UNDERSTANDING THE MOLECULAR MECHANISMS OF FAT STORAGE

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: HATCH
• Reporting Frequency: Annual
• Accession No.: 1004639
• Project Start Date: Feb 16, 2015
• Project End Date: Sep 30, 2019
• Program Code: [(N/A)]- (N/A)

Recipient Organization
UNIVERSITY OF CALIFORNIA, BERKELEY
(N/A)
BERKELEY,CA 94720

Performing Department
Nutritional Sciences

Non Technical Summary
An in depth understanding of human nutrition is important to human health and to food policy. Determining how we store nutrients and what nutrients we need impacts government food policy, food production, and food distribution. Our research aims to understand how we store nutrients, like fats and cholesterols. Interestingly, we now know that fats and cholesterols are stored in a subcellular storage compartment called a lipid droplet. Altered storage of fats and cholesterols in lipid droplets can have devastaing consequences, and are integral to the current metabolic disease epidemic that plagues our country. For many years, researchers thought that lipid droplets were inert compartments. We now know that they are in fact incredibly dynamic compartments that are important in human disease and we know very little about their biology.Our proposed research aims define the molecular basis of how lipid droplets are formed and how they are regulated. Our research program develops and applies cutting edge cell biological methods to understand lipid droplet biology. This understanding will broadly impact medicine and food practice / policy in the United States. The end goals are to: 1) decrease the prevalence of metabolic diseases and 2) ≈to increase the efficiency of food production and distribution for optimal society health by understanding human nutrition.

Animal Health Component

0%

Research Effort Categories

Basic

100%

Applied

0%

Developmental

0%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
702	7010	1010	100%

Knowledge Area
702 - Requirements and Function of Nutrients and Other Food Components;

Subject Of Investigation
7010 - Biological Cell Systems;

Field Of Science
1010 - Nutrition and metabolism;

Keywords

food science

nutrition

diet

fat storage

lipid droplet

Goals / Objectives
Our goal is to employ systems-level and mechanistc cell biological approaches to define the molecular mechanisms underlying lipid droplet biogenesis and function. Achieving this important goal will broadly impact our understanding of cellular fat storage, human nutrition, human diet, and agricultural needs. The primary objectives of the proposed research are to:Goal 1: Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome ­- The functional role of lipid droplets differs drastically depending on the nutrient status, necessitating unique proteome compositions and activities. Ubiquitin is a posttranslational modification that can target proteins for proteasomal degradation and regulates virtually every cellular process. However, few connections between ubiquitin and lipid droplets have been reported. We recently discovered a metabolically regulated lipid droplet ubiquitinaton complex that controls lipid droplet turnover by inhibiting the rate-limiting enzyme in lipolysis, adipose triglyceride lipase (Olzmann et al., Proc. Natl. Acad. Sci. 2013). Employing these findings as a foundation, we are using global proteomic and functional genomic strategies to characterize the ubiquitination pathways and targets that regulate lipid droplet functions.1a. Identify and characterized ubiquitinated lipid droplet proteins1b. Define the components and pathways that mediate lipid droplet protein degradation1c. Elucidate the role of lipid droplet ubiquitination components in the regulation of metabolic pathwaysGoal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites - For years scientists have focused on the advantages instilled by the separation of cellular environments. However, to function as a holistic unit, it is essential that organelles communicate and coordinate activities. Membrane contact sites are important regions of organelle-organelle communication, but have been resistant to study with traditional approaches and for most membrane contact sites we lack a molecular understanding of their structure, constituents, and functions. Lipid droplets make contacts with mitochondria and lipolysis-liberated fatty acids represent important substrates for energy production through b-oxidation in mitochondria. To define the molecular basis of lipid droplet-mitochondrial contact sites we are developing chemical biology labeling methods and synthetic biology-based functional screens. These technologies will have immediate implications for our understanding of the mechanisms underlying cellular energy homeostasis and will provide new methods that can be widely applied to characterize other cellular membrane contact sites.2a. Determine the functional role of lipid droplet-mitochondrial contacts2b. Identify the molecular composition of lipid droplet-mitochondrial contacts

Project Methods
Our research will employ proteomics, functional genomics, metabolomics and traditional cell biological approaches to achieve an in depth understanding of lipid droplet function. Our proposed research involves the development of several novel methods, including the development of new chemical biology methods to label lipid droplet proteomes, of inducible organelle tethers to study organelle-organelle contacts, of global ubiquitin proteomic methods, and of reporter cell lines to define lipid droplet protein degradation pathways.Efforts- To influence our target audience, our research results will be presented at local and international conferences and will be published in peer-reviewed journals.Evaluation- Research will be evaluated:- Short term - At weekly data meetings and large group meetings- Long term - Through annual review of research projects objectives, of progress towards those objectives, and the publication of our results.Review will also be conducted annually through discussions with the chair of the department of Nutritional Sciences & Toxicology, Dr. Joseph Napoli. Dr. Napoli is a leading expert in metabolic biology and his many years of experience will provide an important outside review of our efforts to successfully achieve our stated objectives.Thus we will employ both internal reviews or our progress and well as an outside review of our progress.Progress towards our goals will primarily be based on the presentation of our results at scientific meetings and publication of our results in peer-reviewed journals. In addition, press releases and news that report the significance our findings, thereby enables the education of a larger audience, will also be considered an important metric of progress and success.

