S1-S6 RNA-seq reads on the forward strand, PPARγ and MED1 ChIP-seq over input ratios (33), and H3K4me3 and H3K27ac signals across PPARG1 and PPARG2 TSSs in beige vs white adipocytes. (C) ΔPSI on exon1-exon2 junction for PPARG1 (left panel) and PPARG2 (right panel) in white vs beige adipocytes derived from 6 ASC lines (p < 0.0001; LeafCutter). (D) Total PPARG expression level (FPKM) in white vs beige adipocytes (**p < 0.01, ns, non-significant; eBayes method, limma package; n=3) (E) Relative expression of PPARG1 (upper panel) and PPARG2 (lower panel) assessed by qPCR using isoform specific primers (*p < 0.05, **p < 0.01, ***p < 0.001; two-way ANOVA with Tukey’s multiple comparison; n=3). (F) Representative western blot and (G, H) quantification of PPARγ isoforms and total protein expression normalized to γTubulin (*p < 0.05, **p < 0.01, ***p < 0.001; two-way ANOVA with Sidák’s multiple comparisons test; n=3).
Beige-specific isoform expression of PEMT. (A) Schematic representation of differential splicing pattern across PEMT gene. (B) Integrative genomics viewer (IGV) tracks showing an overlay of S1-S6 RNA-seq reads on the reverse strand, PPARγ and MED1 ChIP-seq over input ratios (33), H3K4me3 and H3K27ac signals and Fantom CAGE-seq peaks across PEMT TSSs in white vs beige adipocytes. (C) ΔPSI on exon1-exon2 junction for PEMT-L (left panel) and PEMT-C (right panel) in beige vs white adipocytes derived from 6 ASC lines (p < 0.0001; LeafCutter). (D) Alphafold models for PEMT-S, PEMT-L and PEMT-C isoforms. The additional N-terminal amino acids in PEMT-L and PEMT-C are highlighted in yellow. (E) Total PEMT expression level (FPKM) in white vs beige adipocytes (***p < 0.0005, eBayes method, limma package, n = 3). (F) Relative expression of PEMT-C (left panel) and PEMT-L (right panel) assessed by qPCR using isoform specific primers (***p < 0.001, ns, non-significant, two-way ANOVA with Tukey’s multiple comparison, n = 3). (G) Representative western blot and (H) quantification of PEMT protein expression normalized to γTubulin (*p < 0.05, **p < 0.01, ***p < 0.001; two-way ANOVA with Sidák’s multiple comparisons test; n=3).
Alternative isoform expression of key thermogenic genes in human beige adipocytes

May 2024

·

28 Reads

·

2 Citations

Background The beneficial effect of thermogenic adipocytes in maintaining body weight and protecting against metabolic disorders has raised interest in understanding the regulatory mechanisms defining white and beige adipocyte identity. Although alternative splicing has been shown to propagate adipose browning signals in mice, this has yet to be thoroughly investigated in human adipocytes. Methods We performed parallel white and beige adipogenic differentiation using primary adipose stem cells from 6 unrelated healthy subjects and assessed differential gene and isoform expression in mature adipocytes by RNA sequencing. Results We find 777 exon junctions with robust differential usage between white and beige adipocytes in all 6 subjects, mapping to 562 genes. Importantly, only 10% of these differentially spliced genes are also differentially expressed, indicating that alternative splicing constitutes an additional layer of gene expression regulation during beige adipocyte differentiation. Functional classification of alternative isoforms points to a gain of function for key thermogenic transcription factors such as PPARG and CITED1, and enzymes such as PEMT, or LPIN1. We find that a large majority of the splice variants arise from differential TSS usage, with beige-specific TSSs being enriched for PPARγ and MED1 binding compared to white-specific TSSs. Finally, we validate beige specific isoform expression at the protein level for two thermogenic regulators, PPARγ and PEMT. Discussion These results suggest that differential isoform expression through alternative TSS usage is an important regulatory mechanism for human adipocyte thermogenic specification.


