Hepatic Peroxisome and Its Implications on Cardiac Function, Inflammation and Aging

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2020

Hepatic peroxisome and its implications on cardiac function, inflammation and aging

Kerui Huang Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd

Recommended Citation Huang, Kerui, "Hepatic peroxisome and its implications on cardiac function, inflammation and aging" (2020). Graduate Theses and Dissertations. 18329. https://lib.dr.iastate.edu/etd/18329

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Hepatic peroxisome and its implications on cardiac function, inflammation and aging

Kerui Huang

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Genetics and Genomics

Program of Study Committee: Hua Bai, Major Professor Stephen Howell Kristen Johansen Maura McGrail Marna Yandeau-Nelson

The student author, whose presentation of the scholarship herein was approved by the program of study committee, is solely responsible for the content of this dissertation. The Graduate College will ensure this dissertation is globally accessible and will not permit alterations after a degree is conferred.

Iowa State University

Ames, Iowa

2020

DEDICATION

I would like to dedicate this thesis to my parents, who always give me full support even under difficult times. Also, my husband, who endured the same level of stress as a PhD student yet managed to give guidance and encouragement. I would also like to dedicate this thesis to my dear friends for making this journey enjoyable. iii

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...... v

ABSTRACT ...... vi

CHAPTER 1. GENERAL INTRODUCTION ...... 1

CHAPTER 2. LIVER HEPATOKINES AND PEROXISOMES AS THERAPEUTIC TARGETS FOR CARDIOVASCULAR DISEASES ...... 2 2.1 Liver dysfunctions that cause cardiac diseases ...... 3 2.2 Hepatokines linked to cardiovascular diseases ...... 5 2.3 Conclusion ...... 7 2.4 References ...... 7

CHAPTER 3. FRIENDS OR FOES? THE ROLE OF PEROXISOMES IN INFLAMMAGING ...... 10 3.1 Introduction ...... 10 3.2 Peroxisomes as a source and sink for reactive oxygen species (ROS) ...... 11 3.3 What happens to redox level when peroxisome is impaired? ...... 13 3.4 Peroxisomes and aging ...... 15 3.5 References ...... 16

CHAPTER 4. RIBOTAG TRANSLATOMIC PROFILING OF DROSOPHILA OENOCYTES UNDER AGING AND OXIDATIVE STRESS ...... 19 4.1 Abstract ...... 20 4.2 Background ...... 21 4.3 Results ...... 23 4.4 Discussion ...... 36 4.5 Conclusion ...... 43 4.6 Methods ...... 43 4.7 References ...... 48 4.8 Figures ...... 56

CHAPTER 5. IMPAIRED PEROXISOMAL IMPORT IN DROSOPHILA OENOCYTES CAUSES CARDIAC DYSFUNCTION BY INDUCING UPD3 AS A PEROXIKINE ...... 69 5.1 Abstract ...... 69 5.2 Introduction ...... 70 5.3 Results ...... 72 5.4 Discussion ...... 85 5.5 Methods ...... 88 5.6 References ...... 95 5.7 Figures ...... 100 5.8 Appendix ...... 119

CHAPTER 6. PEROXISOME IMPORT STRESS IMPAIRS RIBOSOME BIOGENESIS IN DROSOPHILA AND HUMAN CELLS...... 120 6.1 Abstract ...... 120 6.2 Introduction ...... 121 6.3 Results ...... 124 6.4 Discussion ...... 137 6.5 Methods ...... 143 6.6 References ...... 151 6.7 Figures ...... 160 6.8 Appendix ...... 172

CHAPTER 7. CONCLUSIONS ...... 173

ACKNOWLEDGMENTS

I would like to thank my committee chair Dr. Hua Bai for his support, advice, encouragement, and guidance through my PhD. He saw the raw potential in me and guide me to gain confidence and to become a competent scientist. His passion for science will always inspire me and I will take the skills I learned throughout my scientific career.

I would also like to thank my former committee member, Prof. Linda Ambrosio who helped me in so many ways and my committee members: Prof. Maura McGrail, Prof. Marna

Yandeau-Nelson, Prof. Stephen Howell and Prof. Kristen Johansen - for their time and commitment throughout the course of this research. In addition, I would also like to thank the

GDCB department faculty and staff for making my time at Iowa State University a wonderful and unforgettable experience.

ABSTRACT

Aging is characterized by a chronic, low-grade inflammation, which is a major risk factor for cardiovascular diseases. As one of the important metabolic centers, the liver shows age- related dysregulation of lipid metabolism, increased inflammation, and oxidative stress response.

It remains poorly understood whether pro-inflammatory factors released from liver contribute to the non-autonomous regulation of age-related cardiac dysfunction.

We showed that Drosophila oenocytes share molecular similarities and age-related changes with mammalian livers. We perform translatomic profiling on Drosophila oenocytes to understand age- and stress-regulated gene expression changes. We show that aging down- regulates oxidative phosphorylation, ribosome, proteasome and peroxisomal biogenesis in oenocytes, whereas innate immune response and Ras/MAPK signaling pathways are induced.

Next, we identify that age-dependent induction of cytokine unpaired 3 (upd3) in

Drosophila oenocytes is the primary non-autonomous mechanism for cardiac aging. Oenocyte- specific knockdown of upd3 is sufficient to block aging-induced cardiac arrhythmia. We further show that the age-dependent induction of upd3 is triggered by impaired peroxisomal import and elevated JNK signaling in aged oenocytes. Intriguingly, oenocyte-specific overexpression of

Pex5, the key peroxisomal import receptor, blocks age-related upd3 induction and alleviates cardiac arrhythmicity.

Our previous studies prompt us to understand the transcriptomic responses under peroxisome import dysfunction. Thus, we performed RNA-seq analysis to understand cellular stress response caused by peroxisome import defects in oenocytes and human cells. Intriguingly, we identify that endoplasmic reticulum genes are induced, and ribosome biogenesis is impaired vii under peroxisome defects. Altogether, these studies demonstrate the important role of hepatic peroxisome in maintaining cellular homeostasis and whole organismal health. 1

CHAPTER 1. GENERAL INTRODUCTION

It remains poorly understood whether pro-inflammatory factors released from liver contribute to the non-autonomous regulation of age-related cardiac dysfunction. Peroxisomes are important organelles that regulate aging, inflammation, and lipid metabolism. Liver is highly enriched with peroxisomes. However, little is known about whether and how liver peroxisomes regulate other organs non-autonomously. In this thesis, I first summarize current knowledge of liver-heart communications, how liver peroxisomes regulate inflammation and aging. In the following chapter, I present the translatomic profiling on Drosophila oenocytes (hepatocyte-like cells) under aging or oxidative stress. In the translatomic study, we discover that peroxisomes are highly enriched in oenocytes and compromised during aging. Next, we present genetic evidence showing that liver dysfunction can induce cardiac arrhythmia non-autonomously through proinflammatory factor, upd3 (homologous to interleukin 6). We also identify the role of liver peroxisome on liver-aging induced upd3 production and cardiac arrhythmia. Lastly, we conduct transcriptomic analysis on fly oenocytes with peroxisome dysfunction, aiming to identify cellular changes induced by peroxisome defects. We identify that endoplasmic reticulum genes and ribosome biogenesis genes are induced under peroxisome stress. These studies on liver peroxisomes will provide new insights to treat liver related disease and cardiac diseases. 2

CHAPTER 2. LIVER HEPATOKINES AND PEROXISOMES AS THERAPEUTIC TARGETS FOR CARDIOVASCULAR DISEASES

An editorial published in Future Cardiol. 2020 Oct 22. doi: 10.2217/fca-2020-0166.

PMID: 33090051

Kerui Huang 1*, Hua Bai 1*

From the 1Department of Genetics, Development, and Cell Biology, Iowa State University,

Ames, Iowa 50011, USA

* To whom correspondence should be addressed: Kerui Huang: Department of Genetics,

Development, and Cell Biology, Iowa State University, Ames, Iowa 50011; [email protected];

Hua Bai: Department of Genetics, Development, and Cell Biology, Iowa State University, Ames,

Iowa 50011; [email protected]

Keywords: Liver-heart communication, Interleukin-6, Peroxisome

Increasing evidence from recent clinical studies indicates a close connection between liver dysfunction and cardiovascular diseases (CVDs) 1. Although it remains to be determined how patients with liver diseases often develop cardiovascular dysfunction, recent studies using animal models and epidemiology studies have shed light on the key role of hepatokines (a group of liver-derived proteins) and hepatic peroxisome dysfunction in the induction of 3 cardiomyopathy. Here, we summarize the role of representative hepatokines, Interleukin-6 (IL-

6) and fetuin-A, and liver peroxisome in the progression of cardiovascular diseases.

2.1 Liver dysfunctions that cause cardiac diseases

Liver cirrhosis is a well-known cause that alters systemic haemodynamic system and cardiac function. Cirrhotic cardiomyopathy (CCM) designates a cardiac dysfunction that includes impaired cardiac contractility, systolic and diastolic dysfunction, electromechanical abnormalities in the absence of other known causes of cardiac disease 1. These cardiac changes can occur in up to 40-50% of patients with cirrhosis, even in children with various liver diseases

2. Pathogenesis of CCM includes changes in portal pressure and hepatic blood flow, resulting in increased levels of circulating vasodilators, such as adrenomedullin, carbon monoxide, endocannabinoids, nitric oxide and tumor necrosis factor (TNFα) 2. Additionally, downregulation of the density of β-receptors at the cell surface of cardiomyocytes also contribute to the chronotropic incompetence2. Current pharmacological treatment for CCM is nonspecific, and directed towards treating cardiac dysfunction or through liver transplantation 2.

In the patients with liver transplantation (LT), cardiovascular disease has emerged as the major cause of morbidity and mortality. A meta-analysis of 12 studies conclude that 13.6% of patients who received liver transplant developed cardiovascular disease, being responsible for

42% of deaths within 30 days following LT 3. The primary complications of post-LT heart failure include systolic, diastolic, or mixed heart failure. These cardiac dysfunctions are caused by the acute hemodynamic changes that are resulted from reperfusion of the graft and proinflammatory cytokines 4.

NAFLD is a very common disease in developed countries. Approximately 20 to 30% of adults in western countries have nonalcoholic fatty liver disease. The prevalence increases to 70 to 90% among people who are obese or have diabetes. The disease can progress towards more 4 serious stages: non-alcoholic steatohepatitis (NASH) and cirrhosis, where inflammation and damage to the liver occur. Beyond the liver-related morbidity and mortality, there is increasing evidence showing that patients with NAFLD are also at high risk of cardiovascular disease1. A recent meta-analysis involving 34043 adults has shown that patients with NAFLD had a higher risk of fatal and/or non-fatal CVDs comparing to those without NAFLD. Patients with more severe NAFLD were more likely to develop CVDs 5. Based on Framingham scoring method, the mean of cardiovascular risks in subjects with and without NAFLD was 16.0% and 12.7% in men,

6.7% and 4.6% in women6. Manifestations of cardiovascular disease induced by NAFLD include left ventricular dysfunction, atherosclerotic CV diseases, ischemic stroke, and cardiac arrhythmias. Possible mechanisms include NAFLD associated insulin resistance, inflammation, visceral adiposity, dyslipidemia, and oxidative stress. These studies suggest the close connection between liver health and cardiovascular diseases.

NAFLD is a common disease in the elderly, in who it carries a higher chance to develop more serious hepatic diseases (NASH, cirrhosis, and hepatocellular carcinoma). Specific aged- related hepatic changes include increased hepatocyte size and the number of binucleated cells, reduction in mitochondrial number, decreased oxidative phosphorylation, impaired peroxisome function 7, 8, and upregulation of inflammatory signaling. In addition, aging is associated with a physiological increase in lipid accumulation in the liver and other nonadipose tissues. The association between NAFLD and CVDs is particularly significant in elderly people. Data suggest that in middle-aged individuals (45-54 year-old age group), NAFLD is a strong independent risk factor for cardiovascular mortality 9. One study evaluated 810 elderly males and 1273 elderly women in Japan and demonstrated that NAFLD was associated with coronary risk factors independent of obesity in either gender 10. High prevalence of liver diseases and CVDs in elderly 5 might be a result of inflamm-aging, a low-grade chronic inflammation that contributes to the aging pathologies. These studies underscore the possibility of liver aging as a causal factor in

CVDs.

2.2 Hepatokines linked to cardiovascular diseases

The underlying mechanisms by which liver diseases increase the risk of CVDs has not been fully elucidated. One possible explanation is that intrahepatic inflammation or hepatic damage releases a variety of proatherogenic proinflammatory factors, such as high-sensitivity C- reactive protein (hsCRP), fibrinogen, interleukin-6 (IL-6), and plasminogen activator inhibitor-1

(PAI-1). These factors can directly regulate the development of CVDs. Another mechanism is through the secretion of hepatokines that can induce metabolic abnormalities, such as T2DM, hypertriglyceridaemia and glucose intolerance, all of which contribute to development of CVDs

11, 12.

IL-6 is one of the most important proinflammatory cytokines associated with inflammaging and age-related diseases. It is a soluble mediator with a pleiotropic effect on immune response, inflammation, and hematopoiesis. Short-term induction of IL-6 signaling can protect myocytes from injury-induced apoptosis. However, prolonged induction of IL-6 signaling can cause pathological hypertrophy and decrease cardiomyocyte contractility through the activation of Janus Kinases-signal transducer and activator of transcription (JAK-STAT) pathway. Elevated levels of circulating IL-6 are often associated with myocardial damage, heart failure, and atherosclerosis 13. IL-6 is originally discovered to be secreted primarily from immune cells. However, IL-6 can also be produced from a variety of cell types under tissue injury or infection, including cardiomyocytes, skeletal muscle, and hepatocytes. IL-6 expression is increased in animal models of NAFLD and in the livers of patients with nonalcoholic steatohepatitis (NASH), compared to patients with simple steatosis or normal biopsies 14. Using 6

Drosophila oenocytes as a hepatocyte model, we recently discovered that the expression of upd3

(the fly homology of mammalian IL-6) in oenocytes increased more than 60-fold during aging.

Additionally, we found that reducing oenocyte upd3 expression can effectively reduce cardiac arrythmia during aging 8. Collectively, these data suggest that increased production of hepatic

IL-6 plays an important role in cardiovascular diseases of NAFLD and NASH patients, as well as in age-associated CVDs.

Fetuin-A is a major carrier protein of free fatty acids. It’s predominantly made in liver and can be secreted into the bloodstream 15. Emerging evidence suggests that fetuin-A plays a role in the development of the metabolic syndrome 11, as levels of fetuin A are increased in

NAFLD patients. It's also well-established that the level of fetuin A is positively associated with insulin resistance11. Injection of recombinant fetuin A into mice can decrease insulin sensitivity, whereas fetuin A deficient mice are insulin sensitive and resistant to weight gain under high fat diets 11. Given its close association with insulin resistance, it’s not surprising that fetuin-A levels are positively associated with markers of early atherosclerosis, incidence of myocardial infarction, ischemic stroke, independent of other cardiovascular risk parameters 16. A possible mechanism to explain the role of hepatic fetuin A in insulin resistance and CVDs, is that fetuin A might serve as a sensor for nutrient level. In an excessive nutrient state, fetuin A level is increased and is secreted into bloodstream, which contributes to chronic inflammation and metabolic syndrome 17.

Peroxisomes are essential subcellular organelles of all eukaryotic cells that are crucial in the regulation of cellular redox homeostasis, oxidation of very long chain fatty acids (VLCFAs), and biosynthesis of ether phospholipids 18. Impaired peroxisome function often results in disrupted redox homeostasis, elevated lipotoxicity, and induction of systemic inflammation 18, 19. 7

Using the fruit fly Drosophila as the model organism, we recently demonstrated a causal role of hepatic peroxisome dysfunction in the non-autonomous regulation of cardiomyopathy and cardiac aging. In the study, we discovered that hepatocyte specific knockdown (KD) of peroxisomal import receptor (Pex5) triggers significant production of inflammatory cytokine upd3 (the fly homolog of mammalian IL-6). Notably, either reducing upd3 or overexpressing

Pex5 in fly hepatocytes alleviates aging-induced cardiac dysfunction 8. Our studies provide the first direct evidence supporting the emerging role of hepatic peroxisome function in mediating non-autonomous regulation of cardiac function.

2.3 Conclusion

Accumulated evidence from both clinical studies and basic research highlights the importance of liver dysfunction in the development of cardiomyopathy. Hepatokines that are mainly secreted from the liver can directly affect glucose, lipid metabolism and modulate inflammatory signaling. All these factors can contribute to the progression of CVDs. Despite the important role of hepatokines in CVDs, how these hepatokines are regulated is less well-known.

Peroxisomes, as well as mitochondria, are highly enriched organelles in the liver. Impaired peroxisomal function may contribute to the induction of hepatokines and the development of

CVDs. Therefore, it is critical to develop management strategies and interventions to reduce

CVDs risks by targeting liver inflammation, hepatokine secretion, and hepatic peroxisome function. Identification of novel hepatokines and biomarkers will assist such development of strategies to treat CVDs.

2.4 References

1. El Hadi, H., Di Vincenzo, A., Vettor, R. & Rossato, M. Relationship between Heart Disease and Liver Disease: A Two-Way Street. Cells 9 (2020).

2. Wiese, S., Hove, J.D., Bendtsen, F. & Moller, S. Cirrhotic cardiomyopathy: pathogenesis and clinical relevance. Nat Rev Gastroenterol Hepatol 11, 177-186 (2014).

3. VanWagner, L.B. et al. High early cardiovascular mortality after liver transplantation. Liver Transpl 20, 1306-1316 (2014).

4. Bezinover, D. et al. Release of cytokines and hemodynamic instability during the reperfusion of a liver graft. Liver Transpl 17, 324-330 (2011).

5. Targher, G., Day, C.P. & Bonora, E. Risk of cardiovascular disease in patients with nonalcoholic fatty liver disease. N Engl J Med 363, 1341-1350 (2010).

6. Motamed, N. et al. Non-alcoholic fatty liver disease (NAFLD) and 10-year risk of cardiovascular diseases. Clin Res Hepatol Gastroenterol 41, 31-38 (2017).

7. Premoli, A. et al. Characteristics of liver diseases in the elderly: a review. Minerva Gastroenterol Dietol 55, 71-78 (2009).

8. Huang, K. et al. Impaired peroxisomal import in Drosophila oenocytes causes cardiac dysfunction by inducing upd3 as a peroxikine. Nature Communications 11, 2943 (2020).

9. Dunn, W. et al. Suspected nonalcoholic fatty liver disease and mortality risk in a population-based cohort study. Am J Gastroenterol 103, 2263-2271 (2008).

10. Akahoshi, M. et al. Correlation between fatty liver and coronary risk factors: a population study of elderly men and women in Nagasaki, Japan. Hypertens Res 24, 337-343 (2001).

11. Meex, R.C.R. & Watt, M.J. Hepatokines: linking nonalcoholic fatty liver disease and insulin resistance. Nat Rev Endocrinol 13, 509-520 (2017).

12. Yoo, H.J. & Choi, K.M. Hepatokines as a Link between Obesity and Cardiovascular Diseases. Diabetes Metab J 39, 10-15 (2015).

13. Fontes, J.A., Rose, N.R. & Cihakova, D. The varying faces of IL-6: From cardiac protection to cardiac failure. Cytokine 74, 62-68 (2015).

14. Wieckowska, A. et al. Increased hepatic and circulating interleukin-6 levels in human nonalcoholic steatohepatitis. Am J Gastroenterol 103, 1372-1379 (2008).

15. Pal, D. et al. Fetuin-A acts as an endogenous ligand of TLR4 to promote lipid-induced insulin resistance. Nat Med 18, 1279-1285 (2012).

16. Jung, T.W., Yoo, H.J. & Choi, K.M. Implication of hepatokines in metabolic disorders and cardiovascular diseases. BBA Clin 5, 108-113 (2016).

17. Haukeland, J.W. et al. Fetuin A in nonalcoholic fatty liver disease: in vivo and in vitro studies. Eur J Endocrinol 166, 503-510 (2012).

18. Waterham, H.R., Ferdinandusse, S. & Wanders, R.J. Human disorders of peroxisome metabolism and biogenesis. Biochim Biophys Acta 1863, 922-933 (2016).

19. Baes, M. et al. A mouse model for Zellweger syndrome. Nat Genet 17, 49-57 (1997).

CHAPTER 3. FRIENDS OR FOES? THE ROLE OF PEROXISOMES IN INFLAMMAGING

3.1 Introduction

Inflammaging is a chronic low-grade inflammation that develops with advanced age. It is believed to induce age-associated proinflammatory biomarkers and diseases, such as cardiovascular diseases (CVDs). However, the root causes of inflammaging remains to be undetermined. Recently, increasing evidence placed peroxisomes the center of inflammation and related diseases, alongside with mitochondria.

Peroxisomes are eukaryotic subcellular organelles. Generally peroxisomes are responsible for catabolism of branched or very long chain fatty acids (VLCFAs, ≥C22) through

β-oxidation system, purine catabolism, synthesis of plasmalogens, ether-lipids or bile acids, metabolism of cholesterol and phytanic acid, regulation of hydrogen peroxide and other reactive oxygen species (ROS) 1. The volume, number or content of peroxisomes within a cell changes dynamically depending on the surrounding conditions and cell types 2. There are several competing models for peroxisome proliferation in mammalian cells, but all based on two themes:

1) templated assembly by fission and growth of the existing pool; 2) de novo assembly from small membrane vesicles derived from endoplasmic reticulum (ER) and proteins imported from the cytoplasm, through the help of receptor proteins 3. Thirteen peroxisome assembly factors

(Peroxins, or PEX) are required for peroxisome biogenesis in humans. Among them, PEX5 and

PEX7 act as receptors that recognize peroxisome targeting signals (PTS) in soluble peroxisomal proteins to traffic them from the cytosol to the peroxisome matrix 4. PEX5 recognizes the C- terminal PTS1 with the canonical sequence Ser-Lys-Leu (SKL), while PEX7 recognizes an N-

4 terminal nonapeptide PTS2 with the consensus sequence (R/K)(L/V/I)X5(H/Q)(L/A) . 11

The importance of peroxisomal function is highlighted by fatal human genetic diseases: peroxisomal biogenesis disorders (PBDs), including Zellweger spectrum disorders (ASDs),

Heimler syndrome, and rhizomelic chondrodysplasia punctate (RCDP) 1. PBD patients manifest metabolic diseases, as well as develop abnormalities in the central nervous system (CNS).

Mutations in human genes encoding PEX genes, or enzymes that perform metabolic function inside peroxisomes, can lead to these PBDs. The function of many PEX genes are conserved among species. For example, In Drosophila Schneider 2 (S2) cells and oenocytes (hepatocyte- like cells), knockdown of the Pex5 transcript reduces targeting of PTS1-containing proteins to peroxisomes 5, 6. Furthermore, Pex5 mutants’ flies exhibit severe developmental defects in the muscle and embryonic nervous system. Similar to that observed in patients of PBDs with PEX5 mutations 4. However, it is worth noting that the actual function of Drosophila Pex7 remains debatable, as no peroxisomal PTS2-containing proteins have been identified in Drosophila.

Besides focusing on PBD pathologies and understanding basic PEX function, recent studies on peroxisomes have expanded the horizon. Emerging evidence suggest peroxisome’s important role in mounting immune responses to microbial infection7, control of ferroptosis8 and aging 6, 9. Studies support a peroxisomal role in the pathogenesis of aging-related diseases, including neurodegenerative disorders, diabetes, and cancer, although it is not clear whether these pathologies are a cause or consequence of peroxisomal dysfunction10. Contradictory findings also arise regarding to the role of peroxisomes in inflammation and aging. This review will summarize existing evidence supporting the connection between peroxisome and these cellular processes and provide an explanation for these contradictory findings.

3.2 Peroxisomes as a source and sink for reactive oxygen species (ROS)

Peroxisomes have major metabolic function in: lipid metabolism and ROS metabolism.

However, these two processes are highly interconnected and they require close interactions of 12 peroxisomes with other organelles including mitochondria, endoplasmic reticulum (ER), lipid droplets and lysosomes 10. In addition to regulating ROS, peroxisomes are essential for beta oxidation of very long chain fatty acids (VLCFAs), alpha oxidation of branched-chain fatty acids

10. Peroxisomal beta-oxidation differs from mitochondrial beta-oxidation in several ways. The first step of mitochondrial beta oxidation produces FADH2, which is used to generate ATP via electron transport chain. In contrast, the first step of beta oxidation in peroxisomes is to the reduction of O2 to the H2O2. The peroxisome is the major site for cellular oxygen consumption

10 (up to 20%) and H2O2 production (up to 35% of total H2O2 in mammalian tissues) . To counterbalance this massive ROS production, peroxisomes in some cell types contain abundant amount of the antioxidant enzyme: catalase, which can reduce H2O2 to water. In addition, peroxisomes also produce other types of ROS, including nitric oxide radicals and superoxide radicals, from xanthine oxidase and the inducible form of nitric oxide synthase. These radicals are handled with their associated antioxidant enzymes: catalase for degrading H2O2; superoxide dismutase 1 (SOD1) for degrading O2•−; peroxiredoxin 5 (PRDX6) targeting ONOO-, ROOH,

11 H2O2 . Other peroxisome-localizing anti-oxidant enzymes include glutathione S-transferases 9

(GST) and epoxide hydrolase 2 (EPHX2), which have unknown targets 11. Another interesting group of peroxisome specific non-enzymatic antioxidant is plasmalogens. Plasmalogens are specifically synthesized in peroxisomes. Current evidence suggest that these intermediates may act as radical scavengers 11. For example, it has been shown that plasmalogens can protect unsaturated membrane lipids against oxidation. Cells deficient for peroxisome or plasmalogen biosynthesis are more sensitive to UV-induced ROS production than control cells 11. Despite numerous knowledge on the regulation of ROS from peroxisome under basal cellular conditions, no consensus has been made on whether peroxisome dysfunction could increase ROS on the 13 cellular level or organelle level. This knowledge is crucial because it will provide insights on the pathologies of PBDs and peroxisome defect related diseases, such as Alzheimer’s disease and aging 10, 12.

3.3 What happens to redox level when peroxisome is impaired?

To identify major regulators and transcriptomic changes in oenocyte (hepatocyte-like cells in Drosophila) aging, we performed Ribo-Tag translatomic profiling on aged Drosophila oenocytes. We identified that the majority of genes in peroxisome function declined during aging

13. This is consistent with study conducted in c.elegans, where the authors analyzed proteomic changes during aging 14. Surprisingly, in our genetic screen which aims to identify factors that mediate liver-heart communication, we found that knocking down Pex5 in oenocytes (liver) significantly increased cardiac arrhythmia, through up-regulation of inflammatory factor upd3

(IL-6 like factor in mammals). It has been well-established that ROS increase can cause IL-6 elevation 15. Consistently, using dihydroethidium (DHE) which measures total level of cellular

ROS, we found that Pex5 and Pex1 knockdowns significantly induced the level of ROS in oenocytes 6. In addition, Pex5 knockdown induced upd3 expression in a JNK (c-Jun N-terminal kinase)-dependent manner 6. Since oxidative stress in the most notable JNK activator 16, our evidence suggest a high level of oxidative stress in Pex5 knockdown. We also found IL-6 and

JNK increased in PEX1-G843D-PTS1 cell line, which is derived from patients with Zellweger syndrome. Using Amplex Red, which specifically measures the production of H2O2, we showed

G843D 6 that PEX1 mutant cells showed elevated H2O2 levels . Together, our data suggest that peroxisome dysfunction induces proinflammatory factor.

However, there are conflicting data on how peroxisome defects regulate ROS level.

Using a fluorescence resonance energy transfer (FRET) probe, Redoxfluor, which senses the 14 physiological redox state via the internal disulfide bonds and protein conformational change,

Yano et al. measured the redox states in the peroxisome and cytosol in yeast and Chinese hamster ovary (CHO) cells 17. They found that peroxisomes have a more reductive state, even though reactive oxygen species were generated within the peroxisomes. In addition, PEX5 and

PEX2 mutants (CHO cells) also showed a more reductive environment in the cytosol.

Meanwhile, another group of scientists used a redox-sensitive variant of enhanced green fluorescent protein (roGFP2) to monitor redox level in different subcellular compartments18.

Interestingly, when they transfected roGFP2-PTS1 and c-roGFP2 to mouse embryonic fibroblasts (MEFs), they found peroxisomal but not the cytosolic redox level altered when cells raised under different culture media. The intraperoxisomal redox environment is more oxidizing than the cytosol when the cells are cultured in MEM alpha medium, and more reducing when the cells are grown in the F-12 nutrient mixture 18. These data suggest that intraperoxisomal redox status can be strongly influenced by environmental conditions. In addition, it is important to note the subcellular ROS level under peroxisome stress.

Peroxisome dysfunction might lead to an excess ROS leaking to the cytosol. One possibility is that peroxisome defects caused mitochondria dysfunction, which then becomes the major source of ROS production. Over the years, substantial evidence suggest close interaction between peroxisomes and mitochondria. For example, peroxisomes and mitochondria share substrates in the β-oxidation of fatty acids and detoxification of ROS 19. Furthermore, peroxisomes and mitochondria also share key proteins of their division machinery 19. It is not surprising that mitochondria are defective in several peroxisomal disorders. A mouse model of hepatocyte specific Pex5-KO was generated to investigate the consequences of peroxisome deficiency on hepatocytes’ function. The most prominent was the abnormal structure of the inner 15 mitochondrial membrane, accompanied by reduced activities of complex I, III and V 20.

Surprisingly, no change in redox level was observed in Pex5 KO mice, using malondialdehyde and carbonyl content as indicators of oxidative stress damage to lipids and proteins 20. In addition, by measuring oxidized glutathione (GSSG)/glutathione (GSH) ratio as indicator of oxidative stress, they observed lower ratios in liver extracts 20. However, the same group later discovered that using roGFP-mt, mitochondrial DNA was depleted, and peroxisome deficient hepatocytes displayed increased mitochondrial oxidative stress in Pex5 KO hepatocytes 21.

Furthermore, they demonstrated that transient reduction of Pex5 produced these mitochondrial defects. These studies suggest that peroxisome defects have pronounced effects on mitochondrial functions, possibly through altering mitochondrial membrane permeability and membrane potential. Furthermore, the conflicting results from various studies highlight the importance to consider organelle-specific ROS under peroxisome stress. It is possible that peroxisome dysfunction specifically elevates mitochondria ROS level without affecting the cytosol ROS, and that mitochondrial ROS can directly regulate induce signaling pathways such as JNK. In summary, peroxisome dysfunction will disrupt ROS homeostasis, to what extent depending on the cell type and nutrient status.

3.4 Peroxisomes and aging

The role of peroxisomes on aging regulation has been largely ignored, partly because of the similarities in peroxisomes and mitochondria, with much of the attention on the role of mitochondrial dysfunction on oxidative-stress-associated aging 10. Although peroxisomes play a major role in H2O2 production, it is difficult to quantify the contribution of peroxisomal ROS in this process. Nevertheless, growing evidence suggests that peroxisomal function declines with aging and links dysregulated peroxisomal ROS metabolism to diseases 22, 23. 16

Studies have found that dietary restricted flies with enhanced lifespan compared to wild type flies show elevated expression of peroxisomal genes, including Catalase (Cat), Pex16, rosy

(ry), Dodecenoyl-CoA delta-isomerase (Dci) and Sod 24. Similarly, our Ribo-Tag experiment as well as proteomic measurement in C.elegans showed ~30 peroxisomal proteins or genes have decreased abundance with aging 13, 14. Studies in cultured cells also suggest that catalase is increasingly excluded from the peroxisomes as the cells enter senescence, resulting in elevated oxidative stress. Moreover, one of the most important peroxisome biogenesis proteins, Pex5, an import receptor for peroxisomal matrix proteins containing PTS1 signal, showed decreased import activity during aging 6. However, a causal relationship of peroxisome dysfunction with longevity has not been well established. Female flies with non-lethal mutations in Pex1 or Pex13 had reduced peroxisome proliferation, increased stress tolerance to paraquat and increased lifespan. In contrast, in yeast cells, Pex5 knockout mutants show strong reduction in chronological lifespan. Further experiments are needed to decipher peroxisome’s role on lifespan and aging.

Altogether, the evidence suggests that peroxisomal function is compromised in aging, which contribute to the disrupted redox homeostasis and contribute to development of inflammaging. To avoid age-related diseases with the accumulation of defective peroxisomes, it might be crucial to maintain peroxisomal homeostasis during aging.

3.5 References

1. Fujiki, Y. et al. Recent insights into peroxisome biogenesis and associated diseases. Journal of Cell Science 133, jcs236943 (2020).

2. Pridie, C., Ueda, K. & Simmonds, A.J. Rosy Beginnings: Studying Peroxisomes in Drosophila. Frontiers in Cell and Developmental Biology 8 (2020).

3. Mast, F.D., Rachubinski, R.A. & Aitchison, J.D. Peroxisome prognostications: Exploring the birth, life, and death of an organelle. J Cell Biol 219 (2020).

4. Di Cara, F., Rachubinski, R.A. & Simmonds, A.J. Distinct Roles for Peroxisomal Targeting Signal Receptors Pex5 and Pex7 in Drosophila. Genetics 211, 141-149 (2019).

5. Baron, M.N., Klinger, C.M., Rachubinski, R.A. & Simmonds, A.J. A Systematic Cell- Based Analysis of Localization of Predicted Drosophila Peroxisomal Proteins. Traffic 17, 536-553 (2016).

6. Huang, K. et al. Impaired peroxisomal import in Drosophila oenocytes causes cardiac dysfunction by inducing upd3 as a peroxikine. Nat Commun 11, 2943 (2020).

7. Di Cara, F., Sheshachalam, A., Braverman, N.E., Rachubinski, R.A. & Simmonds, A.J. Peroxisome-Mediated Metabolism Is Required for Immune Response to Microbial Infection. Immunity 47, 93-106 e107 (2017).

8. Zou, Y. et al. Plasticity of ether lipids promotes ferroptosis susceptibility and evasion. Nature 585, 603-608 (2020).

9. Weir, H.J. et al. Dietary Restriction and AMPK Increase Lifespan via Mitochondrial Network and Peroxisome Remodeling. Cell Metab 26, 884-896 e885 (2017).

10. Cipolla, C.M. & Lodhi, I.J. Peroxisomal Dysfunction in Age-Related Diseases. Trends Endocrinol Metab 28, 297-308 (2017).

11. Fransen, M., Nordgren, M., Wang, B. & Apanasets, O. Role of peroxisomes in ROS/RNS-metabolism: implications for human disease. Biochim Biophys Acta 1822, 1363-1373 (2012).

12. Kou, J. et al. Peroxisomal alterations in Alzheimer's disease. Acta Neuropathol 122, 271- 283 (2011).

13. Huang, K. et al. RiboTag translatomic profiling of Drosophila oenocytes under aging and induced oxidative stress. BMC Genomics 20, 50 (2019).

14. Narayan, V. et al. Deep Proteome Analysis Identifies Age-Related Processes in C. elegans. Cell Syst 3, 144-159 (2016).

15. Didion, S.P. Cellular and Oxidative Mechanisms Associated with Interleukin-6 Signaling in the Vasculature. Int J Mol Sci 18 (2017).

16. Wu, J., Mei, C., Vlassara, H., Striker, G.E. & Zheng, F. Oxidative stress-induced JNK activation contributes to proinflammatory phenotype of aging diabetic mesangial cells. Am J Physiol Renal Physiol 297, F1622-1631 (2009).

17. Yano, T. et al. A novel fluorescent sensor protein for visualization of redox states in the cytoplasm and in peroxisomes. Mol Cell Biol 30, 3758-3766 (2010).

18. Ivashchenko, O. et al. Intraperoxisomal redox balance in mammalian cells: oxidative stress and interorganellar cross-talk. Mol Biol Cell 22, 1440-1451 (2011).

19. Shai, N., Schuldiner, M. & Zalckvar, E. No peroxisome is an island - Peroxisome contact sites. Biochim Biophys Acta 1863, 1061-1069 (2016).

20. Dirkx, R. et al. Absence of peroxisomes in mouse hepatocytes causes mitochondrial and ER abnormalities. Hepatology 41, 868-878 (2005).

21. Peeters, A. et al. Mitochondria in peroxisome-deficient hepatocytes exhibit impaired respiration, depleted DNA, and PGC-1alpha independent proliferation. Biochim Biophys Acta 1853, 285-298 (2015).

22. Fransen, M., Nordgren, M., Wang, B., Apanasets, O. & Van Veldhoven, P.P. Aging, age- related diseases and peroxisomes. Subcell Biochem 69, 45-65 (2013).

23. Terlecky, S.R., Terlecky, L.J. & Giordano, C.R. Peroxisomes, oxidative stress, and inflammation. World J Biol Chem 3, 93-97 (2012).

24. Li, J.Q. et al. Bioinformatic prediction of critical genes and pathways involved in longevity in Drosophila melanogaster. Mol Genet Genomics 294, 1463-1475 (2019).

CHAPTER 4. RIBOTAG TRANSLATOMIC PROFILING OF DROSOPHILA OENOCYTES UNDER AGING AND OXIDATIVE STRESS

A paper accepted by BMC Genomics 20, 50 (2019).

https://doi.org/10.1186/s12864-018-5404-4

Kerui Huang 1*, Wenhao Chen2, Fang Zhu3, Patrick Wai-Lun Li4, Pankaj Kapahi4 & Hua Bai 1*

From the 1Department of Genetics, Development, and Cell Biology, Iowa State University,

Ames, Iowa, 50011, USA

From the 2Department of Electrical and Computer Engineering, Iowa State University, Ames,

Iowa, 50011, USA

From the 3Department of Entomology, Pennsylvania State University, University Park,

Pennsylvania, 16802, USA

From the 4Buck Institute for Research on Aging, Novato, California, 94945, USA

* To whom correspondence should be addressed: Kerui Huang: Department of Genetics,

Development, and Cell Biology, Iowa State University, Ames, Iowa 50011; [email protected];

Hua Bai: Department of Genetics, Development, and Cell Biology, Iowa State University, Ames,

Iowa 50011; [email protected]

Keywords: oenocytes, fat body, liver, aging, ribosomal profiling, peroxisome, fatty acid beta- oxidation, Ras/MAPK signalling

4.1 Abstract

Background: Aging is accompanied with loss of tissue homeostasis and accumulation of cellular damages. As one of the important metabolic centers, liver shows age-related dysregulation of lipid metabolism, impaired detoxification pathway, increased inflammation and oxidative stress response. However, the mechanisms for these age-related changes still remain unclear. In the fruit fly, Drosophila melanogaster, liver-like functions are controlled by two distinct tissues, fat body and oenocytes. Compared to fat body, little is known about how oenocytes age and what are their roles in aging regulation. To address these questions, we used cell-type-specific ribosome profiling (RiboTag) to study the impacts of aging and oxidative stress on oenocyte translatome in Drosophila.

