UC Berkeley

Abstract

The Study of Adipose Tissue Differentiation and Development

by

Olga Gulyaeva

Doctor of Philosophy in Endocrinology

University of California, Berkeley

Professor Hei Sook Sul, Chair

Two functionally different adipose tissues exist in mammals: white (WAT) and brown

(BAT). WAT is the primary energy storage site and a critical organ for regulation of

energy metabolism through secretion of various adipokines. In contrast, BAT is a key

thermogenic organ, capable of burning chemical substrates for heat generation through

unique protein UCP1. Interestingly, UCP1 is only expressed in brown or beige

adipocytes found in WAT, however all the transcriptional activators of UCP1 promoter

identified by far are not tissue-specific and, therefore, cannot solely explain BAT-specific

expression of UCP1. BAT is believed to form early in embryogenesis from muscle-like

progenitors, while WAT depending on the anatomical locations, develops in late

embryogenesis or postnatally from non-muscle–like progenitors. Although embryogenic

BAT precursors have been well characterized, the molecular identity of WAT precursors

remains controversial. The aim of this dissertation work was to identify novel tissuespecific

transcriptional regulator of UCP1 in BAT as well as to identify and characterize

WAT precursors at various stages of adipogenic commitment and differentiation.

Chapter 1 reviews developmental origin of adipose tissue. Most of subcutaneous WAT

depots develop perinatally, while visceral WAT develops postnatally from adipose

precursors expressing markers such as Myf5, Pref-1, Wt1, and Prx1 depending on the

anatomical location. BAT develops in early embryogenesis form Myf5+, Pax7+ musclelike

precursors. Brown fat-like cells found in WAT upon cold exposure or b3-adrenergic

stimulation may arise from transdifferentiation of white adipocytes or by de-novo

differentiation of yet to be characterized precursors. Finally, transcriptional and

epigenetic control of brown and beige adipocyte differentiation is discussed.

Chapter 2 demonstrates my efforts to characterize the hierarchy of adipose progenitors

in WAT in vivo. By using Pref-1 promoter-mediated conditional ablation of Sox9, one of

Pref-1 targets in inhibiting adipogenesis, coupled with fluorescent labeling, I identified

and characterized a population of early adipose progenitors that express Pref-1, Sox9

and CD24. This early WAT precursor population is highly proliferative and does not yet

express adipogenic markers and rather resembles stem cell-like population. Sox9

appeared to be critical for maintaining these early precursors in highly proliferative and

undifferentiated state. Loss of Sox9 in Pref-1+ cells lead to a depletion of this precursor

population and an increase in more committed PDGFRa+ preadipocytes that do not

proliferate and express early adipogenic markers. This switch in precursor populations

from Pref-1+ to PDGFRa+ in the absence of Sox9 leads to enhanced adipogenesis and

increased adiposity impairing insulin sensitivity in vivo.

Chapter 3 describes a role of novel BAT-enriched transcription factor, Zfp516 in

regulating UCP1 transcription and BAT differentiation and development. I found that

Zfp516 is induced upon cold exposure and directly binds and activates UCP1 and

PGC1a promoters. I also showed that knockdown of Zfp516 in cultured cells impairs

BAT-gene program and induces muscle-like gene signature. Thus, total body Zfp516

KO mice exhibit severe defect in BAT development, while transgenic mice

overexpressing Zfp516 in all adipose tissues show increased browning of WAT. This

leads to improved cold tolerance and reduces body weight accumulation upon high fat

diet improving metabolic parameters.

Chapter 4 focuses on my efforts to identify Zfp516 interacting partners for activation

BAT-gene program resulting in discovery of lysine-specific demethylase 1, Lsd1, as a

direct binding partner of Zfp516. Lsd1 through interaction with Zfp516 is recruited to

UCP1 and PGC1a promoters to demethylate H3K9 and therefore promoter

transcription. Lsd1 is induced during BAT differentiation and its ablation using Myf5-Cre

results in impaired embryonic BAT development. UCP1-Cre mediated ablation of Lsd1

impairs BAT gene program and results in formation of BAT that resembles WAT,

therefore reducing thermogenic capacity and causing obesity in mice. Lastly, Zfp516-

Lsd1 interaction is required for Lsd1 function for BAT gene program in vitro.

Supplement summarizes my effort to identify a novel gene critical for BAT function and

characterizes one such gene, C11orf54. I found that C11orf54 is highly expressed in

murine BAT, kidney and liver and is highly induced during BAT cell line differentiation.

C11orf54 is localized to the nucleus and cytoplasm and appears to be important for BAT

differentiation as CRISPR-mediated KO lines demonstrate dramatically impaired

adipogenic differentiation. Overexpression of C11 in Cos7 cells results in increased

formation of TAG through yet-to be identified mechanism.

Finally, chapter 5 summarizes my work describing importance of novel genes identified

for BAT and WAT development, future directions and remaining questions.

Thus, these studies provide a better understanding of how BAT and WAT development

and differentiation are regulated in vivo, as well as explain the hierarchy of adipose

progenitors with various degree of commitment and differentiation in WAT.