Nuclear remodeling during adipogenesis
a Immunofluorescence of lamin A/C, Phalloidin and DAPI stainings in differentiating ASCs scale bar: 10 µm). b, c, d Relative cytoskeleton volume, coherence, and average branch length measured from phalloidin signal (***p < 0.0001 vs D0, two-way ANOVA with Tukey’s multiple comparison test; n ≥ 5 fields of 2500 µm²). e Lamin A signals, normalized to γTubulin, quantified from Western blots (*p < 0.05, one-way ANOVA with Holm-Šídák’s multiple comparisons; n = 3 experiments). Data are presented as mean ± SD. f Nuclear volumes (voxel) measured from DAPI signal (***p < 0.001 vs D0, two-way ANOVA with Tukey’s multiple comparison; n ≥ 30 cells per time-point from 3 experiments). g, h Scatterplots of nucleus volume (voxel) or nuclear elongation vs maximal lipid droplet (LD) size (pixel), fit with linear regression. i Representative immunofluorescence images of Perilipin1 (PLIN1) and DAPI staining in D15 adipocytes (scale bar: 10 µm).
Nucleolar remodeling during adipogenesis
a Immunofluorescence of Nucleolin (NCL), Perilipin1 (PLIN1) and DAPI staining in differentiating ASCs (scale bar: 10 µm). b Scatter plot of nucleolar volume measured from Nucleolin immunostaining (***p < 0.0001 vs D0, two-way ANOVA with Tukey’s multiple comparison; n ≥ 60 cells per time-point from 3 experiments). c Scatter plot of the number of nucleoli per cell (*p < 0.05, ***p < 0.0001 vs D0, two-way ANOVA with Tukey’s multiple comparison test; n ≥ 60 cells per condition from 3 experiments). Data are presented as mean ± SD. d Scatter plot of nucleolus-to-nucleus volume (No/Nu) ratio (***p < 0.0001 vs D0, two-way ANOVA with Tukey’s multiple comparison; n ≥ 60 cells per condition from 3 experiments; ns, non-significant).
Cell cycle arrest triggers a rearrangement of nucleolar substructure
a Immunofluorescence analysis of Nucleolin (NCL) and UBTF in proliferating (Pro) and D0 conditions (scale bar: 10 µm). b Line profiles from (a) (left panel; *p < 0.05, **p < 0.01 vs nucleolus center, mixed effect analysis with Holm-Šídák’s multiple comparisons; n ≥ 10 nucleoli per condition) and average NCL fluorescence intensity at nucleoli border vs center (right panel) (**p < 0.01 two-tailed paired T test; n ≥ 10 nucleoli per condition). c Immunofluorescence analysis of Nucleophosmin (NPM1) and RNA POL 1 (RPA194) in Pro and D0 conditions (scale bar: 10 µm). d Line profiles from c (left panel; *p < 0.05, **p < 0.01, ***p < 0.001 vs nucleolus center, mixed effect analysis with Holm-Šídák’s multiple comparisons; n ≥ 10 nucleoli per condition) and average NCL fluorescence intensity at nucleoli border vs center (right panel) (**p < 0.01 two-tailed paired T test; n ≥ 10 nucleoli per condition). e SRRF-Stream super-resolution microscopy images of Nucleolin and UBTF immunostainings in differentiating ASCs (scale bar: 10 µm). f Average number of UBTF foci per nuclei in differentiating ASCs (***p < 0.001 vs D0, one-way ANOVA with Tukey’s multiple comparison; n ≥ 3 fields per condition from two independent experiments). g Representative SRRF-Stream super-resolution microscopy images of Nucleolin and UBTF immunostainings in fasted and refed undifferentiated ASCs. h Western blot analysis of NCL expression in fasted and refed conditions. γTubulin is shown as a loading control. i NCL signals normalized to γTubulin, quantified from western blots (non-significant (n.s), Wilcoxon matched-pairs signed rank test; n = 3). Data are presented as mean ± SD.