Results: We show that aging and oxidant paraquat significantly increased the levels of reactive oxygen species (ROS) in adult oenocytes of Drosophila, and aged oenocytes exhibited reduced sensitivity to paraquat treatment. Through RiboTag sequencing, we identified 3324 and 949 differentially expressed genes in oenocytes under aging and paraquat treatment, respectively.

Aging and paraquat exhibit both shared and distinct regulations on oenocyte translatome. Among all age-regulated genes, mitochondrial, proteasome, fatty acid metabolism, and cytochrome P450 pathways were down-regulated, whereas DNA replication and glutathione metabolic pathways were up-regulated. In addition, most of the peroxisomal genes were down-regulated in aged oenocytes, including genes involved in peroxisomal biogenesis factors and fatty acid beta- oxidation. Many age-related transcriptional changes in oenocytes are similar to aging liver, such as up-regulation of Ras/MAPK signaling pathway and down-regulation of peroxisome and fatty acid metabolism. Furthermore, oenocytes highly expressed genes involving in liver-like processes (e.g., ketogenesis). 21

Conclusions: Our oenocyte-specific translatome analysis identified many genes and pathways that are shared between Drosophila oenocytes and mammalian liver, highlighting the molecular and functional similarities between the two tissues. Many of these genes were altered in both oenocytes and liver during aging. Thus, our translatome analysis provide important genomic resource for future dissection of oenocyte function and its role in lipid metabolism, stress response and aging regulation.

4.2 Background

Aging is the major risk factor for many chronic diseases 1. The prevalence of liver diseases, such as non-alcoholic fatty liver disease (NAFLD), increase dramatically in the elderly

2, 3. It is known that aging is associated with alterations of hepatic structure, physiology and function 4. For example, aged liver shows reduced blood flow, loss of regenerative capacity, decreases in detoxification and microsomal proteins synthesis, increases in polyploidy, oxidative stress and mitochondrial damage 5. Additionally, the metabolism for low-density lipoprotein cholesterol decreases by 35% 3. Age-related increases in neutral fat levels and high-density lipoprotein cholesterol predispose aged liver to NAFLD and other liver diseases. Accumulated evidence suggests that age-related decline of liver function can be attributed to increased ROS production, DNA damage, activation of p300-C/EBP-dependent neutral fat synthesis 6, decreases in autophagy, increases in inflammatory responses 7, 8, and activation of nuclear factor-κB (NF-

κB) pathway 4, 9. Despite the genetic and functional analysis of liver aging and liver diseases, only a few studies have looked at the global transcriptional changes during liver aging 10-12.

Similar to mammals, the fruit fly (Drosophila melanogaster, hereafter as Drosophila) also shows age-dependent decline of tissue function and loss of homeostasis (reviewed in 13). In

Drosophila, liver-like functions are shared by two distinct tissues, fat body and oenocytes 14. Fat 22 body is the main tissue for energy storage in insects, and it plays a key role in metabolism, nutrition sensing, growth and immunity (reviewed in 15). Fat body has also been implicated in the regulation of organismal aging 16. Many longevity pathways act on fat body to control lifespan 17-

19. Compared to fat body, little is known about how oenocytes age and what is the role of oenocytes in aging regulation. Oenocytes are specialized hepatocyte-like cells responsible for energy metabolism, biosynthesis of cuticular hydrocarbon and pheromone (14, 20, reviewed in 21,

22). Oenocytes coordinate with fat body in mobilizing lipid storage upon nutrient deprivation 14,

23, 24. Recent studies in the yellow fever mosquito Aedes aegypti showed that pupal oenocytes highly express cytochrome P450 genes, suggesting an important role of oenocytes in detoxification 25. Despite its roles in lipid metabolism and wax production, we know very little about oenocyte’s other physiological functions, including its role in the regulation of aging and longevity. It is known that aging oenocytes undergo dramatic morphological changes (e.g., increases in cell size and pigmented granules 26) and exhibit dysregulation of mitochondrial chaperone Hsp22 27. However, transcriptional characterization of oenocyte aging has not been previously performed.

Here, we utilized RiboTag technique 28 to profile the genome-wide changes in ribosome- associated transcripts (translatome) during oenocyte aging in Drosophila. We show that aging and paraquat (PQ) exhibit common and distinct regulation on adult oenocyte translatome. Gene ontology (GO) and gene set enrichment analysis (GSEA) revealed that ribosome, proteasome, peroxisome, xenobiotic metabolism, fatty acid metabolism, and DNA replication pathways were altered under aging and oxidative stress. Comparing tissue-specific transcriptomes further revealed that oenocytes were enriched with genes involved in liver-like functions (e.g., ketogenesis). Aging oenocytes also shared many molecular signatures with aging liver. Taken 23 together, our translatome analysis revealed a conserved molecular mechanism underlying oenocyte and liver aging. Our study will offer new opportunities for future dissection of novel roles of oenocytes in lipid metabolism, stress response, and aging control.

4.3 Results

Characterization of age-related changes in ROS production in Drosophila oenocytes

In Drosophila, larval and adult oenocytes exhibit distinct morphological characteristics

21. Larval oenocytes are clustering along the lateral body wall 14, while adult oenocytes (used in the present study) appear as segmental dorsal stripes and ventral clusters nearby the abdominal cuticle (Fig. 1A). As oxidative stress is commonly observed in aging tissue, we first examined the age-related changes in ROS production in adult oenocytes. As shown in Figs. 1B&1C, both aging and PQ (an oxidative stress inducer) significantly increased ROS levels in adult oenocytes.

Increases in cell and nuclear sizes were also seen in aged oenocytes (Fig. 1B, Fig. S1). In the present study, oenocytes were dissected from two ages, 10 days (young) and 30 days (middle age). Middle age was used because many epigenetic and transcriptional changes have been previously observed in the midlife 29-31. Since elevated ROS levels were already apparent at middle age, a comparison between young and middle age will allow us to capture the early-onset age-related changes in adult oenocytes. Additionally, we noticed that young oenocytes showed much higher induction of ROS under PQ treatment than the oenocytes from middle age (Fig.

1C), suggesting the response to oxidative stress was altered in aged oenocytes.

Oenocyte-specific translatomic profiling through RiboTag sequencing

Besides their roles in metabolic homeostasis, hydrocarbon and pheromone production

(reviewed in 21), the role of oenocytes in aging regulation has not been carefully examined. 24

Characterization of age-related transcriptional changes in oenocytes is an important step toward our understanding of oenocyte aging. To date, only a few oenocyte transcriptome analyses have been reported 23, 25. Most of these studies used dissected oenocytes, which often have issues with tissue cross-contamination. To overcome this issue, we performed an oenocyte-specific RiboTag analysis. In the analysis, oenocyte-specific driver PromE-Gal4 was used to drive the expression of FLAG-tagged RpL13A. According to RNA-seq database (at Flybase.org) and a recent ribosomal proteome analysis 32, RpL13A is one of the highly and ubiquitously expressed components in Drosophila large ribosomal subunit. Our experimental design facilitates the enrichment of oenocyte-specific ribosome-associated mRNAs and translatomic profiling (Fig.

2A). To verify the efficiency and specificity of our RiboTag profiling, we performed a qRT-PCR analysis to measure the expression of Desaturase 1 (Desat1). Desat1 is a transmembrane fatty acid desaturase and its E isoform (desat1-E) was known to be specifically expressed in female oenocytes 20. We found that the expression of desat1-E was much higher in anti-FLAG immunoprecipitated sample (oenocytes) compared to the input (whole body), suggesting that our

RiboTag approach can effectively detect the gene expression from adult oenocytes (Fig. 2B).

To confirm the specificity of the RiboTag analysis, we measured the expression of a brain-specific gene, insulin-like peptide 2 (Dilp2), and found that Dilp2 expression in oenocyte

RiboTag samples was very low compared to the head samples (Fig. 2C). Thus our RiboTag analysis has very little contamination from other tissues (such as brain). We also set up two control experiments to test the specificity of the reagents used in our pull-down assay: 1)

Immunoprecipitation of PromE>RpL13A-FLAG expressing females using only protein G magnetic beads without adding FLAG antibody. 2) Immunoprecipitation of PromE-gal4 flies using both Protein G magnetic beads and FLAG antibody. No detectable RNAs were pulled 25 down from the two control groups, suggesting there is none or very little non-specific binding from FLAG antibodies or protein G magnetic beads during the immunoprecipitation (Fig. 2D).

Notably, the total RNA pulled down from aged samples were less than those from young oenocytes. This is probably due to age-related decreases in general transcription and translation, because the PromE-gal4 driver activity remained the same during aging (Fig. S1). Due to the variation in RNA quantity among different samples, we used equal amount of RNAs for all library construction. To examine age- and stress-related transcriptional changes in Drosophila oenocytes, we performed RiboTag sequencing on four different experimental groups: H2O-

Young, PQ-Young, H2O-Aged, PQ-Aged (see Methods for more details). Female flies were used in the present study, because previous studies showed that PromE-gal4 drives expression in testis

(additional to oenocytes) in male flies 20.

Differential gene expression (DGE) analysis reveals common and distinct transcriptional regulation by aging and oxidative stress

Using Illumina sequencing (HiSeq 3000, single-end, a read length of 50 base pair), we obtained a total of 402 million reads from 12 library samples (about 11.6X coverage per library).

On average, 82.43% of unique reads were mapped to annotated Drosophila reference genome.

To visualize how gene expression varies under different conditions, we performed principal component analysis (PCA) on the fragments per kilobase million (FPKM) reads. The first component accounts for 50% of the variance and the second component accounts for 9% of variance (Fig. 3A). The PCA analysis showed that three replicates of each condition cluster together, except for one of the H2O-young samples. Two age groups were also well separated.

Interestingly, there was a reduced variation between H2O and paraquat treatment in aged oenocytes compared to the young ones (Fig. 3A). 26

DGE analysis was performed using Cufflinks and Cuffdiff tools (fold change ≥ 2, FDR adjusted p-value ≤ 0.05, only protein-coding genes were analyzed). To compare the impacts of aging and oxidative stress on transcriptional changes in adult oenocytes, we first performed correlation analysis using Log-transformed FPKM reads from all four groups. The coefficient of determination (R2) was 0.861 between H2O-aged and H2O-young groups (Fig. 3B), 0.926 between H2O-young and PQ-young (Fig. 3C), 0.948 between PQ-aged and H2O-aged (Fig. 3D).

Aging induced a bigger transcriptional shift compared to paraquat treatment. Although the change of R2 was relatively small, the total number of age-regulated genes was much higher than that under paraquat treatment (Figs. 3E&3F). Thus, both PCA and correlation analyses suggest that aging and paraquat exhibit different impacts on oenocyte translatome.

DGE analysis identified 3324 genes that were differentially expressed during oenocyte aging (1092 up-regulated and 2232 down-regulated), while 949 genes (198 up-regulated and 751 down-regulated) were regulated by paraquat treatment at young ages (Figs. 3E&3F). About 706

DEGs were commonly regulated by aging and paraquat (127 up-regulated and 579 down- regulated). The genes commonly up-regulated by aging and PQ were involved in DNA metabolism, DNA repair and recombination (Fig. 3E), while those commonly down-regulated genes were involved in immune response and fatty acid elongation (Fig. 3F).

Besides common transcriptional regulation between aging and oxidative stress, many genes were differentially regulated between the two processes. A total of 2618 genes (965 up- regulated and 1653 down-regulated) were only regulated by aging (Figs. 3E&3F). Genes up- regulated in aged oenocytes were enriched in several Gene ontology (GO) terms, including developmental process, glutathione metabolism and metabolism of xenobiotics. The down- regulated genes are enriched in peroxisome, ribosome, proteasome, oxidative phosphorylation, 27 and fatty acid metabolism. About 243 genes (71 up-regulated and 172 down-regulated) were only regulated by paraquat treatment at young ages. These genes are enriched for biological processes like response to bacterium, response to other organisms, and phototransduction.

It is known that stress tolerance declines with age 33, which can be caused by impaired transcriptional regulation of stress signaling pathways 34. Our transcriptome analysis showed that the total number of PQ-regulated genes decreased with aging (Figs. 3G&3H). About 949 genes were differentially expressed under paraquat treatment at young ages (198 up-regulated and 751 down-regulated), while only 385 genes were differentially expressed at middle ages (213 up- regulated and 172 down-regulated). In addition, paraquat treatment targeted a different set of the biological processes and signaling pathways between young and middle ages (Figs. 3G&3H). In young oenocytes, paraquat up-regulated pathways like response to DNA metabolism and DNA recombination, while down-regulating immune response, fatty acid biosynthesis, and fatty acid elongation. In contrast, different sets of pathways were up regulated by paraquat at middle ages, such as pheromone binding and cation channel activity. No pathway was found enriched for genes downregulated by paraquat at middle ages.

Next, we performed hierarchical clustering analysis and identified 11 distinct clusters among four groups (Fig. 3I). Among 11 clusters, cluster 3 and 5 are two major clusters. Cluster 3 includes genes that were up regulated in aged oenocytes compared to young ones. Gene ontology analysis showed that cluster 3 was enriched with genes in endocytosis, hippo, JAK-STAT, fanconi anemia pathway, phosphatidylinositol signaling, and DNA replication (Fig. 3J). Cluster

5 consisted of genes downregulated by aging, and was enriched in fatty acid metabolism, oxidative phosphorylation, and proteasome (Fig. 3K). Taken together, our RiboTag analysis 28 revealed common and distinct transcriptional changes under aging and oxidative stress in adult oenocytes.

Gene set enrichment analysis (GSEA) reveals key pathways that are up- and down- regulated in aged oenocytes

To further characterize oenocyte-specific signaling pathways that were regulated by aging and oxidative stress, we performed gene set enrichment analysis (GSEA) using a collection of pre-defined gene sets retrieved from Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Through GSEA, we discovered five pathways within which genes were up-regulated with age (FDR q-value<0.05) (Figs. 4A&4C). They are mismatch repair, DNA replication, base excision repair, nucleotide excision repair, and fanconi anemia pathways. These pathways were tightly related to the cellular responses to DNA replication stress, suggesting a possible increased

DNA damage during oenocyte aging. Several key players in DNA replication stress response were up-regulated aged oenocytes, such as ATR/mei-41 (ATM- and Rad3-related kinase) and

TopBP1/mus101 (DNA topoisomerase 2-binding protein 1).

On the other hand, GSEA analysis revealed 14 pathways within which most of genes were significantly downregulated during aging, such as oxidative phosphorylation, ribosome, proteasome, and peroxisome (Figs. 4B, 4D, 4E, 4F). These results suggest that the functions of many key cellular organelles/components (e.g., mitochondria and peroxisome) were impaired in aged oenocytes. In aged oenocytes, we found that the key components of all five complexes in mitochondrial electron transport chain were down-regulated, such as NADH dehydrogenase subunits (e.g., ND-13, ND-15, ND-30, ND-B8), succinate dehydrogenase (e.g., SdhC, SdhD), cytochrome bc1 complex (e.g., Cyt-c1, UQCR-14, UQCR-C2, UQCR-Q, ox), cytochrome c oxidase subunits (e.g., COX4, COX5A, COX5B), and ATP synthase subunits (e.g., ATPsynB, 29

ATPsynD, ATPsynF, ATPsynO). Interestingly, we found that aging downregulated many mitochondrial ribosomal subunit genes (44 out of 72 annotated mitochondrial ribosomal proteins) (Fig. 4I). Lastly, we observed an age-related decrease in the expression proteasome subunit genes. These include 20S protein subunits (e.g., Prosalpha2, Prosalpha3, Prosbeta1,

Prosbeta2, Prosbeta3), and 19S regulatory cap subunits (e.g., Rpn1, Rpn11, Rpn12, Rpt1, Rpt2,

Rpt3).

Reduced xenobiotic metabolism is one of the hallmarks of liver aging 35. Xenobiotics metabolism (or detoxification) consists of three major phases 36. The Phase I and II enzymes represent the most abundant classes of detoxification system, including cytochrome P450 (CYPs) and glutathione S-transferases (GSTs). Interestingly, our GSEA analysis revealed distinct expression patterns for these two detoxification enzyme families. We found that almost all GSTs in Delta class were up regulated under aging, while other classes showed mixed expression patterns (Fig. 4J). The microsomal glutathione S-transferase (Mgstl), one of the highly enriched oenocyte genes, was significantly down-regulated during oenocyte aging.

On the other hand, most of the cytochrome P450 genes were downregulated in aged oenocytes (Figs. 4H&4K). Many of the downregulated CYPs have been previously linked to insecticide resistance or xenobiotic metabolism, such as Cyp6a8, Cyp6a21, Cyp308a1, Cyp12a4,

Cyp6a2, Cyp6w1, and Cyp313a1. Besides metabolizing exogenous chemicals, several CYPs catalyze endogenous metabolites and play key roles steroid hormone biosynthesis and fatty acid metabolism. For example, Cyp4g1 is a key CYP gene involved in cuticular hydrocarbon biosynthesis 37 and triglyceride metabolism 14. The expression of Cyp4g1 was decreased in aged oenocytes. About eight CYPs (also known as the Halloween genes) in Drosophila that are known to regulate ecdysteroid metabolism. Two of them, Cyp306a1 (Phantom) and Cyp315a1 30

(Shadow), were highly expressed in oenocytes (32-fold and 12.5-fold enriched respectively).

During oenocyte aging, Phantom was down-regulated, whereas Shadow was up-regulated.

Peroxisome pathway is transcriptionally deregulated in aged oenocytes

Recent studies suggest that peroxisome protein import is impaired during aging 38. Our

GSEA analysis revealed that except for Pex1 (up-regulated), most of the genes involved in peroxisome biogenesis (also called peroxin, PEX) were down-regulated in aged oenocytes (Figs.

4F&5A). Out of 16 peroxin genes, five of them showed significant down-regulation during aging

(fold change ≥ 2, FDR adjusted p-value ≤ 0.05). They are matrix enzyme import components

(Pex5, Pex12), receptor recycling (Pex6), and membrane assembly components (Pex16, Pex19)

(Figs. 5A&5B). In addition, most of the PEX genes were also downregulated by paraquat treatment, but to a less extend comparing to aging (Fig. 5C).

Besides peroxisome biogenesis, genes involved in other peroxisomal functions were also downregulated during oenocyte aging (Figs. 5D&5E). These functions include fatty acid beta- oxidation, ether phospholipid biosynthesis, amino acid metabolism, ROS metabolism, purine metabolism, and retinol metabolism. Several beta-oxidation genes showed significantly decreased expression, including sterol carrier protein X-related thiolases (ScpX and CG17597), enoyl-CoA hydratase (ECH/CG9577), carnitine O-acetyl-transferases (CRAT and CG5122), and nudix hydrolases (CG10194, CG10195, CG18094) (Fig. 5D). Consistently, hepatocyte nuclear factor 4 (HNF4), the major regulator for mitochondrial and peroxisomal beta-oxidation, was significantly downregulated under aging and paraquat. On the other hand, a few other beta- oxidative genes were up-regulated in aged oenocyte, such as ABC transporters (Pmp70,

CG2316) that are responsible for transporting long-chain fatty acids into peroxisome, delta3- delta2-enoyl-CoA isomerase (PECI/CG13890), carnitine O-octanoyltransferase 31

(CROT/CG12428). Acyl-CoA oxidases (Acox) that are involved in the first step of beta-oxidation showed mixed expression pattern (Fig. 5D).

Consistent with increased ROS production during oenocyte aging, most of the genes regulating peroxisomal ROS metabolism were downregulated in aged oenocytes, such as catalase (Cat), superoxide dismutase 1 (SOD1), peroxiredoxin 5 (Prx5). Although the majority of ether phospholipid synthesis genes (e.g., fatty acyl-CoA reductase, FAR) were down- regulated, there are a few genes that showed up-regulation during aging, such as dihydroxyacetone phosphate acyltransferase (DHAPAT or Dhap-at), the key enzyme for the production of acyl-DHAP (the obligate precursor of ether lipids). Additionally, three aldehyde oxidases (Aox1, Aox2, Aox4) in purine metabolism were up-regulated (Fig. 5E).

To verify our RiboTag sequencing results, we performed qRT-PCR analysis on three selected peroxisome genes, Pex5, Pex19, and Cat. Consistent with RNA-Seq results, qRT-PCR showed that all three genes were significantly down-regulated in aged oenocytes (Figs. 5F-5H).

Ketogenesis, fatty acid elongation, and peroxisome pathways are enriched in both oenocytes and liver

Fat body, but not oenocytes, is a long-established tissue model to study liver- and adipose-like functions in Drosophila 39. Although hepatocyte-like functions (e.g., steatosis) have been previously observed in oenocytes 14, it remains unclear whether fat body and oenocytes each perform different aspects of liver-like functions in Drosophila. To address this question, we performed a transcriptome comparison among Drosophila oenocytes, fat body, and human liver

(Table S1). 32

We first identified genes that were enriched in adult oenocytes by comparing our oenocyte RiboTag data (H2O-Young group) with previously published whole body transcriptome data. Fat body-enriched genes were identified based on Drosophila tissue transcriptome database, FlyAtlas 40, 41. The genes with more than 5-fold higher expression in oenocytes (or fat body) comparing to whole body are defined as oenocyte-enriched (or fat body- enriched) genes. A total of 423 oenocyte-enriched genes and 544 fat body-enriched genes were identified through tissue transcriptome comparison. A recent study showed that Drosophila oenocytes express many liver-like lipid metabolic genes/pathways [14]. About 15 of these genes were also found enriched in our oenocyte translatome analysis (e.g., Cpr, Cat, spidey, FarO).

About 463 human liver-enriched genes were retrieved from the Human Protein Atlas 42.

Interestingly, there was very little overlap between oenocyte-enriched and fat body- enriched genes, suggesting that adult fat body and oenocytes may regulate distinct biological processes (Additional file 1: Fig. S3A). Gene ontology analysis revealed that fat body mainly expressed genes in carboxylic acid and amino acid metabolism, whereas oenocytes were enriched with genes in pathways like fatty acid biosynthesis, fatty acid elongation, proteasome- mediated protein catabolism, xenobiotic metabolism, ketone body metabolism, and peroxisome.

Furthermore, we found that two innate immunity pathways, Toll and Imd (Immune deficiency), were differentially enriched in fat body and oenocytes (Additional file 1: Fig. S3B). Several genes in Imd pathway (PRGP-LC, PRGP-LB, Dredd) were enriched in oenocytes, whereas fat body were enriched with genes in Toll pathway (Tl, PGRP-SA, GNBP3, modSP) (Additional file

1: Fig. S3B). Additionally, most of the antimicrobial peptides (AMPs) were enriched in oenocytes, but not in fat body (Additional file 1: Fig. S3B). 33

When comparing oenocyte and liver transcriptomes, we found that several pathways were specifically enriched in both liver and oenocytes, such as long-chain fatty acid metabolism, peroxisome, and xenobiotic metabolism. A close look at the enriched genes shared between oenocytes and liver revealed that HMG-CoA synthase (Hmgs in fly and HMGCS1/2 in human), the key enzyme involved in ketogenesis and production of β-hydroxy-β-methylglutaryl-CoA

(HMG-CoA), was highly expressed in both oenocytes and liver (Figs. 6A&6B). Additionally, two other ketogenesis genes were also highly expressed in both oenocytes and liver. They are

HMG-CoA lyase (CG10399 in fly and HMGCL in human) and D-β-hydroxybutyrate dehydrogenase (shroud in fly and BDH1 in human) (Figs. 6A&6B). Ketogenesis is primarily activated in mammalian liver, especially during fasting. These results suggest that oenocyte may be the fly tissue regulating ketogenesis like mammalian liver.

Microsomal fatty acid elongation and the synthesis of very-long-chain fatty acid

(VLCFA) were also enriched in both oenocytes and liver (Figs. 6C&6D). Liver and oenocytes were enriched for very-long-chain 3-ketoacyl-CoA synthase (ELOVL2 in human and CG18609 in fly), which catalyzes the first step of VLCFA synthesis in smooth endoplasmic reticulum

(smooth ER). Oenocytes also showed high expression of three other key enzymes in this process

(spidey, CG6746, Sc2) (Figs. 6C&6D). The enrichment of fatty acid elongation factors in oenocytes aligns well with previously known oenocyte function in the biosynthesis of VLCFA and hydrocarbons 20, 24. Notably, several key players involved in the production of cuticular hydrocarbons were enriched in adult oenocytes, including cytochrome P450 Cyp4g1 (3.4-fold) and its obligatory redox partner, cytochrome P450 reductase Cpr (5.7-fold), as well as five peroxisome-localized fatty acyl-CoA reductases (FAR) (FarO, CG13091, CG14893, CG17562, 34 and CG4020). In particular, FarO was 123-fold enriched in oenocytes, while CG13091 was 243- fold enriched (Additional file 1: Fig. S4).

Additionally, many oenocyte- and liver-enriched genes belong to peroxisome pathway, especially peroxisomal beta-oxidation (CG17597, CG9577 in oenocytes, ACOX2, BAAT,

EHHADH, ACAA1, SLC27A2, ACSL1, PECR in liver) (Additional file 1: Fig. S4). Genes involved in ROS metabolism (e.g., Cat, Sod1) were also enriched in both oenocytes and liver

(Additional file 1: Fig. S4). Lastly, we found that fibroblast growth factor 21 (bnl in fly and

FGF21 in human), a key hormonal factor that regulates glucose homeostasis, was enriched in both oenocytes and liver. Taken together, our translatome analysis suggests that oenocytes and fat body regulate distinct processes, and oenocytes may participate several liver-like functions

(e.g., ketogenesis, and long-chain fatty acid metabolism).

Conservation in age-related transcriptional changes between oenocytes and liver

Since our analyses suggest that Drosophila oenocytes may perform liver-like functions, we wonder if oenocyte and liver exhibit similar transcriptional changes during aging. To test this, we compared age-related transcriptomic profiles between Drosophila oenocytes and mouse liver

10. We first searched for fly orthologues of mouse liver genes using Drosophila Integrative

Ortholog Prediction Tool (DIOPT) 43. Out of 1052 protein-coding genes that are differentially expressed in aging mouse liver, 735 of them have putative orthologues in Drosophila genome, corresponding to 881 Drosophila genes (Fig. 7A). About 30% of these Drosophila orthologues

(252 out of 881) also showed differential expression during oenocyte aging, suggesting a large conservation between liver and oenocyte aging. 35

Gene ontology analysis showed that several key biological processes were altered in aged liver, including immune response, apoptosis, peroxisome, bile acid biosynthesis, and fatty acid metabolism. Among these biological processes, peroxisome and fatty acid metabolism are shared between liver and oenocyte aging (Fig. 7A). Next, we took a close look at the pathways that contain same orthologues between fly and mouse. Genes up-regulated in both aged oenocytes and liver were enriched in pathways like Mitogen-activated protein kinase (MAPK), Ras signaling, NF-κB, and JAK-STAT (Fig. 7B), while down-regulated genes were found in peroxisome, fatty acid metabolism, and oxidative phosphorylation pathways (Fig. 7C). Using

STRING protein network analysis, we found that large number of Ras/MAPK signaling components were up-regulated under both oenocyte and liver aging (Figs. 7D&7E), suggesting that age-dependent dysregulation of these pathways are conserved between fly and mammal.

Lastly, we examined age-related transcriptomic changes between oenocytes and several other fly tissues, such as fat body, midgut, and heart. The age-related transcriptional profiles in these fly tissues were obtained from recent genomic studies 44-46. Pathway analysis (using

STRING) on these tissue transcriptomes revealed a tissue-specific transcriptional profiles during fly aging (Fig. 7F). Each tissue has its own and unique age-regulated biological processes and pathways (Fig. 7G). For example, genes that were differentially expressed in aged oenocytes are enriched for proteasome and ribosome-related functions, while aged fat body showed transcriptional changes in aminoglycan metabolism, chitin metabolism, and detoxification. In aging heart, immune response, glycolysis and gluconeogenesis were enriched. And ion transport,

DNA replication, and fatty acid degradation were altered in aging midgut (Fig. 7G). Taken together, aged oenocytes share similar transcriptional profiles with aging liver, while they also exhibit unique features compared to other fly tissues. 36

4.4 Discussion

Oenocytes are poorly studied yet important cells in insects 21, 22. Although previous studies show that oenocytes play a crucial role in lipid metabolism (e.g., synthesis of cuticular hydrocarbon and pheromone), many other oenocyte-regulated physiological functions remain to be determined. Among the uncharacterized functions, we know very little about oenocyte aging and the role of oenocytes in aging regulation. To address these issues, we performed RiboTag sequencing to characterize Drosophila oenocyte translatome under aging and oxidative stress.

We show that both aging and paraquat up-regulated DNA repair pathway, while down-regulating immune response and fatty acid elongation. In addition, aged oenocytes were associated with impaired peroxisome, mitochondrial, proteasome, and cytochrome P450 pathways. Our RiboTag sequencing also revealed many shared tissue-specific pathways and age-related transcriptional changes between fly oenocytes and mammalian liver, highlighting evolutionarily conserved mechanisms underlying oenocyte and liver aging and potential functional homologies between the two tissues.

Oenocyte-specific expressed genes are involved in insect-specific and conserved liver-like functions

Previous functional and histological analyses showed that oenocytes contain large amounts of smooth ER and acidophilic cytoplasm (high protein and lipid contents) 47, 48, which is consistent with their roles in lipid synthesis and processing, especially the production of VLCFA and hydrocarbon 20, 24, 49, 50. Interestingly, Drosophila oenocytes uptake and process fatty acids that are released from the storage tissue fat body during food deprivation 14. The coordination between fat body and oenocytes in mobilizing lipid storage during fasting is quite similar to the adipose-liver axis in mammals. Besides lipid metabolism, many other oenocyte-associated 37 functions (e.g., detoxification and ecdysteroid biosynthesis) have not yet been thoroughly examined at the molecular level. It is unclear whether some of these functions are also conserved liver-like functions, or they are merely insect-specific roles.

To better understand oenocyte function, we conducted oenocyte-specific translatome profiling in adult Drosophila and identified 423 genes that were highly expressed in oenocytes

(at least 5-fold higher than whole body expression). These genes were enriched in pathways like fatty acid elongation, proteasome-mediated protein catabolism, xenobiotic metabolism, ketogenesis, and peroxisome pathways. There was only a small overlap between oenocyte- enriched and fat body-enriched genes, suggesting that the two tissues regulate distinct functions in Drosophila. Comparing to the genes and pathways enriched in human liver, we found that oenocytes shared several biological processes with liver, such as ketogenesis, peroxisomal beta- oxidation, ROS metabolism, long-chain fatty acid metabolism, and xenobiotic metabolism. This is consistent with a previous study showing that Drosophila oenocytes expressed high levels of lipid metabolic genes similar to those of mammalian liver 14. One enriched pathway in

Drosophila oenocytes that was not observed in the previous study is the ketogenesis pathway. It is well-known that ketone bodies (acetoacetate, β-hydroxybutyrate, and acetone) are primarily produced by liver when glucose is not available as fuel source 51. Ketogenesis in insects, however, is not well studied. Ketone bodies have been detected in hemolymph, fat body, and thoracic muscle of adult desert locust and cockroach 52-54. It is speculated that ketone bodies are produced in fat body according to the ex vivo tissue culture assay in locust 53. However, fat body

(along with many other tissues) can also oxidize ketone bodies, which is quite different from mammals where the ketogenesis tissue liver cannot oxidize ketone 53. It might be possible that in previous ex vivo tissue culture studies, the ketone production came from a contaminated tissue 38

(like oenocytes), rather than fat body. Based on our oenocyte translatome analysis, most of the ketogenesis genes are highly expressed in oenocytes, but not in fat body. Our data suggest that oenocytes are likely the major ketogenesis tissue. A careful function and genetic analysis, such as cell ablation or tissue-specific gene silencing, will need to be performed to examine whether oenocytes are responsible for ketogenesis in Drosophila and in other insect species.

Insect hydrocarbons serve as important waterproofing components, and species- and sex- specific recognition signals. The biosynthesis of hydrocarbons are involved in fatty acid elongation, desaturation, reduction, and oxidative decarbonylation 55. Our oenocyte translatome analysis revealed an enrichment of genes in microsomal fatty acid elongation system, such as

CG18609, spidey, CG6746, and Sc2. This is consistent with oenocyte’s role in hydrocarbon production and its abundant smooth ER content. In microsomal fatty acid elongation system, spidey (also known as Kar) encodes for the only very-long-chain 3-ketoacyl-CoA reductase in

Drosophila genome, and it has been implicated in oenocyte VLCFA synthesis and waterproof of the trachea system 50, as well as the production of cuticular hydrocarbon, ecdysteroid metabolism, and oenocyte maturation 24, 56. Final two steps of hydrocarbon production in insects are very-long-chain fatty acyl-CoA to aldehydes conversion by FAR and aldehyde oxidative decarbonylation by Cyp4g1 and Cpr 21, 37. Our translatome analysis showed that five different

FARs (including FarO), Cyp4g1, and Cpr are highly expressed in adult oenocytes. The large number of FARs expressed in adult oenocytes suggests that aldehyde-forming FARs may be responsible for the production of a variety of hydrocarbons in oenocytes, and each FAR can catalyze a unique set of very-long-chain fatty acyl-CoA esters that vary in saturation status and chain length. 39

In adult insects (especially in females), ovary is the major tissue for ecdysteroid biosynthesis 57, 58. It remains to be determined whether other adult tissues are also capable to synthesize ecdysteroids. Interestingly, we found two Halloween genes (phantom and shadow) that are highly expressed in adult oenocytes, suggesting that oenocytes may participate in ecdysteroid synthesis in adult females. Our findings are consistent with an early study showing that abdominal oenocytes dissected from Tenebrio molitor larvae can synthesize 20-

Hydroxyecdysone (β-ecdysone) 59. Several recent studies also detected the expression of

Halloween genes in adult tissues other than ovaries, such as brain 60, fat body, muscle, and

Malpighian tubule 61. To functionally verify the role of adult oenocytes in ecdysteroid biosynthesis, direct measurement of ecdysteroid production is needed when Halloween genes are specifically knocked down in oenocytes.

Impaired peroxisome pathway and fatty acid beta-oxidation are the hallmarks of oenocyte aging

Our translatome analysis identified large number of genes (1092 up-regulated and 2232 down-regulated) that were differentially expressed between young and middle ages, suggesting that dramatic cellular and molecular alterations can be observed in oenocytes at the middle age.

Some of these changes are consistent with previous aging transcriptome analysis in Drosophila

30, 31, 62, such as up-regulation of DNA repair and down-regulation of oxidative phosphorylation.

On the other hand, oenocyte aging was specifically associated with the dysregulation of several other pathways, such as down-regulation of peroxisome and fatty acid metabolism pathways.

Peroxisomes are important subcellular organelles that participate in a variety of metabolic pathways, including alpha-oxidation of phytanic acids, beta-oxidation of VLCFA, ether phospholipid synthesis (e.g., plasmalogen biosynthesis), ROS and hydrogen peroxide 40 metabolism, glyoxylate metabolism, catabolism of amino acids and purine 63. There are about 16 peroxisome biogenesis genes (also known as peroxin, or PEX) in Drosophila that are responsible for peroxisome membrane assembly (Pex3, Pex6, Pex9), matrix enzyme import and receptor recycling (Pex5, Pex7, Pex13, Pex14, Pex2, Pex10, Pex12, Pex1, Pex6), and peroxisome proliferation (Pex11) 64. Mutation in many peroxin genes leads to various forms of peroxisome biogenesis disorder (PBD), also known as Zellweger syndrome (ZS) in human 63. Our data revealed that aging and oxidative stress decreased the expression of most of the peroxisome biogenesis and protein import genes, which may lead to reduced peroxisome function, including hydrogen peroxide metabolism. Decreased expression of receptor protein Pex5 and reduced peroxisomal enzyme import were previously observed in aged C. elegans 38 and during human fibroblast senescence 65. Among many key peroxisomal enzymes, the importing of antioxidant catalase was significantly affected during fibroblast senescence, which led to accumulation of hydrogen peroxide and further disruption of peroxisome import 65. Similar to early studies in aging rat liver 66-68, we found that the expression of many peroxisomal antioxidant enzymes (e.g.,

Cat, SOD1, Prx5) decreased in aged oenocytes. The combined dysregulation of peroxisomal gene expression and protein import may attribute to elevated toxic reactive oxygen species, and impaired oenocyte functions. Furthermore, generation of excess peroxisomal ROS could disrupt mitochondria redox balance, leading to mitochondrial dysfunction and tissue aging 69.

Impaired peroxisome biogenesis/protein import during aging not only contributes to reduced antioxidant capacity and elevated ROS levels, but also dysregulation of other peroxisomal functions. Besides ROS metabolism, our translatome analysis revealed that genes involved in peroxisomal beta-oxidation and ether phospholipid were down-regulated under oenocyte aging. This is consistent with previous studies showing that peroxisomal beta-oxidation 41 activity decreased in old mouse liver 70. Peroxisome has been shown to coordinate with mitochondrial fission/fusion pathway to regulate cellular fatty acid oxidation 71, a major metabolic process dysregulated during mouse aging 72. Although the metabolic reactions for fatty acid beta-oxidation are similar in mitochondria and peroxisome, a set of fatty acid substrates can only be processed by peroxisomes, such as VLCFA, pristanic acid, di- and trihydroxycholestanoic acid (DHCA and THCA), long-chain dicarboxylic acids, certain polyunsaturated fatty acids 63, 73. Mutation of peroxisome fatty acid transporter ABCD1 impaired peroxisomal beta-oxidation and caused to accumulation of VLCFAs and neuroinflammation, which is associated with X-link neurodegenerative disease adrenoleukodystrophy (ALD) 74, 75.

Mouse homozygous mutants of ACOX, which catalyzes the first step of peroxisomal beta- oxidation, also showed accumulation of VLCFA and development of microvesicular fatty liver.

Although the expression of two Drosophila ACOX genes were not significantly altered during oenocyte aging, ScpX (peroxisomal thiolase) was significantly downregulated. Mice with ScpX mutation showed defects in peroxisome proliferation, hypolipidemia, motor and peripheral neuropathy, as well as impaired catabolism of methyl-branched fatty acids 76. In addition, reduced peroxisome function can disrupt lipid homeostasis and lipid composition, which could lead to compromised immune response 77, 78.

