If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Dysregulated adipose tissue expansion is a primary defect in Prader-Willi syndrome (PWS) children before obesity-onset.
•
White and beige adipogenesis programs are impaired which might result from PPARγ downregulation in PWS adipose-derived SVFs.
•
SNORD116 deficiency is associated with blunted beige adipogenesis process in adipocyte.
•
SNORD116 loss mediates aberrant transcriptional signatures and alterations in alternative pre-mRNA splicing in PWS AdMSCs.
Abstract
Objective
Prader-Willi syndrome (PWS) is a rare genetic imprinting disorder resulting from the expression loss of genes on the paternally inherited chromosome 15q11–13. Early-onset life-thriving obesity and hyperphagia represent the clinical hallmarks of PWS. The noncoding RNA gene SNORD116 within the minimal PWS genetic lesion plays a critical role in the pathogenesis of the syndrome. Despite advancements in understanding the genetic basis for PWS, the pathophysiology of obesity development in PWS remains largely uncharacterized. Here, we aimed to investigate the signatures of adipose tissue development and expansion pathways and associated adipose biology in PWS children without obesity-onset at an early stage, mainly from the perspective of the adipogenesis process, and further elucidate the underlying molecular mechanisms.
Methods
We collected inguinal (subcutaneous) white adipose tissues (ingWATs) from phase 1 PWS and healthy children with normal weight aged from 6 M to 2 Y. Adipose morphology and histological characteristics were assessed. Primary adipose stromal vascular fractions (SVFs) were isolated, cultured in vitro, and used to determine the capacity and function of white and beige adipogenic differentiation. High-throughput RNA-sequencing (RNA-seq) was performed in adipose-derived mesenchymal stem cells (AdMSCs) to analyze transcriptome signatures in PWS subjects. Transient repression of SNORD116 was conducted to evaluate its functional relevance in adipogenesis. The changes in alternative pre-mRNA splicing were investigated in PWS and SNORD116 deficient cells.
Results
In phase 1 PWS children, impaired white adipose tissue (WAT) development and unusual fat expansion occurred long before obesity onset, which was characterized by the massive enlargement of adipocytes accompanied by increased apoptosis. White and beige adipogenesis programs were impaired and differentiated adipocyte functions were disturbed in PWS-derived SVFs, despite increased proliferation capacity, which were consistent with the results of RNA-seq analysis of PWS AdMSCs. We also experimentally validated disrupted beige adipogenesis in adipocytes with transient SNORD116 downregulation. The transcript and protein levels of PPARγ, the adipogenesis master regulator, were significantly lower in PWS than in control AdMSCs as well as in SNORD116 deficient AdMSCs/adipocytes than in scramble (Scr) cells, resulting in the inhibited adipogenic program. Additionally, through RNA-seq, we observed aberrant transcriptome-wide alterations in alternative RNA splicing patterns in PWS cells mediated by SNORD116 loss and specifically identified a changed PRDM16 gene splicing profile in vitro.
Conclusions
Imbalance in the WAT expansion pathway and developmental disruption are primary defects in PWS displaying aberrant adipocyte hypertrophy and impaired adipogenesis process, in which SNORD116 deficiency plays a part. Our findings suggest that dysregulated adiposity specificity existing at an early phase is a potential pathological mechanism exacerbating hyperphagic obesity onset in PWS. This mechanistic evidence on adipose biology in young PWS patients expands knowledge regarding the pathogenesis of PWS obesity and may aid in developing a new therapeutic strategy targeting disturbed adipogenesis and driving AT plasticity to combat abnormal adiposity and associated metabolic disorders for PWS patients.
]. PWS clinical phenotypes are characterized by a series of complex neurodevelopmental–endocrine–metabolic disorders, including neonatal hypotonia and failure to thrive, childhood-onset hyperphagia and excessive weight gain, global development delay, hypogonadism, and mild to moderate mental retardation [
]. The genetic molecular defects of PWS are classified into four subtypes: (1) paternal deletion, (2) maternal uniparental disomy (matUPD), (3) imprinting defects, and (4) rare chromosomal rearrangements [
]. Collective evidence from human data and mouse models suggests that a eutherian-specific noncoding C/D box small nucleolar RNA (snoRNA) family–SNORD116, located in the minimal PWS genetic lesion, plays a major role in PWS pathogenesis, yet the underlying molecular mechanism remains poorly understood [
Loss of the imprinted, non-coding Snord116 gene cluster in the interval deleted in the Prader Willi syndrome results in murine neuronal and endocrine pancreatic developmental phenotypes.
]. Substantially, the SNORD116 cluster can be grouped into group I (SNORD116-1 to -9), group II (SNORD116-12 and SNORD116-14 to -24), and group III (SNORD116-25 and SNORD116-26) [
PWS represents the most common syndromic life-threatening obesity. Approximately 40 % of children with PWS are overweight or exhibit obesity, but this prevalence increases to 82–98 % in adult PWS patients [
]. Substantially, PWS obesity can develop in a switching trajectory: the initial phase 1 (from birth to 2 Y) displays poor feeding during infancy and then steady weight gain; phase 2 (aged 2–6 Y) is the start of excessive weight gain involving a subphase 2a of no appetite change and a subphase 2b of increased food interest; phase 3 and 4 are characterized by typical hyperphagia and morbid obesity that continues into adolescence and adulthood [
]. Hypothalamic dysfunction in PWS has emerged as a recurrent finding that can cause appetite dysregulation and eating disorders and, finally, hyperphagic obesity [
]. Nevertheless, the exact mechanism involved in the developmental shift from phase 2a to subsequent phases has not been explored clearly. Besides, to date, no murine model, either harboring a large deletion of homologous PWS region or a SNORD116 locus deletion, has fully recapitulated the human PWS phenotype, especially the lack of obesity phenomenon [
]; as such, performing research on the pathophysiological mechanism for PWS obesity has been challenging.
