京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第18647号 / 医博第3946号 / 新制||医||1006(附属図書館) / 31561 / 京都大学大学院医学研究科医学専攻 / (主査)教授 福山 秀直, 教授 平岡 眞寛, 教授 宮本 享 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DGAM

Phosphatase and tensin homolog (PTEN) Hamartoma Tumor Syndrome (PHTS) is linked to heterozygous germline mutations in the tumor suppressor gene PTEN. While clinical features of PHTS are broad, the scientific focus of this work lay on the development of aberrant adipose tissue growth in the form of lipomas in pediatric PHTS patients. Although lipomas are generally of benign nature, obstruction of other organs due to their size can lead to life threatening complications. PTEN antagonizes the growth promoting PI3K/AKT pathway, leading to a hyper activation in PHTS patients. Therefore, inhibitors of this pathway might be considered for pharmacological therapy. Treatment attempts with the mTOR inhibitor rapamycin showed beneficial effects, but reports of adaption to the drug indicated the need for further research. In the first part of the project, published in Cancers 2019, we tested effects of the PI3Kα inhibitor alpelisib on growth, apoptosis, senescence and adipogenesis of lipoma cells from PHTS patients. Alpelisib was previously used to successfully treat patients with PI3K related overgrowth syndrome (PROS). PROS is caused by mosaic somatic activating mutations in the PI3Kα catalytic subunit and shares an increased PI3K signaling and a predisposition for lipoma development as a common feature with PHTS. We compared the effects of alpelisib on PHTS patients’ lipoma cells (LipPD1-3) and PROS patients’ lipoma cells (Lip3 and Lip4) and found a similar dose and time dependent reduction of cell viability. We also tested a combined treatment of alpelisib and rapamycin and observed a synergistic activity of the drug combination on cell viability. Proliferation was reduced in PHTS lipoma cells in a concentration dependent manner. In contrast, alpelisib did not induce cell death as measured via annexin V/PI apoptosis assay and LDH cytotoxicity assay. The reduction in cell number was found to be facilitated via induction of senescence in the lipoma cells as determined via senescence associated β-galactosidase (SA β-gal) assay and detection of senescence marker CDKN2A (p16) expression. Interestingly, alpelisib did not only inhibit cell growth but also adipogenesis of the lipoma cells. We established lipoma cell spheroids in 3D culture as lipoma models. While size of control spheroids continuously increased during 10 days in adipogenic culture medium, the size of spheroids in medium containing 10 µM alpelisib decreased, indicating reduced lipid accumulation. An important advantage of alpelisib compared to rapamycin treatment, was observed on deactivation of the downstream PI3K target AKT. While both drugs inhibited phosphorylation of mTORC1 and ribosomal protein S6, only alpelisib reduced AKT phosphorylation. The observed effects on cell viability were similar in PHTS and PROS lipoma cells. PROS patients, including 15 children and adolescents, were successfully treated with alpelisib and exhibited only mild side effects. The in vivo safety and efficacy in pediatric PROS patients studied Vernot et al. 2018, together with our results on cell viability in PROS and PHTS patients’ lipoma cells, provide hope for a beneficial effect of alpelisib in treatment of PHTS related lipoma formation. Primary cells from the stromal vascular fraction (SVF) of adipose tissue can be used as human in vitro models for adipogenesis. Disadvantages include their limited availability, donor variability, limited proliferation and especially adipogenic potential that declines during continuous culture. Therefore, new cell models with enhanced or prolonged adipogenic potential are a useful tool for adipose tissue biology research. In this respect, we established and characterized a lipoma cell strain termed LipPD1 and published the findings in Adipocyte 2020. LipPD1 cells were isolated from abdominal lipoma tissue of an 11-month old male PHTS patient with a heterozygous deletion of exons 2-9 of the PTEN gene and suffering from a severe lipomatosis. SVF cells from the lipoma tissue and control SVF cells from visceral adipose tissue of obese, but otherwise healthy donors undergoing bariatric surgery were isolated via collagenase digest and selected for plastic adherence. LipPD1 cells were compared to the control SVF cells and SGBS cells, which are one of the few available human preadipocyte cell strains established by Wabitsch et al. in 2001. To date the underlying genetic cause for the retained adipogenic potential in SGBS cells remains unclear. LipPD1 