Conservation between oenocytes and liver aging

The comparison of aging transcriptomes between fly oenocytes and mouse liver revealed many shared pathways between the two tissues. Among these conserved pathways, MAPK and

Ras signaling pathways were significantly up regulated in both aged oenocytes and liver. MAPK signaling is one of the major regulatory pathways involved in stress responses (e.g., oxidative stress). The typical MAPK pathway includes three branches: c-Jun N-terminal kinase (JNK), 42 p38/MAPK, and extracellular signal-regulated kinase (ERK). Previous studies show that all three

MAPK cascades are elevated under aging, probably due to increased oxidative stress 79, 80.

Dysregulated MAPK signaling has been implicated in cancer and neurodegenerative diseases such as Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis (reviewed in

81). In model organisms (e.g., Drosophila and C. elegans), activation of JNK and p38/MAPK extended lifespan and improved tissue functions in late life 82-84. Among many MAPK components identified in our analysis, the activator protein 1 (AP-1) subunit, Drosophila Kay and its mouse orthologue c-Fos, were found significantly induced under aging. Both Kay and c-

Fos are basic leucine zipper transcription factors that mediate MAPK signaling (especially JNK pathway) to regulate cell proliferation, tissue regeneration, stress tolerance 85, 86. Since JNK signaling is the key regulator for the maintenance of tissue homeostasis in response to intrinsic and extrinsic stresses (e.g., UV irradiation, ROS, DNA damage, inflammatory cytokines, infection), the induction of Kay/c-Fos indicates an up-regulation of JNK signaling (and other

MAPK pathways), as well as an elevated cellular stress responses in both aged oenocytes and liver. In addition, Ras small GTPase pathway, the upstream regulator of MAPK kinase cascades, was also up regulated during oenocyte and liver aging. The direct role of Ras signaling pathway in longevity regulation has been previously demonstrated in several model organisms 87-90.

Further studies on Ras/MAPK signaling are needed to advance our understanding on the specific contributions of these pathways in oenocyte and liver aging. Nevertheless, the up regulation of

Ras/MAPK signaling pathways can be used as an important molecular signature and biomarker for oenocyte and liver aging. 43

4.5 Conclusion

Using RiboTag sequencing, we characterized the first oenocyte translatome profiles in

Drosophila. Our analysis uncovered many previously unexplored oenocyte-specific molecular pathways, especially those associated with oxidative stress and aging. Some of these pathways were enriched in both fly oenocytes and mammalian liver, suggesting a functional homolog between the two tissues. We believe that the analysis of oenocyte translatome will contribute significantly to our understanding of oenocyte biology, as well as the molecular mechanisms for its role in stress response and aging regulation.

4.6 Methods

Fly strains, aging and paraquat treatment

Flies are raised in 12h:12 h light:dark cycle at 25 ˚C, 60% relative humidity on agar- based diet with 0.8% cornmeal, 10% sugar, and 2.5% yeast (unless otherwise noted). Fly strains used in the present study include: w*; PromE-Gal4 (also known as Desat1-GAL4.E800)

(Bloomington #65405) 20, PromE-Gal4; UAS-CD8::GFP (a gift from Alex Gould), UAS-

RpL13A-FLAG (a gift from Pankaj Kapahi), To age flies, females were collected two days after eclosion, and twenty females per vial were maintained at 25 °C and transferred to fresh food every 2-3 days. Two ages were tested, young (10-day-old) and middle age (30-day-old). For paraquat treatment, flies were fed on fly food containing 10 mM of paraquat (at the food surface) for 24 hours prior to each assay.

Dihydroethidium (DHE) staining

Young and aged flies were fed on normal food or paraquat (10 mM) for 24 hours prior to the staining with dihydroethidium (Calbiochem, Burlington, MA, USA. Catalog number: 38483- 44

26-0). DHE staining was performed as previously described 91. Briefly, fly abdomen was dissected out (fat body removed) and incubated with 30 µM of DHE in Schneider’s Drosophila

Medium (ThermoFisher Scientific, Catalog number: 21720-024) for 5 minutes in a dark chamber on an orbital shaker. After additional 5 minutes incubation with 1 µg/mL of Hoechst 33342

(ImmunoChemistry Technologies, Bloomington, MN, USA. Catalog number: 639), fly abdomen was mounted with 50% glycerol in PBS. DHE staining was visualized with Olympus BX51WI upright epifluorescence microscopy.

Oenocyte RiboTag

Female progeny from the crosses between PromE-gal4 and UAS-RpL13A-FLAG were collected two days after eclosion. Four different experimental groups were tested: 1). 10-day-old females fed on normal food (H2O-Young); 2). 10-day-old females treated with 10 mM of paraquat for 24 hours (PQ-Young); 3). 30-day-old females fed on normal food (H2O-Aged); 4).

30-day-old females treated with 10 mM of paraquat for 24 hours (PQ-Aged). Three biological replicates (200 females per replicate) were performed for each group. Female flies were used in the present study, because PromE-gal4 drives expression in testis (additional to oenocytes) in male flies 20.

RiboTag was performed following the protocol from McKnight Lab 28. Briefly, flies were first frozen and ground in nitrogen liquid. The fly powder was then further homogenized in a

Dounce tissue grinder containing 5 mL of homogenization buffer (50 mM Tris-HCl, pH 7.4, 100 mM KCl, 12 mM MgCl2, 1 mM DTT, 1% NP-40, 400 units/ml RNAsin RNase inhibitor, 100

µg/ml of cycloheximide, 1 mg/ml heparin, and Protease inhibitors). After centrifuging the homogenate at 10,000 rpm for 10 minutes, the supernatant was first pre-cleaned using

SureBeads™ Protein G Magnetic Beads (Bio Rad, Hercules, CA, USA. Catalog number: 161- 45

4023), and then incubated with 15 µl of anti-FLAG antibody (Sigma-Aldrich, St. Louis, MO,

USA. Catalog number: F1804) for about 19 hours at 4 0C. The antibody/lysate mixture was then incubated with 100 µl of SureBeads for 3 hours at 4 0C. Ribosome-bound RNA was extracted and purified using RNeasy Plus Micro Kit (Qiagen, Hilden, Germany. Catalog number: 74034).

Transcriptome library construction and high-throughput sequencing (RNA-Seq)

RNA-Seq libraries were constructed using 300 ng of total RNA and NEBNext Ultra

Directional RNA Library Prep Kit for Illumina (New England Biolabs (NEB), Ipswich, MA,

USA. Catalog number: E7420). Poly(A) mRNA was isolated using NEBNext Oligo d(T)25 beads and fragmented into 200 nt in size. After first strand and second strand cDNA synthesis, each cDNA library was ligated with a NEBNext adaptor and barcoded with an adaptor-specific index.

Twelve libraries were pooled in equal concentrations, and sequenced using Illumina HiSeq 3000 platform (single-end, 50 bp reads format).

RNA-Seq data processing and differential expression analysis

The RNA-Seq data processing was performed on Galaxy, an open source, web-based bioinformatics platform (https://usegalaxy.org) 92. FastQC was first performed to check the sequencing read quality. Then the raw reads were mapped to D. melanogaster genome (BDGP

Release 6 + ISO1 MT/dm6) using Tophat2 v2.1.0 93. Transcripts were reconstructed using

Cufflinks v2.2.1 with bias correction. Cuffmerge (http://cole-trapnell-lab.github.io/cufflinks/) was used to merge together 12 Cufflinks assemblies to produce a GTF file for further differential expression analysis with Cuffdiff v2.2.1.3 94. After normalization, differentially expressed protein-coding transcripts were obtained using following cut-off values, false discovery rate

(FDR) ≤ 0.05 and fold-change ≥ 2. Non-coding gene and low expressed genes (FPKM<0.01) 46 were excluded from the analysis. RNA-Seq read files have been deposited to NCBI 's Gene

Expression Omnibus (GEO) (Accession # GSE112146). To review GEO files: Go to https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE112146, and enter token: sbmxiisabfcvdyz.

Principal component analysis (PCA), heatmap and expression correlation plot

PCA graph was generated using plotPCA function of R package DESeq2 95. Heatmaps and hierarchy clustering analysis were generated using heatmap.2 function of R package gplots.

(https://cran.r-project.org/web/packages/gplots). Expression data was log2 transformed and all reads were added by a pseudo-value 1. The expression correlation plots were plotted using R package ggplot2 (https://cran.r-project.org/web/packages/ggplot2).

Oenocyte-enriched genes and tissue-specific aging transcriptome analysis

Oenocyte-enriched genes were identified by comparing our oenocyte RiboTag data (H2O

Young group) to the whole body transcriptome profiles from previous studies (two wild-type backgrounds: w1118: GSM2647344, GSM2647345, GSM2647345. yw: GSM694258,

GSM694259).The sequencing reads with FPKM ≥ 0.01 were normalized by quantile normalization function using preprocessCore package.

(https://www.bioconductor.org/packages/release/bioc/html/preprocessCore.html). Oenocyte- enriched genes were defined as those with 5-fold higher FPKM in oenocytes comparing to whole body. Fat body-enriched genes were obtained similarly by comparing the expression values between adult fat body and whole body (data retrieved from FlyAtlas).

The lists of differentially expressed genes in multiple fly tissues were extracted from previous transcriptome analyses, heart 44, posterior midgut 46, fat body 45. Venn diagram analysis 47

(http://bioinformatics.psb.ugent.be/webtools/Venn/) was performed to identify overlapping genes between different tissues.

Gene set enrichment analysis (GSEA)

For GSEA analysis, a complete set of 136 KEGG pathways in Drosophila were downloaded from KEGG. Text were trimmed and organized using Java script. Quantile normalized FPKM values for each group were used as input for parametric analysis, and organized as suggested by GSEA tutorial site (GSEA, http://software.broadinstitute.org/gsea/doc/GSEAUserGuideFrame.html) 96. Collapse dataset to gene symbols was set to false. Permutation type was set to gene set; enrichment statistic used as weighted analysis; metric for ranking genes was set to Signal to Noise.

Gene ontology and pathway analysis

Functional annotation analysis of differentially expressed genes was performed using

STRING. GO terms (Biological Process, Molecular Function, Cellular Component), KEGG pathway, INTERPRO Protein Domains and Features, were retrieved from the analysis. To build

Ras/MAPK protein network in STRING, “kmeans clustering” option was used, and number of clusters was set to 2 or 3.

Quantitative real-time polymerase chain reaction (qRT-PCR)

qRT-PCR was performed using Quantstudio 3 Real-Time PCR system and SYBR green master mixture (Thermo Fisher Scientific, Waltham, MA, USA Catalog number: A25778). To determine the most stable housekeeping gene, the Ct values for four housekeeping genes were examined in all twelve cDNA samples obtained from different treatments. Using an Excel-based tool, Bestkeeper 97, we confirmed that Gapdh1 is the least-variable housekeeping gene across 48 samples. All gene expression levels were normalized to Gapdh1 by the method of comparative

Ct 98. Mean and standard errors for each gene were obtained from the averages of three biological replicates, with one or two technical repeats.

Statistical analysis

GraphPad Prism (GraphPad Software, La Jolla, CA, USA) was used for statistical analysis. To compare the mean value of treatment groups versus that of control, either student t- test or one-way ANOVA was performed using Dunnett’s test for multiple comparison.

4.7 References

1. Kennedy, B.K. et al. Geroscience: linking aging to chronic disease. Cell 159, 709-713 (2014).

2. Koehler, E.M. et al. Prevalence and risk factors of non-alcoholic fatty liver disease in the elderly: results from the Rotterdam study. Journal of hepatology 57, 1305-1311 (2012). 3. Kim, I.H., Kisseleva, T. & Brenner, D.A. Aging and liver disease. Current opinion in gastroenterology 31, 184-191 (2015).

4. Sheedfar, F., Di Biase, S., Koonen, D. & Vinciguerra, M. Liver diseases and aging: friends or foes? Aging Cell 12, 950-954 (2013).

5. Schmucker, D.L. Age-related changes in liver structure and function: Implications for disease ? Experimental gerontology 40, 650-659 (2005).

6. Jin, J. et al. Increased expression of enzymes of triglyceride synthesis is essential for the development of hepatic steatosis. Cell Rep 3, 831-843 (2013).

7. Amir, M. & Czaja, M.J. Autophagy in nonalcoholic steatohepatitis. Expert review of gastroenterology & hepatology 5, 159-166 (2011).

8. Zhang, C. & Cuervo, A.M. Restoration of chaperone-mediated autophagy in aging liver improves cellular maintenance and hepatic function. Nature medicine 14, 959-965 (2008).

9. Franceschi, C. et al. Inflamm-aging. An evolutionary perspective on immunosenescence. Annals of the New York Academy of Sciences 908, 244-254 (2000).

10. White, R.R. et al. Comprehensive transcriptional landscape of aging mouse liver. BMC genomics 16, 899 (2015). 49

11. Tollet-Egnell, P. et al. Gene expression profile of the aging process in rat liver: normalizing effects of growth hormone replacement. Molecular endocrinology 15, 308- 318 (2001).

12. Horvath, S. et al. Obesity accelerates epigenetic aging of human liver. Proc Natl Acad Sci U S A 111, 15538-15543 (2014).

13. He, Y. & Jasper, H. Studying aging in Drosophila. Methods 68, 129-133 (2014). 14. Gutierrez, E., Wiggins, D., Fielding, B. & Gould, A.P. Specialized hepatocyte-like cells regulate Drosophila lipid metabolism. Nature 445, 275-280 (2007).

15. Arrese, E.L. & Soulages, J.L. Insect fat body: energy, metabolism, and regulation. Annual review of entomology 55, 207-225 (2010).

16. Partridge, L., Alic, N., Bjedov, I. & Piper, M.D. Ageing in Drosophila: the role of the insulin/Igf and TOR signalling network. Experimental gerontology 46, 376-381 (2011). 17. Bai, H., Kang, P. & Tatar, M. Drosophila insulin-like peptide-6 (dilp6) expression from fat body extends lifespan and represses secretion of Drosophila insulin-like peptide-2 from the brain. Aging Cell 11, 978-985 (2012).

18. Hwangbo, D.S., Gershman, B., Tu, M.P., Palmer, M. & Tatar, M. Drosophila dFOXO controls lifespan and regulates insulin signalling in brain and fat body. Nature 429, 562- 566 (2004).

19. Giannakou, M.E. et al. Long-lived Drosophila with overexpressed dFOXO in adult fat body. Science 305, 361 (2004).

20. Billeter, J.C., Atallah, J., Krupp, J.J., Millar, J.G. & Levine, J.D. Specialized cells tag sexual and species identity in Drosophila melanogaster. Nature 461, 987-991 (2009). 21. Makki, R., Cinnamon, E. & Gould, A.P. The development and functions of oenocytes. Annual review of entomology 59, 405-425 (2014).

22. Martins, G.F. & Ramalho-Ortigao, J.M. Oenocytes in insects. Invertebrate Survival Journal 9, 139-152 (2012).

23. Chatterjee, D. et al. Control of metabolic adaptation to fasting by dILP6-induced insulin signaling in Drosophila oenocytes. Proc Natl Acad Sci U S A 111, 17959-17964 (2014). 24. Cinnamon, E. et al. Drosophila Spidey/Kar Regulates Oenocyte Growth via PI3-Kinase Signaling. PLoS Genet 12, e1006154 (2016).

25. Martins, G.F. et al. Insights into the transcriptome of oenocytes from Aedes aegypti pupae. Memorias do Instituto Oswaldo Cruz 106, 308-315 (2011).

26. Johnson, M.B. & Butterworth, F.M. Maturation and aging of adult fat body and oenocytes in Drosophila as revealed by light microscopic morphometry. Journal of morphology 184, 51-59 (1985).

27. Tower, J. et al. Variegated expression of Hsp22 transgenic reporters indicates cell- specific patterns of aging in Drosophila oenocytes. The journals of gerontology. Series A, Biological sciences and medical sciences 69, 253-259 (2014).

28. Sanz, E. et al. Cell-type-specific isolation of ribosome-associated mRNA from complex tissues. Proc Natl Acad Sci U S A 106, 13939-13944 (2009).

29. Peleg, S., Feller, C., Ladurner, A.G. & Imhof, A. The Metabolic Impact on Histone Acetylation and Transcription in Ageing. Trends in biochemical sciences 41, 700-711 (2016).

30. Hall, H. et al. Transcriptome profiling of aging Drosophila photoreceptors reveals gene expression trends that correlate with visual senescence. BMC genomics 18, 894 (2017). 31. Pletcher, S.D. et al. Genome-wide transcript profiles in aging and calorically restricted Drosophila melanogaster. Curr Biol 12, 712-723 (2002).

32. Alonso, J. & Santaren, J.F. Characterization of the Drosophila melanogaster ribosomal proteome. Journal of proteome research 5, 2025-2032 (2006).

33. Bloomer, S.A., Zhang, H.J., Brown, K.E. & Kregel, K.C. Differential regulation of hepatic heme oxygenase-1 protein with aging and heat stress. The journals of gerontology. Series A, Biological sciences and medical sciences 64, 419-425 (2009).

34. Suh, J.H. et al. Decline in transcriptional activity of Nrf2 causes age-related loss of glutathione synthesis, which is reversible with lipoic acid. Proc Natl Acad Sci U S A 101, 3381-3386 (2004).

35. Mori, K. et al. Hepatic transcript levels for genes coding for enzymes associated with xenobiotic metabolism are altered with age. Toxicologic pathology 35, 242-251 (2007). 36. Xu, C., Li, C.Y. & Kong, A.N. Induction of phase I, II and III drug metabolism/transport by xenobiotics. Archives of pharmacal research 28, 249-268 (2005).

37. Qiu, Y. et al. An insect-specific P450 oxidative decarbonylase for cuticular hydrocarbon biosynthesis. Proc Natl Acad Sci U S A 109, 14858-14863 (2012).

38. Narayan, V. et al. Deep Proteome Analysis Identifies Age-Related Processes in C. elegans. Cell systems 3, 144-159 (2016).

39. Pandey, U.B. & Nichols, C.D. Human disease models in Drosophila melanogaster and the role of the fly in therapeutic drug discovery. Pharmacological reviews 63, 411-436 (2011).

40. Chintapalli, V.R., Wang, J. & Dow, J.A. Using FlyAtlas to identify better Drosophila melanogaster models of human disease. Nat Genet 39, 715-720 (2007).

41. Leader, D.P., Krause, S.A., Pandit, A., Davies, S.A. & Dow, J.A.T. FlyAtlas 2: a new version of the Drosophila melanogaster expression atlas with RNA-Seq, miRNA-Seq and sex-specific data. Nucleic Acids Res 46, D809-D815 (2018).

42. Uhlen, M. et al. Proteomics. Tissue-based map of the human proteome. Science 347, 1260419 (2015).

43. Hu, Y. et al. An integrative approach to ortholog prediction for disease-focused and other functional studies. BMC bioinformatics 12, 357 (2011).

44. Monnier, V. et al. dJun and Vri/dNFIL3 are major regulators of cardiac aging in Drosophila. PLoS Genet 8, e1003081 (2012).

45. Chen, H., Zheng, X. & Zheng, Y. Age-associated loss of lamin-B leads to systemic inflammation and gut hyperplasia. Cell 159, 829-843 (2014).

46. Resnik-Docampo, M. et al. Tricellular junctions regulate intestinal stem cell behaviour to maintain homeostasis. Nat Cell Biol 19, 52-59 (2017).

47. Martins, G.F. et al. Isolation, primary culture and morphological characterization of oenocytes from Aedes aegypti pupae. Tissue & cell 43, 83-90 (2011).

48. Jackson, A. & Locke, M. The formation of plasma membrane reticular systems in the oenocytes of an insect. Tissue & cell 21, 463-473 (1989).

49. Ferveur, J.F. et al. Genetic feminization of pheromones and its behavioral consequences in Drosophila males. Science 276, 1555-1558 (1997).

50. Parvy, J.P. et al. Drosophila melanogaster Acetyl-CoA-carboxylase sustains a fatty acid- dependent remote signal to waterproof the respiratory system. PLoS Genet 8, e1002925 (2012).

51. Laffel, L. Ketone bodies: a review of physiology, pathophysiology and application of monitoring to diabetes. Diabetes/metabolism research and reviews 15, 412-426 (1999).

52. Bailey, E.J., Horne, J.A., Izatt, M.E. & Hill, L. Concentrations of acetoacetate and 3- hydroxybutyrate in pigeon blood and desert locust haemolymph. Life sciences. Pt. 2: Biochemistry, general and molecular biology 10, 1415-1419 (1971).

53. Hill, L., Izatt, M.E.G., Horne, J.A. & Bailey, E. Factors affecting concentrations of acetoacetate and d-3-hydroxybutyrate in haemolymph and tissues of the adult desert locust. Journal of insect physiology 18, 1265-1285 (1972).

54. Shah, J. & Bailey, E. Enzymes of ketogenesis in the fat body and the thoracic muscle of the adult cockroach. Insect Biochemistry 6, 251-254 (1976).

55. Howard, R.W. & Blomquist, G.J. Ecological, behavioral, and biochemical aspects of insect hydrocarbons. Annual review of entomology 50, 371-393 (2005).

56. Chiang, Y.N. et al. Steroid Hormone Signaling Is Essential for Pheromone Production and Oenocyte Survival. PLoS Genet 12, e1006126 (2016).

57. Gaziova, I., Bonnette, P.C., Henrich, V.C. & Jindra, M. Cell-autonomous roles of the ecdysoneless gene in Drosophila development and oogenesis. Development 131, 2715- 2725 (2004).

58. Uryu, O., Ameku, T. & Niwa, R. Recent progress in understanding the role of ecdysteroids in adult insects: Germline development and circadian clock in the fruit fly Drosophila melanogaster. Zoological letters 1, 32 (2015).

59. Romer, F., Emmerich, H. & Nowock, J. Biosynthesis of ecdysones in isolated prothoracic glands and oenocytes of Tenebrio molitor in vitro. Journal of insect physiology 20, 1975- 1987 (1974).

60. Yamazaki, Y., Kiuchi, M., Takeuchi, H. & Kubo, T. Ecdysteroid biosynthesis in workers of the European honeybee Apis mellifera L. Insect biochemistry and molecular biology 41, 283-293 (2011).

61. Zheng, W. et al. Dehydration triggers ecdysone-mediated recognition-protein priming and elevated anti-bacterial immune responses in Drosophila Malpighian tubule renal cells. BMC biology 16, 60 (2018).

62. Zhan, M. et al. Temporal and spatial transcriptional profiles of aging in Drosophila melanogaster. Genome research 17, 1236-1243 (2007).

63. Wanders, R.J. & Waterham, H.R. Biochemistry of mammalian peroxisomes revisited. Annual review of biochemistry 75, 295-332 (2006).

64. Fujiki, Y., Okumoto, K., Mukai, S., Honsho, M. & Tamura, S. Peroxisome biogenesis in mammalian cells. Front Physiol 5, 307 (2014).

65. Legakis, J.E. et al. Peroxisome senescence in human fibroblasts. Mol Biol Cell 13, 4243- 4255 (2002).

66. Haining, J.L. & Legan, J.S. Catalase turnover in rat liver and kidney as a function of age. Experimental gerontology 8, 85-91 (1973).

67. Semsei, I., Rao, G. & Richardson, A. Changes in the expression of superoxide dismutase and catalase as a function of age and dietary restriction. Biochemical and biophysical research communications 164, 620-625 (1989).

68. Xia, E., Rao, G., Van Remmen, H., Heydari, A.R. & Richardson, A. Activities of antioxidant enzymes in various tissues of male Fischer 344 rats are altered by food restriction. The Journal of nutrition 125, 195-201 (1995).

69. Ivashchenko, O. et al. Intraperoxisomal redox balance in mammalian cells: oxidative stress and interorganellar cross-talk. Mol Biol Cell 22, 1440-1451 (2011).

70. Perichon, R. & Bourre, J.M. Peroxisomal beta-oxidation activity and catalase activity during development and aging in mouse liver. Biochimie 77, 288-293 (1995).

71. Weir, H.J. et al. Dietary Restriction and AMPK Increase Lifespan via Mitochondrial Network and Peroxisome Remodeling. Cell Metab 26, 884-896 e885 (2017).

72. Houtkooper, R.H. et al. The metabolic footprint of aging in mice. Scientific reports 1, 134 (2011).

73. Poirier, Y., Antonenkov, V.D., Glumoff, T. & Hiltunen, J.K. Peroxisomal beta-oxidation- -a metabolic pathway with multiple functions. Biochimica et biophysica acta 1763, 1413- 1426 (2006).

74. Singh, J., Khan, M. & Singh, I. Silencing of Abcd1 and Abcd2 genes sensitizes astrocytes for inflammation: implication for X-adrenoleukodystrophy. Journal of lipid research 50, 135-147 (2009).

75. Pujol, A. et al. Late onset neurological phenotype of the X-ALD gene inactivation in mice: a mouse model for adrenomyeloneuropathy. Human molecular genetics 11, 499- 505 (2002).

76. Seedorf, U. et al. Defective peroxisomal catabolism of branched fatty acyl coenzyme A in mice lacking the sterol carrier protein-2/sterol carrier protein-x gene function. Genes & development 12, 1189-1201 (1998).

77. Im, S.S. et al. Linking lipid metabolism to the innate immune response in macrophages through sterol regulatory element binding protein-1a. Cell Metab 13, 540-549 (2011).

78. Di Cara, F., Sheshachalam, A., Braverman, N.E., Rachubinski, R.A. & Simmonds, A.J. Peroxisome-Mediated Metabolism Is Required for Immune Response to Microbial Infection. Immunity 47, 93-106 e107 (2017).

79. Li, Z. et al. Age-induced augmentation of p38 MAPK phosphorylation in mouse lung. Experimental gerontology 46, 694-702 (2011).

80. Kim, H.J., Jung, K.J., Yu, B.P., Cho, C.G. & Chung, H.Y. Influence of aging and calorie restriction on MAPKs activity in rat kidney. Experimental gerontology 37, 1041-1053 (2002).

81. Kim, E.K. & Choi, E.J. Pathological roles of MAPK signaling pathways in human diseases. Biochimica et biophysica acta 1802, 396-405 (2010).

82. Vrailas-Mortimer, A. et al. A muscle-specific p38 MAPK/Mef2/MnSOD pathway regulates stress, motor function, and life span in Drosophila. Developmental cell 21, 783- 795 (2011).

83. Wang, M.C., Bohmann, D. & Jasper, H. JNK signaling confers tolerance to oxidative stress and extends lifespan in Drosophila. Developmental cell 5, 811-816 (2003).

84. Oh, S.W. et al. JNK regulates lifespan in Caenorhabditis elegans by modulating nuclear translocation of forkhead transcription factor/DAF-16. Proc Natl Acad Sci U S A 102, 4494-4499 (2005).

85. Biteau, B., Karpac, J., Hwangbo, D. & Jasper, H. Regulation of Drosophila lifespan by JNK signaling. Experimental gerontology 46, 349-354 (2011).

86. Johnson, G.L. & Nakamura, K. The c-jun kinase/stress-activated pathway: regulation, function and role in human disease. Biochimica et biophysica acta 1773, 1341-1348 (2007).

87. Sun, J., Kale, S.P., Childress, A.M., Pinswasdi, C. & Jazwinski, S.M. Divergent roles of RAS1 and RAS2 in yeast longevity. J Biol Chem 269, 18638-18645 (1994). 88. Slack, C. et al. The Ras-Erk-ETS-Signaling Pathway Is a Drug Target for Longevity. Cell 162, 72-83 (2015).

89. Nanji, M., Hopper, N.A. & Gems, D. LET-60 RAS modulates effects of insulin/IGF-1 signaling on development and aging in Caenorhabditis elegans. Aging Cell 4, 235-245 (2005).

90. Thyagarajan, B. et al. ETS-4 is a transcriptional regulator of life span in Caenorhabditis elegans. PLoS Genet 6, e1001125 (2010).

91. Owusu-Ansah, E., Yavari, A. & Banerjee, U. A protocol for in vivo detection of reactive oxygen species. Protocol Exchange (2008).

92. Afgan, E. et al. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2016 update. Nucleic Acids Res 44, W3-W10 (2016).

93. Kim, D. et al. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome biology 14, R36 (2013).

94. Trapnell, C. et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nature protocols 7, 562-578 (2012).

95. Love, M.I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550 (2014).

96. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102, 15545- 15550 (2005).

97. Pfaffl, M.W., Tichopad, A., Prgomet, C. & Neuvians, T.P. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper--Excel-based tool using pair-wise correlations. Biotechnology letters 26, 509- 515 (2004).

98. Schmittgen, T.D. & Livak, K.J. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 3, 1101-1108 (2008).

4.8 Figures

A DHE DAPI Merge

Young

H2O

Young -

DHE quantification PQ )

C % 25 ****

(

I ns

Aged R

20 -

i t 15 H2O

c ****

r F

10 a

e **

% DHE staining % DHE

Aged t

5 -

j PQ

b 0 O PQ —O +Q —ld +Q 2 P P H O + Young Agedld O

Figure 1. CharacterizationSource: 11152017 of age- relatedDHE 2-d changes396 in ROS production in Drosophila 11142017 DHE 2-d 496 oenocytes. (A) Fluorescent image of GFP-labeled oenocytes from the abdomen of PromE-Gal4; UAS-CD8::GFP female flies. Scale bar: 100 µm. (B) ROS levels indicated by DHE staining in female oenocytes under aging and paraquat (PQ) treatment. Young: 10-day-old, Aged: 30-day- old. DAPI stains for nuclei. Scale bar: 10 µm. (C) Quantification of DHE staining from Panel (B). One-way ANOVA (*** p<0.001, * p<0.05). N=5.

A Oenocyte Ribo-Tag B Ribo-tag PCR Dilp2 desat1-E C 2 dilp2

3.0 3 0.025

PromE-gal4 X UAS-RpL13A-FLAG *** L INPUT*

p R 0.020 o IP

2.5 t

n o

i 0.015

s s

Expression 2.0

Expression

e r

p 0.010

x e

1.5

-5 e Relative Relative 110

Relative Relative 0.005

t a Homogenization l 0 e 0.000 Anti-FLAG InputDesat1-EOe-IP R Headd OeP -IP a I e Total RNA pulled downH D TotalSource: RNA pull2018-02-08_184839_kh-down BMC verification 1 ) 40

AAA L 2018-03-13_123219_kerui_BMC test 2



g μ

AAA n (

o i

Oenocyte-specific t

a r

RNA isolation t 20

e c

AAA n

o 10

concentration(ng/ c

AAA

AAA A ND ND

N RNA

R 0

O Q O Q ly t 2 P 2 P n w H , H , o > , g , d l4 RNA-Seq g n d l s a n u l O d G u o O a - o Y e E Y B m ro P Figure 2. Oenocyte-specific translatomicSource: Fig1-average profiling through mRNA quantification RiboTag sequencing. (A) Schematic diagram showing RiboTag procedures. FLAG-tagged ribosomal protein RpL13A was first ectopically expressed in oenocytes. Translating RNAs were immunoprecipitated using anti- FLAG antibodies. RNAs were further purified and used in RNA-seq analysis. (B) Oenocyte- specific transcript desat1-E highly expressed in anti-FLAG immunoprecipitated sample (IP) compared to the input (whole body lysate). (C) The transcripts of brain-specific gene Dilp2 was enriched in head samples compared to oenocyte RiboTag samples. One-way ANOVA (**** p<0.0001, *** p<0.001, ** p<0.01, * p<0.05, ns = not significant). N=3. (D) RNA concentrations of various immunoprecipitated samples. ND: Not detected. 200 female flies were used in each condition. Three biological replicates per condition. 58

A B C D 20 y = -0.22x + 0.93 y = -0.28x2 + 0.97 y =2 -0.034x + 0.95 group 2 2 R = 0.926 R2 = 0.948 PQ-Aged H2O-Young RR= 0.86= 10.861 R2 = 0.926 R = 0.948 H2OO_0 10 H2OO_1 H2OO_2 H2O-Aged H2OY_0 H2OY_1

0 H2OY_2 Aged

- Aged

PQO_0 Young -

PQO_1 - PC2: 9% variance PC2: 9% PQO_2 PQY_0 -10 PQ

PQ-Young PQ PQY_1 H2O

PQY_2 PC2: 9% variance 9% PC2: -20 -20 0 20 40 PC1: 50% variance PC1: 50% variance H2O-Young H2O-Young H2O-Aged

E Up-regulated F Down-regulated Aging Aging PQ Aging only Shared PQ only PQ Aging only Shared PQ only

Developmental process OXPHOS Immune response Ribosome Xenobiotics Proteasome 965 127 71 DNA metabolism 1653 579 172 Peroxisome DNA recombination Lipid metabolism Phototransduction Defense response 0 6 12 Fatty acid elongation -log10 (p value) Response to bacterium 0 6 12 G PQ-Up -log10 (p value) Young Aged I J Cluster3 Cluster5 Cluster10

187 11 202 DNA replication 1 2 3 4 5 6 7 8 9 10 11 PI3K signaling H20-Y JAK-STAT Hippo signaling H PQ-Down H20-O Endocytosis Young Aged Proteasome PQ-Y OXPHOS FA metabolism PQ-O Cytoskeleton 722 29 143 Cell cycle Low High 0 2.5 5 -log (p value) 10

Figure 3. Differential gene expression analysis reveals common and distinct transcriptional regulation by aging and oxidative stress. (A) Principal component analysis (PCA) on four oenocyte translatomes. (B-D) Correlation analysis on the gene expression between H2O-Young and H2O-Aged; H2O-Young and PQ-Young; H2O-Aged and PQ-Aged. Log10 (FPKM) was used in the analysis (E-F) Venn diagram and GO terms for the genes commonly and differentially regulated by aging and paraquat. (G-H) Venn diagram and GO terms for the genes commonly and differentially regulated by paraquat at two ages. (I) Hierarchy clustering analysis on oenocyte translatome. (J) Gene ontology analysis on cluster 3 and 5 in panel (I).

C 0.0- A Up-regulated in aged oenocytes DNA replication D 0.6- Oxidative -0.2- 0.4- phosphorylation NAME ES FDR MISMATCH REPAIR -0.65 0.025 -0.4- 0.2- DNA REPLICATION -0.59 0.028 BASE EXCISION REPAIR -0.61 0.035 Enrichment score Enrichment -0.6- Enrichment score Enrichment 0.0- NUCLEOTIDE EXCISION REPAIR -0.53 0.037 FANCONI ANEMIA PATHWAY -0.56 0.039 GLUTATHIONE S TRANSFERASE -0.38 0.055 Down in Aged Up in Aged Down in Aged Up in Aged Down-regulated in aged oenocytes E B 0.6- Proteasome F 0.5- Peroxisome 0.4- NAME ES FDR 0.3- OXIDATIVE PHOSPHORYLATION 0.60 0.001 RIBOSOME 0.57 0.001 0.2- 0.1- PROTEASOME 0.65 0.002

0.0- Enrichment score Enrichment Enrichment score Enrichment 0.0- CARBON METABOLISM 0.52 0.007 THIAMINE METABOLISM 0.70 0.008 PEROXISOME 0.53 0.008 PENTOSE PHOSPHATE 0.66 0.013 Down in Aged Up in Aged Down in Aged Up in Aged NEUROACTIVE LIGAND- G H 0.35- RECEPTOR INTERACTION 0.55 0.015 0.1- Glutathione S-transferase 0.3- Cytochrome P450 GALACTOSE METABOLISM 0.59 0.016 0.0- 0.2- GLYCOLYSIS 0.53 0.033 -0.1- FATTY ACID METABOLISM 0.54 0.033 0.1- GLYOXYLATE METABOLISM 0.56 0.044 -0.2- 0.0-

GLYCINE METABOLISM 0.56 0.045 -0.3- Enrichment score Enrichment

FATTY ACID ELONGATION 0.63 0.049 score Enrichment -0.4- CYTOCHROME P450 0.36 0.095

Down in Aged Up in Aged Down in Aged Up in Aged

Low High Low High K I mRpS9 J mRpS7 mRpS6 mRpS21 mRpS2 mRpS18C Mito mRpS18A mRpS17 Clan mRpS16 Delta mRpS14 mRpS11 1000 mRpS10 Clan 2 mRpL9 mRpL4 mRpL36 mRpL35 mRpL34 mRpL33 mRpL32 mRpL30 mRpL3 Clan 3 mRpL28 Epsilon mRpL27 mRpL24 mRpL23 mRpL22 500 mRpL21 mRpL20 mRpL2 mRpL19 mRpL18 mRpL17 mRpL16 mRpL15 Clan 4 mRpL14 mRpL13 Others mRpL12 mRpL11 mRpL10 mRpL1 H2OYoung PQ AgingH2O FPKM H2O H2O PQ PQ H2O H2O PQ PQ Young Young Aged Young Aged Young Aged Young Aged Young Aged

Figure 4. GSEA analysis revealed up- and down-regulated pathways under aging. (A) List of the pathways up-regulated in aged oenocytes. (B) List of the pathways down-regulated in aged oenocytes. ES: Enrichment score. (C-H) GSEA enrichment profiles of six pathways: DNA replication, oxidative phosphorylation, proteasome, peroxisome, glutathione S-transferase, cytochrome P450. (I-K) Heatmaps for mitochondrial ribosomal subunits, glutathione S- transferase, cytochrome P450.