Progress 02/16/15 to 09/30/19

Outputs
Target Audience:The target audience reached by our efforts during the reporting period was the scientific community. Our research and its significance were communicated to the scientific community through a combination of presentations and publications. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Postdoctoral scholars, graduate students, and undergraduates benefited from one-on-one mentorship in experimental technique and scientific thought. We have weekly subgroup and large group meetings to discuss recentfindings. We also participate in several working groups on campus (e.g. Quality control supergroup) to present and receive feedback. In addition, several members of the lab attended and presented at conferences or symposia (listed below in section regarding dissemination of results). How have the results been disseminated to communities of interest?In addition to our publications (listed in products), members of the lab have been active in disseminating the results of our studies to communities of interest through presentations: James Olzmann (PI) Speaker, UCSF Cancer and Metabolism Symposium. 2019. San Francisco, CA. Speaker, International Congress on Obesity and Metabolic Syndrome (ICOMES). 2019. Seoul, Korea. Speaker, KERN Lipid Conference. 2019. Vail, CO. Speaker, Center for Cellular Construction (NSF Science and Technology Center), San Francisco, CA. Speaker, University of California, San Francisco ­- Life and Death, Symposium. 2019. San Francisco, CA. Seminar, California State University, Northridge - Department of Biology. 2019. Northridge, CA. Seminar, Wayne State University - Lipid Metabolism Seminar Series. 2019. Detroit, MI. Seminar, University of California, Davis - Dept. Molecular and Cell Biology and the Graduate Student Group Equity in Science, Technology, Engineering, Math, and Entrepreneurship (ESTEME). 2019. Davis, CA. Seminar, University of California, San Francisco - Hellen Diller Cancer Center. 2019. San Francisco, CA. Seminar, San Francisco State University - Molecular, Cell, and Microbiology seminar series. 2018. San Francisco, CA. Speaker, Annual Biomedical Research Conference for Minority Students (ABRCMS). 2018. Indianapolis, IN. Seminar, University of Zurich - Institute of Molecular Life Sciences. 2018. Zurich, Switzerland. Student and postdoc presentations: Poster presentation, UCSF Diabetes Center Retreat. 2019. Santa Cruz, CA. Poster presentation, SACNAS Annual Conference. 2019. Honolulu, HI. Oral presentation, UCSF Diabetes Center Retreat. 2019. Santa Cruz, CA. Oral presentation, SACNAS Annual Conference. 2019. Honolulu, HI. Poster presentation, ASCB Annual Conference. 2019. Washington DC. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? The biology of lipid homeostasis is a challenging frontier in science and medicine, and many outstanding questions remain. The importance of this field is underscored by the dysregulation of lipid metabolism in numerous diseases, including prevalent metabolic diseases and cancer. My research program exploits interdisciplinary strategies that integrate systems-level discovery "omics" methods, chemical biology tools, and cell biology approaches to advance our understanding of the mechanisms that regulate lipid homeostasis. Goal 1: Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome The composition of an organelle proteome determines its functions. Attempts to define the lipid droplet (LD)proteome have been limited by non-specific protein contamination from co-purifying organelles. To overcome this obstacle, we developed a chemical biology strategy to label the LD proteome in living cells (Dev Cell 2018, Meth Mol Biol 2019). This approach takes advantage of an engineered ascorbate peroxidase (APEX2) that facilitates the addition of biotin-phenol to proteins in close proximity. Employing this method, we identified high confidence LD proteomes from multiple human cell lines, providing an important resource for the field. We launched an online portal to facilitate access to the data (www.dropletproteome.org). Furthermore, the temporal resolution provided by the APEX2 proximity labeling method allows us to capture a "snapshot" of the LD proteome, and we can compare a series of snapshots over time to explore LD proteome dynamics. Coupling our APEX2 proximity labeling approaches with quantitative proteomics, we discovered aLD protein that istargeted for ubiquitin-dependent proteasomal degradation via ERAD. This protein, c18orf32, inserts into the ER and traffics to LD (Class I LD protein). Our results indicate that c18orf32 is degraded by a canoncial ERAD pathway employing gp78, derlin-1, and VCP. These findings are impactful because they demonstrate the utility of proximity labeling to reveal LD proteome dynamics and identify a basic mechanism of LD proteome regulation in which LDs coopt ERAD for protein degradation. More recently we have developed genetic strategies to dissect the ubiquitination pathways required for the degradation of LD proteins that insert directly into the LD surface (Class II LD proteins). We first generated a fluorescence-based reporter cell line in which the key LD protein PLIN2 was tagged at its endogenous locus with green fluorescent protein (GFP). This allows us to easily quantify PLIN2 levelsby flow cytometry. We confirmed that PLIN2-GFP localizes properly to LDs and is degraded by a ubiquitin-dependent proteasomal pathway. To identify the ubiqutination components involved in PLIN2 clearance we performed a genome-wide CRISPR-Cas9 screen to identify genes that when silenced result in defects in PLIN2 degradation. Our approach has identified several candidate factors, including ERAD E3 ligases (MARCH6, TRC8) and a cytosolic E3 ligase (UBR4). These data suggest that Class II LD proteins are delivery to the ER for degradation via compensatory ERAD pathways. Our genetic screen also identified numerous candidates that potential regulate cellular lipid metabolism. We have also sought to characterize the functions and regulations of LD proteins. Our recent findings identify a surprising role for an LD protein in a lipotoxic form of cell now known as ferroptosis.Ferroptosis is exquisitely sensitive to the cellular lipid landscape since polyunsaturated fatty acids (PUFAs) in phospholipids are highly sensitive to oxidative damage. Ferroptosis can be prevented by the glutathione peroxidase GPX4, which converts lipid peroxides into non-toxic lipid alcohols, and inhibition of GPX4 is being explored as a cancer therapeutic strategy. However, not all cancer cells are sensitive to GPX4 inhibition, suggesting that additional unknown factors govern ferroptosis. To identify these putative factors, we performed a synthetic lethal CRISPR-Cas9 screen in cancer cells that are resistant to GPX4 inhibition and discovered FSP1 (a type II LD protein) as a powerful ferroptosis suppressor (Nature 2019). Overexpression of FSP1 in sensitive lung cancer lines inhibits ferroptosis and knockout of FSP1 sensitizes lung cancer lines to ferroptosis in culture and in tumor xenografts. Mechanistically, we find that FSP1 acts as an NADPH/NADH oxidoreductase at the plasma membrane to reduce CoQ, generating a potent lipid antioxidant that suppresses the propagation of lipid peroxides. This mechanism is remarkable because although CoQ has been observed in non-mitochondrial membranes for ~60-years, its function in these membranes was a mystery. Our findings uncover a function for non-mitochondrial CoQ in protecting membranes, establish a new pathway that prevents ferroptosis parallel to the canonical pathway, and identify FSP1 as a target for chemotherapeutics. Goal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites Membrane contact sites are regions in which two organelles are held in close proximity, enabling the exchange of lipids and ions. These sites also function as organizing platforms for metabolic processes and organelle remodeling (NRCMB 2019, JCB 2019). We and observed that LDs form membrane contacts with mitochondria (Dev Cell 2017), possibly to facilitate the efficient and safe transfer of released fatty acids for breakdown by b-oxidation. Membrane contact sites are maintained through the actions of tethering proteins, which bridge the gap between the two organelles, binding lipids in the adjacent organelle or binding a protein to form a trans-organelle tethering complex. The molecular composition of LD-mitochondrial sites remains mostly unknown. We initially attempted to identify LD-mitochondrial tethering complexes using APEX2 proximity labeling protoemics.In this approach, LD-targeted APEX2induced the biotinylation ofneighboring proteins. This will primary label LD proteins as well as proteins in close proximity, such as at membrane contact sites. We isolated mitochondria and the biotinylated proteins were identified by mass spectrometry. While this identified candidate tethering proteins, they were not validated in subsequent follow up experiments. One possibility moving forward is to use split-APEX2 strategyor to tether APEX2 directly to potential tethering proteins. We are also working towards a fluorescence-based reporter system that would be conducive to genome-wide genetic approaches. In this strategy, we are generating cell lines expressing Förster resonance energy transfer (FRET) pairs to mark contacts. These plasmids have been cloned an initial characterization of proper localization performed. These cell lines express a CFP-tagged LD protein and a YFP-tagged outer mitochondrial protein. Theproteins arealso tagged with FKBP-FRB to enable small molecule inducible organelle tethering. These cell lines will allow us to study the dynamics of LD-mitochondrial contacts under different metabolic conditions (e.g. nutrient deprivation) and will provide a system amenable to genetic screens to identify tethering complexes.