Previous studies considered PWS as an exceptional model of extreme adiposity accompanied by adipose tissue (AT) impairment, especially in subcutaneous fat depots [
]. Peculiar body composition in obese PWS patients has been extensively indicated; these patients exhibit abnormally expanded total fat mass and reduced muscle quantity than those with primary obesity of matched body mass index (BMI) [
]. To the best of our knowledge, scarce data exist on the potential molecular mechanism of the formation of unusual adiposity and disturbed AT remodeling in PWS.
Within an individual, the way white adipose tissue (WAT) expands and remodels profoundly impacts the development of obesity and metabolic disorders [
]. De novo adipogenesis can allow promotion of energy homeostasis and metabolic health and offset the negative effects of disturbed metabolic states in patients with obesity and type 2 diabetes [
]. Mechanistic investigation into AT development and adipogenesis process in PWS at an early nonobese stage may help to further decipher affected adipose biology and elucidate the detailed relevance of disturbed AT remodeling in PWS.
In this study, we collected inguinal (subcutaneous) white adipose tissues (ingWATs) from nonobese phase 1 PWS and healthy children aged from 6 M to 2 Y and found impaired WAT development and unusual fat expansion pathway exhibiting adipocyte hypertrophy and blunted hyperplasia in PWS patients. The capacity of in vitro induced white and beige adipogenesis was inhibited and differentiated adipocyte functions were disturbed in PWS-derived adipose stromal vascular fractions (SVFs), which was also validated in SNORD116-knockdown adipocytes. RNA-seq analysis in adipose-derived mesenchymal stem cells (AdMSCs) highlighted transcriptome-wide alterations in pre-mRNA splicing profiles in PWS cells that were mediated by SNORD116 loss. Overall, our findings established the first mechanistic delineation of the pathological AT expansion pathway and defective adipogenesis in PWS and could provide new experimental evidence for the further development of therapeutic options against processive obesity and associated metabolic disorders in PWS patients.
2. Material and methods
2.1 Clinical samples from PWS and healthy children
Six phase 1 PWS male patients and twelve healthy male subjects ranging in age from 6 M to 2 Y were enrolled at the Children's Hospital of Zhejiang University School of Medicine (Hangzhou, China) during 2019–2022. Given that male PWS children with hypogonadism can present cryptorchidism at birth which requires surgical treatment, we collected ingWAT samples specifically during the cryptorchidism surgery in PWS male patients. Also, ingWATs were collected from the same region in healthy control male children with cryptorchidism. This study was approved by the Ethics Committee of the Children's Hospital of Zhejiang University School of Medicine (2018-IRB-055 and 2020-IRB-080). Written informed consent was obtained from the children's guardians with no financial incentive.
All PWS patients were genetically diagnosed and hadn't received growth hormone therapy. For each child participant, BMI was calculated and transformed into age- and sex-specific BMI Z-score based on the WHO standards. All children subjects were at normal weight at the time point of sample collection. Sample information of PWS subjects was included in Table S1.
2.2 Histology
Collected ingWATs were fixed in 10 % neutral buffered formalin for >24 h. The tissues were then embedded in paraffin, sectioned at 4–5 μm, and stained with hematoxylin and eosin (H&E). Immunohistochemical analysis in ingWAT sections was performed using primary antibodies of fatty acid binding protein 4 (FABP4) (Proteintech, Wuhan, China), KI67 (Proteintech), and uncoupling protein 1 (UCP1) (Proteintech) and visualized with 3, 3′-diaminobenzidine (DAB) staining (ZSGB-Bio, Beijing, China). The slides were examined under a digitalized microscope camera (Olympus, Tokyo, Japan) and further analyzed using the ImageJ software (NIH, Bethesda, MD).
2.3 Isolation of SVF and sorting of adipose stem and progenitor cells (ASPCs)
Collected ingWAT samples were rinsed in PBS, minced, and digested in type I collagenase solution (0.2 % in Hank's Balanced Salt Solution (HBSS)) at 37 °C for 90 min. Digested tissue was filtered through a 100-μm cell filter and centrifuged at 1000 rpm for 10 min. The pellet was washed and resuspended in growth media consisting of Dulbecco's modified Eagle's medium (DMEM) with 15 % fetal bovine serum (FBS) and 1 % penicillin/streptomycin (Pen/Strep). Then the SVF cells were seeded within T25 flasks. The culture media were changed every 3 days.
From grown SVFs of the second passage, the ASPC population (defined as being CD31−/CD45−) consisting of AdMSC (CD90+/CD29+) and white adipocyte progenitor (WAP) (platelet-derived growth factor receptor α+ (PDGFRα+)) subpopulations was sorted by flow cytometry on the BD FACS S ORP ARIA II flow cytometer (BD Biosciences, USA) (details in Supplementary materials and methods). The sorting data were analyzed using FlowJo software (Tree Star, USA). The sorted identified AdMSCs were seeded within T25 flasks and cultured with growth media. Both the SVF cells and AdMSCs were subcultured until 80–90 % confluence and used for subsequent experiments within seven passages.