Translation-related genes expression and translation efficiency during adipogenesis
a Heatmap representation of relative gene expression (log2 FPKM) for genes pertaining to the ribosome biogenesis (GO:0042254) and translation (GO:0006412) gene ontology pathways (differentially expressed genes across time, p < 0.01, eBayes method, limma package). Selected genes of interest are highlighted for each cluster. b SUnSET analysis of protein synthesis during adipogenesis with (+) or without (−)1 µM Puromycin. γTubulin is shown as a loading control. c SUnSET signals normalized to γTubulin, quantified from (b) (*p < 0.05, **p < 0.01, two-way ANOVA with Tukey’s multiple comparison test; n = 3). d SUnSET analysis of protein synthesis rates in fasted and refed conditions during adipogenesis. γTubulin is shown as a loading control. e SUnSET signals normalized to γTubulin, quantified from (d) (*p < 0.05, **p < 0.01, two-way ANOVA with Sidak’s multiple comparisons test; n = 3). f Western blot analysis of P-Thr389 S6K, total S6K and 4EBP1 in fasted and refed conditions during adipogenesis. γTubulin is shown as a loading control. g 4EBP1 signals normalized to γTubulin, quantified from western blots (***p < 0.001, two-way ANOVA with Sidák’s multiple comparisons test; n = 3). Data are presented as mean ± SD.
Cytoskeleton structure defines nucleoli number
a Phalloidin staining only (upper panel) or merged Nucleolin (NCL) immunostaining with DAPI (lower panel) in control condition, and after cytochalasin D (CytoD) 0.5 µM or 5 µM, and Methystat 1 µM treatments in proliferating ASCs (scale bar: 10 µm). b Actin immunostaining only (upper panel) or merged Nucleolin (NCL) immunostaining with DAPI (lower panel) in control condition, and after Japakinolide 2 µM treatment (scale bar: 10 µm). c, d Coherence measurements from (a) and (b), respectively (***p < 0.001 vs Control, Two-way ANOVA with Tukey’s multiple comparison test; n ≥ 20 cells from 3 independent experiments). e Nucleus volume (***p < 0.001 vs Control, Two-way ANOVA with Tukey’s multiple comparison test; n ≥ 35 cells from 3 independent experiments) and (f) cell surface (pairwise t-test with holmberg adjustment for multiple testing; ≥10,000 cells from 3 independent experiments) in control condition, and after CytoD 0.5 µM or 5 µM, Methystat 1 µM and Japakinolide 2 µM treatments. g Nucleoli number, (h) nucleoli volumes and (i) nucleolus-to-nucleus (No/Nu) volume ratio measured from Nucleolin and DAPI in control condition, and after CytoD 0.5 µM or 5 µM, Methylstat 1 µM or Jasplakinolide 2 µM treatments (****p < 0.0001 vs Control, Two-way ANOVA with Tukey’s multiple comparison test; n ≥ 35 cells from 3 independent experiments).
Cytoskeletal rearrangement precedes nucleolar remodeling during adipogenesis