Publications

• Type: Journal Articles Status: Published Year Published: 2019 Citation: Bersuker, K., Hendricks, J., Li, Z., Magtanong, L., Ford, B., Tang, P.H., Roberts, M.A., Tong, B., Maimone, T.J., Zoncu, R., Nomura, D.K., Bassik, M.C., Dixon, S.J., Olzmann, J.A.# (2019) The CoQ oxidoreductase FSP1 acts parallel to GPX4 to inhibit ferroptosis. Nature. 575(7784):688-692. PMID: 31634900.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Bersuker, K. and Olzmann, J.A.# (2019) Identification of lipid droplet proteomes by proximity labeling proteomics using APEX2. Methods Mol. Biol. - Proximity Labeling. 57-72. PMID: 31124088.

Progress 10/01/17 to 09/30/18

Outputs
Target Audience:The target audience reached by our efforts during the reporting period was the scientific community. Our research and its significance were communicated to the scientific community through a combination of presentations and publications. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Postdoctoral scholars, graduate students, and undergraduates benefited from one-on-one mentorship in experimental technique and scientific thought. We have weekly subgroup and large group meetings to discuss recentfindings. We also participate in several working groups on campus (e.g. Membrane supergroup) to present and receive feedback. In addition, several members of the lab attended and presented at conferences or symposia (listed below in section regarding dissemination of results). How have the results been disseminated to communities of interest?James Olzmann (PI) Speaker, EMBO The Endoplasmic Reticulum function in health and disease. 2018. Lucca, Italy. Seminar, Johns Hopkins - Department of Cell Biology. 2018. Baltimore, MD. Seminar, UC Berkeley - Department of Nutritional Sciences & Toxicology. Metabolic Biology Seminar Series. 2018. Berkeley, CA. Speaker, FASEB Summer Conference Series: Lipid Droplets: Metabolic Consequences of the Storage of Neutral Lipids. 2018. Steamboat Springs, CO. Seminar, Tsinghua University - Department of Cell Biology. 2018. Beijing, China. Seminar, National Institute of Health - Protein trafficking interest group. 2018. Bethesda, MD. 36. Speaker, Stanford University - Center for Cell Biology, Think & Drink Series. 2018. Stanford, CA. Speaker, American Society for Cell Biology (ASCB) meeting. 2017. Philadelphia, PA. Kirill Bersuker Speaker, UC Berkeley Quality Control Supergroup. 2018. Berkeley, CA. Poster presenter, FASEB Summer Conference Series: Lipid Droplets: Metabolic Consequences of the Storage of Neutral Lipids. 2018. Steamboat Springs, CO. Melissa Roberts Poster presenter, FASEB Summer Conference Series: Lipid Droplets: Metabolic Consequences of the Storage of Neutral Lipids. 2018. Steamboat Springs, CO. What do you plan to do during the next reporting period to accomplish the goals?Goal 1: Elucidate mechanisms of lipid droplet proteome regulation using functional genomics. Non-alcoholic fatty liver disease (NAFLD) is a prevalent liver disease that increases the risk for cirrhosis and hepatocellular cancer. In NAFLD, lipids accumulate in large hepatic lipid droplets (LDs), which are neutral lipid storage organelles that have emerged as key hubs of cellular lipid metabolism and as central regulators of metabolic diseases. For years, LDs were largely ignored by scientists, and fundamental questions regarding the mechanisms of LD regulation remain unresolved. Elucidating the mechanisms and pathways that control LDs biology in hepatocytes is paramount to understanding NAFLD and to developing new therapeutic strategies. Genetic approaches have been employed to understand LDs in yeast and drosophila, but no genetic screens have been performed in human cells, including hepatocytes. We propose to exploit powerful whole genome CRISPR screens to comprehensively define the mechanisms of LD lipolytic degradation in human cells. These findings have the potential to be transformative by identifying novel factors that can be therapeutically targeted for fatty liver disease. To define the mechanisms of LD degradation in human cells we will employ a whole genome CRISPR/Cas9 library and a new fluorescent reporter cell line of LD abundance.High priority candidates from our genetic screen will be selected and validated using fluorescence microscopy assays. Validated candidates with therapeutic potential will be further characterized using a battery of cell biology and biochemistry assays to reveal the underlying mechanism of regulation. Goal 2: Understanding the cell biologyof lipid metabolism and lipotoxicity. Our findings have revealed a profound connection between LDs and the prevention of lipotoxic damage. For example, in our recent publication in Developmental Cell we determined that loss of LD biogenesis during autophagy can cause fatty acid flux into acylcarnitines, which subsequently damage mitochondria and abrogate cellular respiration. We hypothesize that many additional pathways function to protect lipotoxic damage. A key lipid that has been widely implicated in lipotoxicity is ceramide. Interestingly, ceramide can be convertedinto acylceramide and sequestered in LDs to prevent lipotoxicity. However, the mechanisms by which ceramide disrupt cellular functions and leads to cell death remain incompletely understood.We propose to utilize whole genome CRISPR screens to comprehensively dissect the mechanisms by which ceramide damages cells and causes apoptosis. Immortalised myelogenous leukemia K562 cells expressing Cas9 will be infected with a whole genome sgRNA library, subjected to multiple rounds of ceramide treatment, and the final cellular populations examined by deep sequencing. Key candidates that are enriched (sensitizing factors) and disenriched (resistance factors) will be selected, validated and characterized. These findings have the potential to impact our understanding of lipotoxicity in metabolic diseases and potentially in novel chemotherapeutic strategies to treat cancer.