A PME Peroxisome Matrix Enzyme PMP Peroxisome Membrane Protein

Pex5 Cytosol PME PME Receptor Matrix enzyme Pex5 Pex7 import recycling Pex5 Pex1 Pex10 Pex12 Pex2 PEX26 Pex14 Pex13 Pex6 PMP Pex19 PME Pex3 Peroxisome Pex16 lumen Pex11 Pex19 PMP Proliferation CG13827 CG33474 Membrane assembly

) B g 1 C 1

Aged / Young ) Paraquat / H2O

o H

y 0 0

P g

-1 - -1

(fold (fold change)

(fold change) (fold

(

F 2

( -2 -2

Log

g L

o -3 -3 L 3 6 9 5 7 3 4 2 0 2 1 6 1 7 4 3 6 9 5 7 3 4 2 0 2 1 6 1 7 4 x 1 1 x x 1 1 x 1 1 x x 1 2 7 x 1 1 x x 1 1 x 1 1 x x 1 2 7 e x x e e x x e x x e e x 8 4 e x x e e x x e x x e e x 8 4 P e e P P e e P e e P P e 3 3 P e e P P e e P e e P P e 3 3 P P P P P P P 1 3 P P P P P P P 1 3 G G G G C C C C Membrane Matrix protein Receptor Proliferation Membrane Matrix protein Receptor Proliferation

assembly import recycling assembly import recycling

)

g 1

H Young H2O and Old H2O: t-test ** n

D D Pex5

P 0.8 u F 0.8 **

2 Beta-oxidation A ns Young H2O and Old PQ: t-test *

o y

t 0.60.6

1 e

d t

a 0.40.4

l e

0 e

n 0.20.2

a i

- -1

Relative expression Relative

(fold (fold change)

e 2 C 0.00.0 r

-2 p

F PQ x —O Q+ —O Q+

2 P 2 P (

E H g H d Log l g n ld 2 n uYoung OAged

-3 u o O

4 2

8 o 0

2 Y

g -

- Y

2 7

7 Pex19-GAPDH1 0315

M 1

S Pex19 m

C **

1 1

1 Source: 2018-03-15_164715_BMC refine

G G

P x

x 1.0

G G

G ns

C C

C 1.0

o o

P 2018-03-15_110509

c A

) Young H2O and Old H2O: t-test **

G A A 0.8

0.8

e 0.6 v

n E 0.6

l u Ether phospholipid biosynthesis ROS metabolism e 0.4 r 0.4

n o

Sterol precursor biosynthesis Purine metabolism o i

s 0.2

y 0.2 s

/ Amino acid metabolism

Retinol metabolism e

Relative expression Relative r

2 p 0.00.0 d

E O Q O Q PQ 2 2 e — P+ — P+ H g H d g n d l n u l O

g u o O 0 o Y Young Aged

Y a

- Source: 2018-03-15_164715_BMC refine

1 (fold (fold change)

H 2018-03-15_110509 2

C H 25 -2 D 25 ** Cat

P ns Young H2O and Old H2O: t-test **

A (

G 20

Log 20 o

t Young H2O and Old PQ: t-test ** 2

-4

2 5

1 15

h v

S 15

S S

S 10

A A

A 10

C C

s 5.05

Relative expression Relative r

p 0.00 x E PQ O Q O Q 2— P+ 2— P+ H g H d g n d l n u l O u oYoungO Aged o Y Y Source: 2018-03-15_164715_BMC refine 2018-03-15_110509 Figure 5. Peroxisome pathway is transcriptionally deregulated in aged oenocytes. (A) Schematic diagram showing peroxisome pathway and the role of each peroxin (PEX) genes. (B- C) Log2 (fold change) of the expression of PEX genes under aging and paraquat treatment, based on oenocyte RiboTag sequencing results. (D-E) Log2 (fold change) of the expression of genes involved in other peroxisome functions during oenocyte aging. (F-H) qRT-PCR verification of three peroxisome genes (Pex5, Pex19, Cat). One-way ANOVA (, ** p<0.01, ns = not significant). N=3. 61

A Ketogenesis enzyme Oenocyte Liver 1 HMG-CoA synthase Hmgs (7.7x) HMGCS1 (4.4x), HMGCS2 (16x) 2 HMG-CoA lyase CG10399 (8.2x) HMGCL (3.9x), HMGCLL1 D-β-hydroxybutyrate Shroud (3.6x), 3 BDH1 (12.9x), BDH2 dehydrogenase CG13377

B +

CoA

NAD

NADH + H NADH

CoA

Acetyl Acetyl

Acetoacetyl-CoA HMG-CoA Acetoacetate D-β-hydroxybutyrate ① ② ③

Fatty acid elongation C Oenocyte Liver D Long-chain acyl-CoA (n≥16) (in SER) Malonyl-CoA Elo68beta,Elo68a ELOVL1,ELOVL2 ① Very-long-chain CoA-SH + CO2 lpha,CG17821, (5.2x),ELOVL3, 1 3-ketoacyl-CoA CG18609 (19X), ELOVL4,ELOVL5, Long-chain 3-oxoacyl-CoA synthase Baldspot ELOVL6,ELOVL7 NADPH + H+ ② NADP+ Very-long-chain 2 3-ketoacyl-CoA Spidey (20.7x) HSD17B12 Long-chain 3-hydroxyacyl-CoA reductase ③ H2O Very-long-chain CG6746 (29.6x) HACD1,HACD2, 3 3-hydroxyacyl-CoA Long-chain trans-2,3-dehydroacyl-CoA CG9267 HACD3,HACD4 dehydratase NADPH + H+ ④ NADP+ Very-long-chain 4 Sc2 (21.5x) TECR enoyl-CoA reductase Long-chain acyl-CoA (n+2)

Very-long-chain fatty acid

Figure 6. Ketogenesis and fatty acid elongation are enriched in both oenocytes and liver. (A) List of ketogenesis genes that are enriched in both oenocytes and liver. (B) Schematic diagram showing ketogenesis pathway. (C) List of genes in microsomal fatty acid elongation pathway that are enriched in both oenocytes and liver. (D) Schematic diagram showing microsomal fatty acid elongation pathway (in smooth ER). Oenocyte-enriched genes are highlighted in red. Liver-enriched genes are highlighted in blue. 62

A Human liver Fly oenocyte B Mouse liver Fly oenocyte

1847 744 2580 629 252 3072

Innate Innate

Immune response Immune response

Up - Defense response Up - MAPK cascade Ras signaling

Age NF-κB signaling Age NF-κB signaling Ras signaling 0 10 20 0 10 25 OXPHOS OXPHOS

Down Peroxisome -

Down Lipid metabolism -

Fatty acid metabolism Peroxisome Age

0 15 30 Age 0 2.5 5 -log10 (p value) -log10 (p value)

C Fly D Mouse MAPK MAPK Ras Ras

Figure 7. Conservation of age-related transcriptional changes between oenocytes and liver. (A) Venn diagram comparing genes differentially expressed in aged liver and aged oenocytes. The mouse liver genes were first converted to their putative Drosophila orthologues before comparing to oenocyte aging genes. GO terms were shown in the lower panel. (B) Signaling pathways that were up-regulated under both oenocyte and liver aging. (C) Signaling pathways that were down-regulated under both oenocyte and liver aging. The genes listed in Panel B&C are the orthologues between Drosophila and mouse. (D-E) List of all genes in Ras/MAPK signaling pathway that were down-regulated in aged fly oenocytes and mouse liver. Protein network was generated using STRING (with kmeans clustering option). (F) Venn diagram showing the overlap of differentially expressed genes in aged oenocytes, fat body, heart, and midgut. (G) GO terms enriched in aged oenocytes, fat body, heart, and midgut.

Figure S1 63

A B C 22000 ns

y 20000

t i

s 18000

n e

t 16000

P 14000 F

G 12000 10000 g d n e u g o A Young Aged Y PromE-Gal4; UAS-CD8::GFP

Figure S1. Age-dependent PromE-gal4 expression pattern. (A-B) Fluorescent image PromE- Gal4; UAS-CD8::GFP female flies at two ages: Young (10-day-old), Aged (30-day-old). Scale bar: 50 µm. (C) Quantification of GFP intensity from Panel (A&B). Student t-test (ns = not significant). N=9.

Figure S2 64

Ecdysteroid hormone metabolism pathway

Cholesterol

Neverland 20,26-Dihydroxyecdysone

7-Dehydrocholesterol 26-hydroxylase Cyp18a1 Spook/Cyp307a1 (3x) Spookier/Cyp307a2 Shroud (3.6x) 20-Hydroxyecdysone Cyp6t3 (2.4x)

Diketol Shade Cyp314a1

3β,5β-Ketodiol Ecdysone

Phantom Shadow Cyp306a1 Cyp315a1 (32x) Disembodied (1.7x) (12.5x) 3β,5β-Ketotriol 2-Deoxyecdysome Cyp302a1

Figure S2. Two ecdysteroid biosynthesis genes highly express in oenocytes. Schematic diagram showing ecdysteroid hormone metabolism pathway. Two Halloween genes, phantom and shadow, highly expressed in adult female oenocytes (Highlighted in red).

Figure S3 65

A B Toll Oenocyte Fat body Oenocyte PGRP-SD Fat body Cact PGRP-SA modSP Myd88 Drs Spz Tl Tube Dl/Dif CecA2 GNBP3 nec psh Pelle IM2 387 36 508 IMD PGRP-LE AttC Ank PGRP-LB DptA CecC PGN PGRP-LC Imd Relish PGRP-SC DptB Dro Mtk Dredd

Figure S3. Genes in innate immunity pathway highly express in oenocytes. (A) Genes enriched in oenocytes and fat body show less overlap. (B) Genes in Imd pathway were enriched in oenocytes, while fat body were enriched with genes in Toll pathway (Red arrows denote for age-induced genes. Blue arrows denote for age-repressed gene.).

Figure S4 66

1. Peroxisome biogenesis and protein import Oenocyte Liver

Pex19 PXMP2 (7.4x) (5.1x)

2. Peroxisome function Fatty acid Ether phospholipid ROS metabolism beta-oxidation biosynthesis Oenocyte Liver Oenocyte Liver Oenocyte Liver ACOX2 (7.5x) SCPX/CG17597 FarO (123x) None Cat (11.8x) CAT (5.2x) (11.9x) BAAT (34.6x) EHHADH (12x) CG13091(243x) Sod1(18.7x) SOD1 (3.3x) ECH/CG9577 ACAA1 (6.3x) CG14893 (21x) Prx5 (6.4x) (8.2x) SLC27A2 (9x) CG17562 (40x) ACSL1 (6.4x) CG4020 (51x) PECR (7.8x)

Amino acid metabolism Purine metabolism Retinol metabolism Oenocyte Liver Oenocyte Liver Oenocyte Liver

AGXT/Spat AGXT (36.3x) AOX1 (3.3x) XDH (5.5x) DHRS4/ None (2.2x) DAO (16.1x) CG10672 PIPOX (18.5x) (16.3x) HMGCL/CG10399 HAO1(36.8x) (8.2x) HAO2 (15x)

Figure S4. Peroxisome pathways are enriched in both oenocytes and liver. List of peroxisome genes that are enriched in both oenocytes and liver.

67 Figure S5

A B Oenocyte Fat body Heart Midgut

Aminoglycan Ion Proteasome Immune response metabolism transport

Chitin metabolic Glycolysis / DNA Ribosome process Gluconeogenesis replication

Detoxification of Oxidative Oxidative Fatty acid inorganic phosphorylation phosphorylation degradation compound

Figure S5. (A) Venn diagram showing the overlap of differentially expressed genes in aged oenocytes, fat body, heart, and midgut. (B) GO terms enriched in aged oenocytes, fat body, heart, and midgut.

Table S1. Comparison of the GO terms enriched in oenocyte, fat body and liver

GO Oenocyte Fat body Liver Lipid metabolism Fatty acid biosynthesis Long-chain fatty acid metabolism Fatty acid elongation Carboxylic acid metabolism Biological Oxidation reduction Oxidation reduction Immune response Process Immune response Immune response Amino acid metabolism Xenobiotic metabolism Protein catabolism Bile acid metabolism Oxidoreductase activity Oxidoreductase activity Oxidoreductase activity Endopeptidase activity Molecular Metalloendopeptidase Serine-type peptidase activity Fatty acid synthase activity Function activity Lipid binding Fatty acid elongase activity Heme binding Heme binding Peptidoglycan binding Peroxisome Cellular Peroxisome Extracellular region Extracellular region Component Proteasome Endoplasmic reticulum Metabolism of xenobiotics Metabolism of xenobiotics Glycine, serine and Peroxisome Peroxisome KEGG threonine metabolism Bile secretion Proteasome Pathway Arginine and proline PPAR signaling pathway Ketone body metabolism metabolism Retinol metabolism Unsaturated fatty acid synthesis Biosynthesis of amino acids Protein ELO family Peptidase S1, M13 Serpin family Domain Proteasome subunit EGF-like domain Cytochrome P450

CHAPTER 5. IMPAIRED PEROXISOMAL IMPORT IN DROSOPHILA OENOCYTES CAUSES CARDIAC DYSFUNCTION BY INDUCING UPD3 AS A PEROXIKINE

A paper accepted by Nat Commun 11, 2943 (2020). https://doi.org/10.1038/s41467-020-

16781-w

Kerui Huang 1, Ting Miao 1, Kai Chang 1, Jinoh Kim 1, Ping Kang 1, Qiuhan Jiang 1,

Andrew J. Simmonds 2, Francesca Di Cara 3 & Hua Bai 1*

From the 1Department of Genetics, Development, and Cell Biology, Iowa State University,

Ames, Iowa 50011, USA

From the 2Department of Cell Biology, University of Alberta, Edmonton, AB, T6G 2H7, Canada

From the 3Department of Microbiology & Immunology, Dalhousie University, Halifax, NS,

B3H4R2, Canada

* To whom correspondence should be addressed: Hua Bai: Department of Genetics,

Development, and Cell Biology, Iowa State University, Ames, Iowa 50011; [email protected]

Keywords: Liver-heart communication, Interleukin-6, Peroxisome

5.1 Abstract

Aging is characterized by a chronic, low-grade inflammation, which is a major risk factor for cardiovascular diseases. It remains poorly understood whether pro-inflammatory factors released from non-cardiac tissues contribute to the non-autonomous regulation of age-related cardiac dysfunction. Here, we report that age-dependent induction of cytokine unpaired 3 (upd3) 70 in Drosophila oenocytes (hepatocyte-like cells) is the primary non-autonomous mechanism for cardiac aging. We show that upd3 is significantly up regulated in aged oenocytes. Oenocyte- specific knockdown of upd3 is sufficient to block aging-induced cardiac arrhythmia. We further show that the age-dependent induction of upd3 is triggered by impaired peroxisomal import and elevated JNK signaling in aged oenocytes. We term hormonal factors induced by peroxisome dysfunction as peroxikines. Intriguingly, oenocyte-specific overexpression of Pex5, the key peroxisomal import receptor, blocks age-related upd3 induction and alleviates cardiac arrhythmicity. Thus, our studies identify an important role of hepatocyte-specific peroxisomal import in mediating non-autonomous regulation of cardiac aging.

5.2 Introduction

Age is a major risk factor for a wide range of human diseases1 including cardiovascular diseases (CVD)2. During aging, cardiomyocytes undergo rapid remodeling with a variety of intracellular changes, such as impaired mitochondria, increased reactive oxygen species (ROS), and elevated inflammation1. The low-grade chronic and systemic inflammation (or inflammaging) is often associated with increased levels of circulating proinflammatory biomarkers (e.g., interleukin-6 (IL-6) and C-reactive protein), which are notable risk factors for

CVD3,4. Short-term expression of IL-6 can protect myocytes from injury-induced apoptosis.

However, prolonged production of IL-6 induces pathological hypertrophy and decreases cardiomyocyte contractility through the activation of Janus kinases-signal transducer and activator of transcription (JAK-STAT) signaling5. Elevated levels of circulating IL-6 are often associated with heart failure, myocardial damage, and atherosclerosis5,6,7. IL-6 can be produced not only by cardiomyocytes themselves in response to injury, but also by other neighboring tissues (e.g., endothelial cells), immune cells, and the liver6,8. However, the root causes of 71 inflammaging, its impact on cardiac aging, and the primary sources of these inflammatory factors remain to be determined.

The liver is a major endocrine organ that produces a variety of systemic factors to coordinate body’s physiology and metabolism. It can produce proinflammatory cytokine IL-6 upon infection or injury9. Patients with liver dysfunction, such as cirrhosis, often show increased cardiac arrhythmias. Furthermore, nonalcoholic fatty liver disease is a strong risk factor for cardiomyopathy10. About 30% of alcoholic hepatitis patients develop cardiomyopathy and organ failure. Together, these evidence suggest a potential cross-talk between liver and heart. It is known that aging significantly alters liver morphology and function11. Recently, using

Drosophila oenocytes as a hepatocyte model, we observed a similar downregulation of oxidative phosphorylation, and upregulation of inflammatory signaling in aged fly oenocytes12. However, it remains unclear whether liver inflammation directly influences heart function at old ages.

The liver is known to enrich with the peroxisome, a key organelle for ROS metabolism, alpha and beta oxidation of fatty acids, biosynthesis of ether phospholipids13. The peroxisome assembly and the import of peroxisomal matrix proteins are controlled by a group of peroxisomal proteins called peroxins (PEXs). Mutations in PEXs disrupt normal peroxisome function and cause peroxisome biogenesis disorders, such as Zellweger syndrome14. Several studies suggest that peroxisomal import function declines with age15,16,17. Consistently, our recent translatomic analysis shows that most peroxisome genes are downregulated in aged fly oenocytes12. However, the role of peroxisome in aging regulation is unclear.

Our findings here demonstrate a peroxisome-mediated interorgan communication between the oenocyte and the heart during Drosophila aging. We find that elevated ROS in aged oenocytes promotes cardiac arrhythmia by inducing unpaired 3 (upd3), an IL-6-like 72 proinflammatory cytokine18. Either decreasing the expression of upd3 in oenocytes or blocking the activation of JAK-STAT signaling in cardiomyocytes alleviates aging- and oxidative stress- induced arrhythmia. Finally, we show that peroxisomal import function is disrupted in aged oenocytes. Knockdown (KD) of cargo receptor Pex5 triggers peroxisomal import stress (PIS), which induces upd3 expression through c-Jun N-terminal kinase (JNK) signaling in oenocytes.

On the other hand, oenocyte-specific overexpression of Pex5 restores peroxisomal import blocks age-induced upd3 and cardiac arrhythmicity. Together, our studies reveal a nonautonomous mechanism for cardiac aging that involves in hepatic peroxisomal import-mediated inflammation.

5.3 Results

Oenocyte ROS homeostasis modulates cardiac function

Disrupted ROS homeostasis is one of the hallmarks of aging 19. Our recent translatomic analysis in Drosophila oenocytes (a hepatocyte-like tissue) revealed an overall down-regulation of anti-oxidant genes under aging, which was consistent with elevated oxidative stress in this tissue 12. To determine whether redox imbalance in oenocytes can non-autonomously impact cardiac function, we first induced oxidative stress specifically in oenocytes of female flies by crossing the PromE-Gal4 driver 20 to RNAi lines against ROS scavenger genes Catalase (Cat) and Superoxide dismutase 1 (Sod1) (Supplementary Fig. 1a-b). Heart contractility was then assessed using the Semi-automatic Optical Heartbeat Analysis (SOHA). By crossing to UAS-

GFP lines, we showed that PromE-Gal4 driver is specifically active in oenocytes of female flies

(Supplementary Fig. 1c-e). Interestingly, oenocyte-specific knockdown (KD) of Cat or Sod1 resulted in an increase in cardiac arrhythmicity, as measured by arrhythmia index (AI) (Fig. 1a).

These results suggest that disrupted ROS homeostasis in Drosophila oenocytes can modulate cardiac rhythm through an unknown non-autonomous mechanism. 73

Next, we asked whether heart function could be protected from oxidative stress and aging by maintaining redox balance in oenocytes. We first induced ROS level systemically by feeding flies with paraquat (PQ), an oxidative stress inducing agent. Feeding flies with paraquat for 24 hours induced ROS level in oenocytes, as measured by dihydroethidium (DHE) staining (Fig.

1b-c). Consistent with the previously report 21, paraquat feeding also induced arrhythmicity in fly hearts (Fig. 1d-e). Intriguingly, using an oenocyte-specific GeneSwitch driver (PromEGS-Gal4,

Supplementary Figs. 1d, 2a), overexpression of Sod1 in adult oenocytes (PromEGS-Gal4>UAS-

Sod1OE) was sufficient to block PQ-induced ROS production in oenocytes (Fig. 1b-c), as well as alleviated PQ-induced arrhythmicity in the heart (Fig. 1d-e). Similarly, overexpressing Sod1 in oenocytes attenuated aging-induced cardiac arrhythmicity (Fig. 1g-h). RU486 (mifepristone, or

RU) was used to activate PromEGS-Gal4 driver (+RU), whereas control genotype is the same, but with no RU feeding (-RU) (Supplementary Fig. 2a). RU486 feeding alone did not significantly affect cardiac arrhythmia (Supplementary Fig. 2c-e). To examine whether Sod1-mediated cardiac protection is specific to oenocytes, we crossed Sod1 overexpression line to a fat body/gut- specific GeneSwitch driver S106GS-Gal4 22 (Supplementary Fig. 2b). Overexpression of Sod1 in fat body and gut did not rescue PQ-induced arrhythmia (Fig. 1f). Together, these data suggest that oenocytes play a specific and crucial role in maintaining cardiac health during aging and

PQ-induced oxidative stress, likely through an unknown circulating factor.

Oenocyte upd3 mediates aging- and PQ-induced arrhythmia

To identify factors that are secreted from oenocytes and communicate to the heart to regulate cardiac function during aging and oxidative stress, we first compared the list of

Drosophila secretory proteins 23 with our recent oenocyte translatomic data set 12. We identified

266 secretory factors that are differentially expressed in aged (4-week-old) or PQ-treated 74 oenocytes (Fig. 2a). Among these secretory factors, we selected 27 candidates that encode for cytokines and hormonal factors in a reverse genetic screen to determine their roles in mediating oenocyte-heart communication under oxidative stress. Knockdown of several candidate factors

(e.g., sala, BG642167) in oenocytes induced cardiac arrhythmia (Supplementary Fig. 3a), similar to the knockdown of Cat and Sod1. On the other hand, our genetic screening identified four candidates whose knockdown specifically in oenocytes significantly attenuated paraquat-induced cardiac arrhythmicity (Fig. 2b). The four candidate genes are PGRP-SB1, Ag5r2, TotA, and upd3. We further verified our screening results using oenocyte-specific GeneSwitch driver

(PromEGS-Gal4) and repeated the knockdown experiments for PGRP-SB1 (Supplementary Fig.

3b) and upd3 (Fig. 2c, two independent upd3 RNAi lines used). The knockdown efficiency of upd3 RNAi was verified by QRT-PCR (Supplementary Fig. 5a). Consistent with the screening results, knockdown of PGRP-SB1 and upd3 in adult oenocytes blocked paraquat-induced arrhythmia.

Among four identified secretory factors, upd3 is a pro-inflammatory factor that belongs to four-helix bundle interleukin-6 (IL-6) type cytokine family 18. In Drosophila, upd3 is one of the three ligands that activate JAK/STAT signaling pathway. The expression of upd3 in oenocytes was higher than two other unpaired proteins (upd1 and upd2), and upd3 expression was significantly induced under aging (Fig. 2d). Although the expression of upd2 was slightly induced in aged oenocytes (Fig. 2d), upd2 KD did not block paraquat-induced arrhythmicity

(Fig. 2b). Furthermore, we found that upd3 transcripts can be detected in several other adult tissues besides oenocytes, such as abdominal fat body (FB), gut, and ovary (OV) (Fig. 2e). At young ages, the highest expression of upd3 was found in the gut (Fig. 2e). Intriguingly, upd3 expression in oenocytes increased sharply during normal aging (more than 52-fold) and became 75 the highest among all adult tissues at old ages. upd3 expression did not show age-dependent increases in gut and ovary, and it only slightly increased in aged fat body (Fig. 2e). These findings suggest that oenocytes are the primary source of upd3 production in aged flies.

Because upd3 expressed in multiple adult tissues, we wonder if upd3 produced from tissues other than oenocytes also contributes to the non-autonomous regulation of cardiac function. To test this tissue-specific effect, we knocked down upd3 using the fat body/gut- specific driver (S106GS-Gal4) and found that fat body/gut-specific upd3 KD did not alleviate paraquat-induced arrhythmia (Fig. 2f). Similar to the findings from paraquat treatment (Fig. 2c), oenocyte-specific upd3 KD also blocked aging-induced cardiac arrhythmia (Fig. 2g-h, two independent upd3 RNAi lines showed). Conversely, oenocyte-specific overexpression of upd3 at young ages induced premature cardiac aging phenotypes (high arrhythmia index) (Fig. 2i), which is similar to age-induced cardiac arrhythmia seen in multiple control flies (Supplementary Fig.

3c). In contrast, fat body/gut- or cardiac-specific overexpression of upd3 did not induce arrhythmia (Fig. 2j-k). Interestingly, fat body/gut-overexpression of upd3 slightly decreased arrhythmia (Fig. 2j). Taken together, these results suggest that upd3 is the primary cytokine that is secreted from oenocytes to regulate aging- and stress-induced cardiac arrhythmicity.

Oenocyte upd3 induce arrhythmia through JAK-STAT pathway

upd3 is known to systemically up-regulate JAK-STAT pathway in response to tissue injuries, excess dietary lipid, paraquat treatment, and bacterial infection 24-26. We asked whether oenocyte-produced upd3 can signal to the heart and activate JAK-STAT pathway in cardiomyocytes. To test this idea, we used the nuclear localization of Stat92E (fly homology of mammalian STAT transcription factors) to indicate the activation of JAK-STAT signaling.

Consistent with previous findings 26, 27, paraquat treatment induced the levels of transcription 76 factor Stat92E and promoted its localization near to the nucleus in heart tissue (Fig. 3a, b).

Interestingly, we found that oenocyte-specific upd3 KD attenuated paraquat-induced Stat92E in the heart (Fig. 3a, b). To further confirm the activation of cardiac JAK-STAT by oenocyte- produced upd3, we examined the transcription of Socs36E, a key Stat92E target gene. Consistent with the Stat92E immunostaining, the cardiac expression of Socs36E was induced by paraquat treatment, while oenocyte-specific upd3 KD diminished it (Fig. 3c). Age-dependent induction of

Socs36E in the heart was also attenuated by oenocyte-specific upd3 KD (Fig. 3d). We attempted to verify above findings using Stat93E-GFP reporters (2XStat92E-GFP and 10XStat92E-GFP)

28, however paraquat treatment did not significantly induce the reporter activities (Supplementary

Fig. 4a-c).

Activation of JAK-STAT plays a significant role in the pathogenesis of myocardial ischemia and cardiac hypertrophy 29, 30. We then asked whether blocking JAK-STAT signaling in fly hearts could protect cardiac function under oxidative stress and aging. As expected, we found that heart-specific activation of JAK-STAT signaling by expressing an active form of JAK kinase hopscotch/hop (hopTuml) in young fly heart (using heart-specific driver Hand-gal4) induced cardiac arrhythmia (Fig. 3e). Conversely, cardiac-specific knockdown of either the receptor domeless (dome) or the transcription factor Stat92E blocked PQ-induced cardiac arrhythmia (Fig. 3f-g). Similarly, knockdown of Stat92E and hop in the heart attenuated aging- induced arrhythmia (Fig. 3h-3i).

Next, we asked whether oenocyte-produced upd3 is secreted into the hemolymph and directly targeted cardiomyocytes. We overexpressed upd3-GFP fusion proteins specifically in the oenocytes and analyzed the hemolymph samples using western blotting. Using an anti-GFP antibody (with some non-specific cross-reactivities, Supplementary Fig. 3d), the upd3-GFP 77 fusion proteins were successfully detected in the hemolymph extracted from PromE >upd3-GFP flies (Fig. 3j-k), suggesting the oenocyte-produced upd3 indeed can be secreted into the hemolymph. Interesting, free GFP proteins were also found in the hemolymph, which may be due to a cleavage of the C-terminus of upd3 occurring after its secretion (Fig. 3j-k). It is known that the activities of many mammalian cytokines are regulated by proteolytic processing 31.

Human and murine IL-6 is known to be cleaved by meprin metalloproteases at its c-terminus 32.

Together, these data suggest that upd3 produced from oenocytes is released into the hemolymph and activates JAK-STAT in the heart to regulate cardiac function.

Oenocyte peroxisome import stress induces upd3 and arrhythmia

Next, we asked how aging up-regulates upd3 expression in oenocytes. In our previous oenocyte translatomic analysis, we found that genes involved in oxidative phosphorylation and peroxisome biogenesis are significantly down-regulated during aging, which are consistent with elevated ROS levels in aged oenocytes 12. It is known that mitochondria and peroxisomes are two major ROS contributors. We then investigated whether age-dependent down-regulation of genes in mitochondrial respiratory chain complexes and peroxisome biogenesis contributes to upd3 over-production and oenocyte-heart communication. Interestingly, oenocyte-specific knockdown of mitochondrial complex I core subunit ND-75 and mitochondrial manganese superoxide dismutase Sod2 showed no effects on cardiac arrhythmia (Fig. 4b). The knockdown efficiency of

ND-75 RNAi was verified by QRT-PCR (Supplementary Fig. 5b). On the other hand, knockdown of the key factors (peroxines) involved in peroxisomal import process (Pex5, Pex1, and Pex14) in oenocytes significantly induced cardiac arrhythmia (Fig. 4a-b). The knockdown efficiency of Pex1 RNAi, Pex5 RNAi, and Pex14 RNAi was verified by QRT-PCR

(Supplementary Fig. 5c-e). Pex5 is the key import factor that binds to cargo proteins containing 78 peroxisomal targeting signal type 1 (PTS1) and delivers them to peroxisomal matrix through

Pex13/Pex14 docking complex and Pex1/Pex6 recycling complex 33, 34 (Fig. 4a). Interestingly,

Pex5 itself is the major component of the peroxisomal translocon (also known as importomer, or import pore), which interacts with Pex14 and translocates cargo proteins across peroxisomal membrane through an ATP-independent process 35-37. In addition, we noticed that not all peroxisome genes were involved in oenocyte-heart communication. Oenocyte-specific knockdown of Pex19, the key peroxisomal membrane assembly factor, did not promote cardiac arrhythmia (Fig. 4a-b). The knockdown efficiency of Pex19 RNAi was verified by QRT-PCR

(Supplementary Fig. 5f). Interesting, knocking down of either Pex5 or Pex1, but not Pex19, induced the levels of ROS in oenocytes (Supplementary Fig. 5h-j). The differential regulation on

ROS metabolism by different peroxines might explain their distinct roles on cardiac arrhythmia and oenocyte-heart communication.

Our previous oenocyte translatomic analysis found that many peroxisomal matrix enzymes were significantly down-regulated during aging, such as peroxiredoxin-5 (Prx5, a thioredoxin peroxidase regulating hydrogen peroxide detoxification), dihydroxyacetone phosphate acyltransferase (Dhap-at, the key enzyme catalyzing the first step in the biosynthesis of ether phospholipids, such as plasmalogens), alkyldihydroxyacetonephosphate synthase

(ADPS, another key enzyme involved in ether phospholipid biosynthesis), and acyl-CoA oxidase

(Acox57D-d and Acox57D-p, the rate limiting enzyme for peroxisomal beta-oxidation and its deficiency has been previously linked to IL-6 induction) 38. Interestingly, knockdown of Prx5 induced cardiac arrhythmia, but not Dhap-at, ADPS, Acox57D-d, and Acox57D-p (Fig. 4c).

Thus, these data suggest that peroxisome-mediated ROS homeostasis in oenocytes plays a major role in mediating oenocyte-heart communication. 79

We then asked whether impaired peroxisomal import in oenocytes influences upd3 expression. As expected, knockdown of either Pex5 or Pex1 in oenocytes induced the mRNA levels of upd3 (Fig. 4d). Similarly, knockdown of peroxisomal antioxidant gene Cat significantly up-regulated upd3 transcription, whereas knockdown of Pex19 and ND-75 did not alter upd3 expression (Fig. 4d). Interestingly, two other unpaired proteins upd1 and upd2 remained unchanged under Pex5 KD (Fig. 4e). Given that impaired peroxisomal function induces upd3 expression in oenocytes, we speculate that Pex5 KD in oenocytes could remotely regulate JAK-

STAT activity in the heart. Indeed, we found that oenocyte-specific Pex5 KD induced Stat92E expression and Stat92E nuclear localization in the heart (Fig. 4f-g). Together, our findings showed that impaired peroxisomal import promotes upd3 production in oenocytes and non- autonomously induces cardiac JAK-STAT and arrhythmia.

Lastly, we characterized peroxisome tissue distribution in adult female flies. Peroxisomes are marked by Pmp70, a peroxisomal membrane protein involved in the transport of long-chain acyl-CoA across peroxisomal membrane 39. We noticed that peroxisome showed tissue-specific enrichment in adult Drosophila. The peroxisome abundance was the lowest in the heart and highest in oenocytes, while the number of peroxisomes in fat body (adipose-like) and pericardial cells (podocyte-like) was in the middle range (Fig. 4h-i). High enrichment of peroxisome is a key feature of mammalian liver 13. Interestingly, unlike oenocyte-specific manipulation, knocking down of Pex5 in fat body and gut using S106GS-Gal4 did not induce cardiac arrhythmia (Fig. 4j).

These results suggest that oenocyte-specific peroxisomal import contributes significantly to the hyper-production of pro-inflammatory cytokine and non-autonomous regulation of cardiac function. Here, we refer to the cellular stress responses (e.g., production of inflammatory cytokines) caused by impaired peroxisomal import as peroxisomal import stress (PIS). We term 80

PIS-induced hormonal factors as peroxikines, which are produced and released in response to peroxisomal import stresses to modulate cellular homeostasis in distant tissues.

Peroxisome import stress induces upd3 through JNK pathway

We next examined the molecular mechanism by which Pex5 KD-mediated peroxisomal import stress (PIS) induces the expression of peroxikine upd3. It is known that upd3 plays an important role in activating innate immunity and tissue repair upon infection. A recent genetic screen identified several transcription factors that are responsible for infection-induced upd3 transcription, such as Mothers against dpp (Mad), the AP-1 complex (kayak/kay and Jun-related antigen/Jra), and yorkie (yki) 40. Interestingly, we found that oenocyte-specific knockdown of

Pex5 induced the expression of kay and Jra, but not Mad and yki (Fig. 5a), suggesting kay/Jra may be the transcription factors regulating upd3 expression upon PIS. In Drosophila, kay and Jra form a transcription complex, similar to the AP-1 complex in mammals, to mediate JNK signaling 41, 42. We then asked whether PIS can also activate JNK signaling. Through immunostaining analysis, we found that Pex5 KD induced the phosphorylation of JNK in oenocytes near the nucleus (Fig. 5b-c). To further confirm above findings, we performed western blots on dissected oenocytes to assess the levels of P-JNK using an anti-phospho-JNK

(Thr183/Tyr185) antibody. We found that Pex5 KD slightly increased the phosphorylation of

JNK (Fig. 5d-e). On the other hand, we utilized a JNK reporter (TRE-DsRedT4) to monitor the transcription activity of AP-1 complex 43. Pex5 KD significantly induced the transcription activity of AP-1 (kay/Jra) in oenocytes (Fig. 5f). In addition, we confirmed this finding by measuring the mRNA expression of puckered (puc), the downstream target gene of kay.

Consistently, puc expression was also induced by oenocyte-specific Pex5 KD (Fig. 5g).

Interestingly, the activation of JNK pathway, indicated by the up-regulation of kay and puc, was 81 also observed in flies with oenocyte-specific Cat KD, but not ND-75 KD (Fig. 5h). This further suggests that peroxisome dysfunction plays a key role in activating JNK pathway. Finally, to directly examine the role of kay in mediating PIS-induced upd3 transcription, we generated fly lines with Pex5; kay double knockdown. The knockdown efficiency of kay RNAi was verified by QRT-PCR (Supplementary Fig. 5g). As expected, knockdown of kay completely blocked

Pex5 KD-induced upd3 transcription in oenocytes (Fig. 5i), suggesting that Pex5 KD-mediated

PIS up-regulates peroxikine upd3 through JNK signaling.

To determine whether upd3 and JNK signaling are required for oenocyte PIS-induced cardiac dysfunction, we analyzed the genetic interaction between Pex5 and upd3 (or kay, Jra) in non-autonomous regulation of cardiac arrhythmia. Knockdown of either upd3, or kay, or Jra specifically in oenocytes attenuated Pex5 KD-induced cardiac arrhythmia (Fig. 5j-k). upd3 KD alone did not affect cardiac arrhythmicity (Fig. 5l). Thus, these data confirm that upd3 is the primary factor produced in response to oenocyte-specific PIS and JNK activation to mediate oenocyte-heart communication.

Patients with Zellweger syndrome, due to the mutations in peroxins (e.g., PEX1 and

PEX5), often develop severe hepatic dysfunction and show elevated inflammation 44. To test whether peroxisomal dysfunction in patients with Zellweger syndrome also induces the production of pro-inflammatory cytokines, we measured the expression of IL-6 and the phosphorylation of JNK in PEX1-G843D-PTS1 cell line. This cell line is a human fibroblast cell line that was isolated, transformed and immortalized from patients with PEX1-p.G843D allele.

Similar to Pex5 KD flies, PEX1-G843D-PTS1 cells showed significant induction of IL-6 transcripts (Fig. 5m) and elevated phospho-JNK levels (Fig. 5n-o). We further examined the effects of PEX1G843D mutations on the production of hydrogen peroxide (H2O2) and peroxidase 82 activity, instead of general ROS, using Amplex Red assay kit. Consistently, PEX1G843D mutant cells showed elevated H2O2 levels (Supplementary Fig. 5k). Together, our data suggest an evolutionarily conserved mechanism for PIS-mediated induction of pro-inflammatory cytokines and activation of JNK signaling.

Peroxisomal import function is disrupted in aged oenocytes

Previous studies by us and others show that genes involved in peroxisomal import are down-regulated during aging 12, 15, suggesting an age-dependent impairment of peroxisomal import function. To monitor peroxisomal import during oenocyte aging, we first performed immunostaining using an anti-SKL antibody that recognizes peroxisomal matrix proteins containing the peroxisome targeting sequence (PTS), a SKL tripeptide sequence. We found that the number of SKL-positive punctae were significantly reduced in aged oenocytes, suggesting that the import of endogenous peroxisomal matrix proteins was impaired at old ages (Fig. 6a-b).

However, the reduced anti-SKL immunostaining might be caused by the decreased expression of the peroxisomal matrix proteins during aging 12, 15. To address this confounding effect, we directly examined peroxisomal import using a peroxisomal reporter in which YFP is engineered with a C-terminal PTS sequence (YFP-PTS). We transiently induced YFP-PTS reporter expression in young and aged oenocytes using PromEGS-Gal4 (with 1-day RU486 feeding).

Consistent with the results from anti-SKL immunostaining, aged oenocytes showed less YFP-

PTS punctae, which indicates that fewer YFP-PTS proteins were imported into peroxisomes in aged flies (Fig. 6c).