2.4 Cell transfection
Cell transfection of AdMSCs and adipocytes was conducted by the primary cell 4D-nucleofectorTM X kit (Lonza Group Ltd., Basel, Switzerland) on a 4D-Nucleofector (Lonza Group Ltd), following the manufacturer's protocol. Optimal programs and solutions for the nucleofection delivery system were tested. A pool of ten specific phosphorothioate-modified antisense oligodeoxynucleotides (ASOs) was designed to silence the SNORD116 cluster of three groups altogether and synthesized by RiboBio (Guangzhou, China). Primers for ASO are listed in Table S4. The SNORD116-overexpression plasmid via artificial construct, pcDNA3.1-MBII85-5'SS mut [
2.5 RNA-seq analysis and alternative splicing event (ASE) analysis
Total RNA of AdMSCs (second passage) were collected and added Trizol reagent (Invitrogen, USA), and then handed over to LC-Bio Technology Co., Ltd. (Hangzhou, China) for the subsequent cDNA library construction and next-generation RNA-seq analysis on an Illumina sequence platform. After the final transcriptome was generated, StringTie and Ballgown were used to estimate the expression levels of all transcripts and perform expression levels for mRNAs by calculating FPKM. The differentially expressed genes (DEGs) were selected by R (version 3.6.3) using the DESeq2 package version 1.26.0. p-value and Benjamini–Hochberg false discovery rate (FDR) was calculated for each DEG.
Analysis of differential exon usage was performed by the R package DEXSeq. ASEs were systematically detected and analyzed by rMATS (4.1.2) automatically corresponding to all major categories of ASEs (skipped exon, alternative 3′ and 5′ splice site, mutually exclusive exons, and retained intron). Each ASE with an alternatively spliced cassette exon was assigned a ψ value to represent the related exon inclusion level. p-value and FDR were calculated for each ASE. Cassette exons with |Δψ| ≥ 0.2 were identified as significantly changed ASE. Validation of ASEs was performed by RT-PCR with primers listed in Table S3.
2.6 Statistical analysis
Numeral data were presented as mean ± standard deviation (SD). Data normality was determined by Shapiro-Wilk test. Statistical analyses for two groups were performed by unpaired two-tailed Student's t-test (for normally distributed variables) or nonparametric Mann-Whitney U test (for non-normally distributed variables); statistical analyses for more than two groups were examined using one-way ANOVA followed by post hoc Tukey's test (Graphpad Software Inc., La Jolla, CA). p < 0.05 was regarded as statistical difference. All experiments were performed on at least three separate subjects and independently repeated at least three times with similar results.
3. Results
3.1 Nonobese PWS children show a dysregulation in the WAT expansion pathway and inhibited adipogenic differentiation
Histological analysis showed that there was a substantial enlargement of existing adipocytes with a lower adipocyte number in PWS ingWATs than in healthy controls (Fig. 1A ). The terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay indicated more apoptotic adipocytes in PWS ingWATs (Fig. 1B), but no significant difference in adipose proliferation was detected by Ki67 immunostaining (Fig. S1). Moreover, through immunofluorescence staining of the macrophage marker F4/80, we observed greater macrophage infiltration into PWS-derived ingWAT (Fig. 1C), which suggested a changed adipose microenvironment. Then, we sought to assess the characteristics of adipose hyperplasia in isolated SVFs derived from PWS ingWATs from the perspectives of ASPC population and differentiation capacity. Flow cytometry analysis of SVF cells demonstrated no significant variation in the percentage of hematopoietic non-Lin-committed cell populations (CD31−/CD45−), AdMSC population (CD29+/CD90+) and WAP population (PDGFRα+) between the two groups (Fig. 1D).
Fig. 1PWS children without obesity-onset show dysregulated adipose tissue remodeling features. The inguinal (subcutaneous) white adipose tissue (ingWAT) samples were collected from nonobese PWS and healthy control (Con) children. (A) Representative images of H&E staining with the average adipocyte diameter (6 PWS and 12 Con samples, 3 sections per sample) and FABP4 immune-histochemical staining in ingWAT sections (n = 3) in ingWAT samples. Scale bar = 100 μm. (B) Representative TUNEL staining of ingWAT sections (n = 3). Scale bar = 500 μm. (C) Immuno-fluorescence staining of F4/80 (macrophage marker) (green) in ingWAT sections (n = 3). The nuclei (blue) were stained with DAPI. Scale bar = 20 μm. (D) Representative images of ASPC profiles analyzed by flow cytometry and AdMSC population percentage in SVFs from ingWATs (n = 3). Values are mean ± SD. *p < 0.05 and ns stands for not significant. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
The expression loss of genes located in the PWS region on chromosome 15 in PWS SVFs was validated by quantitative real-time PCR (qRT–PCR) (Fig. 2A ). Through a 14-day in vitro adipogenic induction (Fig. 2B), PWS SVF-differentiated adipocytes showed less formation of lipid droplets (Fig. 2C) and a lower level of triglyceride content (Fig. 2D) than control cells. The transcript levels of white adipogenesis markers, such as peroxisome proliferator activated receptor γ (PPARγ), CCAAT/enhancer binding protein-α (C/EBP-α), EBF transcription factor 2 (EBF2), bone morphogenetic protein 2 (BMP2), FABP4, and fatty acid synthase (FAS), as well as lipid droplet genes, such as perilipin 1 and 2 (PLIN1 and PLIN2), were significantly decreased in PWS differentiated adipocytes than in control adipocytes (Fig. 2E). Consistently, the protein expression levels of PPARγ, C/EBP-α, and FABP4 were reduced in PWS white adipocytes (Fig. 2F). Additionally, the function of releasing adipokines, including leptin and adiponectin, in PWS mature adipocytes was also found to be impeded (Fig. 2G, H).