Impacts
What was accomplished under these goals? We have made significant progress towards our stated goals over the past two years. A major portion of this progress has been published. A proximity labeling strategy provides insights into the composition and dynamics of lipid droplet proteomes(Published in: Bersuker et al.,Dev Cell. 2018 44, 97-112. PMID: 29275994) The composition of an organelle proteome determines its cellular functions. Previous studies to characterize the composition of LD proteomes relied on proteomic analysis of LD-enriched buoyant fractions and were limited by non-specific protein contamination from co-purifying organelles (i.e. false positives). The inability to accurately define LD proteomes has been a major obstacle impeding our understanding LD functions. During the last funding period we developed a new chemical biology strategy to identify the LD proteome. This research was recently published inDevelopmental Cell. In this study, we took advantage of an engineered ascorbate peroxidase (APEX2), a genetically encoded enzyme that facilitates the addition of biotin-phenol to proteins in close proximity in living cells. We adapted this method to label the LD proteome by fusing APEX2 to known LD proteins, effectively targeting APEX2 to the LD surface. Proteomic analyses of affinity purified biotinylated proteins identified high confidence LD proteomes from two different human cell lines (U2OS and Huh7) that consisted of ~150 proteins.Our new technology provides the most advanced and specific method to define the LD proteome.The high confidence LD proteomes we identified provide a resource for researchers and are available via an online portal that we created (www.dropletproteome.org). We and others previously discovered that a subset of proteins involved in ER-associated degradation (ERAD) exhibit a dual ER and LD localization, including ubiquitination machinery and factors that recruit the AAA ATPase VCP. Coupling our APEX2 proximity labeling approaches with quantitative proteomics, we discovered that VCP inhibition stabilizes c18orf32, a protein of unknown function that traffics from the ER into nascent LDs. Depletion of c18orf32 alters the lipid composition of LDs, suggesting that c18orf32 abundance may regulate lipid metabolism. Importantly, further analyses revealed that c18orf32 levels are controlled by a canonical ERAD pathway involving the E3 ligase gp78, the dislocation factor derlin-1, and VCP. Together, these findings demonstrate the utility of our proximity labeling approach to uncover mechanisms of LD protein dynamics and uncover a fundamental mechanism of LD protein regulation involving the degradation of LD proteins transiting the ER via ERAD.The novel LD protein AIFM2, identified by our proteomics method, and its role in lipotoxicity are a focus of the current proposed research. A VCP inhibitor substrate trapping approach (VISTA) enables proteomic profiling of endogenous ERAD substrates(Published in Huang, To, et al.,Mol. Biol. Cell.29(9), 1021-1030. PMID: 29514927) ERAD regulates the levels of important proteins (e.g. HMG CoA reductase). However, our understanding of the endogenous proteins targeted by ERAD remains limited, in part due to the lack of generalizable methods to identify endogenous ERAD substrates. To overcome this obstacle, we developed a new proteomic method that exploits a small molecular inhibitor of the AAA ATPase VCP to trap ubiquitinated ERAD substrates. We refer to this approach as theVCPinhibitorsubstratetrappingapproach (VISTA) and our findings were recently published inMolecular Biology of the Cell. Coupling VISTA with ubiquitin proteomics, which takes advantage of an antibody against the diglycine (diGly) remnant of ubiquitin that remains on tryptic fragments, enabled global identification of accumulating ubiquitinated ERAD candidates in HepG2 liver cells. Our results identified several known substrates involved in lipid metabolism (ApoB100, Insig2, and 7-Dehydrocholesterol reductase) and novel substrates (SCD1 and RNF5). Our data establish a new method to globally profile endogenous ERAD substrates and identify candidate ERAD substrates. These technologies could be adapted to identify VCP-dependent substrates in other compartments such as LDs.