cells cultured for 40-60 days showed a high adipogenic potential comparable to SGBS cells, while control SVF cells lost their capacity for adipogenesis during this period as shown via Oil red O and Nile red lipid staining. Expression of adipogenic marker genes PPARγ, aP2, FASN, leptin and adiponectin was markedly increased after 8 days of adipogenesis compared to undifferentiated cells in LipPD1 and SGBS but was only slightly elevated in control SVF cells. Adipocyte functional properties were similar in LipPD1 and SGBS cells. Both cell strains formed spheroids in 3D culture, which increased in size during culture in adipogenic medium, reflecting lipid accumulation. A major difference between the cell strains was an increased basal and stimulated PI3K pathway activation in LipPD1 cells, reflecting their PTEN haploinsuffciency. In conclusion, LipPD1 cells are comparable to SGBS cells as a human adipocyte model, with the advantage of knowing the genetic lesion responsible for the enhanced adipogenic potential compared to wild-type SVF cells. A main goal of this project was to investigate the underlying mechanism of lipoma development in PHTS patients and the role of PTEN in adipose tissue. Therefore, PTEN was downregulated in SVF cells transiently via siRNA (PTEN KD) or stable with the CRISPR method (PTEN CR) and the observed effects were published in The Journal of Biological Chemistry in 2021. PTEN KD cells were compared to controls simultaneously transfected with scramble siRNA and PTEN CR cells were compared to cells simultaneously transfected with control guideRNA with no genomic target. An advantage of this approach compared to the previous analysis of PHTS lipoma cells was that controls and PTEN mutant/knockdown cells had the same genetic background. Using both methods we achieved downregulation of PTEN to similar extends as observed in the PHTS lipoma cells and confirmed PI3K pathway activation by detecting enhanced AKT and ribosomal protein S6 phosphorylation. The permanent downregulation of PTEN in PTEN CR cells allowed for the analysis of effects in long-term culture. Both, PTEN KD and PTEN CR cells, showed an increased proliferation rate compared to control cells, reflecting the growth promoting nature of the PI3K pathway. We downregulated PTEN transiently in SVF cells that already lost their capacity for adipogenesis and found that the adipogenic potential was restored partially as shown by Nile red staining and observations of an increased spheroid size in 3D culture. Cultures of PTEN CR and control cells also revealed marked differences in adipogenic potential after 2-6 weeks in culture. Following the notion that PTEN downregulation restored the adipogenic potential lost in long-term culture, we investigated if PTEN protein levels change during continuous culture of PTEN wild-type SVF cells. We found an increase of PTEN levels and downregulation of AKT phosphorylation in high-passage SVF cells. In addition, the senescence marker p21, p16, p15 and HIPK2 were downregulated at the mRNA level. After long-term culture, there was less senescence in PTEN CR cells compared to controls as determined via SA β-gal staining. We performed RNA sequencing of PTEN KD versus control SVF cells as an untargeted approach and identified 1379 genes regulated in conditions of PTEN downregulation. Gene set enrichment analysis identified 18 significantly enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways with Cellular Senescence most significantly enriched. To find genes responsible for the enhanced adipogenesis observed in PTEN KD cells, we compared our gene set with results from two other RNA sequencing studies of lipid accumulation models. We found 36 overlapping genes and chose the downregulated FOXO1 and RNF144B for further analysis. To check whether the effects of PTEN downregulation on adipogenesis could be attenuated by reintroducing these factors into PTEN CR cells, we overexpressed RNF144B and constitutively active FOXO1 in these cells. While RNF144B had no influence, neither on proliferation nor on adipogenesis of the PTEN CR cells, FOXO1 overexpression reduced adipogenesis in the PTEN CR cells. FOXO1 phosphorylation, which is known to induce adipogenesis, was induced in PTEN KD cells. The lipogenesis inducing factor SREBP1, which is transcriptionally repressed by unphosphorylated FOXO1, was upregulated. In contrast, SREBP1 protein was reduced in the FOXO1 overexpressing PTEN CR cells, explaining the reduced lipid accumulation. These results indicate that the observed adipose tissue overgrowth in PHTS patients is caused by an induction of adipose progenitor growth and adipogenesis, mediated at least partially