To further confirm whether the YFP-PTS punctae are indeed localized with peroxisomes, we performed co-localization analysis between YFP-PTS and peroxisome marker Pmp70. We found that the number of YFP-PTS punctae after normalized to Pmp70 signals was significantly 83 reduced in aged oenocytes (Fig. 6e), while the number of Pmp70-positive peroxisomes did not change during oenocyte aging (Fig. 6f). Consistently, the co-localization of Pmp70 and YFP-

PTS was significantly reduced in aged oenocytes according to the line scan analysis (Fig. 6d) and

Pearson’s correlation measurement (Fig. 6g). Although the number of peroxisomes (Pmp70- positive) remained unchanged during aging, there were higher number of peroxisomes that showed either none or reduced YFP-PTS signals (Fig. 6c-d). The peroxisomes with no PTS- positive matrix proteins are known as peroxisomal ghosts(or non-functional peroxisomes), which is one of the cellular hallmarks of Zellweger Syndrome 45, 46. Together, these results suggest that peroxisomal import is impaired in aged oenocytes.

Similarly, the decreased peroxisomal import was observed in oenocyte-specific Pex5 KD flies (Fig. 6h). The knockdown of Pex5 significantly reduced the number of SKL-positive punctae and co-localization between SKL and Pmp70 (Fig. 6j, l), while the number of Pmp70- positive peroxisomes was only weakly reduced (Fig. 6k). In contrast, the number of SKL containing punctae in Pex19 KD oenocytes remained unchanged (Fig. 6i-j), while Pex19 KD significantly decreased the number of Pmp70-positive peroxisomes (Fig. 6k). These results suggest that Pex5, but not Pex19, plays an important role in peroxisomal import, which may explain why Pex5 KD, but not Pex19 KD, induces ROS, upd3 production, and cardiac dysfunction.

Oenocyte Pex5 activation alleviates PIS and arrhythmia in aging

Although peroxisomal import function declines with age, the causal relationship between impaired peroxisomal import and tissue aging remains elusive. To determine the physiological significance of peroxisomal import in aging and cardiac health, we next asked whether preserving peroxisomal import function in oenocytes could attenuate the production of pro- 84 inflammatory cytokines and protect hearts from paraquat- and aging-induced arrhythmicity. Our previous oenocyte translatomic profiling showed that Pex5 is down-regulated upon paraquat treatment and aging 12. We thus used a CRISPR/Cas9 transcriptional activation system 47 to drive physiologically relevant expression of Pex5 from its endogenous locus. To express Pex5 specifically in oenocytes, we first combined a catalytically dead Cas9 line (UAS-dCas9-VPR) with oenocyte-specific driver line (PromEGS-Gal4), and then cross it with flies expressing Pex5 guide RNA. We predict that overexpression of Pex5 in oenocytes could rescue impaired peroxisomal import, block age-related induction of upd3, and preserve cardiac health.

As expected, Pex5 transcripts in oenocytes were induced about 2-fold using the dCas9-

VPR system (Fig. 7a). Interestingly, Pex5 overexpression attenuated the age-related upd3 induction in aged oenocytes (Fig. 7b). The overexpression of Pex5 also restored peroxisomal import function in aged oenocytes, as indicated by higher number of anti-SKL immunostaining

(functional peroxisomes) (Fig. 7c-d) and lower percentage of peroxisomal ghosts (non-functional peroxisomes) (Fig. 7c, e). Pex5 overexpression did not change the total peroxisome numbers

(Pmp70-positive punctae) (Fig. 7c, f). Our result is consistent with the previous study showing that overexpression of Pex5p partially restores the peroxisomal import function of pex14Δ mutants in yeast 48. Intriguingly, oenocyte-specific overexpression of Pex5 blocked paraquat- induced arrhythmicity (Fig. 7g-h) and preserved cardiac function at old ages (Fig. 7i-j). Besides protecting cardiac function, oenocyte-specific overexpression of Pex5 alleviated age-related ROS production in oenocytes (Fig. 7k-l). Together, our data showed that preserving peroxisomal import function via overexpression of endogenous Pex5 can block the production of ROS and peroxikine upd3 and protect hearts from oxidative stress- and aging-induced cardiomyopathy 85

(Fig. 7m). Thus, our studies provide strong evidence suggesting that age-related impairment of peroxisomal import is an understudied cause of tissue aging.

5.4 Discussion

Here our studies provide direct evidence linking oenocyte/liver dysfunction to cardiac health. We also demonstrate that impaired peroxisomal import at old ages is the cause of inflammaging, whereas maintaining peroxisomal import slows tissue aging. Our genetic analyses show that aging-induced pro-inflammatory cytokine upd3 from oenocytes/liver, in response to impaired peroxisomal import, activates JAK-STAT signaling in the heart to induce cardiac arrhythmia. Interestingly, either reducing upd3 or overexpressing Pex5 in oenocytes alleviates aging- or paraquat-induced cardiac dysfunction. Together, our studies established the vital role of hepatic peroxisomal import in activating systemic inflammation and non-autonomous regulation of cardiac aging. Our findings suggest that protecting oenocyte peroxisome function is critical for prolonging cardiac health span.

IL-6 is the most important pro-inflammatory cytokine that is associated with inflammaging and age-related diseases 4. The up-regulation of IL-6 is often seen upon tissue injury, such as ischemia-reperfusion 49. IL-6 is also found to be secreted as a myokine from skeletal muscle during exercise 50 to mediate the beneficial effects. However, the studies using IL-6 knockout mouse models suggest that IL-6 is pathogenic and pro-inflammatory. For example, IL-6-deficient mice show reduce expansion of CD4+ T cells and autoimmune response 51. In Drosophila, IL-6-like cytokine upd3 plays a key role in intestinal stem cell homeostasis 52-54, reproductive aging 55, glucose homeostasis and lifespan 24. Although unpaired family members, especially upd2, have been shown to relays nutrient signals from fat body to brain 56, the role of upd3 in inter-tissue communication and systemic aging regulation has not been fully established. It is not known which 86 tissue is the main source of upd3 production, and which tissuesupd3 targets during normal aging.

In the present study, we show that upd3, but not upd1 and upd2, mediates the communication between oenocytes and hearts. Among all tissues tested, oenocytes produce highest amount of upd3 at old ages. Interestingly, oenocyte specific upd3, not gut/adipocyte, modulates cardiac function during aging. Thus, our studies demonstrate that oenocytes are the main source of upd3 in aged animals. It remains to be determined why upd3 produced from different tissues exhibits distinct roles, and whether oenocyte-produced upd3 can target tissues beyond the heart (e.g., the central nervous system).

It is well-known that mitochondria play an important role in regulating inflammaging 4.

Mitochondrial stresses (such as mitochondrial unfolded protein response, UPRmt) can lead to the induction and secretion of hormonal factors, termed mitokines, which in turn activate mitochondrial stress responses in distant tissues 57, 58. There is a strong interplay between mitochondria and peroxisomes in the maintenance of intracellular redox homeostasis 59. In fact, peroxisome deficiency can alter mitochondrial morphology and respiration 60, 61, and induce mitochondria-mediated cell death 62. However, the role of peroxisome in inflammaging still remains elusive. In the present study, we discover that knockdown of peroxisomal import receptor

Pex5 triggers a significant production of cytokine upd3 in a JNK-dependent manner. Although it remains to be solved how impaired peroxisomal import activates JNK signaling pathway.

Interestingly, PIS in both flies and human fibroblasts of Zellweger spectrum patients activates JNK pathway and the production of IL-6-type cytokines. These findings suggest the existence of a conserved cellular stress response to impaired peroxisomal import. Similar to the production of mitokines upon mitochondrial stresses, impaired peroxisomal import can also induce the expression and release of a group of inflammatory factors (such as upd3/IL-6) to modulate cellular 87 function of distant tissues. We term PIS-induced hormonal factors peroxikines. The peroxikines play essential roles in maintaining tissue homeostasis upon PIS, since we can modulate the production of these factors in one tissue to protect distant tissues. The expression of the peroxikines could potentially serve as biomarkers for cellular peroxisomal import stress and inflammaging.

Age-related impairment of peroxisomal import has been previously reported in human senescence cells and nematodes 15-17, although the underlying mechanism is poorly understood.

Possible mechanisms include increased Pex5 oxidation and degradation 63, dampened Pex5 recycling activity 17, decreased ATP production 64 and decreased Pex5 expression during aging 12.

Our data suggests that functional Pex5 are significantly reduced during normal aging, which explains why overexpressing Pex5 can restore the import function and delay age-related pathologies. It should be noticed that excess expression of PEX5 can block the normal import function 65. We utilized a CRISPR/Cas9 activation system to achieve optimal expression of endogenous Pex5, which has proven to be an effective way to restore impaired peroxisomal import function and attenuate aging-induced cardiac dysfunction. Thus, our genetic analysis of Pex5 establish peroxisomal import as a cause of tissue aging.

Besides the regulation of redox homeostasis, peroxisome performs several other essential functions, such as fatty acid beta-oxidation and ether phospholipid biosynthesis. Although our studies cannot fully exclude the role of peroxisomal beta-oxidation and ether phospholipid biosynthesis in oenocyte-heart communication and cardiac aging, impaired peroxisomal import could also lead to the accumulation of very long chain fatty acids and reduced production of anti- inflammatory lipids, such as docosahexaenoic acids (DHA) 66. Thus, the decreased of DHA production, due to impaired peroxisomal import, can promote chronic inflammation in the liver 88 and induce cardiomyopathy at old ages. Further studies are needed to carefully examine these possibilities.

In summary, our studies reveal that Drosophila oenocytes (hepatocyte-like tissue) play a vital role in non-autonomous regulation of cardiac aging. In response to peroxisomal import stress, oenocytes produce pro-inflammatory peroxikine upd3, to modulate heart function. Protecting oenocytes from PIS and upd3 production can alleviate cardiomyopathy under oxidative stress and aging. Our findings suggest that peroxisome a central regulator of inflammaging and inter-tissue communication. Future work will be of interests to examine the role of liver peroxisomes in age- related cardiac diseases in mammalian systems.

5.5 Methods

Drosophila husbandry and strains

The following genotypes were used as control in the knockdown or overexpression experiments: ywR, w1118, Gal4RNAi (BDSC#35783), y1 v1; P[CaryP]attP2 (BDSC #36303), or y1 v1; P[CaryP]attP40 (BDSC # 36304). Two independent RNAi lines were used in the knockdown of upd3, Pex5, and ND-75.

Female flies were used in all experiments. Flies were maintained at 250C, 60% relative humidity and 12-hour light/dark cycle. Adults and larvae were reared on a standard cornmeal and yeast-based diet, unless otherwise noted. The standard cornmeal diet consists of the following materials: 0.8% cornmeal, 10% sugar, and 2.5% yeast. RU486 (mifepristone, Fisher Scientific) was dissolved in 95% ethanol, and added to standard food at a final concentration of 100 µM for all the experiments except for Pex5 overexpression, where 200 µM of RU486 was used. For the use of GeneSwitch driver, gene knockdown or overexpression was achieved by feeding flies on 89

RU486 food for 5 days, unless otherwise noted. For paraquat feeding assay, paraquat (dichloride hydrate pestanal, Sigma) was dissolved in distilled water to a final concentration of 10 mM

(Figs. 1-3) or 20 mM (Figs. 4-7). About 150 µl of paraquat working solution was prepared freshly and added onto the surface of fly food. Paraquat solution was air dried for 20 minutes at room temperature prior to use. In Fig. 3c, flies were fed on either 5% sucrose or 5% sucrose plus

20 mM of PQ on the agar vials. Fly heart tissues were dissected after 24 hours of feeding.

Human PEX1-G843D-PTS1 cell culture

The human PEX1-G843D-PTS1 cell lines were from the Braverman laboratory 67. It is a fibroblast cell line that was originally isolated, transformed and immortalized from patients with

PEX1-p.G843D, a missense allele that accounts for one-third of all Zellweger spectrum disorder alleles. It stably expresses GFP-peroxisome targeting signal 1 (PTS1) reporter. The PEX1-

G843D-PTS1 cell lines were cultured in DMEM with 10% FBS at 5% CO2. The wild-type cell lines were established from the fibroblasts of healthy donors.

Fly heartbeat analysis

Cardiac contractility was measured using semi-intact female Drosophila68. Female flies were dissected in oxygenated artificial hemolymph to measure heart contractions within the abdomen. Artificial adult hemolymph containing 108 mM NaCl2, 5 mM KCl, 2 mM CaCl2, 8 mM MgCl2,1 mM NaH2PO4, 4 mM NaHCO3, 15 mM 4-(2-hydroxyethyl)-1- piperazineethanesulfonic acid (HEPES), 10 mM sucrose, and 5 mM trehalose, at pH 7.1. High speed movies of heart contraction (30s, 100 frames/second) were recorded using a Hamamatsu

ORCA-Flash 4.0 digital CMOS camera (Hamamatsu) on an Olympus BX51WI microscope with a 10X water immersion lens (Olympus). The live images taken from the heart tub within 90 abdominal A3 segment were processed using HCI imaging software. M-modes and cardiac parameters were generated using Semi-automatic Optical Heartbeat Analysis (SOHA), a

MATLAB-based image application. A fifteen-second of representative M-mode was presented in the figures to show the snapshot of the movement of heart wall over time. Arrhythmia index (AI) is calculated as the standard deviation of all heart periods in each fly normalized to the median heart period. Heart period is the pause time between the two consecutive diastole 68. At least 15 fly hearts were dissected and analyzed for each group in all the experiments.

DHE staining

ROS detection was performed using DHE dye (dihydroethidium, Fisher Scientific).

Female oenocytes were dissected in 1x PBS according to a published protocol 69 with modification. Dorsal side of abdomen of female flies are placed on Vaseline, spread evenly on a culture dish. After removing unwanted organs, such as ovaries and intestine, abdominal fat body was removed by liposuction and fly carcass with intact oenocytes was incubated in 30 μM of

DHE solution for 5 minutes in the dark. After washing with PBS three times, the oenocytes were stained with Hoechst 33342 (1 μg/ml) (ImmunoChemistry Technologies) for 10 minutes, mounted in ProLong Gold antifade reagent (Thermo Fisher Scientific), and imaged with an epifluorescence-equipped BX51WI microscope (Olympus).

Measurement of H2O2 by Amplex Red

The human PEX1-G843D-PTS1 cells were homogenized in PBS and clarified by centrifugation. H2O2 amounts in the resultant supernatants were measured using the Amplex Red

Hydrogen Peroxide/Peroxidase Assay Kit (Thermo Fisher) and normalized to protein amounts. 91

Protein amounts were measured using a Qubit II fluorimeter (Thermo Fisher). Experiments were done in triplicate.

Secretory factor screen

Drosophila secretory proteins were first identified from the Gene List Annotation for

Drosophila (GLAD) 23 and compared to our recent oenocyte translatomic analysis 12. Candidate genes that are differentially expressed in aged or PQ-treated oenocytes were selected in a RNAi screening using oenocyte-specific driver PromE-Gal4. About 7-25 mated females (3~5-day-old) per genotype were treated with 10 mM paraquat (as described above) for 24 hours before heart beat analysis (SOHA). The statistical number for Fig2b is Nleft-right= 17, 35, 18, 24, 15, 18, 19, 22,

21, 15, 15, 17, 14, 21, 14, 17, 25, 14, 20, 18, 20, 18, 10, 10, 24, 17, 24, 17.

RNA extraction and quantitative RT-PCR

Adult tissues (oenocyte, heart, gut, fat body, pericardial cells) were all dissected in 1 x

PBS before RNA extraction. For oenocyte dissection, we first removed fat body through liposuction and then detached oenocytes from the cuticle using a small glass needle. Tissue lysis,

RNA extraction, and cDNA synthesis were performed using Cells-to-CT kit (Thermo Scientific).

Quantitative RT-PCR (QRT-PCR) was performed with a Quantstudio 3 Real-Time PCR system and PowerUp SYBR Green Master Mix (Thermo Fisher Scientific). Two to three independent biological replicates were performed with two technical replicates. The mRNA abundance of each candidate gene was normalized to the expression of RpL32 for fly samples and GAPDH for human samples, by the comparative CT methods. Primer sequences are listed in the Supplemental Table 2.

Antibody and immunostaining 92

The following antibodies were used in immunostaining: anti-GFP (Cell Signaling

Technology, #2956S, 1:200), anti-P-JNK (Thr183/Tyr185) (Cell Signaling Technology, #4668S,

1: 100), anti-Stat92E (1:900, a gift from Steven X. Hou), anti-Pmp70 rabbit polyclonal antibody

(1:200, generated against the Drosophila Pmp70 C-terminal region 646-665

(DGRGSYEFATIDQDKDHFGS) by Pacific Immunology), and anti-SKL rabbit polyclonal antibody (1:250, raised by Richard Rachubinski). An anti-Pmp70 Guinea Pig polyclonal antibody (A gift from Kyu-Sun Lee, 1:500) was used in peroxisome import assay (Fig. 7).

Secondary antibodies were obtained from Jackson ImmunoResearch.

Adult tissues were dissected and fixed in 4% paraformaldehyde for 15 min at room temperature. Tissues were washed with 1x PBS with 0.3% Triton X-100 (PBST) for three times

(~5 minutes each time) and blocked in PBST with 5% normal goat serum (NGS) for 30 minutes.

Tissues were then incubated overnight at 40C with primary antibodies diluted in PBST, followed by the incubation with secondary antibodies for one hour at room temperature. After washes, tissues were mounted using ProLong Gold antifade reagent (Thermo Fisher Scientific) and imaged with an epifluorescence equipped BX51WI microscope and a FV3000 Confocal Laser

Scanning Microscope (Olympus). DAPI or Hoechst 33342 was used for nuclear staining.

Image analysis and quantification

Fluorescence images were first processed and deconvoluted using Olympus CellSens

Dimensions software (Olympus). The number of punctae or fluorescent intensity/area in a selected region of interest (ROI, 918.33 µm2 for Figs. 5-6, 367.29 µm2 for Figs. 4, 7) was measured using the CellSens Measure and Count module after adjusted the intensity threshold to remove background signals. To quantify the punctae near the nucleus, we first selected ROIs surround the nucleus according to Hoechst signal, and then counted the punctae number within 93 each ROI using the CellSens Measure and Count module. Two to four ROIs were analyzed for each image. The imaging quantifications were done single or double blind.

Co-localization analysis was performed in Image J. Coloc 2 plugin function was used to calculate Pearson's correlation (r). Line scan analysis in Fig. 6 was performed by creating composite images containing SKL and Pmp70 immunostaining. We then drew a line crossing three peroxisomes in a selected ROI, and intensity plots were generated using multi-channel plot profile in the BAR package (https://imagej.net/BAR).

To identify peroxisomal ghost, images were processed, and thresholds were adjusted following the procedures described above. Image segmentation was performed for each channel

(SKL: green, Pmp70: red) using Otsu thresholding. Then SKL and Pmp70 channels were merged in ImageJ. For each merged image, three ROIs (166.41 µm2) were selected and the total number of peroxisome and peroxisomal ghosts (Pmp70-positive punctae with no SKL signals) were manually counted. The percentage of peroxisomal ghosts was presented.

Hemolymph extraction

Hemolymph was extracted by piercing into the thoraces of 30 adult female flies (each genotype) with glass needles made of capillary tubes using Sutter Puller (Sutter Instrument,

Model P-97). The following settings were used to prepare glass needle: Heat = 302, Pull =75,

Vel = 75, Time = 155, P = 435. Vacuum was used to facility the hemolymph extraction. About

0.5 µl of hemolymph extracted from 30 flies was pooled in a 1.5 ml microcentrifuge tube containing 10 µl of 1x PBS and protein inhibitor cocktail. After quick centrifuge at 402 xg for 2 minutes at 40C, the hemolymph samples were snap-freeze with liquid nitrogen and stored at - 94

800C for western blot analysis. Equal amount of hemolymph (0.5 µl in 10 µl of 1x PBS) was denatured and loaded on SDS-PAGE gels.

Western blot analysis

Proteins samples were denatured with Laemmli sample buffer (Bio-Rad, Cat# 161-0737) at 950C for 5 minutes. Then proteins were separated by Mini-PROTEAN® TGX Precast Gels

(Bio-Rad). Following incubation with primary and secondary antibodies, the blots were visualized with Pierce ECL Western Blotting Substrate (Thermo Scientific). The following antibodies were used: anti-GFP (Cell Signaling Technology, #2956S, rabbit, 1:1000), anti-P-JNK

(Cell Signaling Technology, #9255, 1:2000), anti-JNK (Cell Signaling Technology, #9252,

1:1000), anti-Tubulin (Sigma, #T5168, 1:2000). For hemolymph samples, 4–15% Mini-

PROTEAN® TGX Stain-Free™ Precast Gels (Bio-Rad, Cat# 456-8085) was used. The Stain-

FreeTM gel was UV activated (Bio-Rad) using ChemiDoc MP Imager to visualize total protein loading before immunoblotting.

Statistical analysis

GraphPad Prism (GraphPad Software, version 6.07) was used for statistical analysis.

Unpaired two-tailed Student’s t-test or one-way ANOVA (Holm-Sidak comparison) was performed to compare the mean value between control and treatment groups. The effects of mutants during aging or oxidative stress were analyzed by two-way ANOVA, followed by

Holm-Sidak multiple comparisons. The outliers were excluded using Robust regression and

Outlier removal (ROUT) method (Q=1%) prior to the data analysis. Outlier values can be found in the Source Date file. 95

5.6 References

1. Barzilai, N. & Rennert, G. The rationale for delaying aging and the prevention of age- related diseases. Rambam Maimonides medical journal 3, e0020 (2012).

2. Dai, D.F., Chen, T., Johnson, S.C., Szeto, H. & Rabinovitch, P.S. Cardiac aging: from molecular mechanisms to significance in human health and disease. Antioxidants & redox signaling 16, 1492-1526 (2012).

3. Ferrucci, L. & Fabbri, E. Inflammageing: chronic inflammation in ageing, cardiovascular disease, and frailty. Nature reviews. Cardiology 15, 505-522 (2018).

4. Franceschi, C. & Campisi, J. Chronic inflammation (inflammaging) and its potential contribution to age-associated diseases. The journals of gerontology. Series A, Biological sciences and medical sciences 69 Suppl 1, S4-9 (2014).

5. Fontes, J.A., Rose, N.R. & Cihakova, D. The varying faces of IL-6: From cardiac protection to cardiac failure. Cytokine 74, 62-68 (2015).

6. Kanda, T. & Takahashi, T. Interleukin-6 and cardiovascular diseases. Japanese heart journal 45, 183-193 (2004).

7. Bacchiega, B.C. et al. Interleukin 6 Inhibition and Coronary Artery Disease in a High- Risk Population: A Prospective Community-Based Clinical Study. J Am Heart Assoc 6 (2017).

8. Schmidt-Arras, D. & Rose-John, S. IL-6 pathway in the liver: From physiopathology to therapy. Journal of hepatology 64, 1403-1415 (2016).

9. Norris, C.A. et al. Synthesis of IL-6 by hepatocytes is a normal response to common hepatic stimuli. PloS one 9, e96053 (2014).

10. Mantovani, A. Nonalcoholic Fatty Liver Disease (NAFLD) and Risk of Cardiac Arrhythmias: A New Aspect of the Liver-heart Axis. Journal of clinical and translational hepatology 5, 134-141 (2017).

11. Sheedfar, F., Di Biase, S., Koonen, D. & Vinciguerra, M. Liver diseases and aging: friends or foes? Aging cell 12, 950-954 (2013).

12. Huang, K. et al. RiboTag translatomic profiling of Drosophila oenocytes under aging and induced oxidative stress. BMC genomics 20, 50 (2019).

13. Wanders, R.J. & Waterham, H.R. Biochemistry of mammalian peroxisomes revisited. Annual review of biochemistry 75, 295-332 (2006).

14. Wanders, R.J. & Waterham, H.R. Peroxisomal disorders: the single peroxisomal enzyme deficiencies. Biochimica et biophysica acta 1763, 1707-1720 (2006). 96

15. Narayan, V. et al. Deep Proteome Analysis Identifies Age-Related Processes in C. elegans. Cell systems 3, 144-159 (2016).

16. Weir, H.J. et al. Dietary Restriction and AMPK Increase Lifespan via Mitochondrial Network and Peroxisome Remodeling. Cell metabolism 26, 884-896 e885 (2017). 17. Legakis, J.E. et al. Peroxisome senescence in human fibroblasts. Molecular biology of the cell 13, 4243-4255 (2002).

18. Oldefest, M. et al. Upd3--an ancestor of the four-helix bundle cytokines. Biochemical and biophysical research communications 436, 66-72 (2013).

19. Lopez-Otin, C., Blasco, M.A., Partridge, L., Serrano, M. & Kroemer, G. The hallmarks of aging. Cell 153, 1194-1217 (2013).

20. Billeter, J.C., Atallah, J., Krupp, J.J., Millar, J.G. & Levine, J.D. Specialized cells tag sexual and species identity in Drosophila melanogaster. Nature 461, 987-991 (2009).

21. Lim, H.Y., Wang, W., Chen, J., Ocorr, K. & Bodmer, R. ROS regulate cardiac function via a distinct paracrine mechanism. Cell reports 7, 35-44 (2014).

22. Rera, M. et al. Modulation of longevity and tissue homeostasis by the Drosophila PGC-1 homolog. Cell metabolism 14, 623-634 (2011).

23. Hu, Y., Comjean, A., Perkins, L.A., Perrimon, N. & Mohr, S.E. GLAD: an Online Database of Gene List Annotation for Drosophila. Journal of genomics 3, 75-81 (2015).

24. Woodcock, K.J. et al. Macrophage-derived upd3 cytokine causes impaired glucose homeostasis and reduced lifespan in Drosophila fed a lipid-rich diet. Immunity 42, 133- 144 (2015).

25. Pastor-Pareja, J.C., Wu, M. & Xu, T. An innate immune response of blood cells to tumors and tissue damage in Drosophila. Dis Model Mech 1, 144-154; discussion 153 (2008).

26. Buchon, N., Broderick, N.A., Poidevin, M., Pradervand, S. & Lemaitre, B. Drosophila intestinal response to bacterial infection: activation of host defense and stem cell proliferation. Cell Host Microbe 5, 200-211 (2009).

27. Simon, A.R., Rai, U., Fanburg, B.L. & Cochran, B.H. Activation of the JAK-STAT pathway by reactive oxygen species. Am J Physiol 275, C1640-1652 (1998).

28. Bach, E.A. et al. GFP reporters detect the activation of the Drosophila JAK/STAT pathway in vivo. Gene Expr Patterns 7, 323-331 (2007).

29. Mascareno, E. et al. JAK/STAT signaling is associated with cardiac dysfunction during ischemia and reperfusion. Circulation 104, 325-329 (2001).

30. Yue, H., Li, W., Desnoyer, R. & Karnik, S.S. Role of nuclear unphosphorylated STAT3 in angiotensin II type 1 receptor-induced cardiac hypertrophy. Cardiovasc Res 85, 90-99 (2010).

31. Fu, Z. et al. Highly Selective Cleavage of Cytokines and Chemokines by the Human Mast Cell Chymase and Neutrophil Cathepsin G. Journal of immunology 198, 1474-1483 (2017).

32. Keiffer, T.R. & Bond, J.S. Meprin metalloproteases inactivate interleukin 6. The Journal of biological chemistry 289, 7580-7588 (2014).

33. Platta, H.W. & Erdmann, R. The peroxisomal protein import machinery. FEBS letters 581, 2811-2819 (2007).

34. Faust, J.E., Verma, A., Peng, C. & McNew, J.A. An inventory of peroxisomal proteins and pathways in Drosophila melanogaster. Traffic 13, 1378-1392 (2012).

35. Meinecke, M. et al. The peroxisomal importomer constitutes a large and highly dynamic pore. Nature cell biology 12, 273-277 (2010).

36. Francisco, T. et al. Protein transport into peroxisomes: Knowns and unknowns. BioEssays : news and reviews in molecular, cellular and developmental biology 39 (2017).

37. Meinecke, M., Bartsch, P. & Wagner, R. Peroxisomal protein import pores. Biochimica et biophysica acta 1863, 821-827 (2016).

38. El Hajj, H.I. et al. The inflammatory response in acyl-CoA oxidase 1 deficiency (pseudoneonatal adrenoleukodystrophy). Endocrinology 153, 2568-2575 (2012).

39. Imanaka, T., Aihara, K., Suzuki, Y., Yokota, S. & Osumi, T. The 70-kDa peroxisomal membrane protein (PMP70), an ATP-binding cassette transporter. Cell biochemistry and biophysics 32 Spring, 131-138 (2000).

40. Houtz, P. et al. Hippo, TGF-beta, and Src-MAPK pathways regulate transcription of the upd3 cytokine in Drosophila enterocytes upon bacterial infection. PLoS genetics 13, e1007091 (2017).

41. Rallis, A., Moore, C. & Ng, J. Signal strength and signal duration define two distinct aspects of JNK-regulated axon stability. Developmental biology 339, 65-77 (2010).

42. Ciapponi, L., Jackson, D.B., Mlodzik, M. & Bohmann, D. Drosophila Fos mediates ERK and JNK signals via distinct phosphorylation sites. Genes & development 15, 1540-1553 (2001).

43. Chatterjee, N. & Bohmann, D. A versatile PhiC31 based reporter system for measuring AP-1 and Nrf2 signaling in Drosophila and in tissue culture. PloS one 7, e34063 (2012).

44. Baes, M. & Van Veldhoven, P.P. Hepatic dysfunction in peroxisomal disorders. Biochimica et biophysica acta 1863, 956-970 (2016).

45. Santos, M.J., Henderson, S.C., Moser, A.B., Moser, H.W. & Lazarow, P.B. Peroxisomal ghosts are intracellular structures distinct from lysosomal compartments in Zellweger syndrome: a confocal laser scanning microscopy study. Biology of the cell 92, 85-94 (2000).

46. Soliman, K., Gottfert, F., Rosewich, H., Thoms, S. & Gartner, J. Super-resolution imaging reveals the sub-diffraction phenotype of Zellweger Syndrome ghosts and wild- type peroxisomes. Scientific reports 8, 7809 (2018).

47. Ewen-Campen, B. et al. Optimized strategy for in vivo Cas9-activation in Drosophila. Proc Natl Acad Sci U S A 114, 9409-9414 (2017).

48. Salomons, F.A., Kiel, J.A., Faber, K.N., Veenhuis, M. & van der Klei, I.J. Overproduction of Pex5p stimulates import of alcohol oxidase and dihydroxyacetone synthase in a Hansenula polymorpha Pex14 null mutant. The Journal of biological chemistry 275, 12603-12611 (2000).

49. Chandrasekar, B., Mitchell, D.H., Colston, J.T. & Freeman, G.L. Regulation of CCAAT/Enhancer binding protein, interleukin-6, interleukin-6 receptor, and gp130 expression during myocardial ischemia/reperfusion. Circulation 99, 427-433 (1999).

50. Bruunsgaard, H. et al. Exercise-induced increase in serum interleukin-6 in humans is related to muscle damage. The Journal of physiology 499 ( Pt 3), 833-841 (1997).

51. Eriksson, U. et al. Interleukin-6-deficient mice resist development of autoimmune myocarditis associated with impaired upregulation of complement C3. Circulation 107, 320-325 (2003).

52. Obata, F. et al. Nutritional Control of Stem Cell Division through S-Adenosylmethionine in Drosophila Intestine. Developmental cell 44, 741-751 e743 (2018).

53. Jiang, H. et al. Cytokine/Jak/Stat signaling mediates regeneration and homeostasis in the Drosophila midgut. Cell 137, 1343-1355 (2009).

54. Osman, D. et al. Autocrine and paracrine unpaired signaling regulate intestinal stem cell maintenance and division. Journal of cell science 125, 5944-5949 (2012). 99

55. Wang, L. et al. Pleiotropy of the Drosophila JAK pathway cytokine Unpaired 3 in development and aging. Developmental biology 395, 218-231 (2014).

56. Rajan, A. & Perrimon, N. Drosophila cytokine unpaired 2 regulates physiological homeostasis by remotely controlling insulin secretion. Cell 151, 123-137 (2012).

57. Zhang, Q. et al. The Mitochondrial Unfolded Protein Response Is Mediated Cell-Non- autonomously by Retromer-Dependent Wnt Signaling. Cell 174, 870-883 e817 (2018).

58. Kim, K.H. et al. Autophagy deficiency leads to protection from obesity and insulin resistance by inducing Fgf21 as a mitokine. Nature medicine 19, 83-92 (2013).

59. Lismont, C., Nordgren, M., Van Veldhoven, P.P. & Fransen, M. Redox interplay between mitochondria and peroxisomes. Frontiers in cell and developmental biology 3, 35 (2015).

60. Baumgart, E. et al. Mitochondrial alterations caused by defective peroxisomal biogenesis in a mouse model for Zellweger syndrome (PEX5 knockout mouse). The American journal of pathology 159, 1477-1494 (2001).

61. Peeters, A. et al. Mitochondria in peroxisome-deficient hepatocytes exhibit impaired respiration, depleted DNA, and PGC-1alpha independent proliferation. Biochimica et biophysica acta 1853, 285-298 (2015).

62. Wang, B. et al. Mitochondria are targets for peroxisome-derived oxidative stress in cultured mammalian cells. Free radical biology & medicine 65, 882-894 (2013).

63. Ma, C., Hagstrom, D., Polley, S.G. & Subramani, S. Redox-regulated cargo binding and release by the peroxisomal targeting signal receptor, Pex5. The Journal of biological chemistry 288, 27220-27231 (2013).

64. Dodt, G. & Gould, S.J. Multiple PEX genes are required for proper subcellular distribution and stability of Pex5p, the PTS1 receptor: evidence that PTS1 protein import is mediated by a cycling receptor. The Journal of cell biology 135, 1763-1774 (1996).

65. Knott, T.G. et al. The peroxisomal targeting sequence type 1 receptor, Pex5p, and the peroxisomal import efficiency of alanine:glyoxylate aminotransferase. The Biochemical journal 352 Pt 2, 409-418 (2000).

66. Terlecky, S.R., Terlecky, L.J. & Giordano, C.R. Peroxisomes, oxidative stress, and inflammation. World journal of biological chemistry 3, 93-97 (2012).

67. Zhang, R. et al. Recovery of PEX1-Gly843Asp peroxisome dysfunction by small- molecule compounds. Proc Natl Acad Sci U S A 107, 5569-5574 (2010).

100

68. Fink, M. et al. A new method for detection and quantification of heartbeat parameters in Drosophila, zebrafish, and embryonic mouse hearts. BioTechniques 46, 101-113 (2009).

69. Krupp, J.J. & Levine, J.D. Dissection of oenocytes from adult Drosophila melanogaster. Journal of visualized experiments : JoVE (2010).