Publications

• Type: Journal Articles Status: Published Year Published: 2018 Citation: Bersuker, K., Peterson, C.W., To, M., Sahl, S.J., Savikhin, V., Grossman, E.A., Nomura, D.K., Olzmann, J.A.# (2018) A proximity labeling strategy provides insights into the composition and dynamics of lipid droplet proteomes. Dev. Cell. 44, 97-112. PMID: 29275994
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Huang, E.Y.*, To, M.*, Tran, E., Dionisio, L., Cho, H.J., Baney, K.L.M., Pataki, C.I., Olzmann, J.A.# (2018) A VCP inhibitor substrate trapping approach (VISTA) enables proteomic profiling of endogenous ERAD substrates. Mol. Biol. Cell. 29(9), 1021-1030. PMID: 29514927
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Bersuker, K. and Olzmann, J.A.# (2018) In close proximity: The lipid droplet proteome and crosstalk with the ER. Contact. (1), 1-3.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Li, Z. and Olzmann, J.A.# (2018) A proteomic map to navigate subcellular reorganization in fatty liver disease. Dev. Cell. 47, 139-141.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Olzmann, J.A.# and Carvalho, P.# (2018) Dynamics and functions of lipid droplets. Nat. Rev. Mol. Cell Biol. (In Press, Advanced version online)

Progress 10/01/16 to 09/30/17

Outputs
Target Audience:The target audience reached by our efforts during the reporting period was the scientific community. Our research and its significance were communicated to the scientific community through a combination of presentations and publications. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Postdoctoral scholars, graduate students, and undergraduates benefited from one-on-one mentorship in experimental technique and scientific thought. We have weekly subgroup and large group meetings to discuss recentfindings. We also participate in several working groups on campus (e.g. Membrane supergroup) to present and receive feedback.In addition, several members of the lab attended and presented at conferences or symposia (listed below in section regarding dissemination of results). How have the results been disseminated to communities of interest?In addition to the listed publications, we have also presented to scientific audiences at conferences and Universities. James Olzmann (PI) Speaker, University of California San Francisco Diabetes Center Retreat. Santa Cruz, CA. Speaker, FASEB Summer Conference Series: From Unfolded Proteins in the Endoplasmic Reticulum to Disease. 2017. Saxtons River, VT. Speaker, Bay Area Meeting on Organelle Biology. 2017. Berkeley, CA. Speaker, University of Delaware: Membrane Protein Symposium. 2017. Newark, DE. Seminar, University of the Sciences: Department of Biological Sciences. 2017. Philadelphia, PA. Seminar, Medical College of Wisconsin: Department of Biochemistry. 2017. Milwaukee, WI. Seminar, University of Texas Southwestern Medical Center: Department of Cell Biology, Leading Edge of Cell Biology Seminar Series. 2017. Dallas, TX. Speaker, EMBO The Endoplasmic Reticulum (ER) as a hub for organelle communication. 2016. Girona, Spain. Kirill Bersuker (Postdoc) Speaker, Bay Area Stress & Chaperone Symposium. 2017. Berkeley, Lorraine Ador Dionisio (Undergrad) Speaker, BSP Undergraduate research symposium. UC Berkeley. Berkeley, CA. Melissa Roberts (Graduate student) Poster presenter, Bay Area Metabolism Meeting. 2017. Berkeley, CA Poster presenter,Bay Area Organelle Meeting. 2017. Berkeley, CA. What do you plan to do during the next reporting period to accomplish the goals?Goal 1:Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome - We have made tremendous progress towards this goal over the past year, but much remains unknown. Over the course of the next funding period we will institute two important aims: 1) We will establish our proximity labeling system in more physiologically relevant cell lines to further explore lipid droplet proteome dynamics and regulation. For example, brown adipocytes are an extremely exciting model system because they utilize fatty acids and glucose to generate a proton gradient in the mitochondria, and then they dissipate that gradient and release the stored energy as heat through uncoupling protein 1 (UCP1). This is exciting because increase brown adipocyte amount has been found to greatly ameliorate symptoms in models of metabolic diseases. 2) We will perform CRISPR/Cas9 functional genomic screens to determine the mechanisms and pathways that mediate lipid droplet protein degradation through the ubiquitin-proteasome system. Goal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites - This a challenging, but important goal. To identify lipid droplet-mitochondrial tethering complexes we will exploit our proximity labeling approach. In theory, our proximity labeling fusion proteins will primarily biotinylate proteins on lipid droplets, but they should also biotinylate other proteins that are in extremely close proximity (e.g. mitochondrial tethers). Thus, by stimulating biotinylation of lipid droplets and examining biotinylated proteins on mitochondria, we should identify the tethering complexes. We will also take a parallel approach using PLIN5 as a candidate tether. This protein, when overexpressed, induces mitochondrial recruitment to lipid droplets. We use a combination of chemical crosslinking and quantitative proteomics to identify PLIN5 interacting proteins. We will then employ time-lapse fluorescence microscopy to test their roles as tethering proteins.