April 2024

·

117 Reads

·

11 Citations

Communications Biology

Differentiation of adipose progenitor cells into mature adipocytes entails a dramatic reorganization of the cellular architecture to accommodate lipid storage into cytoplasmic lipid droplets. Lipid droplets occupy most of the adipocyte volume, compressing the nucleus beneath the plasma membrane. How this cellular remodeling affects sub-nuclear structure, including size and number of nucleoli, remains unclear. We describe the morphological remodeling of the nucleus and the nucleolus during in vitro adipogenic differentiation of primary human adipose stem cells. We find that cell cycle arrest elicits a remodeling of nucleolar structure which correlates with a decrease in protein synthesis. Strikingly, triggering cytoskeletal rearrangements mimics the nucleolar remodeling observed during adipogenesis. Our results point to nucleolar remodeling as an active, mechano-regulated mechanism during adipogenic differentiation and demonstrate a key role of the actin cytoskeleton in defining nuclear and nucleolar architecture in differentiating human adipose stem cells.


Pediatric glioma histone H3.3 K27M/G34R mutations drive abnormalities in PML nuclear bodies

December 2023

·

130 Reads

·

13 Citations

Genome Biology

Background Point mutations in histone variant H3.3 (H3.3K27M, H3.3G34R) and the H3.3-specific ATRX/DAXX chaperone complex are frequent events in pediatric gliomas. These H3.3 point mutations affect many chromatin modifications but the exact oncogenic mechanisms are currently unclear. Histone H3.3 is known to localize to nuclear compartments known as promyelocytic leukemia (PML) nuclear bodies, which are frequently mutated and confirmed as oncogenic drivers in acute promyelocytic leukemia. Results We find that the pediatric glioma-associated H3.3 point mutations disrupt the formation of PML nuclear bodies and this prevents differentiation down glial lineages. Similar to leukemias driven by PML mutations, H3.3-mutated glioma cells are sensitive to drugs that target PML bodies. We also find that point mutations in IDH1/2—which are common events in adult gliomas and myeloid leukemias—also disrupt the formation of PML bodies. Conclusions We identify PML as a contributor to oncogenesis in a subset of gliomas and show that targeting PML bodies is effective in treating these H3.3-mutated pediatric gliomas.


Fig. 1. Alterations in genome compartmental properties during breast cancer progression stages. A: Example of Hi-C data for a region on chromosome 1. TAD delineations are shown as black vertical lines below each track. B: Alluvial plot showing A/B compartment conservation and switching during progression from 10A (left), via T1 (middle) and to C1 (right). C: Example region on chromosome 11 showing a statistically significant switch in the first principal component (PC1) in the three stages. Corresponding subcompartments in the three stages (10A, T1, C1) are shown below. D: Subcompartment switches across 10A, T1 and C1 shown as separate alluvial plots starting at each of the eight different subcompartments in 10A. E: Left: Tomographic view of exemplary Chrom3D model from 10A cells with chromosomes colored individually. Right: The same model with regions colored by their subcompartment associations. F: Plot of median distance from the nuclear center for each subcompartment in each condition.
Fig. 2. Coordinated changes in subcompartments and gene expression. A: Gene expression levels (TPM) by subcompartment state. Vertical axis in log scale. B: Example of subcompartment switching on chromosome 9. C: Heatmaps showing number of DE downregulated genes (left panel), non-differential genes (middle panel) and upregulated genes (right panel) in subcompartments switching between 10A (vertical axis) and T1 (horizontal axis). Dotted lines highlight regions with enrichment relative to non-differential and upregulated genes. D: Log2-ratio of number of genes in subdiagonal sums in the upper vs. lower triangular of the corresponding heatmap from C. E: Heatmaps as in C but contrasting 10A with C1. Dotted lines highlight regions with enrichment relative to non-differential and downregulated genes. F: Log2-ratio plots as in D, but contrasting 10A with C1.
Fig. 4. De novo enhancer formation upon MYC locus insertion on chromosome 10. A: Plot showing lost/gained copies for the entire chromosome 8 in the three cell types. Blue dotted region highlights specific amplification of the chromosome 10 region. B: Plot showing lost/gained copies for the entire chromosome 10 in the three cell types. Red dotted region highlights specific amplification of the chromosome 8 region. C: Interchromosomal C1 Hi-C contacts between the region on chromosome 8 (vertical axis) and the entire chromosome 8. Highlighted region from B indicated in red. End of chromosome 9 and beginning of chromosome 11 shown on the left and right side, respectively. D: Zoom-in on the C1 Hi-C map of the chromosome 8 -10 amplification unit. Dotted circles indicate enriched "dots" of contacts involving enhancers and MYC. Dotted line shows position of MYC gene on chromosome 8 relative to contacts on chromosome 10 within the amplification unit. Positions of enhancers from MCF10A indicated as green segments. Genes shown in red, with names for selected genes.
Multi-level 3D genome organization deteriorates during breast cancer progression