Fig. 2PWS-derived SVF cells exhibit attenuated adipogenic differentiation capacity. (A) The expression of genes located in the PWS region on chromosome 15 was measured by qRT–PCR (n = 6 PWS, n = 12 control (Con)). Relative mRNA levels normalized to β-actin are presented, relative to control levels. (B) Schematic illustration of in vitro cocktail (Insulin/Dexamethasone (DEX)/3-isobutyl-1-methylxanthine (IBMX)/Rosiglitazone (Rosg)/Indomethacin (Indo)) induction of adipogenic differentiation process. (C–H) The adipogenesis process of PWS and Con SVFs was induced by incubating with the differentiation cocktail on Day 0. The adipogenesis status of mature adipocytes was determined and analyzed on Day 14. (C) Oil Red O-stained differentiated adipocytes (n = 4). Scale bar = 100 μm. (D) Triglyceride contents of differentiated adipocytes, quantified and normalized to protein contents (n = 4). (E) Relative mRNA levels of adipogenesis marker genes and lipid droplet genes in differentiated adipocytes (n = 4), as detected by qRT–PCR analysis. β-actin serves as an internal control. (F) Western blot analysis of the protein contents of PPARγ, C/EBP-α, and FABP4 in differentiated adipocytes (n = 4). α-Tubulin was used as a loading control. (G, H) Leptin levels (G) and adiponectin levels (H) in cell culture supernatants of mature adipocytes at Day 14 after white adipogenesis induction (n = 4), as assessed by enzyme-linked immunosorbent assays (ELISAs). Values are mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3.2 Beige adipogenic differentiation program is impaired in PWS adipocytes
The biogenesis of beige adipocytes, a specific subset of brown-like cells, can occur in subcutaneous WAT to act a vital role in regulating energy and metabolism homeostasis [
]. We found that the expression of UCP1, a marker of beige AT, was substantially reduced in PWS ingWATs compared to controls (Fig. 3A ). To further investigate the beige adipocyte differentiation process in PWS subjects, we successfully established beige adipogenesis induction in SVF cells (Fig. S2). Subsequently, qRT–PCR analysis showed reduced transcription of browning marker genes, including PR domain-containing protein 16 (PRDM16), peroxisome proliferator activated receptor γ co-activator 1-α (PGC1-α), UCP1, peroxisome proliferator-activated receptor α (PPARα), and fatty acid binding protein 3 (FABP3), and mitochondrion-associated genes, such as carnitine palmitoyltransferase 1B (CPT1B), cytochrome c oxidase subunit 7A1 (COX7A1), and ATP synthase subunit α (ATP5A), in PWS beige adipocytes after in vitro beiging differentiation than in controls (Fig. 3B). The protein levels of PRDM16, PGC-1α, PPARγ, and UCP1 were also significantly decreased in PWS beiged adipocytes (Fig. 3C). UCP1 immunoreactivity and mitochondrial density were also lower in differentiated PWS beige adipocytes (Fig. 3D, E). Upon an induction of 3-day browning transdifferentiation, mitochondrial respiration and oxygen consumption were suppressed in PWS differentiating beige adipocytes compared with control cells (Fig. 3F).
Fig. 3PWS-derived SVF cells show disturbed beige adipogenic differentiation process in vitro. (A) Representative images of immuno-histochemical staining of UCP1 in PWS and healthy control (Con) ingWATs sections (n = 3). Scale bar = 100 μm. (B–E) The beige adipogenesis process of SVFs was induced by incubating with the beiging differentiation cocktail on Day 0. The beige adipogenesis status of mature adipocytes was determined and analyzed on Day 14. (B) Relative mRNA levels of beige adipocyte marker genes and mitochondrion-associated genes in differentiated beige adipocytes (n = 4), as indicated by qRT–PCR analysis. β-actin serves as an internal control. The mitochondria gene expression is normalized to PPIA. (C) Western blot analysis of the protein contents of PRDM16, PGC-1α, PPARγ, and UCP1 in differentiated beige adipocytes (n = 4). α-Tubulin was used as a loading control. (D) Immuno-fluorescence staining of UCP1 (red) and fluorescence quantification in differentiated beige adipocytes (n = 4). The nuclei (blue) were stained with DAPI. Scale bar = 20 μm. (E) MitoTracker staining (green) and fluorescence quantification in differentiated beige adipocytes (n = 4). The nuclei (blue) were stained with Hoechst. Scale bar = 20 μm. (F) Oxygen consumption rate (OCR) analysis in beige adipocytes at Day 3 of differentiation (n = 3). Oligomycin, carbonyl-cyanide-4-(trifluoromethoxy)-phenylhydrazone (FCCP), and rotenone/antimycin A (Rot/AA) were sequentially added into the cell pool at the time points indicated by dashed lines. Values are mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3.3 Transcriptome-wide profiling is suggestive of aberrant adipogenesis and cell proliferation pathways in PWS AdMSCs
Through unbiased transcriptome-wide RNA-seq data from PWS and control AdMSCs, we characterized transcriptomic signatures in PWS subjects. Among a total of 56,465 examined genes, we identified 1203 DEGs between PWS AdMSCs and controls (Benjamini–Hochberg FDR < 0.25; 444 with FDR < 0.05); 944 genes were upregulated (fold change (FC) >1.5; 413 with FDR < 0.05); 259 genes were downregulated (FC < 0.5; 65 with FDR < 0.05) (Figs. 4A and S3A, B; Table S5). A random subset of upregulated and downregulated DEGs was validated by qRT–PCR (Fig. S3C, D). The volcano plot representing the significant DEGs appeared dissymmetric, suggesting a stronger transcript abundance change in upregulated genes against downregulated ones in PWS cells (Fig. 4A). The 31 most significantly changed genes (FDR < 5 × 10−5; 27 increased and 4 decreased) included genes located in the PWS critical region (NDN, SNRPN and SNORD116 cluster) and 29 other affected genes (e.g., PDK4, MMD, ADM2, ASPN, and JPH2) (Fig. 4B; Table S2).