Impacts
What was accomplished under these goals? Goal 1: Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome­- Lipid droplets are neutral lipid (fat and cholesteryl esters) storage organelles that provide an on demand source of fatty acids taht can be utilized for membrane biogenesis and energy. Dysregulation of lipid droplets is associated with numerous emtabolic diseases, including obesity, diabetes, and fatty liver disease, highlighting the importance of achieving a detailed understanding lipid droplet biology.Nearly all lipid droplet functions are regulated by proteins that decorate the lipid droplet surface. For example, lipid droplet associated lipase mediate the breakdown of stored fat. Despite the importance of the lipid droplet protein coat, the composition of the lipid droplet proteome has remained elusive. This is part because of the lack of sufficient methods to specifically identify lipid droplet proteins. Indeed, previous approaches using biochemically purified lipid droplet fractions were plagued by high amounts of non-specific proteins, likely due to co-purifying membrane fragments. To overcome this considerable obstacle, we developed a novel proximity labeling approach. This approach exploited lipid droplet-target ascorbate peroxidase (APEX2) to biotinylate proteins in close proximity (~10-20nm) of the lipid droplet surface. Employing this method, we mapped high confidence lipid droplet proteomes in two human cell lines, the U2OS osteosarcoma and Huh7 hepatocellular carcinoma cell lines. Moreover, we combined this method with quantitative proteomics to elucidate mechanisms that regulate lipid droplet proteome remodeling. Our findings reveal a role for the endoplasmic reticulum-associated degradation (ERAD) pathway in controlling the abundance of some lipid droplet proteins. Our findings provide a useful resource for the scientific community and identify a fundmental mechanism of lipid droplet proteome regulation. These findings were recently published (Bersuker et al 2018). Goal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites- Lipolytic degradation of lipid droplets releases fatty acids that can be utilized by mitochondria as substrates to generate energy during periods of starvation. A complementary cellular process that is employed by cells during starvation is autophagy. Autophagy mediates the breakdown of cellular components during starvation, allowing recycling of components for essential processes. Our analyses revealed an unexpected regulationship between autophagy and lipid droplets. Surprisingly, we observed increases in the amounts of lipid droplets during starvation. These new lipid droplets required autophagy, suggesting that lipid released during the autophagic breakdown of membranous organelles were being repackaged into new lipid droplets. To explore the functional importance of these lipid droplets, we utilized inhibitors of diacylglycerol acyltransferases, which block triacylglycerol synthesis and the generation of lipid droplets. Under these conditions, fatty acids were still delivered to mitochondria and accumulated as acylcarntine, resulting in defects in mitochondrial respiration and membrane potential. Thus, our results reveal that lipid droplets function to sequester fatty acids and prevent lipotoxic cellular damage during autophagy. To further examine the mechanisms underlying lipid droplet communication with mitochondria, we have been characterizing our proximity labeling proteomic methods as a strategy to identify lipid droplet-mitochondrial protein tethering complexes.

Publications

• Type: Journal Articles Status: Published Year Published: 2017 Citation: Nguyen, T.B. and Olzmann, J.A.# (2017) Lipid droplets and lipotoxicity during autophagy. Autophagy. 13:11, 2002-2003. PMID: 28806138.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Nguyen, T.B., Louie, S.M., Daniele, J., Tran, Q., Dillin, A., Zoncu, R., Nomura, D.K., Olzmann, J.A.# (2017) DGAT1-dependent lipid droplet biogenesis protects mitochondrial function during starvation-induced autophagy. Developmental Cell. 42, 921. PMID: 28697336.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Bersuker, K., Peterson, C.W., To, M., Roberts, E., Sahl, S., Nomura, D.K., Olzmann, J.A.# (2018) A proximity labeling strategy provides insights into the composition and dynamics of lipid droplet proteomes. Developmental Cell. 44, 97-112. PMID: 29275994
• Type: Journal Articles Status: Published Year Published: 2017 Citation: To, M., Peterson, C.W., Roberts, M.A., Counihan, J.L., Wu, T.T., Forster, M.S., Nomura, D.K., Olzmann, J.A.# (2017) Lipid disequilibrium disrupts ER proteostasis by impairing ERAD substrate glycan trimming and dislocation. Mol. Biol. Cell (28), 270-284. PMID: 27881664.

Progress 10/01/15 to 09/30/16

Outputs
Target Audience:The target audience reached by our efforts during this reporting period was the scientific community. Our work and its significance was conveyed to the scientific community through presentations and publications Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Postdoctoral scholars, graduate students, and undergraduates benefited from one-on-one mentorship in experimental technique and scientific thought. In addition, several members of the lab attended and presented at conferences or symposia (listed below in section regarding dissemination of results). How have the results been disseminated to communities of interest?James Olzmann (PI): Speaker, FASEB Summer Conference Series: Lipid Droplets: Metabolic Consequences of the Storage of Neutral Lipids. 2016. Snowmass, CO. Speaker, American Society for Cell Biology (ASCB) meeting. 2015. San Diego, CA. Speaker, American Society for Biochemistry and Molecular Biology (ASBMB) meeting. 2015. San Diego, CA. Seminar, Oxford University / Ludwig Cancer Institute. 2015. Oxford, United Kingdom. Speaker, Biochemical Society: Organelle Crosstalk in Membrane Dynamics and Cell Signaling. 2015. Saxtons River, VT. Quan Tran(Undergraduate Student): Poster presentation, UC Berkeley Molecular and Cell Biology Undergraduate Honors Symposium Kirill Bersuker (Postdoctoral Scholar): Poster presentation,FASEB Summer Conference Series: Lipid Droplets: Metabolic Consequences of the Storage of Neutral Lipids. 2016. Snowmass, CO. Speaker, UC San Francisco Diabetes Center retreat, 2016. Santa Cruz. CA Truc Nguyen (Graduate Student) EMBO Workshop: Organelle Contact Sites: Intracellular Communicaiton and Role in Disease. 2016. Domus De Maria CA, Italy. What do you plan to do during the next reporting period to accomplish the goals?Goal 1: Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome During the past funding period we have continued to develop quantitative proximity labeling approaches to study lipid droplet (LD) proteome dynamics. These will be applied in different cell types, under different metabolic conditions to examine the role of proteome dynamics in organelle adaptation to environmental conditions. In addition, we have generated a reporter cell for fluorescence-based analysis of LD protein degradation. This cell line will be characterized and developed for use in whole genome CRISPR/Cas9 screens to identify the components of LD protein degradation pathways. Goal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites We have established that LDs transfer fatty acids to mitochondria under starvation conditions. We have also discovered that starvation stabilized LD-mitochondrial contacts, which can be observed by coordinate movement of the organelles in time-lapse microscopy experiments. We will couple our proximity labeling approaches with starvation conditions to label LD-mitochondrial tethering complex. Candidate components will be characterized using biochemistry and genetics. Publication: A major goal for this funding period is to publish our recent discoveries, including the discovery that LDs protect against lipotoxicity during starvation (submitted) and the initial description of our proximity labeling approach to study LD proteome dynamics (in preparation). We are also currently preparing an invited review article on mechanisms of LD protein targeting and degradation.