November 2023

·

84 Reads

Breast cancer entails intricate alterations in genome organization and expression. However, how three-dimensional (3D) chromatin structure changes in the progression from a normal to a breast cancer malignant state remains unknown. To address this, we conducted an analysis combining Hi-C data with lamina-associated domains (LADs), epigenomic marks, and gene expression in an in vitro model of breast cancer progression. Our results reveal that while the fundamental properties of topologically associating domains (TADs) remain largely stable, significant changes occur in the organization of compartments and subcompartments. These changes are closely correlated with alterations in the expression of oncogenic genes. We also observe a restructuring of TAD-TAD interactions, coinciding with a loss of spatial compartmentalization and radial positioning of the 3D genome. Notably, we identify a previously unrecognized interchromosomal insertion event, wherein a locus on chromosome 8 housing the MYC oncogene is inserted into a highly active subcompartment on chromosome 10. This insertion leads to the formation of de novo enhancer contacts and activation of the oncogene, illustrating how structural variants can interact with the 3D genome to drive oncogenic states. In summary, our findings provide evidence for the degradation of genome organization at multiple scales during breast cancer progression revealing novel relationships between genome 3D structure and oncogenic processes.


PML modulates epigenetic composition of chromatin to regulate expression of pro-metastatic genes in triple-negative breast cancer

November 2023

·

77 Reads

·

6 Citations

Nucleic Acids Research

The promyelocytic leukemia (PML) protein organizes nuclear aggregates known as PML nuclear bodies (PML-NBs) , where many transcription factors localize to be regulated. In addition, associations of PML and PML-NBs with chromatin are described in various cell types, further implicating PML in transcriptional regulation. However, a complete understanding of the functional consequences of PML association to DNA in cellular contexts where it promotes relevant phenotypes is still lacking. We examined PML chromatin association in triple-negative breast cancer (TNBC) cell lines, where it exerts important oncogenic functions. We find that PML associates discontinuously with large heterochromatic PML-associated domains (PADs) that contain discrete gene rich euchromatic sub-domains locally depleted of PML. PML promotes heterochromatic organization in PADs and expression of pro-met ast atic genes embedded in these sub-domains. Importantly, this occurs outside PML-NBs, suggesting that nucleoplasmic PML exerts a relevant gene regulatory function. We also find that PML plays indirect regulatory roles in TNBC cells by promoting the expression of pro-metastatic genes outside PADs. Our findings suggest that PML is an important transcriptional regulator of pro-oncogenic metagenes in TNBC cells, via transcriptional regulation and epigenetic organization of heterochromatin domains that embed regions of local transcriptional activity.



Cellular and Genomic Features of Muscle Differentiation from Isogenic Fibroblasts and Myoblasts

August 2023

·

175 Reads

·

5 Citations

The ability to recapitulate muscle differentiation in vitro enables the exploration of mechanisms underlying myogenesis and muscle diseases. However, obtaining myoblasts from patients with neuromuscular diseases or from healthy subjects poses ethical and procedural challenges that limit such investigations. An alternative consists in converting skin fibroblasts into myogenic cells by forcing the expression of the myogenic regulator MYOD. Here, we directly compared cellular phenotype, transcriptome, and nuclear lamina-associated domains (LADs) in myo-converted human fibroblasts and myotubes differentiated from myoblasts. We used isogenic cells from a 16-year-old donor, ruling out, for the first time to our knowledge, genetic factors as a source of variations between the two myogenic models. We show that myo-conversion of fibroblasts upregulates genes controlling myogenic pathways leading to multinucleated cells expressing muscle cell markers. However, myotubes are more advanced in myogenesis than myo-converted fibroblasts at the phenotypic and transcriptomic levels. While most LADs are shared between the two cell types, each also displays unique domains of lamin A/C interactions. Furthermore, myotube-specific LADs are more gene-rich and less heterochromatic than shared LADs or LADs unique to myo-converted fibroblasts, and they uniquely sequester developmental genes. Thus, myo-converted fibroblasts and myotubes retain cell type-specific features of radial and functional genome organization. Our results favor a view of myo-converted fibroblasts as a practical model to investigate the phenotypic and genomic properties of muscle cell differentiation in normal and pathological contexts, but also highlight current limitations in using fibroblasts as a source of myogenic cells.