5.7 Figures

H O a P=0.01 b DHE Oeno-GFP Merge c P=0.005 2 1.0 PQ P=0.04 10

H2O Ctrl

Index ns PQ 0.5 5

H2O

Arrhythmia

% DHE staining DHE %

% DHE%staining Sod1 0.0 PQ 0 Ctrl Cat Sod1 Ctrl Sod1OE T i i o 6 o E A A 2 9 2 O PromE>RNAiW N N PromE h 3 -hGal4> R R 6 Q E D e 1 9 P O s 3 OS a D D l O Q ta O S e H2O PQ S P H2O PQ d ca f P=0.029 P=0.004 H2O 2.0 1.0 ns P=0.05 Ctrl 1.5

Index Index 1.0 0.5 H O

2 0.5

ArrhythmiaIndex

Arrhythmia

Arrhythmia ArrhythmiaIndex

Sod1 PQ 0.0 0.0 -RU +RU -RU +RU O Q O Q O Q O Q 2 PGS 2 P OE PromE2 PGS>Sod12 P OE S106H H>Sod1 H U H U U U U R U R U R U R R 0 R 0 R 0 R 0 0 0 0 0 0 0 0 2 0 2 2 2 g h Young Old P=0.001 Y 2.0 ns

Ctrl 1.5

O Index 1.0

OE 0.5

Arrhythmia Arrhythmia index Arrhythmia

Sod1 O 0.0 -RU +RU O ldGSg ld OE PromE2 n>Sod1O H -O u - U o U U R -Y R R 0 U 0 0 R 0 0 2 0 Fig. 1 Oenocyte ROS homeostasis2 non-autonomously modulates cardiac function. a Arrhythmia index of oenocyte-specific Cat (n=9) and Sod1 (n=13) knockdown flies (1-week- 101 old). Ctrl genotype is PromE>attP40 (n=16). b Representative images of ROS levels in dissected oenocytes from flies fed on normal diet (white bar) or 10mM paraquat (grey bar). All flies express mCD8::GFP under PromE-Gal4. Sod1 was specifically overexpressed in the oenocytes (Sod1OE). Scale bar: 20 µm. c Quantification of the percentage of DHE-positive staining in region of interest ROIs from 5 flies (nleft-right= 13, 8, 12, 16 ROIs). d Representative M-mode showing heart contraction in control and Sod1 overexpression flies fed on normal or 10mM paraquat food. Sod1 was expressed using the GeneSwitch PromEGS-Gal4 (+RU). Ctrl genotype is PromEGS>Sod1OE with no RU (-RU). e Arrhythmia index of control and oenocyte- specific Sod1 overexpression flies fed on normal or 10mM paraquat diets (nleft-right = 17, 16, 19, 15 flies). f Arrhythmia index of control and fat body/gut-specific Sod1 (S106- Gal4>Sod1OE)overexpression flies fed on normal or 10mM paraquat diets. Overexpression specifically in fat body and gut (nleft-right = 15, 18, 21, 17 flies). g Representative M-mode showing heart contraction in young (2 weeks, white bar) and old (6 weeks, purple bar) flies with or without oenocyte-specific Sod1 overexpression. Ctrl genotype is PromEGS>Sod1OE with no RU. h Arrhythmia index of control and oenocyte-specific Sod1OE flies at young and old ages (nleft-right = 17, 19, 14, 18 flies). Data are represented as mean ± SEM. Panel P values are calculated using either two-way ANOVA (c,e,f,h) or one-way ANOVA (a), followed by Holm- sidak multiple comparisons. ns: not significant. Source data are provided as a Source Data file.

wt PQ vs 102 PGRP-bs1 * wt PQ vs Ag5r2 *

a 0.6 Aging b H O Paraquat 2 wt PQ vs TotA * P = 0.01 0.03 0.05 0.05 PQ 0.4 2652 706 362 wt PQ vs Upd3 * 93 130 43

0.2 ArrythmiaIndex

518 Index Arrhythmia

Secretory factor 0.0

O q 1 r2 tA 3 m 1 8 2 7 F C D 1 9 b 3 8 2 la 4 b 4 1 6 2 2 p b 5 d u c 1 c 6 t tC c tM 0 5 2 3 2 d a z c B 5 0 5 -s g o p t -s 6 S 1 o o e th o 2 2 c 6 6 p S p -b 0 8 8 H T U o 3 2 T T C B 5 4 p 7 8 U S im 4 3 1 P A N P 1 P 4 T 1 1 N 1 1 tin N 3 1 1 R R G R 6 G G G G c G G G G G C G G C C C C e C C C P P P B L Young PromE>RNAi Old H2O 1 week PQ 4 weeks e 1 week 6 weeks H O c d Upd3 Tissue-MS-new f 2 PQ P=0.0012 1.0 0.150.15 P=0.02 P=0.002 0.30.3 P=0.0011 ns 0.8 P=0.06 ns 0.100.10 0.6 0.20.2 0.5 0.4 0.05 0.05 expression mRNA

0.10.1 ns ArrhythmiaIndex

mRNA expression mRNA P=0.024 0.2

Arrhythmia Index Arrhythmia ArrhythmiaIndex Upd3 expressionUpd3 P=0.01 ns upd3 ns 0.0 #1 #2 0.000 0.0 Ctrl RNAi RNAi 0.000.0 -RU +RU upd3 t 1 1 2 2 upd1 upd2 u y ry te O Q Gut FBd OVa yOe 2 L L RNAiL L G o v c O Q O Q PromE P ->upd3- - - g ld g ld g ld b o GS RNAi H n n n t O n S1062 P >upd32 P U O Q O Q OenocyteO O expressionO a e 2 P 2 P u - u - u - F H U H U U R H H o 1 o 2 o 3 O U R U R R 0 U U -Y d -Y d -Y d R 0 R 0 0 U R U R 1 p 2 p 3 p 0 0 0 R 0 R 0 d U d U d U 0 2 g 0 0 0 0 p p h p Young Old i j P=0.03 2k 0 2 0 2 U U U P<0.0001 ns 2 2 1.0 1.0 0.8 P=0.008 ns 2.5 KH168 Y ns 0.8 0.8

0.6 2.0 Ctrl O 1.5 0.6 0.6 0.4

1.0 0.4 0.4 Arrhythmia Index Arrhythmia

Y 0.2 Index Arrhythmia Arrhythmia Index Arrhythmia

Arrhythmia Index Arrhythmia 0.2 Arrhythmia Index Arrhythmia

ArrhythmiaIndex 0.5 0.2 RNAi 0.0 0.0 0.0 0.0 upd3 O OE OE Ctrl upd3#1 upd3#2 Ctrl upd3 -RU+RU Ctrl upd3 g g O Q 1 1 2 2 n n GS OE Hand>0 P 2 P -L -L -L -L u u S106E >upd3E 4 F PromE>RNAi oPromEo > O O tp H U O Q O Q Y 3 3 t -G U 2 P 2 P Y d d A 3 R H H 8 P p p > d R 0 U U 1 F d 0 U R U R 1 U U n p 0 0 1 -G > > U R R w 3 S S a > 0 0 0 0 > d h d 0 2 0 2 6 p G G n Fig. 2 Pro-inflammatory upd32 produced2 from9 oenocytes mediates6 6 arrhythmia.a A Venn 3 -U 0 0 h S 1 1 A S S diagram showing the number of the predicted secretoryU proteins that are differentially expressed > 6 9 (≥ 2-fold, FDR<0.05) under aging and paraquat treatment.3 Aging and paraquat RNA-Seq data were from our previous studies. Fly ages: 10-day-old vs. 30-day-old. b Genetic screening on 27 candidate genes for their role in paraquat-induced arrhythmia. WT: Wild-type (attP2 or attP40 RNAi control lines). For statistical numbers, refer to the Methods section. c PQ-induced arrhythmia measured by SOHA for two independent upd3 RNAi lines under oenocyte-specific GeneSwitch driver (PromEGS-Gal4). Ctrl genotype is PromE>attP40. (nleft-right= 20, 18, 23, 17, 22, 18 flies). d Relative mRNA expression of upd1, upd2 and upd3 from isolated oenocytes at ages of 1 week or 4 weeks. N = 3 biologically independent samples. e Relative mRNA expression of upd3 in different tissues dissected from young (1 week) and old (6 weeks) female flies. FB: fat body, OV: ovary, oe: oenocytes. N = 4 biological samples, results pooled from 2 independent experiments. f PQ-induced arrhythmia measured by SOHA for upd3 RNAi under fat body/gut-specific GeneSwitch driver, (nleft-right= 52, 27, 17, 18 flies). g Representative M-mode traces of wild-type and oenocyte-specific upd3 knockdown flies at young and old ages. h 103

Arrhythmia index of wild-type and oenocyte-specific upd3 knockdown flies. Two independent RNAi lines used. Ctrl genotype is PromE>attP40, (nleft-right= 14, 17, 19, 20, 18, 19 flies). i Arrhythmia index for flies with ectopic upd3 expression (UAS-upd3-GFP) specifically in oenocytes. Flies are 1-week-old. Ctrl genotype is PromE>attP40 (nleft-right= 22, 28 flies). j Arrhythmia index of flies overexpressing upd3 in fat body and gut (S106GS-Gal4), nleft-right=16, 17 flies. k Arrhythmia index of flies overexpressing upd3 in the heart (Hand-Gal4), nleft-right=16, 21 flies. Data are represented as mean ± SEM. Source data are provided as a Source Data file. P values are calculated using two-tailed unpaired t-test (i,j,k), one-way ANOVA (b), or two-way ANOVA (c,d,e,f,h) followed by Holm-Sidak multiple comparisons. ns: not significant.

104

a H2O PQ b H2O PQ Stat92E DNA 80 P=0.04

-RU 60

RNAi ns

>upd3 GS 20

PromE +RU # Punctae near nucleus 0

-RU +RU # # Stat92E punctae near nucleus O Q O Q 2 PGS 2 P RNAi PromEH _ H>upd3_ _ U _ U U R U R R 0 R 0 0 0 0 (Heart) (Heart) 0 2 c d e 2 P=0.0016 P=0.008 0.6 H2O 1.5 Young 0.8 P=0.033 PQ Old P=0.02 0.6

0.4 1.0 expression expression 0.4 ns ns

0.2 0.5

rrhythmia Index rrhythmia A

Socs36E 0.2

Socs36E Arrhythmia Index Arrhythmia

Relativeexpression Rpl32 to 0.0 upd3RNAi 0.0 0.0 Ctrl -RU +RU Ctrl hopTuml

O Q O Q Expressionlevel relativeRpl32 to 2 P 2 P U GSU U U RNAi w l HPromE0 H-Gal4>3 PromER R >upd3R R Handy -mGal4> 0 4 3 d 0 0 0 0 tu 4 tp d p 0 0 > - tp t p U g ld 2 2 l4 p t A U n O a o A u g ld H O PQ h PromE-Gal4> o n O 2 j -G > f PQ y g u1.5 d o n kDal4 y P=0.002 a a Ctrl upd3-GFP H -G d Non-specific Ctrl n 75 a 1.0 H upd3-GFP

ns ns 50 GFP domeRNAi -

nti Free GFP A 25

0.5

rrhythmia Index rrhythmia A

Stat92ERNAi ArrhythmiaIndex 75 0.0 Ctrl dome Stat92E 50 O Q O Q O Q 2 PHand>RNAi2 P 2 P H H i H i 2 i i Total Protein 2 tp A A tp t A N A N 25 t A PN=0.036 R N R h A R e R T 6 weeks e m T A i 1.2 m o A PT=0.07 o D T S D S Ctrl 0.8 Stat92ERNAi

0.4

hopRNAi Index rrhythmia

A Arrhythmia Index Arrhythmia

0.0 CtrlWT STATStat92E Hophop Hand>RNAi

Fig. 3 Oenocyte upd3 activates JAK-STAT pathway in cardiomyocytes to induce arrhythmia. a Representative images of Stat92E immunostaining in the heart of oenocyte- specific upd3 KD (+RU) and control flies (-RU). Flies were treated with normal or paraquat diet for 24 h. Arrows indicate cardiomyocyte nuclei. White boxes indicate the regions shown in the right insets. Scale bar: 20 µm (inset: 5 µm). b Quantification of the Stat92E-positive punctae near cardiomyocyte nucleus. N = 7 flies. Data presented are representative of two independent experiments. c, d QRT-PCR analysis on Socs36E mRNA expression in the heart of control (attP40) and flies with oenocyte-specific upd3 KD under either paraquat treatment or aging. In paraquat treatment (Panel c), 4 independent biological samples are shown, results are pooled from 3 experiments. In aging study (Panel d), flies with two ages (1 week vs 5 weeks) are collected. N = 4 biological samples from 2 independent experiments. e Arrhythmia index of young flies (1-week-old) with heart-specific expression of an activated form of hop (hopTuml) (nleft-right= 21, 26 flies). f Representative M-mode of paraquat-treated wild-type, heart-specific dome and Stat92E RNAi flies. g Arrhythmia Index of paraquat-treated wild-type, heart-specific 105

dome and Stat92E KD flies. Ctrl genotype is Hand>attP40 (nleft-right= 20, 18, 14, 17, 15, 17 flies). h Representative M-mode of heart-specific Stat92E and hop RNAi flies at old ages. i Arrhythmia Index of heart-specific Stat92E and hop RNAi flies at old ages. Ctrl genotype is Hand>attP40. (nleft-right= 28, 33, 18 flies). j Western blot analysis on the hemolymph samples extracted from flies expressing upd3-GFP fusion proteins in oenocytes. Two biological replicates are shown. Total protein loaded onto the Bio-Rad Stain-Free gel was visualized using ChemiDoc MP Imagers after UV activation. Ctrl genotype is PromE>attP40. k Quantification of western blots in Panel j. The data represent the intensity of GFP bands normalized to the total protein. Data are represented as mean ± SEM. P values are calculated using two-tailed unpaired t-test (e,k) or one-way ANOVA (i), or two-way ANOVA (b,c,d,g) followed by Holm-Sidak multiple comparisons. ns: not significant. Source data are provided as a Source Data file.

106

a b -RU +RU c -RU +RU PTS1 P=0.002 1.6 ns Pex5 1.2 ns ns P=0.0009 ns P=0.01 ns 1.2 ns ns P=0.0003 Pex14 ns 0.8 Pex1 ns Pex13 Pex6 0.8

Pmp70 0.4 0.4 Arrhythmia Index Arrhythmia

Pex19 PTS1 ArrhythmiaIndex

Arrhythmia Index Arrhythmia Arrhythmia Index Arrhythmia Peroxisome 0.0 Pmp70 matrix 0.0 l ) l ) l ) l i l l i tr 1 tr 2 tr 2 O 2 tr A tr 4 tr A C ( C ( C d 2 ( C C 1 C rl rl rl 5 i i o H i N x N U U U U U U t U t U t x Pex19 A A S A R e R R R R R R R c R c R C r N N U N 1 P 9 0 0 0 0 0 0 d 0 p 0 P R R R R Peroxisomex 1 t 0 2 0 S 0 - 0 - 0 Mitochondrion 0 e x -a 2 _ 2 2 D 2 D 2 5 5 5 P e p t t 2 P S 7 d 7 p 7 7 x P a -a -a _ D 5 - 5 - D D e h p p t A P x D x D PromEP GS>RNAi a a -a PromED oGS>RNAi7 o 7 N N D h h p A c 5 c 5 a A x A x D D h o o D c c PromEGS> Pex5RNAi A A d -RU +RU e -RU +RU f g 0.15 P=0.01 P=0.026 0.10 P=0.0001

s 6 0

u e 0.08 l

P=0.004 -RU c u

0.10 Stat92E n

P=0.02 r 4 0

0.06 DNA a

ns a

0.04 t

0.05 Index Arrhythmia

c mRNA expression mRNA

n 2 0

ns ns u

mRNAexpression p

0.02

mRNAexpression t

upd3 ns a

+RU t S

0.00 0 0.00 # ) ) ) ) ) ) ) ) ) ) 2 2 U U U U U U U U ) ) # Stat92E punctae near nucleus -RU +RU (L (L R R R R R R R R 1 1 U U U U U U U U (- + (- + 0 0 (- + L L R R R R PromER GS>Pex5R RNAi ( ( ( 0 ( ( ( 0 0 0 0 0 0 -R R 1 1 e e 9 2 5 U U 0 0 0 + x Peroxisomex s s 1 ( 7 Mito5 1 2GS 2 2 RNAi 2 e e la a x 9 - -7 -R PromER d >Pex5d P a l e 1 D + p 1 p 2 P t ta P xGS N D U d U d a aPromEe >RNAiN p p C C P U U h i P<0.0001 j -RU +RU P<0.0001 P=0.03 Pmp70 Fat body Heart 1.2 DNA 2 0 0

n P<0.0001

e r

1 5 0 r

e 0.8 ns

m 1 0 0

s i

Pericardial cell x

Oenocyte Index Arrhythmia o 0.4

r 5 0

Arrhythmia Index

# 0 # Pmp70 punctae # Pmp70punctae region per

t 0.0 r y ls e GS GS a d l t PromE S106 e o e y b c c l ) H t l o r U U ia n t 1 a e C ( R RNAiR F rd i Pex50 0 a O A c b 0 ri N f 2 e R b P f x5 e P Fig. 4 Impaired peroxisomal import in oenocytes induces upd3 and promotes cardiac arrhythmia. a Schematic diagram showing the key genes involved in peroxisomal import and membrane assembly. b Arrhythmia index of oenocyte-specific knockdown of mitochondrial complex I subunit ND-75 (two independent RNAi lines), mitochondrial Mn superoxide dismutase Sod2, peroxisomal import factors (Pex5, Pex1, Pex14), and peroxisomal membrane assembly factor (Pex19), (nleft-right= 14, 15, 14, 11, 16, 17, 18, 18 flies). Oenocyte-specific GeneSwitch driver (PromEGS-Gal4) was used. c Arrhythmia index of oenocyte-specific knockdown of peroxisomal matrix enzymes, Dhap-at, ADPS, Acox57D-d, Acox57D-p, Prx5 (nleft-right= 17, 18, 18, 18, 18, 17, 21, 20, 18, 16, 17, 14, 17 flies). d QRT-PCR analysis showing relative mRNA levels of upd3 from dissected oenocytes of Pex5, Pex1, Cat, Pex19 and ND-75 KD flies. Mito: Mitochondrion. N = 4 biological samples from 2 independent experiments. e QRT-PCR analysis showing relative mRNA levels of upd1, upd2, upd3 from dissected oenocytes of oenocyte-specific Pex5 KD flies. N = 3 biological samples, from 2 independent experiments. f Representative images of Stat92E immunostaining of heart tissues dissected from flies with (+RU) or without (-RU) oenocyte-specific Pex5 KD. Arrows indicate cardiomyocyte nuclei. White boxes indicate the regions shown in the right insets. Scale bar: 20 µm (inset: 5 µm). g 107

Quantification on the number of Stat92E punctae around cardiomyocyte nuclei of oenocyte- specific Pex5 KD flies (blue dots). Dot plot shows the quantifications of 6 biological replicates, 3-4 selected regions of interest (ROIs) per replicate. h Representative images of Pmp70 immunostaining in fly tissues. Arrows indicate cardiomyocyte and pericardial cell nuclei. Scale bar: 6.7 µm. i Quantification of Pmp70-positive peroxisomes per region of interest. Dot plot shows the quantifications of 6 biological replicates, 4 ROIs per replicate. j Arrhythmia of flies GS GS with Pex5 KD in either oenocytes (PromE -Gal4) or fat body/gut (S106 -Gal4) (nleft-right= 19, 22, 13, 17 flies). Data are represented as mean ± SEM. P values are calculated using two-sided unpaired t-test (g), one-way ANOVA followed (i), or two-way ANOVA (b,c,d,e,j) by Holm- Sidak multiple comparisons. ns: not significant. Source data are provided as a Source Data file.

To better show the result, Jra's values are divided by 10. 108

-RU +RU a c b PromEGS>Pex5RNAi d e 0.15 P=0.008 GS RNAi

s 4 0 PromE >Pex5

P-JNK u 1.4

e P<0.0001 l P = 0.1

DNA c kDa -RU +RU u

n 3 0 0.10 1.2 r 50

-RU a

JNK

P 37 1.0

e 2 0 a

0.05 t

ns / JNK JNK c

P=0.001 -

n P u 50 0.8

p 1 0

mRNA expression mRNA

K JNK N 37

J 0.6 mRNAexpression

ns - 0 P -RU +RU +RU U U -RU +RU R R 0 0 U U 0 0.00 R R GS 2 0PromE0 > Mad kay Jra yki 0 d d y y a a i i 2 RNAi a a a a r r k k Pex5 K K GS J J Y YRNAi M MPromE U U>Pex5U U U U U U R R R R g h -RU +RU R R R R 0 0 0 0 0 0 0 0 0 0 0 0 2 2 0.6 P=0.034 2 2 0.15 80000 f PromEGS>Pex5RNAi; TRE-DsRed P<0.0001 P=0.04

-RU +RU -RU 60000 0.10 0.4

y t Oe i P=0.03

Oe intensity s 40000

n P=0.04

e expression t P<0.0001 n 0.05 0.2 +RU I Oe puc

DsRed 20000

Oe expression mRNA

mRNA expressionmRNA mRNAexpression

TRE-DsRed 0 0.00 0.0 -RU +RU -RU+RU kay puc kay puc U U U U R R GS GS ) ) ) ) y y c c PromE0 0 > RPromER > y y c c a a u u 0 0 0 a a u GSu K PromEK P PGS> Pex52 RNAi c 0 RNAi (K PromE(K (P (P > - - - - u Pex52 U U URNAiU P c U U NDU -U75RNAi R R CatR R u R R R R 0 0 0 0 P 0 0 0 0 t- 0 t- 0 -RU +RU 0 0 a 2 a 2 -RU +RU 2 2 t- t- i j P=0.01 1.0C a C ans k l C C 20 1.0 ns P=0.006

0.8 .)

f.c 15 Pex5RNAi 0.6 ns 10 AI Pex5RNAi; 0.5 ns 0.4 RNAi

expression ( expression upd3 Arrhythmia Index Arrhythmia 5 Index Arrhythmia ns 0.2

upd3 RNAi Pex5 ; ArrhythmiaIndex kayRNAi 0 RNAi 0.0 Ctrl kay 0.0 -RU +RU rl i rl i ) ) t AGSct A RNAi 3 3 PromEC N >Pex5N 3 3 ) ) PromEd GSd >upd3RNAi R R E E d d y y U U p p 5 y m m a a R R x a o o p p K K 0 0 (U (U WT e K r r ;U U ( ( 0 P P P ; GS ra 2 RNAi U U i; ; ; 5 PromE5 U U J>Pex5 A 5 5 x x R R a R R N x x e e 0 0 Jr 0 0 ildtype G843D P 0 Pex1 null m R6 n WT PEX1 o e e P 0 2

5 P P U 2

W Pex1null x P=0.0067 Pex7null U G843D e U U RR WT PEX1 P R R 0 0 PromE(GS) > Upd3i 15 0 0 0 0 2 2 P-SAPK/JNK P-JNK 50 kDa P<0.0001 4 37 kDa 10

expression 2

6 -

50 kDa Band intensity IL SAPK/JNK JNK 5 37 kDa ns ns 0

t llTubulin Tubulin 0 w u 50 kDa P-JNK JNK Tubulin n 1 K K lin x N N u e -J J b P P tu

Fig. 5 Pex5 KD-mediated PIS induces upd3 through JNK signaling. a Expression of Mad, kay, Jra, yki in oenocytes of PromEGS>Pex5RNAi flies, n = 2 biological samples from 2 independent experiments. b P-JNK immunostaining of Pex5 KD oenocytes with (+RU, blue dots) or without (-RU, grey dots). Scale bar: 6.7 µm. c Quantification of b. Quantifications of 6 biological replicates (8 nuclei per replicate). d. Western blots showing the levels of P-JNK and 109

JNK (arrows) in control (-RU) and Pex5 KD (+RU) oenocytes. e Quantification of Panel d. N = 2 independent biological samples. f Fluorescence imaging of JNK reporter TRE-DsRed from control (-RU) and oenocyte Pex5 KD (+RU). Scale bar = 20 μm. Quantification is shown on the right. N=9 flies. g Expression of puc in oenocytes from flies with Pex5kd oenocyte. N = 4 independent biological samples. h Expression of kay and puc in oenocytes dissected from flies with oenocyte-specific knockdown of ND-75 and Cat. N = 4 independent biological samples. i Expression of upd3 in oenocytes dissected from flies with Pex5 KD or Pex5; kay double KD. N = 4 biological samples. Ctrl genotype is UAS-Pex5RNAi/attP40; PromEGS-Gal4/+. + indicates wild type. j Representative M-mode of flies with oenocyte Pex5 KD, Pex5; upd3 double KD, or Pex5; kay double KD. k Arrhythmia of flies with oenocyte Pex5 KD, or Pex5; upd3 double KD, or Pex5; kay double KD, or Pex5; Jra double KD. Ctrl genotype is the same as i (nleft-right= 34, 30, 16, 17, 20, 16, 20, 13, 13, 14 flies). l Arrhythmia of flies with oenocyte upd3 KD (nleft-right= 20,17). m Expression of IL-6 in human PEX1-G843D-PTS1 cells. N = 3 independent biological samples. n Western blots showing P-JNK and JNK in wild-type and human PEX1-G843D-PTS1 cells. o Quantification of Panel n. nJNK/tubulin = 2, nPJNK = 3 biological samples. Data are represented as mean ± SEM. P values are calculated using two-sided unpaired t-test (c, e, f, g, l, m), two-way ANOVA followed by Holm-sidak multiple comparisons (a, h, i, k, o) ns: not significant. Source data are provided as a Source Data file.

110

a Young Old b P=0.035 Anti-SKL 1 0 0 0

DAPI n

o i

g 8 0 0

e 6 0 0

L K

S 4 0 0

n a

2 0 0

# # # SKL punctae per region 0 Young Old g ld n u O o c Y Pmp70 Young Pmp70 YFP-PTS Merge Zoom-in d 300300 e Young f Young g YFP-PTS0 Old Old YFP-SKL 7 Old

P 0.90 Pmp70 P<0.0001

200 0.8 800 ns M

200 )

t P

DAPI i

(

s P<0.0001

n n

o 0.85

i g

100 d t

Intensity 0.6 600

100 e

Young

i 0.80

0 0 m c r 400

0.4 0

0 9 18 0.75

o 7 0 320 640 n

Distance n

P o

150

s e

150 M

a a

P 0.70 t

0.2 200

n o

P 0.4

100 u

100

s n

P 0.0 0 0.0

Old

n F I # YFP # of Pmp70 5050 Y

g d d d Intensity n l g l g l

# n n upunctaeO peru regionO u O o o o Y /Pmp70 Y Y 00 0 5 10 15 320Distance640 960 Distance (nm) h Pmp70 SKL Merge i Pmp70 SKL Merge Pmp70 Pmp70 Anti-SKL Anti-SKL

DAPI DAPI

+RU +RU

PromEGS >Pex5RNAi PromEGS >Pex19RNAi

0 0

j 0 k l

7 7

P<0.00017 ns

P P

1.5 1.5 P 1.5 500 500 500 P=0.00141.0 1.0 P<0.00011.0

M M

M P=0.022 P<0.0001

t t

t 0.8 0.8 0.8

400 400 400

z z

i 1.0

r i

1.0 i 1.0

0.6 0.6 0.6

r r

a 300 300 300

0 0

0 0.4 0.4 0.4

7 7

7 200 200 200

e e

0.5 0.5 e 0.5

c c

c 0.2 0.2 0.2

100 100 100

# of Pmp70 per # of per Pmp70region

Pearson correlationPearson (r)

Pearson (r) Pearson correlation

# # SKL punctae / Pmp70

L L

0.0 0.0 L 0.0 0 0 0 0.0 0.0 0.0

K K

K -RU +RU -RU +RU

S S S -RU +RU -RU +RU

-RU +RU

-RU +RU U U UU UU UU UU U U

U U U U UU UU UU UU UU UU UU UU U U

# # # R R RR RR RR RR R R R R R R RR RR RR RR RR RR RR RR -0 R 0 R -0-0 00RNAi -0-0 0 0RNAi -0 0 -0 0 --00 RNAi0 --00 00RNAi --00 00 -0-0 00RNAi -0-0 00 RNAi9 -0 0 0 95 Pex500 5 9 Pex190 0 5 0 9 0 95 Pex50 59 Pex1900 59 00 95 Pex500 59 Pex1900 1 5 -2 0 1x -2-2 x 1 -2-2 x -2 1 -2 1x --2 x1 --22 x1 --22 1x --22 x1 -2-2 x x 9 -2 xe 95 e x 5 9 e 5 x 9 xe 95 ex 59 ex 59 xe 95 ex 59 e e 1 5 eP 1x P e xGS1 P x e 1 eP 1x Pe GSx1 Pe x1 eP 1x PromEPe GSx1 >P P x x P xe PromEP e x > e P x P xe PromEP ex > P ex P xe P ex e e eP P e P e Pe Pe Pe eP Pe P P P P P P P P P P

Fig. 6 Peroxisomal import function is disrupted in aged oenocytes. a Representative images of anti-SKL immunostaining of young and aged wild-type oenocytes. Scale bar: 6.7 µm. b Quantification on the number of SKL-positive punctae in Panel a. Dot plot shows the quantifications of 6 biological replicates, 2 ROIs per replicate. c Representative images to show co-localization of Pmp70 and YFP-PTS in young and aged oenocytes (Scale bar: 6.7 µm). Insets 111 on the right show zoom-in peroxisome structures (the regions indicated by the white boxes in the merged panels, Scale bar: 125 nm). d Line scan analysis to show the fluorescence intensity of Pmp70 (red) and YFP-PTS (green) crossing peroxisomes in the insets (dashed line). e Quantification of the number of YFP-positive punctae normalized Pmp70 in Panel c. N = 6. Dot plot shows the quantifications of 6 biological replicates, 4 ROIs per replicate. f Quantification of the number of Pmp70-positive punctae in Panel c, N = 6 biological replicates, 4 ROIs per replicate. g Pearson correlation quantification measuring the correlation coefficiency of the co- localization between Pmp70 and YFP-PTS in Panel c. N = 6 biological replicates. h Representative images of anti-SKL and anti-Pmp70 immunostaining of PromEGS>Pex5RNAi oenocytes. Scale bar: 6.7 µm. i Representative images of anti-SKL and anti-Pmp70 immunostaining of PromEGS>Pex19RNAi oenocytes. Scale bar: 6.7 µm. j Quantification of the number of SKL-positive punctae normalized Pmp70 in Panel h, i. k Quantification of the number of Pmp70-positive punctae in Panel h, i. l Quantification of Pearson correlation coefficiency of the co-localization between Pmp70 and SKL in Panel h, i. 6 biological replicates, 3 ROIs per replicate. Data are represented as mean ± SEM. P values are calculated using two-sided unpaired t-test (b,e,f,g). or two-way ANOVA by Holm-Sidak multiple comparisons (j,k,l), ns: not significant. Source data are provided as a Source Data file. For specific statistical number, please refer to the source data.

112

(4 weeks) (6 weeks) a b c Pmp70 SKL Merge d e P=0.039 f 0.04 P=0.03 P=0.003 0.15 3 0 0 1 0 0 2 5 0 ns

n P=0.002

o i

e 2 0 0

0.03 m

Ctrl

0.10 e 2 0 0

e 1 5 0

0.02 a 5 0

7 t

n 1 0 0

P s

u 1 0 0

o M

0.05 p

mRNA expression mRNA

mRNA expression mRNA L

0.01 f

5 0

Pex5 mRNA level

mRNAexpression

% Peroxisomal ghost Peroxisomal %

Pex5

Pex5 #

upd3 0 0 0 0.00 0.00 OE OE OE Ctrl OE Ctrl Pex5 Ctrl Pex5OE Ctrl Pex5 Ctrl Pex5 U Pex5U U U U U U R R R R R R U U U 0 0 GS 0 0gRNA 0 0 RGS R gRNA R R (PromE0 -Gal4/Pex50 ;UAS-dCas9-VPR/+0 ) (PromE 0 -Gal4/Pex50 ;UAS0 -dCas90 -VPR/+) 2 2 2 5 0 0 x 2 2 e 5 PQ x h H2O 30 P e (6 weeks) P<0.0001 g P k l 1.5 P=0.013 DHE DAPI Merge H2O 20

Ctrl Ctrl

PQ 1.0 10

ns OE

% DHE staining % staining DHE

Pex5 0 H O 0.5 Ctrl Pex5OE 2 U U Arrhythmia Index Arrhythmia R R 0 0 Arrhythmia Index 0 Pex5OE m 2 OE PQ 0.0 Pex5 OE Peroxisomal import Ctrl Pex5 Peroxisomal O Q O Q 2 P 2 P ROS import H _ H _ U _ U U R U R R 0 RC0o m b in e d JNK i j 0 0 0 0 2 Y 0 .6 0.6 2 P=0.053 ROS 2 0 0 R U Ctrl upd3 Peroxikine JNK O 0 .4 0.4

I upd3 A

0 .2 0.2 Y Ctrl JAK-STAT ArrhythmiaIndex Pex5OE OE Pex5 JAK-STAT 0 .0 0.0 Cardiac O 0 22 44 6 8 Cardiac arrhythmia A g e arrhythmia Age (weeks)

Fig. 7 Oenocyte-specific Pex5 activation alleviates age-related PIS and preserves cardiac function. a QRT-PCR analysis showing expression of Pex5 in oenocytes dissected from flies with oenocyte-specific overexpression of Pex5. N=3 biological samples from 2 independent experiments. b QRT-PCR analysis showing relative mRNA expression of upd3 in 4-week-old oenocytes dissected from flies with oenocyte-specific overexpression of Pex5. N=4 independent biological samples. c Representative images to show co-localization of anti-Pmp70 and anti-SKL immunostaining in 6-week-old oenocytes dissected from flies with oenocyte-specific overexpression of Pex5 (yellow dots). Scale bar: 6.7 µm. d Quantification of SKL-positive punctae in Panel e Quantification of the percentage of peroxisomal ghosts (Pmp70-punctae with no SKL signals) per region of interest (166.41 µm2) in Panel d. f Quantification of Pmp70- positive punctae in Panel. d-f. Dot plot shows the quantifications of 6 biological replicates, 6 ROIs per replicate. g Representative M-mode of flies with oenocyte-specific Pex5 overexpression under paraquat treatment. h Arrhythmia index of flies with oenocyte-specific Pex5 overexpression under paraquat treatment. (nleft-right= 15,14,14,21 flies). i Representative M- mode of flies with oenocyte-specific Pex5 overexpression during normal aging. j Arrhythmia index of flies with oenocyte-specific Pex5 overexpression (red dots) during normal aging (2- week, 4-week and 6-week-old), n0RU, y-o= 35, 15, 15 flies; n200RU, y-o= 9, 15, 18 flies. k DHE staining in 6-week-old oenocytes from flies with oenocyte-specific overexpression of Pex5. 113

Scale bar: 20 µm. l Quantification of the percentage of DHE-positive staining in a Panel k. N = 6 flies, 3 ROIs per replicate. Ctrl genotype is PromEGS>Pex5OE with no RU feeding. Data are represented as mean ± SEM. P values are calculated using two-sided unpaired t-test (a,b,d,e,f,l,j) or two-way ANOVA, followed by Holm-sidak multiple comparisons (h). m Proposed model to show that impaired hepatic peroxisomal import promotes ROS production, JNK activation, and peroxikine upd3 expression, which non-autonomously controls cardiac JAK-STAT and arrhythmia. Overexpression of endogenous Pex5 preserves cardiac health. Source data are provided as a Source Data file.

Supplementary Fig. 1

a d b c GFP.nls DHE GFP DAPI DAPI P=0.007

Ctrl 2 .5 GFP.nls

> GS

2 .0

PromE n

i 1 .5 a

t P=0.002

E 1 .0 e

CatRNAi H mCD8::GFP D DAPI % 0 .5

0 .0

i i T A A >mCD8::GFP W N N RNAi R R Sod1 t 1 a d C o PromE PromES -Gal4> PromEGS>GFP.nls PromE-Gal4>

Supplementary Fig. 1 a DHE staining and PromE-gal4 specificity. Representative images to show DHE staining in oenocytes dissected from control and oenocyte-specific Sod1 KD flies (PromE-Gal4>UAS-Sod1RNAi). Hoechst 33342 was used for nuclear staining. Scale bar: 20 µm. b Quantification of relative DHE staining in control and oenocyte-specific Sod1 KD flies (PromE-Gal4>UAS-CatRNAi). Data are represented as mean ± SEM. P values are calculated using one-way ANOVA followed by Holm-sidak multiple comparisons, ns: not significant. n = 5 flies, 2 ROI per replicate. c, d, e Verification of oenocyte-specific driver (PromE-Gal4), oenocyte- specific GeneSwitch driver (PromEGS-Gal4). n = 5 flies. RU: mifepristone (RU486). Dashed line delineate cardiac cells. Scale bar: 20 µm. Source data are provided as a Source Data file.

114 Supplementary Fig. 2

a Oenocyte b Gut Fat body

-RU

+RU S106GS > UAS-GFP.nls (+ RU)

PromEGS >YFP-PTS

c 0.8 d 0.8 e 0.8

0.6 0.6 0.6 ns

ns ns

I I

0.4 0.4 I 0.4

A Arrhythmia Index Arrhythmia 0.2 Index Arrhythmia 0.2 Arrhythmia Index Arrhythmia 0.2

0.0 0.0 0.0 -RU +RU -RU +RU -RU +RU U U U U U U R R R R R R 0 0 GS R 0 0GS 0 0GS RNAi PromE 0 >yw PromE0 >attP40 PromE0 >Gal4 yw 2 2 2 yw Supplementary Fig. 2 a Verification of oenocyte-specific GeneSwitch driver (PromEGS- Gal4). RU: mifepristone (RU486). b Verification of fat body/gut-specific GeneSwitch driver (S106GS-Gal4). Scale bar: 20 µm. n = 5 flies. Data presented here are representative of two independent experiments. c, d, e The effect of RU486 feeding on arrhythmia of three wild-type flies (nleft-right = 20, 16, 11, 16, 17, 19 flies). P values are calculated using two-sided unpaired t- test. Source data are provided as a Source Data file. For specific statistical number, please refer to the source data.

115 Supplementary Fig. 3

a P<0.001 b c H2O Young Old P=0.0061 1 . 0 P=0.0013 PQ P=0.016 P<0.001 1 . 0 0 . 8

0 . 6 ns

A A

0 . 5 0 . 4

Arrhythmia Index Arrhythmia 0 . 2 Arrhythmia Index Arrhythmia

0 . 0 0 . 0

O O O O O O Q O Q RNAi 1118 2 2 2 2 2 2 -RUP 2 +RUP Gal4 W H H H H H H H 7 U U T 2 0 l a p 4 6 U R GS U R RNAi W t a PromE >PGRP-SB1 PromE-Gal4> t p 1 R 0 R 0 t S 2 0 0 0 A PromEt -Gal4> 1 4 0 2 A 6 1 B 2 1 G B S S - 1 B B - P B S P R S - R - P G P G R P R d P G G P P ywR w1118 ywR w1118 kDa kDa 75 75

50 50 37 37

25 25

Anti-GFP Total Protein

Supplementary Fig. 3a Arrhythmia index of wild-type flies (PromE-Gal4>w1118, PromE-Gal4>Attp2, PromE-Gal4>attP40) and oenocyte-specific knockdown of sala and BG642167 (nleft-right = 14, 14, 22, 11, 9 flies). b Paraquat (PQ)-induced arrhythmia measured by SOHA for PGRP-SB1 knockdown under oenocyte-specific GeneSwitch driver (PromEGS- RNAi Gal4>PGRP-SB1 ) (nleft-right = 20, 15, 24, 15 flies). c Arrhythmia index of wild-type flies (PromE-Gal4>UAS-Gal4RNAi and PromE-Gal4>w1118) at young (2-week-old) and old age (6- week-old), (nleft-right = 14, 17, 22, 17 flies). Data are represented as mean ± SEM. P values are calculated using one-way ANOVA, followed by Holm-sidak multiple comparisons ns: not significant. d Western blot analysis on the protein extracts from the whole body of two wild-type flies, ywR and w1118. Total protein loaded onto the Bio-Rad Stain-Free gel was visualized using ChemiDoc MP Imagers after UV activation. Source data are provided as a Source Data file. Data presented here are representative of two independent experiments. For specific statistical number, please refer to the source data.

116 Supplementary Fig. 4

a H2O PQ b

2 5 0 0 0 ns

2 0 0 0 0

u l

a 1 5 0 0 0

t i

2XStat92E-GFP s 1 0 0 0 0

t n I 5 0 0 0

0 H2O PQ O Q 2 P H - - X X 2 c 2 ns 2 5 0 0 0

2 0 0 0 0

10XStat92E-GFP e

u l

a 1 5 0 0 0

i s

n 1 0 0 0 0

t n I 5 0 0 0

0 H2O PQ O Q 2 -P -H X X 0 0 1 Supplementary Fig. 4 a Representative images of 2XStat92E1 -GFP or 10XStat92E-GFP reporters under paraquat feeding condition or water fed conditions. Scale bar: 20 µm. b, c Quantification of reporter staining in Panel, n = 4 flies. P values are calculated using two-sided unpaired t-test. a. Source data are provided as a Source Data file.