through repression of FOXO1 transcriptional activity. In summary, these findings provide evidence for a role of PTEN in regulation of adipose tissue expansion and adipogenesis, which could account for the observed lipoma formation in PHTS patients.:Table of Contents 2 Abbreviations 4 Introduction 6 PTEN Hamartoma Tumor Syndrome (PHTS) 6 Phosphoinositide 3-kinase pathway 6 Current PHTS therapy 9 Alpelisib 9 Replicative aging and senescence 10 Adipose tissue and SVF cells 11 Adipose tissue redistribution and PTEN 13 Small interfering RNA (siRNA) and clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 14 Assays to determine cell viability and cell death 16 Rationale 17 Publications 19 1) The Novel Phosphatidylinositol-3-Kinase (PI3K) Inhibitor Alpelisib Effectively Inhibits Growth of PTEN-Haploinsufficient Lipoma Cells 19 Supplementary figures 36 Supplementary tables 44 2) A new human adipocyte model with PTEN haploinsufficiency 46 Supplementary methods 58 Supplementary figures 61 3) PTEN regulates adipose progenitor cell growth, differentiation, and replicative aging 65 Supplementary figures 78 Supplementary tables 86 Summary 90 Publication bibliography 97 Appendix 103 Darstellung des eigenen Beitrags 103 The Novel Phosphatidylinositol-3-Kinase (PI3K) Inhibitor Alpelisib Effectively Inhibits Growth of PTEN-Haploinsufficient Lipoma Cells. 103 A new human adipocyte model with PTEN haploinsufficiency. 104 PTEN regulates adipose progenitor cell growth, differentiation, and replicative aging 105 Erklärung über die eigenständige Abfassung der Arbeit 106 Lebenslauf 107 Bildungsweg 107 Beruflicher und wissenschaftlicher Werdegang 107 Preise und Förderungen 108 Publikationen 109 Konferenzbeiträge 110 Danksagung 112

Phospholipid scramblase 4 (PLSCR4) is a member of a conserved enzyme family with high relevance for the remodeling of phospholipid distribution in the plasma membrane and the regulation of cellular signaling. While PLSCR1 and -3 are involved in the regulation of adipose-tissue expansion, the role of PLSCR4 is so far unknown. PLSCR4 is significantly downregulated in an adipose-progenitor-cell model of deficiency for phosphatase and tensin homolog (PTEN). PTEN acts as a tumor suppressor and antagonist of the growth and survival signaling phosphoinositide 3-kinase (PI3K)/AKT cascade by dephosphorylating phosphatidylinositol-3,4,5-trisphosphate (PIP3). Patients with PTEN germline deletion frequently develop lipomas. The underlying mechanism for this aberrant adipose-tissue growth is incompletely understood. PLSCR4 is most highly expressed in human adipose tissue, compared with other phospholipid scramblases, suggesting a specific role of PLSCR4 in adipose-tissue biology. In cell and mouse models of lipid accumulation, we found PLSCR4 to be downregulated. We observed increased adipogenesis in PLSCR4-knockdown adipose progenitor cells, while PLSCR4 overexpression attenuated lipid accumulation. PLSCR4 knockdown was associated with increased PIP3 levels and the activation of AKT. Our results indicated that PLSCR4 is a regulator of PI3K/AKT signaling and adipogenesis and may play a role in PTENassociated adipose-tissue overgrowth and lipoma formation.

Germline mutations in the tumor suppressor gene PTEN cause PTEN Hamartoma Tumor Syndrome (PHTS). Pediatric patients with PHTS frequently develop lipomas. Treatment attempts with the mTORC1 inhibitor rapamycin were unable to reverse lipoma growth. Recently, lipomas associated with PIK3CA-related overgrowth syndrome were successfully treated with the novel PI3K inhibitor alpelisib. Here, we tested whether alpelisib has growth-restrictive effects and induces cell death in lipoma cells. We used PTEN-haploinsufficient lipoma cells from three patients and treated them with alpelisib alone or in combination with rapamycin. We tested the effect of alpelisib on viability, proliferation, cell death, induction of senescence, adipocyte differentiation, and signaling at 1–100 M alpelisib. Alpelisib alone or in combination with rapamycin reduced proliferation in a concentrationand time-dependent manner. No cell death but an induction of senescence was detected after alpelisib incubation for 72 h. Alpelisib treatment led to a reduced phosphorylation of AKT, mTOR, and ribosomal protein S6. Rapamycin treatment alone led to increased AKT phosphorylation. This effect could be reversed by combining rapamycin with alpelisib. Alpelisib reduced the size of lipoma spheroids by attenuating adipocyte differentiation. Since alpelisib was well tolerated in first clinical trials, this drug alone or in combination with rapamycin is a potential new treatment option for PHTS-related adipose tissue overgrowth.