Figure 4
Cytoskeletal remodeling defines nucleolar architecture during adipogenesis

March 2023

·

65 Reads

Differentiation of adipose progenitor cells into mature adipocytes entails a dramatic reorganization of the cellular architecture to accommodate lipid storage into cytoplasmic lipid droplets. Lipid droplets occupy most of the adipocyte volume, compressing the nucleus beneath the plasma membrane. How this cellular remodeling affects sub-nuclear structure, including size and number of nucleoli, remains unclear. We describe the morphological remodeling of the nucleus and the nucleolus during in vitro adipogenic differentiation of primary human adipose stem cells. We find that cell cycle arrest elicits a remodeling of nucleolar structure which correlates with a decrease in protein synthesis. Strikingly, triggering cytoskeletal rearrangements mimics the nucleolar remodeling observed during adipogenesis. Our results point to nucleolar remodeling as an active, mechano-regulated mechanism during adipogenic differentiation and demonstrate a key role of the actin cytoskeleton in defining nuclear and nucleolar architecture in differentiating human adipose stem cells.


Citations (60)


... Could an increase in FFA locally in the WAT contribute to preparing progenitors for browning during EODF, via, e.g., activation of PPARγ? The primary cells used in the current study are able to beige using rosiglitazone, a PPARγ agonist [80] but did not show signs of browning/beiging by induction of brown/beige markers, e.g., UCP1, CIDEA, PPARGC1A, PPARG, CITED1, and SLC27A1 [81,82], upon in vitro EODF (Supplementary Table S2). Two exceptions to this are PDK4 and PLIN2, which showed an increase in LGHF100 relative to LG-cycled cells on day 0. Both genes play an important role in brown adipocytes and are overexpressed in brown as compared to white adipocytes and induced during cold stimulation [83,84]. ...

Reference:

Transcriptional and Metabolic Changes Following Repeated Fasting and Refeeding of Adipose Stem Cells Highlight Adipose Tissue Resilience
Alternative isoform expression of key thermogenic genes in human beige adipocytes

... Cytoskeletal changes play an important role in BAT development. During adipogenesis, cytoskeletal rearrangement precedes nucleolus remodeling, and as adipocytes differentiate and mature, their cytoskeleton gradually becomes more relaxed [19]. A more relaxed cytoskeleton also facilitates the development of thermogenic programs in BAT [20]. ...

Cytoskeletal rearrangement precedes nucleolar remodeling during adipogenesis

Communications Biology

... Moreover, it was reported that ~30% of pediatric gliomas contained mutations in H3.3, including those at residue G34 (H3.3G34V/R) in the H3 tail (4) . Further studies have also found the G34R mutation in adult gliomas and osteosarcomas (4)(5)(6)(7)(8). Work in yeast has revealed recombination defects and genomic instability in the presence of H3.3G34R, and a mouse model revealed impaired DNA damage repair and activation of the cGAS/STING pathway (9). ...

Pediatric glioma histone H3.3 K27M/G34R mutations drive abnormalities in PML nuclear bodies

Genome Biology

... In this scenario, the genomic characterization of non NB-bound PML remains poorly described and only few studies have reported an association of PML to DNA outside the PML-NBs via ChIP-seq and DNA FISH approaches [6,7]. Of note, it was reported that common peak calling algorithms designed for transcription factors could not identify regions of PML chromatin association from ChIP-seq experiments [5], suggesting that the fraction of PML that is immunoprecipitated with this biochemical approach associates to DNA differently. ...