Moreover, gene ontology (GO) functional clustering analysis of the 1203 aberrant DEGs clearly indicated the topmost remarkably affected categories in biological processes, including the processes of fat cell differentiation, cell proliferation regulation, and glucose metabolic process (Fig. 4C), and those in cellular components and molecular functions (Fig. S3E). The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed the pathways that were the most significantly different between PWS AdMSCs and controls, including Wnt signaling pathway, PPAR signaling pathway, insulin resistance pathway, and adipocyte lipolysis regulation pathway (Fig. S3F). Further DEG clustering analysis was conducted in the regulatory pathways of adipogenesis; cell proliferation, an important step during the early phase of the adipogenesis process [
]; and glucose metabolic process, and the results revealed abnormal expression patterns of genes involved in these processes (Fig. 4D). A higher portion of inhibitory adipogenic genes and a lower portion of positive adipogenic genes in PWS AdMSCs, as also validated by qRT–PCR (Fig. 4E, F), suggested a disordered adipogenesis regulation pathway and adipocyte related biological processes in PWS. Specifically, PPARγ, the master regulator of adipocyte biology involving adipogenesis, was notably reduced in PWS AdMSCs than in control AdMSCs (Fig. 4D). We then validated the decrease in both the transcriptional and protein levels of PPARγ in PWS AdMSCs (Fig. 4G, H). In addition, adipocyte proliferation capacity was observed to be enhanced in PWS AdMSCs compared with controls (Fig. 4I).
Fig. 4RNA-seq analysis suggests dysregulated adipogenesis and proliferation pathways in PWS AdMSCs. (A) The volcano plot illustrating differentially regulated DEGs from RNA-seq data between PWS and control (Con) AdMSCs. See also Table S5. (B) Heatmap visualization of the representative DEGs whose expressions were increased by >1.5-fold or decreased by <0.5-fold. n = 3 PWS, n = 3 Con. Color code refers to the z-score. (C) The Gene Ontology (GO) functional clustering analysis of DEGs for biological processes. (D) Heatmaps of the representative DEGs involved in the specific pathways of adipogenesis, cell proliferation regulation, and glucose metabolic process in PWS and Con AdMSCs with calculated z-score. (E) Relative mRNA levels of genes of positive adipogenesis regulation in PWS and Con AdMSCs by qRT–PCR analysis (n = 4). β-actin serves as an internal control. (F) Relative mRNA levels of anti-adipogenesis genes in PWS and Con AdMSCs by qRT–PCR analysis (n = 4). β-actin serves as an internal control. (G) Relative mRNA levels of PPARγ gene in PWS and Con AdMSCs by qRT–PCR analysis (n = 4). β-actin serves as an internal control. (H) Western blot analysis of PPARγ protein contents in PWS and Con AdMSCs (n = 4). α-Tubulin was used as a loading control. (I) Representative immunocytochemical co-stainings of 5-ethynyl-2′-deoxyuridine (EdU) (green) and Hoechst (blue) in AdMSCs (n = 3). The proliferative cells of EdU-positive are quantified. Scale bar = 100 μm. Values are mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3.4 Knockdown of SNORD116 is associated with PPARγ downregulation and impaired beige adipogenesis in adipocytes
To explore whether SNORD116 deficiency exerted regulatory effects on AdMSC proliferation and beige adipocyte differentiation as tested in PWS adipocytes, we depleted the SNORD116 cluster by transiently transfecting control AdMSCs with chimeric RNA–DNA ASOs (Fig. 5A ). Upon SNORD116 depletion, adipocyte proliferation significantly increased (Fig. 5B). Then, we measured the expression of SNORD116 at three differentiation stages (Day 0/4/14) during the beige adipogenesis process and found that the snoRNAs of SNORD116 cluster increased during the browning differentiation time course and reached the peak level in mature beige adipocytes, with a marked expression variation in distinct snoRNAs (Fig. 5C). Then, we induced the beige adipogenesis in ASO-SNORD116 AdMSCs for 14 days and monitored decreased lipid contents (Fig. 5D), reduced expression levels of beige adipocyte markers, such as PPARγ, PRDM16, PGC-1α, UCP1, PPARα, cell death inducing DFFA like effector a (CIDEA), COX7A1, and ELOVL fatty acid elongase 3 (ELOVL3) (Fig. 5E), and lower mitochondria density (Fig. 5F). Moreover, SNORD116 knockdown mediated PPARγ downregulation both in pre-differentiated AdMSCs and in post-differentiated mature adipocytes (Fig. 5G), which might have been responsible for the inhibited adipogenic gene program in PWS AdMSCs that was linked with the impaired termination of adipogenesis.