Supplementary Fig. 5 117

-RU +RU

a b c d d e P<0.0001 f g P=0.02 P<0.001 0.004 0.04 P<0.0001 P=0.028

P=0.009 n

P<0.001 n

o o

P=0.001 i 0.03

i 0.003

p p

x 0.02

x 0.002

expression

x x

e 0.01 kay

e 0.001

Pex14

Pex19

upd3

Pex5

P Pex1

0.000 0.00 upd3RNAi ND-75RNAi Pex1RNAi RNAi: Ctrl kay Pex14RNAi Pex19RNAi kayRNAi U U U U GS GS GS R RGS R RGS GS PromE > PromE > PromE > PromE0 0 > PromE0 0 > PromE > GS RNAi 0 0 PromE >Pex5 2 2

h DHE DAPI Merge 5 P=0.034 1.5

P=0.002 y

4 t i

-RU g

n n

i 1.0

n i

3 t

P E

2 A

D /

D 0.5

% H

+RU staining DHE % 1

D DHE / intensity DAPIDHE

0 0.0 -RU+RU -RU+RU GS RNAi U U U U PromE >Pex5 R R R R 0 0 0 0 0 0 2 2 DHE DAPI Merge i 4 0.4

P=0.021 P=0.0016

i g

3 s 0.3

i e

-RU n

I s

2 0.2

1 E 0.1

H % DHE staining DHE %

+RU D DHE / DHE DAPIintensity 0 0.0 -RU+RU -RU+RU GS RNAi U U U U PromE >Pex1 R R R R 0 0 0 0 0 0 2 2

j DHE DAPI Merge k Bai-normalized 4 1.5 P=0.043

ns ns 1200

t i

g 3

-RU s

/mg

i n

n 1.0

e i

t 800

I pmol

2 / D 0.5

400 H

1 E

% H

+RU staining DHE %

D DHE / intensity DAPIDHE 0 0.0 0 -RU+RU -RU+RU l) ) GS RNAi U U U g g PromE >Pex19 U m /m R R R R l/ l 0 0 0 0 o o 0 0 m 2 2 m p (p ( T ll W u n 1 x e Supplementary Fig. 5 RNAi knockdown verification by QRT-PCR for upd3 (a),P ND-75 (b), Pex1 (c), Pex5 (d)Supplementary, Pex14 (e), Pex19 (f) Fig., kay 5(g) RNAi. Oenocyte knockdown-specific GeneSwitch verificationdriver by(PromE QRTGS-PCR-Gal4) forwas upd3used. (a)RU:, ND-75 mifepristone (RU486). ROS quantification in oenocyte-specific Pex5 KD (h), Pex1 KD (i), Pex19 KD (j). % (b), Pex1DHE -(c)positive, Pex5area (d)per, Pex14ROI or DHE (e), intensityPex19 normalized(f), kay (g)to .DAPI Oenocytewas analyzed-specific.Scale GeneSwitchbar: 20 µm. k Hdriver2O2 GS (PromElevels-Gal4measured) wasby used.Amplex RU:Red mifepristonein WT and PEX (RU486).1-G843D humana-e, g fibroblastn = 4 biologicalcells. N = samples.6. Data are f n = 3 biologicalrepresented samples.as mean ROS± SEMquantification. P values are incalculated oenocyteusing-specificunpaired tPex5-test, ns KD: not (significanth), Pex1. SourceKD (idata), Pex19 KD (j).are %provided DHE-aspositivea Source areaData perfile. ROI or DHE intensity normalized to DAPI was analyzed.Scale bar: 20 µm. n = 4 flies. k H2O2 levels measured by Amplex Red in WT and PEX1-G843D human fibroblast cells. N = 6 biological samples. Data are represented as mean ± SEM. P values are calculated using unpaired t-test (a-c, e-g, h-k), one-way ANOVA, followed by Holm-sidak multiple comparisons (d). ns: not significant. Source data are provided as a Source Data file. For statistical number, please refer to the source data.

118

Supplementary Fig. 6 Uncropped and unprocessed gel or blots for fig 3j, fig5d, 5n and supplementary fig3d.

119

5.8 Appendix

Supplementary Table 1: Primer list

Oligonucleotides Sequence 5'-3' Species upd1 F CGCAGCCTAAACAGTAGCCA Drosophila melanogaster upd1 R CGCTTTAGGGCAATCGTGGA Drosophila melanogaster upd2 F CTTAAACGCCAGCCAACAGAG Drosophila melanogaster upd2 R TGAATGGCATCACGACGCT Drosophila melanogaster upd3#1 F AAAACGGCCAGAACCAGGAA Drosophila melanogaster upd3#1 R CATGGCCAAGGCGAGTAAGA Drosophila melanogaster upd3#2 F AATGCCAGCAGTACGCATCT Drosophila melanogaster upd3#2 R TTCTGCAGGATCCTTTGGCG Drosophila melanogaster puc F ATCGAAGATGCACGGAAAAC Drosophila melanogaster puc R CAGGGAGAGCGACTTGTACC Drosophila melanogaster yki F AACTAGGCGCCTTGCCG Drosophila melanogaster yki R TCGCTCGGCCATCAAGATTT Drosophila melanogaster Mad F GTGCGTGTGAGTGAAAGCTA Drosophila melanogaster Mad R GGTATTGGAGTAGCTGCCGT Drosophila melanogaster kay F CGCAACATTGCGCTATTTTCAA Drosophila melanogaster kay R GCTTTTGTTGAATCGTTTTGGGT Drosophila melanogaster Jra F ATTCCGCCGCCAATAACA Drosophila melanogaster Jra R CTCGTCCTTAATCACCGAGAAG Drosophila melanogaster Socs36E F AGTCGCAGCAGTAAAGCACT Drosophila melanogaster Socs36E R TTAATCCTCGGATGGCGTCG Drosophila melanogaster Pex5 F CAACCTTACACACCCACATGAC Drosophila melanogaster Pex5 R GCAGCGATCTCCAGAGTTAT Drosophila melanogaster Pex1 F GATCTGGTCAAGTGTGCGCT Drosophila melanogaster Pex1 R AGCACACTGCCCGATATCTTT Drosophila melanogaster Pex14 F CATTGGATCCCAATTGCACGC Drosophila melanogaster Pex14 R AACAGGTAGGGCGCAATGTA Drosophila melanogaster Pex19 F TTCATGGAGGGCATGATGCAG Drosophila melanogaster Pex19 R TGTCCTCAGCGGAGAGCTT Drosophila melanogaster ND-75 F GTACCCGGCACCACTGTC Drosophila melanogaster ND-75 R AGCAGAATCTGGGTATCTCCAC Drosophila melanogaster RpL32 F AAGAAGCGCACCAAGCACTTCATC Drosophila melanogaster RpL32 R TCTGTTGTCGATACCCTTGGGCTT Drosophila melanogaster IL-6 F CCACACAGACAGCCACTCA Homo sapiens IL-6 R CATCCATCTTTTTCAGCCATCT Homo sapiens GAPDH F ACCCACTCCTCCACCTTTG Homo sapiens GAPDH R CTCTTGTGCTCTTGCTGGG Homo sapiens

120

CHAPTER 6. PEROXISOME IMPORT STRESS IMPAIRS RIBOSOME BIOGENESIS IN DROSOPHILA AND HUMAN CELLS

Manuscript in preparation

Kerui Huang 1†, Jinoh Kim1†, Hua Bai 1*

From the 1Department of Genetics, Development, and Cell Biology, Iowa State University,

Ames, Iowa, 50011, USA

†These authors contributed equally to this work

* To whom correspondence should be addressed: Hua Bai: Department of Genetics,

Development, and Cell Biology, Iowa State University, Ames, Iowa 50011; [email protected]

Keywords: oenocytes, peroxisome, ER stress, ribosome biogenesis

6.1 Abstract

Peroxisome biogenesis diseases (PBDs) are characterized by global defects in peroxisomal function and can result in severe brain, liver, kidney, and bone malfunctions. PBDs are due to mutations in peroxisome biogenesis factors (PEX genes) that are responsible for peroxisome assembly and function. There is increasing evidence suggesting that peroxisome import functions decline during aging and regulates this process. The transcriptome profiling of 121 peroxisome import defects is still lacking. To identify conserved responses, we undertook a bioinformatic transcriptomic analysis on Drosophila oenocyte specific Pex1, Pex12 and Pex5 knockdowns. In addition, we performed analysis on human cells with induced peroxisome import stress. We uncovered that oenocyte-specific Pex1, Pex12 and Pex5 have distinct transcriptional profiles with each other, where only 53 genes are commonly regulated by all three

Pexs in Drosophila. Using gene set enrichment analysis (GSEA), we identify protein processing in endoplasmic reticulum pathway, specifically ERAD is enriched and induced in all PEX knockdowns in Drosophila. Moreover, we uncovered decreased expression in ribosome biogenesis genes in flies and human cells. Indeed, we identified a stall at the 5’-ETS cleavage sites during the ribosome biogenesis and impaired 40S small ribosomal export in both flies and human. Our data indicates an unexpected link between peroxisome and ribosome biogenesis and suggests that reduced ribosome biogenesis could be conserved cellular response to reduce peroxisome import stress.

6.2 Introduction

Peroxisomes are single-membrane organelles found in almost all eukaryotic cells.

Peroxisomes are highly dynamic and versatile, as their composition and activity varies between organisms, different cell types and under different environmental stresses1. Generally peroxisomes are responsible for metabolizing branched-chain fatty acids (BCFA) or very long chain fatty acids (VLCFAs), purine catabolism, synthesis of plasmalogens, ether-lipids and bile acids, regulation of hydrogen peroxide and other reactive oxygen species (ROS) 1, 2. In mammals, β-oxidation occurs largely in the mitochondria. Peroxisomes are normally required for catabolism of BCFA and VLCFAs, however it can perform β-oxidation on other fatty acids when mitochondria are compromised 3. Peroxisomes have close interaction with many other organelles. Peroxisomes can physically interact with other organelles including the endoplasmic 122 reticulum (ER), mitochondria, lipid droplets, lysosomes and chloroplasts, to perform specific functions 4.

Peroxisomes are formed by the action of 14 peroxisome assembly factors (Peroxins). The majority of which are involved in translocation of peroxisomal enzymes into the peroxisome matrix. Others are responsible for transporting peroxisome membrane proteins5-8. Unlike mitochondria, the peroxisome is composed of single membrane and it could import fully folded and oligomeric proteins into the matrix 9. Peroxisomal matrix proteins contain one of two types of intrinsic peroxisomal targeting signals (PTSs), PTS1 or PTS2, which can direct import into the organelle 9. Most matrix proteins are targeted through the PTS1 pathway. PTS1 consists of a tripeptide (-SKL or conserved variant) at the C-terminal and it can be recognized by a soluble receptor, Pex5 9. Pex7 will recognize PTS2 containing proteins near the N terminus. Pex5 and

Pex7 bind their respective PTSs in the cytosol, and the receptor-cargo complexes will dock to the peroxisome surface, containing Pex13 and Pex14 docking complexes. Cargo is released from

Pex5 and imported into peroxisomes. Then Pex5 is mono-ubiquitylated by Pex4 and Pex12, enables Pex5 recycling back to the cytosol for another round of import. Pex5 can also undergo polyubiquitylation (by the ubiquitin-conjugating and ligase enzymes, Ubc4 and Pex2 respectively), which directs Pex5 to the proteasome for degradation (see excellent review in 10).

Mutations in these peroxin genes and peroxisome resident enzymes can lead to a variety of disorders, known as peroxisome biogenesis disorders (PBD) with involvement of kidney, brain, bone and liver, and death in infants 11-14.

Recently there are increasing evidence placing peroxisome, especially peroxisomal import function, as an important regulator for aging. Several studies suggest that peroxisomal import function declines with age 15-17. Consistently, our translatomic study shows that the 123 majority of peroxisome genes are downregulated in aged fly oenocytes 18. Our previous study identified that peroxisome import activities in fly oenocytes can non-autonomously regulate cardiac arrhythmia during aging 19. Several other studies have also implicated that peroxisome is involved in longevity20-22. However, it remains elusive how impaired peroxisome import activity contributes to cellular processes such as aging.

Ribosome biogenesis is one of the regulators of aging. The reduced level of ribosomal genes, ribosome biogenesis factors or nutrient sensing pathways (such as TOR signaling), which stimulate ribosome production, can increase the lifespan in multiple organisms, including C. elegans, D. melanogaster, yeast, mice and human 23. Because the rate of protein translation is proportional to the rate of ribosome biogenesis 24, 25, it was suggested that reduced ribosome biogenesis can reduce protein synthesis thus maintaining global proteostasis 23. The biogenesis of the 60S and 40S ribosomal subunits follows a complicated pathway in eukaryotic cells, which requires the assembly of four rRNAs and ~80 ribosomal proteins (RPs) 26. Ribosome biogenesis occurs in the nucleolus and is initiated by transcription of a large precursor rRNA (pre-rRNA) through RNA polymerase I (Pol I), from which the mature 18S, 5.8S, and 25S rRNAs are generated 27. The 5S pre-rRNA is transcribed by RNA polymerase III and incorporated into nascent pre-ribosome in later step 28. The pre-rRNA will assemble co-transcriptionally with numerous trans-acting factors and early binding ribosomal proteins for the small subunit, forming a macromolecular complex, named 90S pre-ribosome or small-subunit (SSU) processome 29, 30. In the firs 90S assembly intermediate, the pre-rRNA undergoes site-specific base modifications at conserved sites and cleavage reactions. These early cleavages will produce

20S pre-rRNA (precursor to 18S), which becomes part of the early pre-40S particle. The pre-40S particle will follow a simple maturation route to the cytoplasm where mature 40S is produced. 124

Ribosome biogenesis is tightly controlled by nutrient availability and extracellular conditions.

Myc and TOR pathways are known regulators of ribosome biogenesis 31, 32, however it remains elusive whether there are additional regulators of this complex biological process.

Using human HEK293 and HepG2 cell lines, in combination of Drosophila tissue specific RNAi, we sought to characterize conserved cellular responses under peroxisome import stress. Our study discovered distinct transcriptional responses from knocking down different components involved in peroxisome import pathways, suggesting PEXs proteins perform additional activities besides facilitating Pex5. Genes participate in protein process and in endoplasmic-reticulum-associated protein degradation (ERAD) are highly induced across all

PEXs knockdowns in flies, suggesting an induced ER stress under peroxisome defects. We also identified repression of oxidative phosphorylation genes and induced inflammation pathways across fly and humans, suggesting close interplay of peroxisome and mitochondria. Finally, for the first time our study discovered a conserved reduction of ribosome biogenesis genes and reduced ribosome biogenesis activity, under peroxisome import stress across fly and human.

Together, these findings shed lights on mechanisms of PBD pathology and aging.

6.3 Results

Transcriptomic analysis for peroxisomal stress response in Drosophila oenocytes and human cells

Despite the emerging importance of peroxisome in the regulation of aging and metabolism, little is known how cells mount defensive response to maintain cellular homeostasis.

Oenocytes are hepatocyte-like cells that are highly enriched with peroxisomes 19, 33. In addition, the oenocyte peroxisome was identified as an important regulator for age-related production of inflammatory factor, unpaired 3 (upd3) 19, which could dampen cardiac function non- autonomously. To study cellular responses induced by peroxisomal stress, we performed 125 transcriptomic analysis from Drosophila oenocytes and human cell cultures. We target on peroxisome import stress by knocking down key factors (peroxins) involved in peroxisomal import process (Pex5, Pex12 and Pex1) specifically in oenocytes, which were shown to efficiently induce the ROS level and upd3 production 19. Pex5 is one of the major peroxisome protein import receptors, which can bind cargo proteins containing peroxisomal targeting signal type 1 (PTS1) and deliver them to peroxisomal matrix through Pex13/Pex14 docking complex.

After releasing its cargo in peroxisomal matrix, Pex5 will be poly- or mono-ubiquitinated through complex Pex2/Pex10/Pex12. Monoubiquitylated receptor is then released from the membrane by the ATPases of the ATPases associated with diverse cellular activities (AAA) family, Pex1 and Pex6 34 (Fig. 1A). In the cytosol, the ubiquitin moiety is removed and Pex5 becomes available for another round of import. The polyubiquitylated Pex5 receptor is released from peroxisome similar to monoubiquitylated Pex5, except that it is directed for proteasomal degradation 34. Knocking down these genes have been shown to greatly impair peroxisome import in oenocytes and S2 cells 19, 35. We utilized oenocyte-specific GeneSwitch driver

(PromEGS-Gal4) 19 to transiently knock down peroxin genes in adult stage. RU486 (mifepristone, or RU) was used to activate PromEGS-Gal4 (+RU), whereas control genotype is the same, but with no RU feeding (-RU). After 5 days of RU activation, fly oenocytes are dissected and polyA- tailed RNA was isolated for downstream RNA sequencing and data analysis (Fig. 1A). To eliminate the transcriptional changes induced by RU feeding, we included a control group

(PromE-GS-gal4>yw) for RNA-seq analysis.

To identify conserved pathways involved in peroxisomal stress responses, we have recently developed an inducible peroxisomal stress system in the mammalian cell culture. In this system, we knocked in a Tet-ON 3G tetracycline-inducible expression construct into the AAVS1 126 safe harbor locus in human embryonic kidney-derived HEK293 cells (Fig. 2B). The expression construct contained a FLAG-tagged full-length human PEX5 with a single amino acid substitution at position 11 (cysteine to alanine, C11A), a conserved ubiquitination site involved in PEX5 recycling 36. Stable expression of PEX5C11A mutant exerts dominant-negative effect on wild type PEX5 recycling and efficiently blocks peroxisomal import in mammalian cell culture

36. PEX5C11A protein level can be robustly induced by treating cells with doxycycline (Dox) at the concentration of 1 μg/ml, but not in wild type (Fig. 2C). Induction of PEX5C11A mutant was able to block GFP-PTS1 import into peroxisomes (marked by PEX14), without significantly affecting peroxisomal number (Fig. 1D-1F). To conduct RNA-seq analysis, PEX5C11A cells were treated with or without doxycycline for 72 hours prior for total RNA extraction. To directly compare similarity with Pex5 knockdown in Drosophila oenocytes, we transfected PEX5-targeting siRNA or non-targeting siRNA into HepG2 cells for 72 hours prior to total RNA isolation. PolyA-tailed

RNA was then isolated and RNA-seq libraries were constructed and pooled for the following analysis (Fig. 1G-1H).

Distinct transcriptional profiling in Pex1, Pex12 and Pex5 knockdown in oenocytes

To understand how peroxisomal stress induced by different peroxins’ knockdown affect transcriptome, we conducted hierarchical cluster analysis across Pex1RNAi, Pex12RNAi, Pex5RNAi and yw control samples, using log2 fold change value (RU treatment / no RU treatment) (Fig.

2A). Intriguingly, we found three peroxin knockdowns produced distinct transcriptomic changes with very little similarities, as reflected on the heatmap (Fig. 2A) and PCA analysis

(Supplementary Fig. 1A). The transcriptomic profile also varies between PEX5C11A and PEX5- siRNA in human cell culture (Supplementary Fig.1C). The little similarity between each gene knockdowns is possibly due to different methods of inducing peroxisome stress and different cell 127 types used. Cluster 2 in Fig.2A represents genes that are highly induced by both Pex5 and Pex12 and contains GO terms such as response to ER stress, carbohydrate metabolic process, glycogen biosynthetic process and lipid particle (Supplementary Fig 1C). Cluster 4 is enriched with genes specifically induced by RU feeding, which are involved in oogenesis, transcription, histone modification (Supplementary Fig.1D). These results are consistent with previous findings that mifepristone can regulate genes involved in X-chromosome gene expression and oogenesis 37.

Cluster 9 and 11, which contains genes commonly induced by Pex5 and Pex1, are enriched with

GO terms including fatty acid beta-oxidation, mRNA splicing, protein folding, mitochondrial translation, and peroxisome. Induction of peroxisomal genes suggests there is retrograde signaling between peroxisome and nucleus. In cluster 10 and 12 contains genes induced by both

Pex1 and Pex12. These clusters contain genes involved in proteasome assembly, translation, mitochondrial electron transport and ribosome (Supplementary Fig1f – 1E).

There are 248 differentially expressed genes (DEGs) in Pex12 knock down from oenocytes (│fold change│ ≥ 1.2, FDR ≤ 0.1), 254 DEGs induced by RU feeding. Both Pex5 knockdown and Pex1 knockdown induced great transcriptional changes: Pex1 knockdown produced 1055 DEGs and Pex5 knockdown produced 1056 DEGs (Fig. 2B). Next, we ask what genes are commonly regulated by all Pexs’ knockdowns and what genes are specifically controlled by individual Pexs. We compiled venn diagram using up-regulated DEGs across all

Pex-knockdown samples and yw control group. Consistent with the heatmap analysis (Fig.2A), only 40 genes are commonly induced by all Pex knockdowns, but not in yw control group (Fig.

2C), confirming the finding that these Pex knockdowns generate different transcriptional responses. To understand how each Pex genes specifically regulate transcriptions, we conducted

GO term analysis on genes specifically induced by individual Pex knock downs. Pex5 128 knockdown specifically induced 337 DEGs, which are enriched in glutathione, proteolysis, and oxidation-reduction processes. Pex12 knockdown specifically induced genes involved in glucose homeostasis and lipid metabolism. In contrast, Pex1 specifically induced genes involved in protein homeostasis process, such as proteasome, ribosome, and translation (Fig. 2D).

Next, we sought to understand the function of all the commonly induced genes. When we analyzed the function of these genes, we found 10% of them function in mitochondria (Fig. 2E),

7% of them function in ER, 5% in Golgi, whereas only 1 of them functions in peroxisome.

Localization predication was based on published datasets on peroxisome and mitochondria 38, 39, predicated localization on Flybase and human orthologues predication. From those mitochondrial genes, COX7A (cytochrome c oxidase subunit 7A) is a subunit for mitochondrial complex IV;

CG9512 encodes a glucose-methanol-choline oxidoreductase, its human orthologue CHDH

(choline dehydrogenase) localizes to the outer membrane of mitochondria in a potential- dependent manner40. CHDH interacts with SQSTM1, a mitophagy adaptor molecule, and CHDH is required for mitophagy 40. This suggests mitochondrial abnormalities and possibly mitophagies are induced under peroxisome import stress. Spidey, an oenocyte enriched gene essential for its growth and lipid metabolism, is also induced by peroxisome stress. Its human orthologue (very- long-chain 3-oxoacyl-CoA reductase, HSD17B12) is predicted to localize in the ER. Spidey is reported to be a 3-ketoacyl-CoA reduction for elongation of long chain fatty acids into VLCFAs is also induced by peroxisome stress41 (Fig. 2E). Together, these data suggest that other organelles’ homeostasis are tightly controlled by peroxisomal and these changes can lead collaborative response against disruption of homeostasis.

Comparing to commonly induced genes, Pex knockdowns only commonly repressed the expression of 10 genes (Fig. 2F and Supplementary Fig. 1F). Among these 10 genes, 3 of them 129 encode transferases (CG1941, CG2781, CG14615), which can transfer acyl groups other than amino acyl groups. CG2781 (ELOVL fatty acid elongase) is predicated to mediate elongation of fatty acids into very-long-chain fatty acids. Eaat1 (excitatory amino acid transporter 1), which encodes a transmembrane protein with glutamate-sodium symporter activity 42, is also commonly repressed by all Pex knockdowns. Its function in oenocytes remain unknown. But glutamate transport could indirectly regulates the synthesis of antioxidant glutathione 43, Eaat1 might also regulate glutathione synthesis, thus regulating the redox level in oenocytes. We further analyzed the genes specifically repressed by individual Pex knockdowns. Pex5 knockdown specifically repressed genes involved in maturation of small subunit ribosomal ribonucleic acid (SSU- rRNA). Pex12 knockdown specifically induced glutamine metabolic process and nucleolus genes. Pex1 knockdown repressed genes involved in mitotic cytokinesis, and defense response.

Why Pex1 regulates genes in mitotic cytokinesis? It is possibly because they perform other functions in oenocytes, such as cytoskeleton organization. Taken together, our Pex specific transcriptomic analysis revealed common distinct changes elicited by peroxisome stress in adult oenocytes.

Peroxisomal stress induced endoplasmic reticulum changes in oenocytes

To further characterize peroxisome-stress induced signaling pathways, we performed

Gene Set Enrichment Analysis (GSEA) on three Pex knockdown samples, using a collection of pre-defined gene sets retrieved from Kyoto Encyclopedia of Genes and Genomes (KEGG) database on Drosophila Melanogaster. Interestingly, Pex12 knockdown affected many pathways involved in RNA metabolism and processing, including RNA polymerase, RNA transport, RNA degradation and spliceosome (Table 1). Spliceosome can carry intron removal to form the mature mRNA. Spliceosome is a massive complex that includes hundreds of proteins and small nuclear 130 ribonucleoproteins (snRNPs), which are crucial in defining intron borders and facilitating alternative splicing 44. Many RNA binding proteins act as splicing regulators to facilitate or inhibit splice site recognition by spliceosome45. Alternative splicing defects can arise when the levels of spliceosome components are altered. Our GSEA analysis identified that small ribonucleoprotein particle protein B (SmB), which is orthologous to human SNRPN that can modulates functions of the splicing machinery and lead to alternative splicing 46. In addition, snRNP components pre-mRNA processing factor 8 (Prp8), U2 small nuclear riboprotein auxiliary factor 38 (U2af38) and splicing factor 3b subunit 1 (sf3b1) are also repressed under

Pex12RNAi. Alterations of these genes’ orthologues in human are associated with broad changes in the production of splice variant and diseases 47. It has been reported that splicing changes can regulate cellular responses under stressed condition such as aging 44. Because majority of RNA processing pathways are downregulated, it is possible that Pex12RNAi induces cellular responses primarily through regulation of alternative splicing.

GSEA results showed all three Pex knock down induced genes involved in DNA replication, which might be attributable to disrupted reactive oxygen species (ROS) level, because ROS are well-known mediators of DNA and mitochondrial DNA damage48, 49. We also observed a decrease of ribosomal subunits, including mitochondrial ribosomes, in Pex5 knockdown. Consistent with this, down-regulation of Pex5 and Pex12 had a significant reduction on ribosome biogenesis (FDR < 0.0001, Table 1), suggesting dampened ribosome biogenesis. In addition, proteasome pathway has been also induced in Pex1 and Pex12 knockdowns (Table 1), which might be possibly because of increasing needs to degrade poly-ubiquitinated Pex5 50, 51.

Among all three different Pex knockdowns, we found protein processing in endoplasmic reticulum pathway is up-regulated (Table 1). Density plot showed that comparing to total gene 131 expression in respective RNAi, Pex5RNAi, Pex1RNAi and Pex12RNAi all showed a consistent, albeit modest increase of the expression level involved in the ER pathway (Fig. 4A-4C). After examining the pathway closely, we identified nine genes that are commonly induced by all Pex

RNAi (Fig. 4D). Some of these genes were not identified in previously venn analysis possibly because they did not reach the cut-off (FC ≥ 1.2, FDR ≤ 0.1). Interestingly, CG30156, CG3061,

Ubc7, p47, Hsp70Ba and Hsp70Bb (highlighted in red) all participate in ER-associated degradation (ERAD). Among them, p47, Hsp70Ba and Hsp70Bb are highly induced in all Pex knockdowns, by which Hsp70Bb is induced by more than 6-fold in Pex1RNAi. p47 is predicted to have ubiquitin binding activity and its human orthologue can interact with ubiquitinated substrates during ERAD and are elevated upon ER stress 52. Similarly, several ubiquitin ligase complex components are also up regulated, such as sip3 (belongs to the HRD1 family) and Der-

1. The evidence suggests that there is higher level of ER stress which induced ERAD activity under peroxisomal stress. Further investigations are needed to verify this possibility.

Peroxisome dysfunction represses oxidative phosphorylation and induced inflammation, metabolic pathways in human cells

To understand how peroxisome stress changes transcriptions in human cell cultures, we performed GO term analysis on DEGs (│fold change│ ≥ 1.2, FDR ≤ 0.1) from PEX5C11A and

PEX5-siRNA treated cells. PEX5-siRNA treatment produced 587 DEGs, whereas PEX5C11A elicited dramatic transcriptional response, with 7393 DEGs. Interestingly, these two treatments share few common transcriptional changes: 26 genes commonly induced versus only 6 genes commonly repressed by two treatments (Fig. 3A). Among these 26 commonly up-regulated genes (Supplementary Fig.1G), majority of them regulates cell proliferation (CCNA2, FGFR3,

CDCA5, PSMB9 and UBA7) and mitochondrial metabolism (PCK1, ACSM3, ALDH8A1 and 132

BTG2). Then we looked at the specific genes induced by PEX5C11A and PEX5-siRNA. PEX5C11A specifically induced immune response, RNA metabolism, nitrogen metabolism (Fig. 3B), whereas PEX5-siRNA specifically induced peroxisome genes, cholesterol homeostasis, steroid metabolism, carboxylic acid metabolism and lipid metabolism (Fig. 3C). PEX5C11A represses mitochondrial genes, translation, ribosome biogenesis. PEX5-siRNA represses negative regulation of protein phosphorylation and MAPK cascade, indicating an induction of MAPK signaling under PEX5 knockdown, which is consistent with previous findings in oenocytes 19.

To further characterize peroxisome-stress induced signaling pathways, we performed

GSEA, using a collection of pre-defined gene sets retrieved from KEGG database on Homo

Sapiens. PEX5C11A significantly regulated 103 signaling pathways, whereas only 2 pathways are significant under PEX5-siRNA (FDR ≤ 0.05). Through GSEA analysis, we found oxidative phosphorylation pathway and valine leucine and isoleucine degradation pathway are downregulated under PEX5C11A treatment (Table 2 and 3G). In oxidative phosphorylation pathway, we found that key components of Complex I-IV of the respiratory chain are down- regulated. For example, NADH dehydrogenase: NDUFV1, NDUFV2, NDUFS2, NDUFS3,

NDUFS4, NDUFS6, NDUFS7, NDUFS8 and NDUFSA11, which are components of Complex I and deficiency of which is the most common cause of mitochondrial disease53. Cytochrome C1

(CYC1) which is part of the metal centers responsible for electron transfer in complex III, is also downregulated. Almost all cytochrome c oxidase (COX, complex IV) subunits are repressed

(COX4l1, COX4l2, COX5A, COX5B, COX6A1, COX6B1, COX6C, COX7A1, COX7A2,

COX7A2L, COX7B, COX7C, COX8A and COX15) and many ATPase subunits (ATP6V1G2,

ATP6V0C, ATP6V1G1 etc.) are repressed as well. These results suggest that mitochondrial electron transport activity is greatly impaired under peroxisomal stress, which is consistent with 133 previous publications54-56. Even though we did not observe decreased oxidative phosphorylation genes in Drosophila oenocytes, we observed an altered mitochondria morphology with higher fusion activity under Pex5 knockdown in oenocytes (unpublished observation). This indicates peroxisome and mitochondria are dynamically connected organelles and the response to peroxisome stress is conserved between flies and human.

It has been well established that defects in electron transport chain leads to elevated reactive oxygen species (ROS) and can induce inflammatory pathways57 58. Interestingly, we also observed activated MAPK signaling, type II diabetes mellitus under PEX5C11A treatment (Table

1). Pex5 knockdown was previously identified to activate MAPK/JNK signaling in Drosophila

41 Hawthorne St #1 oenocytes 19, this suggests a conserved cellular response from peroxisome stress between flies and human. In addition, we also observed an upregulation of HIPPO pathway, including key regulators such as MOB1B, LATS1/2, YAP1. HIPPO pathway which is originally discovered for its regulation on organ size. However, recent evidence has discovered that Hippo signaling is also involved in oncogenesis 59, 60 as well as apoptosis 61-67. Peroxins mutants exhibit enhanced cell apoptosis and cell death 68 (based on unpublished data) , our evidence suggest that peroxisome stress induced by PEX5C11A activates apoptosis through

HIPPO signaling activation.

PEX5-siRNA treatment produced less significant GSEA pathways comparing to

PEX5C11A, with only spliceosome showing significant negative correlation with age (NES =

1.85, FDR = 0.03). Spliceosome and RNA processing are also negatively correlated with age in human69. which suggests that peroxisome defects are one of the causes for tissue aging.

According to GSEA analysis, under PEX5-siRNA induced genes involved in cholesterol metabolism pathway (NES = -1.45, P-value = 0.019) (Fig. 3F). Among the upregulated genes in 134 the pathway, apolipoprotein genes are highly induced, including APOC1, APPOA1, APOA5 and

APOC3 (Fig. 3M). APOC3 had a significant 5-fold increase after PEX5 knockdown (Fig. 3M).

Apolipoproteins are crucial for lipoprotein metabolism. Not only can guide lipoprotein formation, they also serve as activator or inhibitors of enzymes involved in lipoprotein metabolism 70. Notably, single-nucleotide polymorphisms within the apoliprotein (APO) gene cluster, which includes APOA1/C3/A4/5 genes, are strong risk alleles associated with hypertriglyceridemia and increased coronary heart disease (CHD) risk in humans 71, 72. APOA5 encodes a protein which is secreted into plasma to control plasma triglyceride (TG) metabolism by acting as an activator of lipoprotein lipase, thus promoting TG catabolism73. In contrast,

APOC3 encodes a protein that acts as an inhibitor of lipoprotein lipase and high expression of

APOC3 have been associated with increased plasma level of TG and hypertriglyceridemia 74.

Interestingly, APOC3 protein is also a modulator for pro-inflammatory pathways, such as PKCβ and NF-κB in endothelial cells. It is interesting that both APOC3 and APOA5, which have opposing effects on TG, all are induced under PEX5-siRNA. Pex5 perturbation’s effects on plasma TG require further investigations. Many of pathway regulations are cell type or treatment specific. Despite of these differences, we were able to identify several pathways, including

MAPK, mitochondrial oxidative phosphorylation, potentially PERK-mediated ER stress as conserved response to peroxisome dysfunction. It is yet unknown whether there are additional conserved cellular responses.

GSEA uncovers ribosome biogenesis declines as a conserved response in the peroxisomal defects

To identify conserved response to peroxisomal stress, we compared the GSEA results in

Drosophila oenocytes and in human cells. GSEA showed a conserved downregulation of genes 135 functioning in the KEGG pathway ribosome biogenesis in Pex5RNAi, Pex12RNAi (Drosophila oenocytes) and PEX5siRNA in human cells (Fig. 5A and Supplementary Fig. 2A). As shown by density plots (Fig. 5A), most of genes in the ribosome biogenesis gene set showed a consistent, although modest, decrease in the expression level between knockdown and control, compared to total gene expression (Fig. 4A-B and Supplementary Fig. 2B-C). A similar trend was observed with ribosome gene set (Fig. 5B), with a decrease in ribosome genes in PEX5C11A and Pex5RNAi, including majority of genes encoding mitochondrial ribosomal proteins and ribosomal proteins.

Even though RU feeding also induced the ribosome pathway, however the heatmap showed RU and Pex5RNAi showed few similarities on the expression level (Supplementary Fig. 2B).

A closer look on ribosome biogenesis pathway revealed that multiple steps regulating the process are repressed. Peroxisomal stress from three different knockdowns all target on 60S ribosome processing and 90S pre-ribosome components (Fig. 5C). The eukaryotes contain two ribosome subunits, the 40S (small) and the 60S (large), which contain four rRNA species (18S,

5.8S, 25S and 5S) and ribosomal proteins (RPs). Ribosome assembly starts in the nucleolus by transcribing a large precursor rRNA (47S pre-rRNA in Drosophila and mammals) by RNA polymerase I, from which the mature 18S, 5.8S, and 25S rRNAs are generated 27. Our results show multiple steps during ribosome processing are affected under Pex knockdown. For example, NOP58, a core component of the box C/D small nucleolar ribonucleoprotein complex, which controls specific pre-ribosomal RNA processing, is downregulated in PEX5RNAi in human cells. Its fly orthologue, nop5, is also repressed under Pex12RNAi in oenocytes. Ribosome biogenesis protein BMS1 homolog (BMS1), which is crucial for 40S maturation are repressed in human PEX5RNAi, same with its orthologue is flies. Additionally, elongation factor like GTPase 1

(EFL1) can assist the release of eukaryotic translation initiation factor 6 (eIF6) release from 60S- 136 ribosome 75, 76. Thus, it plays an important role in translational activation. CG33158 is the fly orthologue of EFL1. Interestingly, we found reduced expression level of CG33158 in Pex5RNAi and Pex12RNAi in oenocytes, suggesting the translation activity is dampened under peroxisome stress.

During ribosome biogenesis, the 47S pre-rRNA assembles co-transcriptionally with various trans-acting factors and early binding ribosomal proteins of the small subunit. Together a huge macromolecular complex, named 90S pre-ribosome or small-subunit (SSU) processome will form26, 29, 30. 90S pre-ribosome is composed of 60-70 non-ribosomal factors, with most being called U three proteins (Utp) 26, and they can form structurally autonomous subcomplexes (UTP-

A, UTP-B, UTP-C, Mpp10-Imp3-Imp4, U3 snoRNP, and Bms1-Rcl1 modules). These modules will associate with pre-rRNA in a sequential and hierarchical order 77-79. UTP-A and UTP-B are the earliest 90S modules that assemble on the 5’-external transcribed spacer (ETS) of the pre- rRNA, whereas UTP-C and other complexes bind later. It has been reported that UTP-A complex is critical for the initial cleavage of 5’ETS site on 47S pre-rRNA, thus regulating the ribosome biogenesis 80. We found multiple genes in SSU processome are downregulated under peroxisome stress. UTP15 small subunit processome component (UTP15), nucleolar protein 6 (NOL6),

PWP2 small subunit processome (PWP2), are all repressed in PEX5RNAi in human cells or their fly orthologues are repressed in oenocytes. Because UTP-A, UTP-B and UTP-C components are largely dampened under peroxisomal stress, we sought to verify whether peroxisomal stress also caused pre-rRNA processing, especially at the 5’-ETS removal steps.

pre-rRNA processing starts with the removal of the 5’ and 3’-ETS, and the internal transcribed spacers 1 and 2 (ITS1 and ITS2, respectively) (Fig. 6A). To test whether pre-rRNA processing is impaired under peroxisomal stress, we designed primers against the 5’-ETS region 137 and 3’-ETS regions in Pex5RNAi oenocytes and PEX5C11A induced human cells (Fig 6A). We discovered strong accumulation of higher level of 5’-ETS and ITS1 could indicate a defect in rRNA processing, or induced level of rRNA transcription (Fig. 6B-6C). Next, we measured uncleaved ETS using primer 1 and 2 as shown in figure 6A, normalized to 18S (primer 3 and 4) which represents total level of rRNA transcribed 81. Higher relative expression value is indicative of accumulated unprocessed rRNA. Pex5 knockdown specifically in oenocytes can increase the fraction of 5’ETS-18S unprocessed rRNA (Fig. 6D), without significantly affecting the total rRNA level (Fig.6E). We also found a consistent increase of unprocessed rRNA in human cells after PEX5C11A induction, but not 18S level (Fig. 6G-6H). To further verify whether ribosome synthesis is impaired, we measured the localization of ribosomal protein S6 (Rps6) under peroxisomal stress. The rationale for this assay is based on the observation that defects if biogenesis pathway can cause ribosomal subunit export defects 82, 83. Rps6 is an early assembling

40S subunit ribosomal protein 84-86, allowing us to score early defects in 40S synthesis. Indeed, we found an accumulation of RPS6 in the nucleus in Pex5RNAi oenocytes, after normalized to total intensity (Fig. 6I – 6J). We also identified higher percentage of cells contain nucleus rpS6 signal in PEX5C11A expressing human cells, comparing to control (Fig 6K-6L). These results indicate a defect in rRNA processing, instead of increased rRNA transcription, under peroxisome disruption.