PML modulates epigenetic composition of chromatin to regulate expression of pro-metastatic genes in triple-negative breast cancer

Nucleic Acids Research

... Additionally, the mutation-related phenotype in humans significantly depends on genetic background [16][17][18]. It is possible to preserve it using an approach based on MyoD overexpression in the patient's fibroblasts to obtain myoblasts [19], but this model does not allow for studying myogenic precursors as well as MyoD-related signaling pathways may be influenced. The absence of accessible human-derived muscle tissue is a significant barrier for investigating EDMD1 in vitro. ...

Cellular and Genomic Features of Muscle Differentiation from Isogenic Fibroblasts and Myoblasts

... In addition to their well-established role in providing mechanical support to the NE, lamins, particularly lamin A/C, also influence other critical cellular processes such as peripheral chromatin tethering and gene expression. Lamin A/C interacts with the nuclear lamina to anchor peripheral chromatin, which is crucial for maintaining chromatin organization and regulating gene expression (Burke & Stewart, 2013, Buxboim et al., 2023, Madsen-Østerbye et al., 2023, Shevelyov, 2023. This tethering plays a key role in processes such as DNA replication, DNA damage repair, and the regulation of transcription, all of which can impact nuclear morphology (Butin-Israeli et al., 2015, Herr et al., 2024, Kovacs et al., 2023, Maynard et al., 2019. ...

Gene Regulatory Interactions at Lamina-Associated Domains

... This approach contributed significantly to the field but did not lead to the isolation of loci, containing genes with similar regulatory mechanisms. More recent studies, which sort chromatin based on the transcriptional regulators distribution rather than modifications, made it possible to link the chromatin state of the regions with their functional properties-for example, maintaining the transcription of housekeeping or tissue-specific genes [55,56]. This information suggests that the presence of a coregulator on chromatin may not necessarily mean its active functional state. ...

Eight principal chromatin states functionally segregate the fly genome into developmental and housekeeping roles

... However, due to the rarity of lipodystrophy, cases are frequently unrecognized or misdiagnosed [3,14]. Many patients receive a diagnosis only after they are referred to specialist medical centers with multidisciplinary clinical teams; however, the availability of these centers is limited in most countries [17,20,26,27]. For some patients, severe comorbidities may have developed by the time a definitive diagnosis has been established [14,28,29]. ...

Proceedings of the annual meeting of the European Consortium of Lipodystrophies (ECLip) Cambridge, UK, 7-8 April 2022
  • Citing Article
  • October 2022

Annales d Endocrinologie

... Previous studies have demonstrated that in some cases, a non-protein-coding locus can give rise to functionally distinct transcript isoforms [21][22][23][24]. Recently, it was shown that the switch in the lncRNA HOTAIR start site after the induction of differentiation promotes the inclusion of HOTAIR exon 3, containing a protein-binding domain, which likely changes its function [25]. We showed that TMEM244 is necessary for the growth of cells where its expression is at a relatively high level, such as in CTCL cell lines. ...

De novo annotation of lncRNA HOTAIR transcripts by long-read RNA capture-seq reveals a differentiation-driven isoform switch

BMC Genomics

... Direct or indirect binding of lncRNA to DNA usually correlates with changes in chromatin structure and regulation of gene expression [39][40][41]. For example, HOTAIR has been shown to directly bind to DNA in a triple-helix structure, which helps tether PRC2 to specific genomic loci, contributing to gene repression [42,43]. Furthermore, ChIRP-seq was performed confirmed that HOTAIR associates with chromatin independently of the PRC2 complex [21]. ...

HOTAIR interacts with PRC2 complex regulating the regional preadipocyte transcriptome and human fat distribution
  • Citing Article
  • July 2022

Cell Reports