Fig. 5SNORD116 deficiency affects beige adipogenesis and cell proliferation in adipocytes. (A) qRT–PCR analysis of the relative snoRNA expression levels of snoRNAs in SNORD116-SNORD116-6, 19, 26-in AdMSCs transiently transfected with anti-SNORD116 (ASO-SNORD116) or scramble (Scr) ASOs (n = 3). U6 small nuclear RNA (snRNA) serves as an internal control. (B) Representative immunocytochemical co-stainings of EdU (green) and Hoechst (blue) in ASO-SNORD116 and Scr AdMSCs (n = 3). The proliferative cells of EdU-positive are quantified. Scale bar = 100 μm. (C) The relative snoRNA expression levels of snoRNAs in SNORD116 cluster, involving SNORD116-6, 19, 26, at different time points (Day 0/4/14) during the beige adipogenic differentiation process, detected by qRT–PCR (n = 3). U6 serves as an internal control. (D) Oil Red O-stained differentiated Scr and ASO-SNORD116 beige adipocytes (n = 3). Scale bar = 100 μm. (E) Relative mRNA levels of beige adipocyte markers in differentiated ASO-SNORD116 and Scr beige adipocytes (n = 3), as indicated by qRT–PCR analysis. β-actin serves as an internal control. (F) MitoTracker staining (green) and fluorescence quantification of differentiated ASO-SNORD116 and Scr beige adipocytes (n = 3). The nuclei (blue) were stained with Hoechst. Scale bar = 20 μm. (G) Relative mRNA levels of PPARγ gene detected by qRT–PCR in ASO-SNORD116 and Scr AdMSCs as well as in post-differentiated beige adipocytes followed by 48 h treatment with ASO-SNORD116 or Scr (n = 3). β-actin serves as an internal control. Values are mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
3.5 SNORD116 loss mediates extensive alterations in pre-mRNA splicing profiles in PWS adipocytes
Since two previous studies have reported exceptionally different splicing patterns and reduced splicing efficiency in the brain tissue of PWS individuals [
], we conducted a transcriptome-wide search for potential evidence of altered pre-mRNA splicing profiles in PWS adipocytes (Table S6). The RNA-seq data analysis revealed 2481 significantly changed ASEs in 1763 genes (FDR < 0.05) in PWS AdMSCs, among which 748 ASEs in 594 gene loci had |Δψ| ≥ 0.2 (Fig. 6A; Table S6). The most frequently observed splicing mode in PWS was skipped exon (Fig. S3G). Additionally, the differential expression and splicing of genes was decoupled that only a fraction of genes with putative differential splicing tended to be differentially expressed (Fig. 6B). Several alternative splice transcripts in PWS cell samples were randomly selected and applied for validation (Fig. 6C).
Recently, a mechanistic understanding of the molecular role of SNORD116 in modulating alternative RNA splicing has emerged [
]; therefore, we examined whether SNORD116 deficiency would affect splicing choices in PWS adipocytes. Due to the lack of high-quality microarray data on splicing analysis in PWS AT for comparison, we overlapped our splicing analysis result with those obtained from two previous independent microarray studies in human PWS hypothalamus [
]. However, minimal overlap between the differentially spliced genes was found, and calpastatin (CAST) was the only dysregulated gene that overlapped between these data sets (Fig. 6D). Then, we selected several recurrent differentially spliced genes and validated their splicing patterns in ASO-SNORD116 adipocytes (Fig. 6E).
Fig. 6Loss of SNORD116 can mediate the altered patterns of alternative RNA splicing in PWS adipocytes. (A) Scatter plot of altered pre-mRNA splicing of cassette exons in PWS AdMSCs. Cassette exons with significant ψ value (|Δψ| ≥ 0.2, FDR < 0.05) were selected. Red points mean cassette exons with higher inclusion levels in PWS cells; blue points mean cassette exons with higher exclusion levels in PWS cells. See also Table S6. (B) Overlap between putative DEGs and differentially spliced genes in PWS cells. (C) Validation of altered pre-mRNA splicing in PWS cells compared with healthy control (Con) cells by semi-quantitative RT-PCR. Exclusion/inclusion levels of the internal exons were calculated. The expression of GAPDH was detected to ensure equal loading of PCR products. (D) Overlap of the genes with altered splicing pattern in PWS AdMSCs with the computationally predicted SNORD116 targets realized by snoTARGET and those differentially spliced genes implicated in PWS hypothalamus and one human embryonic stem cell (hESC) line with the PWS region lncRNAs (SPAs) knockout. (E) Validation of altered pre-mRNA splicing in ASO-SNORD116 cells compared with Scr cells by semi-quantitative RT-PCR. Exclusion/inclusion levels of the internal exons were calculated. The expression of GAPDH was detected to ensure equal loading of PCR products. (F) Schematic representation of the PRDM16-isoform11 (PRDM16iso11) transcript construct generated by the inclusion of intron13b (I13b) and confirmation by 5′RACE and Sanger sequencing. 5′RACE displayed the alternative transcription start site (TSS) within I13. Sanger sequencing confirmed the junction fragment in the I13b and exon14 (E14). (G) qRT–PCR analysis of the relative mRNA expression of PRDM16iso11 transcript and the I13b inclusion level in PWS cells and Con cells (n = 3). β-actin serves as an internal control. (H) qRT–PCR analysis of the relative mRNA expression of PRDM16iso11 transcript and I13b inclusion level in ASO-SNORD116 cells and Scr cells (n = 3). β-actin serves as an internal control. (I) The relative mRNA expression level of PRDM16iso11 transcript in PWS and Con AdMSCs transfected with SNORD116-expressing vector and a controlled GFP-expressing construct, detected by qRT–PCR (n = 3). β-actin serves as an internal control. Values are mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
To specifically identify the biological relevance of SNORD116 in modulating alternative pre-mRNA splicing in PWS adipocytes, we chose to examine the splicing alteration in one alternatively spliced gene, PRDM16. PRDM16-isoform11 (PRDM16iso11), is a splicing variant with two constituted exons and is transcribed from an alternative transcription start site (TSS) within a retained intron (Fig. 6F). Rapid amplification of cDNA 5′ end (5′ RACE) and Sanger sequencing revealed part of the amplification product of the PRDM16iso11 transcript that mapped to the TSS within the retained intron (Fig. 6F). We found that the expression of the PRDM16iso11 transcript and the intron inclusion level was higher in PWS cells compared with controls (Fig. 6G) and also in ASO-SNORD116 cells compared with scramble (Scr)-treated cells (Fig. 6H). SNORD116 overexpression led to a decreased expression level of the PRDM16iso11 variant in PWS cells than in the cells transfected with a controlled GFP-expressing plasmid but also unexpectedly increased the PRDM16iso11 level in control cells (Fig. 6I), suggesting that SNORD116 could act in the alternative splicing of PRDM16 pre-mRNA and drive intron retention in the PRDM16iso11 isoform.
