Theses Doctoral

Investigations of the neuro-molecular physiology of obesity using hypothalamic neurons derived from human pluripotent stem cells

Wang, Liheng

The hypothalamus is the central regulator of systemic energy homeostasis, and its dysfunction can result in extreme body weight alterations. This small (3-4 mm in thickness in human) neuro-endocrine brain region, located just above the median eminence, is comprised of cell types that subserve specific metabolic and behavioral aspects of the control of body weight, as well as hepatic glucose production, body temperature, autonomic physiology, neuroendocrine axes, serum osmolarity and circadian rhythms. Insights into the complex cellular physiology of this region are critical to the understanding of obesity pathogenesis and its prevention and treatment; however, human hypothalamic cells are largely inaccessible for direct study. My thesis research focused on establishing an in vitro model for understanding the molecular neurophysiology of obesity using, as "proof-of-principle", neurons derived from human pluripotent stem cells (hPSCs) derived from individuals with monogenic forms of obesity. Three related projects are described in details:
I. Differentiation of hypothalamic-like neurons from human pluripotent stem cells (Chapter 2)
This project was designed to establish an in vitro model for studying hypothalamic cell-molecular physiology in neurons derived from hPSCs. After screening several morphogens and other molecules affecting neuronal differentiation, we developed a protocol that combined early activation of sonic hedgehog signaling followed by timed NOTCH inhibition resulting in the generation of hypothalamic arcuate nucleus (ARC)-like neurons. Neuronal cells expressing pro-opiomelanocortin (POMC), neuropeptide-Y/agouti-related protein (NPY/AgRP) were generated from human embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) obtained from patients with monogenic forms of obesity. These hypothalamic-like neurons accounted for over 90% of differentiated cells and exhibited transcriptional profiles characteristic of specific hypothalamic neurons (and explicitly lacking pituitary markers). Importantly, these cells displayed hypothalamic neuronal characteristics, including production and secretion of neuropeptides and responsiveness to metabolic hormones such as insulin and leptin. Nkx2.1 progenitor cells at 12 days of differentiation from iPSC integrated into the hypothalamus following injection into the lateral ventricle of NSG mice. Single cell transcriptome analysis of day 27 hESC-derived hypothalamic neurons enabled us to identify specific hypothalamic cell types (e.g. POMC, NPY, MC4R) based on transcript signatures. These findings, in the aggregate, supported the utility of these cells for elucidation of aspects of the cellular/molecular neurophysiology of body weight regulation.
II. Using stem cell-derived hypothalamic neurons to investigate the neurophysiology of obesity caused by prohormone convertase 1/3 deficiency (Chapter 3).
My second project investigated the use the hPSC-differentiated hypothalamic neurons to assess the cellular physiology of hESC-derived hypothalamic neurons with induced knockdown or mutations of proprotein convertase subtilisin/kexin type 1 (PCSK1, encodes prohormone covertase 1/3 (PC1/3)). Congenital hypomorphism for this gene causes a rare autosomal disorder that impairs the processing of specific proproteins to their more bioactive derivatives, affecting, for example, the processing of POMC, proinsulin and proglucagon. The consequences of inactivating mutations of PCSK1 include obesity, possibly due to impaired function of anorexigenic POMC arcuate neurons. To understand the molecular neurophysiology of the obesity in PC1/3-deficient subjects, we generated PCSK1 deficient hESC lines with CRISPR or by knocking down PCSK1 with shRNA, and assessed the POMC processing in the hypothalamic ARC-like neurons made from these lines. The ratios of adrenocorticotropic hormone (ACTH)/POMC, αMSH/POMC and β endorphin (BEP)/POMC proteins were significantly decreased, while total quantities of POMC peptides were greatly increased in PCSK1-deficient hESC-derived neurons, indicating impaired POMC processing caused by reduced PC1/3 protein. These results are consistent with the elevated plasma POMC and ACTH intermediates levels of in humans segregating for hypomorphic mutations of PCSK1, and the impaired pituitary POMC processing in the PC1/3 mutant mice. Interestingly, in day 28 PC1/3-deficient neurons, in addition to upregulation of POMC gene expression and protein, we found increases in some of the "downstream" proteolytic enzymes for POMC processing and the "upstream" transcription factor that regulates PCSK1 expression. The molecular mechanisms underlying the invocation of these possibly compensatory processes are under study.
These findings provide confidence that the hypothalamic neurons generated by the techniques described in Chapter 2 display molecular phenotypes consistent with a mutation in one of the important neuropeptide processing pathways.
III. Using iPSC-derived neurons to investigate the molecular pathogenesis of obesity in Bardet-Biedl Syndrome (Chapter 4).
To further investigate the use of iPSC-derived neurons in the study of the neurobiology of obesity, I analyzed structural and molecular physiologic phenotypes cells derived from patients with Bardet-Biedl Syndrome (BBS). BBS is a rare autosomal recessive disease characterized by multiorgan dysfunction, including polydactyly, hyperphagic obesity, retinal degeneration, renal cysts and cognitive impairments. Eighteen discrete genes have been implicated in specific instances of BBS, and all cognate proteins that have been identified encode constituents of the basal body of the primary cilium. The primary cilium has also been implicated in other clinical obesities, including the Alstrom syndrome, and the effects of a highly prevalent FTO allele on body weight. We found that ciliogenesis and neurite outgrowth were affected in both BBS1 and BBS10 mutant iPSC-derived neurons as reflected by longer primary cilia, shorter neurite length, and fewer processes. Furthermore, insulin-induced AKT phosphorylation at Thr308 was greatly reduced in both BBS1 and BBS10 mutant neurons compared to controls. Overexpression of BBS10 fully restored insulin signaling in BBS10 mutant neurons by rescue of the tyrosine phosphorylation of insulin receptor. Co-immunoprecipitation assays indicated that both BBS1 and BBS10 interacted physically with the insulin receptor. Leptin signaling was also investigated in BBS mutant fibroblasts and neurons. Both BBS mutations impaired leptin-mediated pSTAT3 activation in both cell lines by affecting either the trafficking or the quantities of leptin receptor in these cells. These data demonstrate that BBS proteins are essential for insulin and leptin signaling in neurons and fibroblasts, in a cellular context independent of the effects of obesity.
These studies further confirm the ability of iPSC-derived neurons to reflect aspects of the molecular pathophysiology of the patients from whom they are obtained, and to enable studies of these phenotypes in circumstances isolated from the secondary effects of adiposity per se.


  • thumnail for Wang_columbia_0054D_12627.pdf Wang_columbia_0054D_12627.pdf binary/octet-stream 20.4 MB Download File

More About This Work

Academic Units
Nutritional and Metabolic Biology
Thesis Advisors
Leibel, Rudolph L.
Ph.D., Columbia University
Published Here
April 28, 2015