Impacts
What was accomplished under these goals? Goal 1:The composition of an organelle's proteome provides considerable information regarding its cellular functions. A major goal our proposed research is to understand how ubiquitin pathways regulate the lipid droplet (LD) proteome. Numerous papers have attempted to define LD proteomes through proteomic analysis of buoyant biochemical fractions, which are enriched in LDs. These studies have been plagued by false positives derived from co-fractionating organelles. To overcome this obstacle, we have developed a novel proximity label proteomics approach. Leveraging our new proximity labeling technology, we have defined a high confidence LD proteomes with unprecedented accuracy. Follow up validation experiments using GFP-tagged proteins has provided microscopy validation of an additional ten new LD proteins identified using our approach. A key component of the LD-associated ubiquitination complex is the AAA ATPase VCP. In ERAD, VCP functions to extract ubiquitinated substrates from the membrane for proteasomal degradation. We hypothesized that VCP may play a similar function on LDs. To examine the role of VCP in regulating LD proteome dynamics we combined our proximity labeling approach with quantitative proteomics methods (stable isotope labeling in cell culture - SILAC) and analyzed oleate-induced LDs that were isolated from cells incubated in the presence and absence of a potent VCP inhibitor (CB5083). Our results indicated that an uncharacterized ORF, c18orf32, accumulated on LDs treated with VCP inhibitor. Immunofluorescence microscopy and biochemical fractionation experiments confirm that endogenous c18orf localized to the ER and to LDs. Employing GFP-tagged versions of c18orf32 we found that c18orf32 is extremely dynamic within the ER and localizes to early sites of LD biogenesis following addition of oleate. An N-terminal hydrophobic region is necessary and sufficient for c18orf32 trafficking to LDs, a structure that is consistent with its insertion into and association with the phospholipid monolayer of the LD. Finally, analysis of c18orf32 turnover confirm that it is degraded via a pathway that requires VCP and ubiquitin. Thus, we have developed a robust new method to characterize LD proteome composition and dynamics. We have applied this tool to study the role of VCP in LD proteome dynamics and have identified c18orf32 as a VCP-regulated ER-LD protein. Our new proximity labeling technology and our resulting datasets will be widely useful to the scientific community. A manuscript describing these findings is currently in preparation. Goal 2: Definingthe molecular mechanisms underlying LD-mitochondrial transfer of fatty acids is critical to advance our understanding of cellular lipid and energy homeostasis.Highly proliferating cells utilize anaerobic glycolysis for energy and shunt their fatty acids towards phospholipid biosynthesis for membrane production. Under starvation conditions cells initiate global programs that coordinately alter their metabolism, shifting from glycolysis to mitochondrial β-oxidation for energy. Essential to this process is the lipolytic breakdown of LDs, which liberates stored fatty acids and providing the substrates for β-oxidation. During prolonged starvation, autophagy is upregulated to recycle cellular components for essential processes. Paradoxically, starvation also results in increased numbers of LDs. A recent manuscript suggests a model in which autophagic breakdown of organelles releases lipids that are stored in new LDs. However, the nutrient sensing pathways that control LD biogenesis under these conditions as well as the function of this new pool of LDs remained unknown. Why expend energy to form LDs when cells lack nutrients and are in an energy crisis? Our recent data answers this mystery in the field. Employing a combination of chemical inhibitors and CRISPR/Cas9 genome-engineered cell lines our findings indicate that the nutrient sensing kinase mTOR monitors amino acid levels and controls the generation of autophagy-dependent LDs during nutrient deprivation. Moreover, we find that autophagy-released fatty acids are channeled into DGAT1, but not DGAT2, -dependent LDs. Inhibition of DGAT1-dependent LD biogenesis during starvation impaired mitochondrial β-oxidation and cell survival. Interestingly, using quantitative lipidomic approaches, we find that in the absence of DGAT1-dependent LD biogenesis autophagy-released fatty acids no longer flux into triacylglycerol.Instead, fatty are channeled to mitochondria where they are converted into acylcaritines and cause lipotoxic damage. These results suggest that the purpose of this "paradoxical" LD biogenesis during starvation conditions is function as a lipid buffering system that protects cells from lipotoxic damage. A manuscript describing these findings is currently in review at Developmental Cell.