4. Discussion
In PWS patients, excessive weight gain and severe obesity develop gradually after the initial phase of poor feeding and lack of weight gain at infancy, and the patients then go through an excessive progression in weight-gain with marked hyperphagia and food addiction continuing from childhood and adolescence into adulthood [
]. Our result that phase 1 PWS children exhibited hypertrophy and increased apoptosis in white adipocytes supported that impaired WAT and imbalanced fat accumulation could be a primary pathogenic process in the syndrome, which may further engage in orchestrating local aberrations of body composition. The finding that adipocytes were substantially larger in PWS children than in healthy children is consistent with that observed in comparisons between obese PWS adults and primary obese subjects [
Within an individual, AT is the most significant energetic metabolism organ and active endocrine organ, lying in the regulatory hub of systemic metabolic states and energy homeostasis [
]. Dysfunctional adiposity can intimately affect multiple physiological processes, such as insulin sensitivity, glucose and lipid metabolism, and inflammation [
]. Indeed, adipocyte hypertrophy, the pathologic characteristic of WAT remodeling, is associated with rapid growth of fat mass accompanied by systemic metabolism disorders [
]. When hypertrophic adipocytes reach their lipid-laden capacity, cell death occurs, resulting in the activation of fibrosis and a chronic low-grade inflammatory phenotype [
]. In contrast, adipocyte hyperplasia represents healthy fat mass expansion manner and appropriate AT remodeling through the effective recruitment of ASPCs into the adipogenic axis, along with subsequent angiogenesis, minimal fibrosis induction, and a lower degree of inflammation [
], during which PPARγ coordinates with C/EBP transcription factors to fully activate the transcriptional cascade, resulting in the maturation of functional insulin-responsive adipocytes [
]. The observed PPARγ downregulation in both PWS and ASO-SNORD116 AdMSCs/adipocytes implicated that SNORD116 deficiency could mediate a disrupted differentiation gene program engaging in impaired adipogenic formation capacity and adipocyte functions in PWS.
Here, we noticed that our finding that there was no significant variation in the resident ASPC population in isolated ingWAT SVF cells obtained from phase 1 nonobese PWS was not in alignment with one previous study in which WAP content was depleted within PWS AT aged from 2 M to 19Y [
]. However, given that the possibility of associative interference from age- and obesity occurrence-dependent regulation in SVF WAP status was not excluded in that study, we cannot explain these contradictory results. Further investigation is required to evaluate the possibility of age-, gender-, and BMI-related effects on the SVF ASPC population in PWS AT.
White adipocytes can undergo transdifferentiation into brown-like adipocytes dependent on PPARγ [
]. Active beige adipocytes exhibit increased expression of UCP1, PRDM16, PGC-1α, and CIDEA, improved mitochondria density and activity, and enhanced thermogenesis ability [
]. In the present study, we also provided direct evidence of the impaired beige adipogenesis process and suppressed uncoupling respiration chain reaction and oxygen consumption rate (OCR) of mitochondria in differentiating beige adipocytes of PWS. In the same line, two earlier studies have reported mitochondrial dysfunction in one PWS mouse model (the PWS-imprinted center deletion mice) [
]. These results may constitute a mechanistic understanding of disrupted mitochondrial function and bioenergetics in PWS, and they drive further investigation to identify whether mitochondrial dysfunction has pathogenic relevance to impairments in AT, affects whole-body energy metabolism, and further acts as a contributing factor in obesity development within PWS individuals. Furthermore, as metabolically active beige AT is substantially involved in body energy expenditure and metabolism-related profiles [
], the impeded beiging differentiation process may exert negative effects in PWS. This solicits further studies, for example, fluorodeoxyglucose positron emission tomography (FDG-PET), to examine beige AT deposits and distribution features in PWS subjects.
Through transcriptional analysis in PWS and control AdMSCs, we characterized aberrant DEG profiles in PWS. Noticeably, among the 31 most significant DEGs (FDR < 5 × 10−5), several genes are known to be linked with adipocyte biology and metabolism homeostasis, such as TIMP metallopeptidase inhibitor 3 (TIMP3), adrenomedullin 2 (ADM2), and pyruvate dehydrogenase kinase 4 (PDK4). TIMP3, markedly increased in PWS AdMSCs, is a secreted metalloproteinase inhibitor and has been implicated in blocking adipogenesis and affecting human adipose tissue remodeling [
The role of metalloproteinases and their tissue inhibitors in adipose tissue remodelling and whole-body lipid distribution: a cross-sectional clinical study.
]. These findings may lead to a hypothesis regarding whether they also mechanistically participate in dysfunctional PWS AT remodeling, which should be explored further.
The minimal genetic lesion in PWS contains the snoRNA SNORD116 family, which is suspected to play a vital role in driving the whole PWS phenotypes but with unclear pathogenic mechanisms [
Loss of the imprinted, non-coding Snord116 gene cluster in the interval deleted in the Prader Willi syndrome results in murine neuronal and endocrine pancreatic developmental phenotypes.