Publications

Progress 02/16/15 to 09/30/15

Outputs
Target Audience:The target audience reached by our efforts during this reporting period was the scientific community. Our work and its significance was conveyed to the scientific community through presentations and publications. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Postdoctoral scholars, graduate students, and undergraduates benefited from one-on-one mentorship in experimental technique and scientific thought. In addition, several members of the lab attended and presented at conferencesor symposia (listed below in section regarding dissemination of results). How have the results been disseminated to communities of interest?Results were disseminated to the scientifc community through publications and presentations: Stevenson, J., Huang, E.Y., Olzmann, J.A.# (2016) Endoplasmic reticulum-associated degradation and lipid homeostasis. Annu. Rev. Nutr. In Press. Riley, B.E., Olzmann, J.A.# (2015) A polyubiquitin chain reaction: Parkin recruitment to damaged mitochondria. PLOS Genet. 11(1), e1004952. James Olzmann (PI): Speaker, American Society for Cell Biology (ASCB) meeting. 2015. San Diego, CA Seminar, Oxford University / Ludwig Cancer Institute. 2015. Oxford, United Kingdom. Speaker, Biochemical Society: Organelle Crosstalk in Membrane Dynamics and Cell Signaling. 2015. Saxtons River, VT. Seminar, University of Seoul. 2015. Seoul, South Korea. Speaker, FASEB Summer Conference Series: From Unfolded Proteins in the Endoplasmic Reticulum to Disease. 2015. Saxtons River, VT. Seminar, University of California, San Francisco. Diabetes and Liver Center. 2015. San Francisco, CA. Melissa Roberts (summer student): Speaker,Understanding the Role of Lipid Modifications in ER Protein Quality Control. Sierra Systems and SynBio Symposium. 2015. Reno NV.(won first place in poster competition) Poster presentation,Understanding the Role of Lipid Modifications in ER Protein Quality Control. Society for Advancement of Chicanos/Hispanics and Native Americans in Science (SACNAS). 2015. Washington DC.(awarded SACNAS travel award) Milton To (Graduate student): Poster presentation, ER protein quality control and lipid homeostasis: unexpected functional connections. UC Berkeley Membrane Supergroup. 2015. Berkeley CA. What do you plan to do during the next reporting period to accomplish the goals?During the past funding period we have developed new, cutting-edge methods that will help us achieve our stated goals and objectives. Goal 1:Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome Objective 1a. Our new proximity labelingtechnology will be combined with genetic and pharmacological tools targeting lipid droplet ubiquitination pathways to facilitate proteomic identification of proteins regulated by ubiquitin-dependent degradation. Objective 1b. Our analysis of the lipid droplet proteome identified new components of a putative ubiquitination complex. In addition to previous components (UBXD8, UBXD2, AUP1, UBE2G2, VCP, SPG20), we identified a lipid droplet associated deubiquitinating enzyme (Usp33) and a portion of an E3 ubiquitin-protein ligase complex (FBX50). It will be important to examine the function of these new components in nutrient storage using siRNA and CRISPR/Cas9 approaches. Objective 1c. The functional role of lipid droplet ubiquitination components will be assessed by genetic deletion and time-lapse analysis of lipid droplet dynamics using a fluorescence deconvolution microscope. Goal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites Objective 2a. Cell lines expressing an synthetic system enabling temporal control of lipid droplet-mitochondrial contacts will be analyzed using microscopy assays of fatty acid transfer and flow cytometry assays of cell viability during nutrient deprivation. Objective 2b. Our proximity labeling technology will be leveraged to label and identify the endogenous lipid droplet - mitochondrial tethering complexes. These will be genetically disrupted and the consequences measured using structured illumination microscopy (super resolution).

Impacts
What was accomplished under these goals? On a basic level, we all recognize that the food we consume impacts our health. The storage of nutrients in our body as fat is essential because it provides important energy reservoirs that we can access during periods of need, such as between meals or during exercise. However, excess nutrient storage underlies the epidemic of metabolic diseases that our society and much of the world now faces. These disease include diabetes, obesity, heart diseases, and others. These conditions not only negatively impact individuals, but they also impose an increasingly large financial strain on our society's healthcare system. Therefore, research into the basic biological mechanisms that underlie nutrient storage and utilization is desperately needed. During the past funding period, I worked together withpostdoctoral scholars, graduate students, and undergraduate students in my lab to understand how fat is stored and mobilized from compartments in the cell called lipid droplets. Our researchidentified new cellular pathways that regulate these compartments. These pathways may facilitate the development of new therapeutics to combat metabolic diseases. In addition, wedeveloped novel methods that will improve the manner in which scientists around the world willstudy these compartments. Goal 1:Understand the role of ubiquitin in the dynamic regulation of the lipid droplet proteome Objective 1a. To identify ubiquitinated lipid droplet proteins that undergo degradation we developed a new proximity labeling approach to define all proteins on lipid droplets. This advanced method allowed us to define the most high confidence lipid droplet proteome in human cells to date. This new technology will allow us to applying quantitative proteomics together with inhibitors of ubiquitination pathways to identify substrates that are degraded through ubiqutiin dependent pathways. Objective 1b. Our analysis of the lipid droplet proteome identified new components of a putative ubiquitination complex. In addition to previously known components (UBXD8, UBXD2, AUP1, UBE2G2, VCP, SPG20), we identified a lipid droplet-associated deubiquitinating enzyme (Usp33) and a subunit of an E3 ubiquitin-protein ligase complex (FBX50). It will be important to examine the function of these new components in nutrient storage using siRNA and CRISPR/Cas9 approaches. Objective 1c. Our results indicate that inhibition of the ubiquitin-dependent AAA ATPase VCP results in the accumulation of ubiquinated proteins on lipid droplets and may influence lipid droplet clustering, an important precursor to lipid droplet fusion and growth. Additional studies are necessary to determine the contribution of other components of the lipid droplet ubiquitination complex and the ubiquitianted targets. We have generated cell lines expressing a His-tagged ubiquitin, which will facilitate global ubiquitin proteomic studies to define ubiquitinated proteins. Goal 2: Determine the molecular composition and functional role of lipid droplet-mitochondrial contact sites Objective 2a. We have generated an unique small molecule system that enables temporal control of lipid droplet - mitochondrial contact formation. This unique system will allow us to test whether induction of such contacts increases fatty acid transfer and viability. Objective 2b. Towards this objective we are characterizing a new proteomic approach that will allow us to label and indenify the protein complexes present at these contact sites. Briefly, this approach leverages our novel proximity biotinylation method that labels neighboring proteins (within ~10nm-40nm) with biotin, facilitating proteomic idenfication. Following biotinylation of the lipid droplet proteome with this strategy we will purify and analyze biotinylated proteins from mitochondrial fractions. These labeled proteins will be high priority candidates for characterization as lipid droplet - mitochondrial tethers.

Publications

• Type: Journal Articles Status: Accepted Year Published: 2016 Citation: Stevenson, J., Huang, E.Y., Olzmann, J.A.# (2016) Endoplasmic reticulum-associated degradation and lipid homeostasis. Annu. Rev. Nutr. In Press.
• Type: Journal Articles Status: Published Year Published: 2015 Citation: Riley, B.E., Olzmann, J.A.# (2015) A polyubiquitin chain reaction: Parkin recruitment to damaged mitochondria. PLOS Genet. 11(1), e1004952.