]. Recent work in the field has revealed that SNORD116 can function in the post-transcriptional processing of mRNA, including alternative RNA splicing [
]. Our transcriptome-wide search in AdMSCs supported the evidence of the extensively differential utilization of alternatively spliced variants in PWS AdMSCs. Additionally, Cast, the only overlapping gene identified in our splicing analysis, previous data performed in human PWS hypothalamus [
], is an endogenous cysteine protease inhibitor, which composes a multifunctional proteolytic system-calpain-calpastatin system and can be involved in numerous physiological processes, such as cytoskeleton remodeling, cell cycle, and angiogenesis, as well as cancer cell invasion and tumorigenesis [
]. Besides, the altered splicing pattern of transcription factor 7 like 2 (TCF7L2) was validated in ASO-SNORD116 adipocytes. The TCF7L2 gene encodes a transcription factor that plays a key role in the Wnt/β-catenin signaling pathway and has been implicated in AT biology, glucose homeostasis, and lipid metabolism [
]. It is worthwhile to further mechanistically investigate the absolute splicing changes in TCF7L2 and other molecules and explore whether these SNORD116 deficiency-mediated alterations in the splicing network have underlying functional relevance to the anti-adipogenesis phenotype in PWS adipocytes.
For the fact that currently available SNORD116-depletion mouse model poorly recapitulates the major phenotypes in human PWS [
], even showing less body fat content opposite to the human PWS, the absence of an effective animal model remains a formidable challenge in the exploration of the role of SNORD116 in AT development. One previous study knocked out the SNORD116 cluster in the SH-SY5Y neuroblastoma human cell and found reduced neuronal differentiation and cell proliferation [
]. With limited mechanistic data on SNORD116, we speculated that SNORD116 loss might exert an inhibitory effect on adipogenic and neuronal differentiation but affect proliferation in different ways depending on the cell type and origin; while further functional verification is needed to determine the role of SNORD116 in cell differentiation and proliferation under different cell-origin background relevant to PWS physiopathology. Moreover, it is worth noting that following the demonstrated pathological role of the snoRNA SNORD116 family in PWS obesity, the data also offered the insight that SNORD116 snoRNAs or other related molecules have the potential to be developed as a novel screening and diagnostic biomarker in laboratory genetic molecular testing for early PWS diagnosis during infancy.
Nevertheless, there are noteworthy limitations in our study. Given that PWS is a rare genetic disease, we only had a finite number of studied PWS subjects; and only male PWS children were studied. Although no significant sex difference in the clinical PWS phenotypes has been reported, we require a broader subject cohort consisting of both sexes to replicate our findings so as to further acquire a general consensus regarding these abnormal adiposity findings in PWS. Additionally, since dysregulated AT remodeling and impaired adipogenesis were demonstrated in phase 1 PWS children, it is crucial to understand whether and how these primary defects in the abnormal fat accumulation pathway are mechanistically associated with the progressive trajectory of obesity development in subsequent PWS phases; further studies are needed to explore the potential relevance.
Taken together, our findings constitute a new mechanistic understanding of the etiology of progressive obesity in PWS, already from pediatric age, that WAT expansion pathway imbalance and disruptions in white and beige adipogenesis programs could be primary defects in phase 1 PWS patients without obesity onset. Hopefully, on the basis of this study, novel therapeutically relevant options, such as metabolic interventions targeting disturbed adipogenesis and promoting healthy AT expansion, are warranted to be initially explored for managing PWS obesity to fill the gap between basic research and clinical care of PWS patients from a translational perspective.
The following are the supplementary data related to this article.
Differential alternatively splicing events (ASEs) of genes in PWS cases versus healthy controls.
CRediT authorship contribution statement
Yunqi Chao and Chaochun Zou: conceptualization and methodology. Lei Gao, Yangli Dai, and Mingqiang Zhu: providing human samples and contributing to clinical information. Yunqi Chao, Xiangzhi Wang, Yuqing Cai, Yingying Shu, Xinyi Zou, and Zheng Shen: experimental investigation and validation, statistical analysis, and visualization. Yunqi Chao, Yifang Qin and Chenxi Hu: literature organization and original manuscript preparation. Chaochun Zou: supervision, reviewing/editing the manuscript, and funding acquisition.
Funding
Our study was funded by the National Natural Science Foundation of China (81371215 & 81670786) and Key R&D Projects of Zhejiang Provincial Department of Science and Technology (2021 C03094).
Declaration of competing interest
The authors declare that they have no conflict of interest.
Acknowledgments
We would like to thank the core facility platform of Zhejiang University School of Medicine for experimental apparatus support. We thank BioRender.com for picture creation.
References
Tan Q.
Orsso C.E.
Deehan E.C.
Triador L.
Field C.J.
Tun H.M.
et al.
Current and emerging therapies for managing hyperphagia and obesity in Prader-Willi syndrome: a narrative review.
Loss of the imprinted, non-coding Snord116 gene cluster in the interval deleted in the Prader Willi syndrome results in murine neuronal and endocrine pancreatic developmental phenotypes.
The role of metalloproteinases and their tissue inhibitors in adipose tissue remodelling and whole-body lipid distribution: a cross-sectional clinical study.
00173-1&id=ga1.jpg){kind=link}
00173-1&id=gr1.jpg){kind=link}
00173-1&id=gr2.jpg){kind=link}
00173-1&id=gr3.jpg){kind=link}
00173-1&id=gr4.jpg){kind=link}
00173-1&id=gr5.jpg){kind=link}
00173-1&id=gr6.jpg){kind=link}