6.4 Discussion

The peroxisome is the important metabolic site to perform β-oxidation of fatty acids and the degradation of toxic hydrogen peroxide, yet it has received little respect until now. Although peroxisomal dysfunction has been linked to severe diseases in man and aging 15, 17, 19, 87, whether cells can mount defensive mechanism and how cells respond to peroxisome stress remain to be determined. Here we characterized a variety of transcriptional alterations under multiple 138 peroxisomal stress in both Drosophila oenocytes and human cell cultures. We show that defects in peroxisome import process can elicits distinct transcriptional response in Drosophila oenocytes. Peroxisomes dysfunction will affect the function of organelles, such as ER and mitochondria. In turn, these organelles will activate stress-response genes to resolve damages caused by peroxisome dysfunction. Specifically, peroxisome defects in oenocytes up-regulates genes involved in ER and ERAD. PEX5C11A induction induced genes involved in oxidative phosphorylation and altered mitochondrial ribosome genes’ expression. PEX5RNAi specifically up-regulated genes involved in cholesterol metabolism and transport. We also identified conserved decrease in ribosome biogenesis and ribosome between Drosophila and human cells.

Different importing peroxins control distinct transcriptional processes

Maintaining peroxisomal activity is a highly dynamic and complicated process, facilitated by peroxisomal biogenesis factors (peroxins). In fact, most of the peroxins identified so far are known to be directly involved in different stages of peroxisomal matrix protein import

88. Newly synthesized and folded peroxisome proteins from cytosol are targeted to the organelle in a post-translational manner88. The protein import process into peroxisome is involved with multiple steps, including cargo recognition, docking of the cargo/receptor- complex at the peroxisome membrane, cargo translocation, cargo release into the peroxisomal matrix and receptor recycling (Fig. 1A). Disrupting key regulators during these processes can compromise the peroxisome import activity and lead to severe physiological consequences19, 35, including development of PBDs. Even though these Peroxin genes work cooperatively to maintain peroxisome activity, it remains unknown whether their sole activity is to regulate peroxisome biogenesis. 139

Interestingly, our transcriptomic analysis on multiple import-regulatory genes (Pex5,

Pex12 and Pex1) suggests that they may regulate other cellular processes. First, we found that

Pex1 and Pex5 RNAi produced high level of transcriptional changes comparing to Pex12 (1055 and 1056 DEGs versus 248 DEGs). This might be because the severity of peroxisome import is different, because Pex12 dsRNA in S2 human cells show less diffusion of cytosolic GFP-PTS comparing to Pex1 and Pex5 35. However, both our GO ontology and GSEA analysis suggest that even Pex1 and Pex5 knockdown induced different genetic pathways. Pex5 knock down induced

GO terms related to oxidation-reduction and glutathione pathways, whereas Pex1 dramatically induced genes involved in translation, ribosome, and proteasome (Table 1). The discrepancies in their transcriptomic profiles, despite sharing the same phenotype in peroxisome import, suggest that they affect different processes under knockdown. Pex1 and Pex6 both belong to the AAA

ATPase family, which is known to be involved in protein quality control, including folding, assembly, transport, and protein degradation 89. This could explain the induction of translation and proteasome in Pex1RNAi. Pex1, Pex15 and Pex6 will form an exportomer and deficiency of these components can trigger pexophagy in mammalian cells 90. Another study found that a missense mutation of Pex1 can ameliorated the defects of Pex6 mutant without restoring PEX5 recycling, possibly through prevention of pexophagy, suggesting other functions Pex1-Pex6 complex 91. In support of this view, studies have unexpectedly identified PEX6 regulating mitochondrial inheritance in daughter cells in yeast. Seo et al. showed that PEX6 overexpression

92 was able to rescue the deficit in Atp2p (the β-subunit of mitochondrial F1, F0-ATPase ) mutant by improving its import into mitochondria. Studies suggest that Pex1 was able to translocate between mitochondria and peroxisomes93, where they showed that mitochondria localized PEX6-

PEX1-PEX26 complex can rescue PEX26 deficient cells. Altogether, our findings provide new 140 insights in peroxisome biogenesis disorders, peroxins functions and their significance of other organelles’ activities.

Peroxisome and inter-organelle communications

No peroxisome is in isolation. A growing body of evidence showed that peroxisome share physical membrane contact sites with other organelles (see an excellent review here 4).

This suggests there is intimate interaction between peroxisomes with these organelles, which is important for controlling processes such as metabolism and organelle proliferation. In agreement of this concept, we identified changes of gene expression for other organelles, especially for ER and mitochondria, under various Pex knockdowns.

Our results of PEX5RNAi in HepG show significant up-regulation of cholesterol metabolism (Fig. 3C and 3F). Previous study shows that cholesterol accumulates drastically in animals and human patients with peroxisomal disorders, as a result of disrupted peroxisome- lysosome contact sites 94. The majority of cellular cholesterol (60% - 80%) is localized at the plasma membrane (PM) 95. Cholesterol is synthesized in the ER or it can be imported exogenously and stored in the lysosome. Cholesterol can be transported from the ER and lysosome to the PM. Recently it was shown that peroxisome mediates the transfer of cholesterol from lysosome to PM. A peroxisome-lysosome contact site was shown to be required for this transport 94. An up-regulation of cholesterol metabolism genes could be a compensatory mechanism to degrade excess cholesterol accumulated inside the cells.

Besides lysosome, peroxisomes also actively interact with ER and mitochondria. Our results show that mitochondrial ribosomal subunits are repressed under Pex5 knockdown in oenocytes (Table 1), as well as in PEX5C11A mutant (Table 2). Consistently, genes involved in oxidative phosphorylation are largely repressed in PEX5C11A, which was also been observed in 141

Pex16 KO 96. Our results suggest that mitochondria function is altered under peroxisome stress.

In fact, previous studies as well as our observation have found that mitochondria underwent functional and morphological changes under Pex5 and Pex16 deficiencies 96, 97. In these studies, they consistently observed a more fused mitochondria structure, reduced level of the respiratory chain activities. Park et. al identified that peroxisome-derived phospholipids are responsible peroxisome dysfunction induced mitochondrial abnormalities 96. Park et. al inhibited plasmalogen synthesis by knocking down expression of GNPAT, an enzyme responsible for the first step in plasmalogen production. GNPAT knockdown was able to recapitulate phenotypes with Pex16 knockdown, including altered mitochondrial morphology and decreased mtDNA.

Interestingly, they also found dietary supplementation of plasmalogen precursors can rescue thermogenesis in Pex16-KO mice. However, whether plasmalogen level is the sole contributor to mitochondrial dysfunction and why there are tissue specific mitochondria abnormalities under peroxisome deficiency, these questions still require investigation.

Our results suggest different Pex knockdowns commonly induce genes VLCFA synthesis, whose protein products can localize on ER (e.g. spidey and Cyp6g1). Biosynthesis of

VLCFA takes place at the ER and can be metabolized in peroxisomes 98, 99. Our findings suggest two organelles work closely to regulate metabolism of VLCFA. EM images showed that two organelles are adjacent to each other and that ER membrane can wrap around peroxisomes 100-103.

In fact, peroxisome-ER crosstalk is important for many lipid-related metabolic pathways, including the biosynthesis of ether-phospholipids, production of polyunsaturated fatty acids, bile acids, isoprenoids, and cholesterol.

Interestingly, we found peroxisome dysfunction in oenocyte cells can trigger protein process pathway in the ER and ERAD (including CG30156, CG30161, Ubc7, p47, Hsp70Ba and 142

Hsp70Bb), which is possibly due to elevated ER stress (Fig. 3). Consistently, it has been previously identified that peroxisome-deficient Pex2 and Pex5 knockout mice had elevated level of ERAD gene expression, ER stress and ER abnormalities, prominently through PERK activation 97, 104. How does peroxisome dysfunction lead to the ER stress? Possible reasons include perturbed flux of mevalonate metabolites, changes in fatty acid levels or composition and increased oxidative stress 105. Specific induction of PERK under peroxisome stress, not activation of Xbp1 splicing, suggesting peroxisome stress induced a specific cellular response distinct from protein misfolding induced ER stress. Genetic experiments and biochemical assays should be utilized to discern the nature of peroxisome induced ER stress.

Peroxisome in ribosome biogenesis and protein homeostasis

Our results identified decreased transcription on ribosomal proteins and ribosome biogenesis factors, as a conserved response to peroxisome stress between humans and

Drosophila. This is also the first study to verify the reduced ribosome biogenesis under Pex5RNAi in oenocytes and PEX5C11A mutant in human cells. Our results show that the 5’ cleavage process is inhibited under Pex5 knockdown. Our result indicates that peroxisome defects can affect ribosome biogenesis, which might be more efficient in dealing with transient stress comparing to transcriptional control. In addition, we also observed an accumulation of RPS6 in the nucleus under Pex5 stress, which is often an indicator of ribosome biogenesis defects, specifically involved with 40S maturation process 82. However, it is highly possible that 60S maturation is also affected, judged by down-regulated expression of several crucial genes: Midasin

(MDN1)106, GTPBP4 107, 108 and EFL175, 76. Overall, our results indicate a down-regulation of ribosome availabilities, and possibly lower translational activities under peroxisome stress. 143

Why multiple peroxisome defects converge to impair ribosome biogenesis? It is possible that nucleolus is sensitive to peroxisomal or mitochondrial ROS level, which can hamper the function of ribosome biogenesis proteins that primarily localize in the nucleolus. However, very little is known on what stresses can alter nucleolus protein activity or composition. Further research is needed to investigate this area.

It has been well established that reduced ribosome proteins and ribosome biogenesis factors can increase longevity in model organisms (see 23 for review). The evidence also suggest that reduced ribosome biogenesis is a protective mechanism to confer peroxisomal stresses.

Being a regulator of ribosome biogenesis, Pex5’s effect on lifespan had been controversial. In

Δpex5 yeast cells, there was a strong reduction in chronological lifespan20. However, post developmental knockdown of prx-5, a C. elegans homolog for mouse Pex5, increased the worm’s lifespan 21, 22. The paradoxically increased lifespan by peroxisome disruption could be attributed to decreased level of ribosome and translational activities. It will be interesting to see whether post developmental knockdown of Pex5 in oenocytes can also prolong lifespan through the regulation on ribosome biogenesis. In addition, how peroxisome import defect reduces ribosome biogenesis will be an interesting avenue to investigate.

6.5 Methods

Plasmid Construction

Human Pex5 cDNA from the Mammalian Gene Collection (MGC) was purchased from

Dharmacon. The hPEX5 was amplified by PCR using the forward and reverse primers (5’-

CACTATAGGGAGACCCAAGCTTATCTAGACATGGCAATGCGGGAGCT-3’ and 5’-

TCTTACTTGTCATCGTCGTCCTTGTAGTCGCCCTGGGGCAGGCC-3’) and introduced between XhoI and BamHI sites in c-Flag pcDNA3 (addgene #20011) to generate Flag-tagged hPEX5 using NEBuilder® HiFi DNA assembly Master mix (New England Biolabs). Site- 144 directed mutagenesis for amino acid substitution was performed using the Q5® Site-directed mutagenesis kit (New England Biolabs) according to the manufacture’s instruction. The primers for C11A mutant were 5’-GGAGGCCGAAgctGGGGGTGCCAACC-3’ and 5’-

ACCAGCTCCCGCATTGCC-3’.

To generate Tetracycline inducible PEX5C11A plasmid, we modified pMK243 (Tet-

OsTIR1-PURO) from Masato Kanemaki (Addgene #72835). pMK243 was digested by BgIII and

MluI to remove OsTIR sequence. Flag-PEX5C11A was amplified by PCR using the forward and reverse primers (5’- gattatgatcctctagacatatgctgcagattacttgtcatcgtcgtccttgtagt-3’ and 5’- tcctaccctcgtaaagaattcgcggccgcaatggcaatgcgggagctggt-3’) and introduced between BgIII and

MluI sites in digested pMK243 plasmid to generate Tet-PEX5C11A-PURO plasmid. All plasmids are confirmed by Sanger sequencing.

Generation of CRISPR Knock-in HEK293 cells expressing PEX5C11A

HEK293 cells were maintained in Dulbecco’s Modified Eagle Medium containing 10% fetal bovine serum (FBS), with penicillin and streptomycin. To generate stable cell line, we followed the protocol described in Natsume et al. 1x10^6 HEK293 cells were plated in one well of a 6-well plate. After 24hrs, 800ng of AAVS1 T2 CRISPR in pX330 (Addgene #72833) and

1ug of Tet-PEX5C11A-PURO were transfected using Effectene (Qiagen) according to the manufacturer’s instructions. After 48hrs, the cells were detached and diluted at 10 to 100 times in 10ml of medium containing 1ug/ml of puromycin and transferred to 10 cm dishes. The selection medium was exchanged every 3 to 4 days. After 8 to 10 days, colonies were marked using a marker pen under a microscope, picked by pipetting with 10ul of trypsin-EDTA, and subsequently transferred to a 96-well plate. 100ul of the selection medium was added. When the cells were confluent, they were transferred to a 24-well plate. After 2-3 days, the cells were 145 transferred to a 6-well plate. After 2-3 days, cells were detached and half of the cells were frozen, and the rest were used for genomic DNA preparation.

Genomic DNA isolation and PCR

To prepare genomic DNA, cells were lysed in buffer A solution (100mM Tris-HCl

[pH7.5], 100mM EDTA, 100mM NaCl, 0.5% SDS) followed by incubation at 65°C for 30min.

Buffer B (1.43M potassium acetate, 4.28M lithium chloride) was added and incubated on ice for

10min. After centrifuge at 12,000 rpm for 15 min, the supernatant was transferred to new microtube and added isopropanol. After isopropanol precipitation, DNA pellets were washed in

70% ethanol and resuspended with DNAse-free Water.

To verify Tet-PEX5C11A-PURO insertion into AAVS1 locus, genomic PCR was performed using Q5® High-Fidelity DNA polymerase (New England BioLabs) according to the manufacturer’s instruction. 5’-cgtttcttaggatggccttc-3’ and 5’-agaaggatggagaaagagaa-3’ were used for WT cell validation. 5’-cgtttcttaggatggccttc-3’ and 5’-ccgggtaaatctccagagga-3’ were used for

Tet-PEX5C11A-PURO integration.

Fly husbandry and strains

The following RNAi lines were used in the KD experiments: y[1] v[1]; P{y[+t7.7] v[+t1.8]=TRiP.HMJ21920}attP40 (BDSC # 58064, Pex5 RNAi), y[1] sc[*] v[1] sev[21];

P{y[+t7.7] v[+t1.8]=TRiP.HMC03536}attP2/TM3, Sb[1] (BDSC # 53308), y[1] v[1]; P{y[+t7.7] v[+t1.8]=TRiP.HM05190}attP2 (BDSC # 28979). The control line used for the KD experiments is yw, a gift from Rochele Lab. All fly lines are crossed to oenocyte specific gene-switch driver, yw; PromE800-GS-gal4/Cyo;+, a gift from Heinrich Jasper.

Female flies were used in all experiments. Flies were maintained at 25°C, 60% relative humidity, and 12-h light/dark cycle. Adults and larvae were reared on a standard cornmeal and 146 yeast-based diet, unless otherwise noted. The standard cornmeal diet consists of the following materials: 0.8% cornmeal, 10% sugar, and 2.5% yeast. RU486 (mifepristone, Fisher Scientific) was dissolved in 95% ethanol, and added to standard food at a final concentration of 100 µM for all the experiments. Gene KD or overexpression was achieved by feeding flies on RU486 food for 5-6 days, before RNA isolation.

Fly oenocyte RNA isolation

Adult tissues oenocytes (20 females per replicate) were all dissected in cold 1 × PBS before RNA extraction. For oenocyte dissection, we first removed FB through liposuction and then detached oenocytes from the cuticle using a small glass needle. Tissue lysis, RNA extraction was performed using RNeasy Micro kit from QIAGEN (catalog number 74034) with the following modifications. Samples were first collected in 1.7ml centrifuge tubes. Dissected tissues were saved in 150ul of Buffer RLT (a component of RNeasy kits) with 143mM β- mercaptoethanol on ice during the dissection. Samples were then put at room temperature for 3 minutes prior to lysis and centrifuged at 7.5xg for 3minutes. Additional 150ul buffer RLT was added and pellet pestle grinder (Kimble pellet pestles, catalog number 749540-0000) was used to lyse the tissue for 1 minutes. 20ng carrier RNA was added to the cell lysates. Freezed samples was thawed in 37°C water bath for 1 minute. Total RNA was extracted using RNeasy Plus Micro columns following company manual.

Drosophila oenocyte RNA-seq library construction and sequencing

RNA-seq libraries were constructed using 100 ng of total RNA and NEBNext Ultra II

RNA Lib Prep (NEB, Ipswich, MA, USA. Catalog number: E7770L). RNA concentrations were measured using Qubit RNA BR Assay Kit (Thermo Fisher Scientific, catalog number: 10210).

Poly(A) mRNA was isolated using NEBNext Oligo d(T)25 beads and fragmented into 200 nt in 147 size. Purification of the ligation products are performed using Beckman Coulter AMPURE XP

(BECKMAN COULTER, catalog number: A63880). After cDNA synthesis, each cDNA library was ligated with a NEBNext adaptor and barcoded with an adaptor-specific index (NEBNext®

Multiplex Oligos for Illumina, NEB, catalog number: E7335S). Twelve libraries were pooled in equal concentration and sequenced using Illumina HiSeq 3000 platform (single end, 150bp reads format).

Human cells library preparation

HEK293 expressing Tet-PEX5C11A-Puro cells were cultured in Dulbecco’s Modified

Eagle Medium containing 10% fetal bovine serum (FBS), with penicillin and streptomycin.

HepG2 cells were cultured in Minimum Essential Medium containing 10% FBS, 1mM sodium pyruvate, 1X MEM Non-Essential Amino Acids solution (ThermoFisher, #11140050) with penicillin and streptomycin. Cells were incubated in a 37°C incubator in an atmosphere of 5%

CO2 in air.

To prepare RNA-seq libraries, 4x10^5 HEK293 expressing Tet-PEX5C11A-Puro cells were seeded in a 12 well plate. After 1 day, 1ug/ml of doxycycline was added to the cells to induce PEX5C11A in the cells. After 72hrs, total RNA was isolated from the cells with or without doxycycline treatment. 3x10^5 HepG2 cells were seeded in a 6-well plate. Transfection of siRNA into HepG2 cells was performed with lipofectamine RNAiMAX (ThermoFisher) according to the manufacturer’s instructions. On-TARGET plus human PEX5 siRNA

(Dharmacon # L-015788-00-0005) was used for knockdown of PEX5 in HepG2 cells. For negative control, On-TARGET plus non-targeting siRNA (Dharmacon # D-001810-02-05) was used. After 72hrs, total RNA was isolated from the cells. All samples were treated TURBOTM

DNase (Thermofisher #AM1907) to remove all traces of DNA. RNA-seq libraries were 148 constructed using 1ug DNA-free total RNA with NEBNext Ultra II RNA Library Prep Kit for

Illumina (New England Biolabs, Ipswich, MA, USA. NEB#E7770). Poly (A) mRNA was isolated using NEBNext Poly(A) mRNA Magnetic isolation module (New England Biolabs,

Ipswich, MA, USA. NEB#E7490). After first strand and second strand and second strand cDNA synthesis, each cDNA library was ligated with a NEBNext adaptor and barcoded with an adaptor-specific index. Twelve libraries were pooled in equal concentrations and sequenced using Illumina HiSeq 3000 platform.

RNA-Seq data processing and differential expression analysis

FaastQC (v0.11.8) was first performed to check the sequencing read quality. Then sequence alignment and mapping was performed using the Spliced Transcripts Alignment to a

Reference (STAR) software (v2.7.3a) 110. The raw reads were mapped to D. melanogaster genome (BDGP Release 6) or Genome Reference Consortium Human Build 38 (GRCh38).

Reads mapped were then counted with summarizedOverlaps function using “Union” mode in R.

Counts are then analyzed in DESeq2 (v1.26.0) 111 for batch control analysis and test for differential expression.

Principal component analysis (PCA), heatmap and expression correlation plot

PCA graph was generated using plotPCA function of R package DESeq2111. Heatmaps and hierarchy clustering analysis were generated using heatmap.2 function of R package gplots.

The density plots were plotted using R package ggplot.

Gene set enrichment analysis (GSEA)

For GSEA analysis, all pre-defined set of 132 KEGG pathways in Drosophila were downloaded from KEGG. Text were trimmed and organized using Java script. Normalized counts were used as input for parametric analysis and organized as suggested by GSEA tutorial 149 site (GSEA 112, 113). Collapse dataset to gene symbols was set to false. Permutation type was set to gene set; number of permutations was set to 1000; enrichment statistic used as weighted analysis; metric for ranking genes was set to Signal to Noise.

Gene ontology and pathway analysis

Functional annotation analysis of differentially expressed genes was performed using

STRING114 or DAVID115, 116. GO terms (Biological Process, Molecular Function, Cellular

Component), KEGG pathway, INTERPRO Protein Domains and Features, were retrieved from the analysis.

Quantitative real-time polymerase chain reaction (qRT-PCR)

qRT-PCR was performed using Quantstudio 3 Real-Time PCR system and SYBR green master mixture (Thermo Fisher Scientific, USA Catalog number: A25778). All gene expression levels were normalized to Rpl32 (in Drosophila), GAPDH (in humans) by the method of comparative Ct117. Mean and standard errors for each gene were obtained from the averages of three biological replicates, with two technical repeats.

Immunostaining

Anti-RpS6 (Cell Signaling, catalog number 2317) at concentration of 1:100 was used to stain oenocytes’ RpS6118. Anti-RpS6 at concentration of 1:50 was used to stain human cells.

Secondary antibodies were obtained from Jackson ImmunoResearch.

Adult oenocyte tissues were dissected in PBS and fixed in 4% paraformaldehyde for

15 min at room temperature. Tissues were washed with 1x PBS with 0.3% Triton X-100 (PBST) for three times (~5 min each time), and blocked in PBST with 5% normal goat serum for 30 min.

Tissues were then incubated overnight at 4 °C with primary antibodies diluted in PBST, followed by the incubation with secondary antibodies for 1 h at room temperature. After washes, tissues 150 were mounted using ProLong Gold antifade reagent (Thermo Fisher Scientific) and imaged with an FV3000 Confocal Laser Scanning Microscope (Olympus). DAPI or Hoechst 33342 was used for nuclear staining.

1.2 x 105 HEK293-PEX5C11A cells were seeded in 24-well plates on coverslips

(Neuvitro #GG1215PLL). After 1day, cells were treated with or without doxycycline (1ug/ml) for 3 days. The cells were rinsed with PBS, fixed in 4% paraformaldehyde for 10 min, and rinsed with PBS again. Cells were permeabilized in 0.5% Triton X-100 in PBS for 10 min. Cells were treated with PBS containing 1% bovine serum albumin (BSA) for 1 hr at room temperature. The cells were incubated with antibody against RPS6 (1:50) diluted in PBS for overnight at 4°C.

Next day, cells were washed with 0.05% Triton X-100 in PBS for three times and incubated with

Alexa Flour® 488 donkey anti-mouse IgG (1:500) and Hoechst (1:1000) diluted in PBS for 1hr at room temperature in the dark. After that, cells were washed with 0.05% Triton X-100 in PBS for three times. Cells were briefly washed with PBS and mounted on glass slides using mounting medium (Thermo Fisher Scientific, #P36961). Images were visualized by confocal microscope.

Image analysis and quantification

Confocal images were quantified using Olympus CellSense Dimension software

(Olympus, v 1.16) and ImageJ (v1.49). For oenocyte RpS6 signaling quantification, five oenocyte nucleus were randomly selected from each image. Each single nucleus was set as a region of interest (ROI). A background ROI was also selected. The intensity of RpS6 signal was calculated for nucleus intensity and normalized to total intensity (total intensity = nucleus intensity + background). For human RpS6 quantification, we counted the number of cells that that contain RpS6 signal inside the nucleus versus total number of cells in one image.

Statistical analysis 151

GraphPad Prism (GraphPad Software, La Jolla, CA, USA, v6.07) or Deseq2 was used for statistical analysis. To compare the mean value of treatment groups versus that of control, student t-test was used. Log2 fold change and FDR values were calculated by Deseq2 using

Benjamini-Hochberg method for sequencing data.

6.6 References

1. Pridie, C., Ueda, K. & Simmonds, A.J. Rosy Beginnings: Studying Peroxisomes in Drosophila. Frontiers in Cell and Developmental Biology 8 (2020).

2. Fujiki, Y. et al. Recent insights into peroxisome biogenesis and associated diseases. J Cell Sci 133 (2020).

3. Violante, S. et al. Peroxisomes can oxidize medium- and long-chain fatty acids through a pathway involving ABCD3 and HSD17B4. FASEB J 33, 4355-4364 (2019).

4. Shai, N., Schuldiner, M. & Zalckvar, E. No peroxisome is an island - Peroxisome contact sites. Biochim Biophys Acta 1863, 1061-1069 (2016).

5. Fujiki, Y., Okumoto, K., Mukai, S., Honsho, M. & Tamura, S. Peroxisome biogenesis in mammalian cells. Front Physiol 5, 307 (2014).

6. Steinberg, S.J. et al. Peroxisome biogenesis disorders. Biochim Biophys Acta 1763, 1733- 1748 (2006).

7. Subramani, S. Components involved in peroxisome import, biogenesis, proliferation, turnover, and movement. Physiol Rev 78, 171-188 (1998).

8. Waterham, H.R. & Cregg, J.M. Peroxisome biogenesis. Bioessays 19, 57-66 (1997).

9. Smith, M.D. & Schnell, D.J. Peroxisomal protein import. the paradigm shifts. Cell 105, 293-296 (2001).

10. Smith, J.J. & Aitchison, J.D. Peroxisomes take shape. Nat Rev Mol Cell Biol 14, 803-817 (2013).

11. Wanders, R.J. Metabolic and molecular basis of peroxisomal disorders: a review. Am J Med Genet A 126A, 355-375 (2004).

12. Wanders, R.J. & Waterham, H.R. Peroxisomal disorders I: biochemistry and genetics of peroxisome biogenesis disorders. Clin Genet 67, 107-133 (2005).

13. Baumgartner, M.R. & Saudubray, J.M. Peroxisomal disorders. Semin Neonatol 7, 85-94 (2002). 152

14. Gould, S.J. & Valle, D. Peroxisome biogenesis disorders: genetics and cell biology. Trends Genet 16, 340-345 (2000).

15. Narayan, V. et al. Deep Proteome Analysis Identifies Age-Related Processes in C. elegans. Cell Syst 3, 144-159 (2016).

16. Weir, H.J. et al. Dietary Restriction and AMPK Increase Lifespan via Mitochondrial Network and Peroxisome Remodeling. Cell Metab 26, 884-896 e885 (2017).

17. Legakis, J.E. et al. Peroxisome senescence in human fibroblasts. Mol Biol Cell 13, 4243- 4255 (2002).

18. Huang, K. et al. RiboTag translatomic profiling of Drosophila oenocytes under aging and induced oxidative stress. BMC Genomics 20, 50 (2019).

19. Huang, K. et al. Impaired peroxisomal import in Drosophila oenocytes causes cardiac dysfunction by inducing upd3 as a peroxikine. Nat Commun 11, 2943 (2020).

20. Lefevre, S.D., van Roermund, C.W., Wanders, R.J., Veenhuis, M. & van der Klei, I.J. The significance of peroxisome function in chronological aging of Saccharomyces cerevisiae. Aging Cell 12, 784-793 (2013).

21. Zhou, B. et al. Midlife gene expressions identify modulators of aging through dietary interventions. Proc Natl Acad Sci U S A 109, E1201-1209 (2012).

22. Curran, S.P. & Ruvkun, G. Lifespan regulation by evolutionarily conserved genes essential for viability. PLoS Genet 3, e56 (2007).

23. Turi, Z., Lacey, M., Mistrik, M. & Moudry, P. Impaired ribosome biogenesis: mechanisms and relevance to cancer and aging. Aging (Albany NY) 11, 2512-2540 (2019).

24. Russell, J. & Zomerdijk, J.C. RNA-polymerase-I-directed rDNA transcription, life and works. Trends Biochem Sci 30, 87-96 (2005).

25. Ruggero, D. & Pandolfi, P.P. Does the ribosome translate cancer? Nat Rev Cancer 3, 179-192 (2003).

26. Kornprobst, M. et al. Architecture of the 90S Pre-ribosome: A Structural View on the Birth of the Eukaryotic Ribosome. Cell 166, 380-393 (2016).

27. Woolford, J.L., Jr. & Baserga, S.J. Ribosome biogenesis in the yeast Saccharomyces cerevisiae. Genetics 195, 643-681 (2013).

153

28. Zhang, J. et al. Assembly factors Rpf2 and Rrs1 recruit 5S rRNA and ribosomal proteins rpL5 and rpL11 into nascent ribosomes. Genes Dev 21, 2580-2592 (2007).

29. Dragon, F. et al. A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 417, 967-970 (2002).

30. Grandi, P. et al. 90S pre-ribosomes include the 35S pre-rRNA, the U3 snoRNP, and 40S subunit processing factors but predominantly lack 60S synthesis factors. Mol Cell 10, 105-115 (2002).

31. Mayer, C. & Grummt, I. Ribosome biogenesis and cell growth: mTOR coordinates transcription by all three classes of nuclear RNA polymerases. Oncogene 25, 6384-6391 (2006).

32. van Riggelen, J., Yetil, A. & Felsher, D.W. MYC as a regulator of ribosome biogenesis and protein synthesis. Nat Rev Cancer 10, 301-309 (2010).

33. Gutierrez, E., Wiggins, D., Fielding, B. & Gould, A.P. Specialized hepatocyte-like cells regulate Drosophila lipid metabolism. Nature 445, 275-280 (2007).

34. Schliebs, W., Girzalsky, W. & Erdmann, R. Peroxisomal protein import and ERAD: variations on a common theme. Nat Rev Mol Cell Biol 11, 885-890 (2010).

35. Mast, F.D. et al. A Drosophila model for the Zellweger spectrum of peroxisome biogenesis disorders. Dis Model Mech 4, 659-672 (2011).

36. Okumoto, K. et al. Cysteine ubiquitination of PTS1 receptor Pex5p regulates Pex5p recycling. Traffic 12, 1067-1083 (2011).

37. Landis, G.N. et al. The progesterone antagonist mifepristone/RU486 blocks the negative effect on life span caused by mating in female Drosophila. Aging (Albany NY) 7, 53-69 (2015).

38. Chen, C.L. et al. Proteomic mapping in live Drosophila tissues using an engineered ascorbate peroxidase. Proc Natl Acad Sci U S A 112, 12093-12098 (2015).

39. Faust, J.E., Verma, A., Peng, C. & McNew, J.A. An inventory of peroxisomal proteins and pathways in Drosophila melanogaster. Traffic 13, 1378-1392 (2012).

40. Park, S., Choi, S.G., Yoo, S.M., Son, J.H. & Jung, Y.K. Choline dehydrogenase interacts with SQSTM1/p62 to recruit LC3 and stimulate mitophagy. Autophagy 10, 1906-1920 (2014).

41. Cinnamon, E. et al. Drosophila Spidey/Kar Regulates Oenocyte Growth via PI3-Kinase Signaling. PLoS Genet 12, e1006154 (2016).

154

42. Seal, R.P., Daniels, G.M., Wolfgang, W.J., Forte, M.A. & Amara, S.G. Identification and characterization of a cDNA encoding a neuronal glutamate transporter from Drosophila melanogaster. Recept Channels 6, 51-64 (1998).

43. Gonzalez, E.A., Garg, A., Tang, J., Nazario-Toole, A.E. & Wu, L.P. A glutamate- dependent redox system in blood cells is integral for phagocytosis in Drosophila melanogaster. Curr Biol 23, 2319-2324 (2013).

44. Deschenes, M. & Chabot, B. The emerging role of alternative splicing in senescence and aging. Aging Cell 16, 918-933 (2017).

45. Fu, X.D. & Ares, M., Jr. Context-dependent control of alternative splicing by RNA- binding proteins. Nat Rev Genet 15, 689-701 (2014).

46. Lee, M.S. et al. Modulation of alternative splicing by expression of small nuclear ribonucleoprotein polypeptide N. FEBS J 281, 5194-5207 (2014).

47. Chabot, B. & Shkreta, L. Defective control of pre-messenger RNA splicing in human disease. J Cell Biol 212, 13-27 (2016).

48. Srinivas, U.S., Tan, B.W.Q., Vellayappan, B.A. & Jeyasekharan, A.D. ROS and the DNA damage response in cancer. Redox Biol 25, 101084 (2019).

49. Shokolenko, I., Venediktova, N., Bochkareva, A., Wilson, G.L. & Alexeyev, M.F. Oxidative stress induces degradation of mitochondrial DNA. Nucleic Acids Res 37, 2539- 2548 (2009).

50. Gonzalez, K.L. et al. Disparate peroxisome-related defects in Arabidopsis pex6 and pex26 mutants link peroxisomal retrotranslocation and oil body utilization. Plant J 92, 110-128 (2017).

51. Kao, Y.T., Fleming, W.A., Ventura, M.J. & Bartel, B. Genetic Interactions between PEROXIN12 and Other Peroxisome-Associated Ubiquitination Components. Plant Physiol 172, 1643-1656 (2016).

52. Park, E.S., Yoo, Y.J. & Elangovan, M. The opposite role of two UBA-UBX containing proteins, p47 and SAKS1 in the degradation of a single ERAD substrate, alpha-TCR. Mol Cell Biochem 425, 37-45 (2017).

53. Fernandez-Vizarra, E., Tiranti, V. & Zeviani, M. Assembly of the oxidative phosphorylation system in humans: what we have learned by studying its defects. Biochim Biophys Acta 1793, 200-211 (2009).

54. Peeters, A. et al. Carbohydrate metabolism is perturbed in peroxisome-deficient hepatocytes due to mitochondrial dysfunction, AMP-activated protein kinase (AMPK) 155

activation, and peroxisome proliferator-activated receptor gamma coactivator 1alpha (PGC-1alpha) suppression. J Biol Chem 286, 42162-42179 (2011).

55. Peeters, A. et al. Mitochondria in peroxisome-deficient hepatocytes exhibit impaired respiration, depleted DNA, and PGC-1alpha independent proliferation. Biochim Biophys Acta 1853, 285-298 (2015).

56. Shinde, A.B. et al. Mitochondrial disruption in peroxisome deficient cells is hepatocyte selective but is not mediated by common hepatic peroxisomal metabolites. Mitochondrion 39, 51-59 (2018).

57. Mittal, M., Siddiqui, M.R., Tran, K., Reddy, S.P. & Malik, A.B. Reactive oxygen species in inflammation and tissue injury. Antioxid Redox Signal 20, 1126-1167 (2014).

58. Murphy, M.P. How mitochondria produce reactive oxygen species. Biochem J 417, 1-13 (2009).

59. Zhou, W. & Zhao, M. How Hippo Signaling Pathway Modulates Cardiovascular Development and Diseases. J Immunol Res 2018, 3696914-3696914 (2018).

60. Johnson, R. & Halder, G. The two faces of Hippo: targeting the Hippo pathway for regenerative medicine and cancer treatment. Nat Rev Drug Discov 13, 63-79 (2014).

61. Lapi, E. et al. PML, YAP, and p73 are components of a proapoptotic autoregulatory feedback loop.

62. Bertini, E., Oka T Fau - Sudol, M., Sudol M Fau - Strano, S., Strano S Fau - Blandino, G. & Blandino, G. YAP: at the crossroad between transformation and tumor suppression.

63. Strano, S. et al. The transcriptional coactivator Yes-associated protein drives p73 gene- target specificity in response to DNA Damage.

64. Cottini, F., Anderson, K.C. & Tonon, G. Awakening the Hippo co-activator YAP1, a mercurial cancer gene, in hematologic cancers.

65. Matallanas, D. et al. RASSF1A elicits apoptosis through an MST2 pathway directing proapoptotic transcription by the p73 tumor suppressor protein. Mol Cell 27, 962-975 (2007).

66. Cottini, F. et al. Rescue of Hippo coactivator YAP1 triggers DNA damage-induced apoptosis in hematological cancers. Nat Med 20, 599-606 (2014).

67. Xiao, Q. et al. Depletion of CABYR-a/b sensitizes lung cancer cells to TRAIL-induced apoptosis through YAP/p73-mediated DR5 upregulation. Oncotarget 7, 9513-9524 (2016).

156

68. Di Cara, F., Bulow, M.H., Simmonds, A.J. & Rachubinski, R.A. Dysfunctional peroxisomes compromise gut structure and host defense by increased cell death and Tor- dependent autophagy. Mol Biol Cell 29, 2766-2783 (2018).

69. Harries, L.W. et al. Human aging is characterized by focused changes in gene expression and deregulation of alternative splicing. Aging Cell 10, 868-878 (2011).

70. Feingold, K.R. & Grunfeld, C. Introduction to Lipids and Lipoproteins, in Endotext. (eds. K.R. Feingold et al.) (South Dartmouth (MA); 2000).

71. Lai, C.Q., Parnell, L.D. & Ordovas, J.M. The APOA1/C3/A4/A5 gene cluster, lipid metabolism and cardiovascular disease risk. Curr Opin Lipidol 16, 153-166 (2005).

72. van Dijk, K.W., Rensen, P.C., Voshol, P.J. & Havekes, L.M. The role and mode of action of apolipoproteins CIII and AV: synergistic actors in triglyceride metabolism? Curr Opin Lipidol 15, 239-246 (2004).

73. Grosskopf, I. et al. Apolipoprotein A-V deficiency results in marked hypertriglyceridemia attributable to decreased lipolysis of triglyceride-rich lipoproteins and removal of their remnants. Arterioscler Thromb Vasc Biol 25, 2573-2579 (2005).

74. Kohan, A.B. Apolipoprotein C-III: a potent modulator of hypertriglyceridemia and cardiovascular disease. Curr Opin Endocrinol Diabetes Obes 22, 119-125 (2015).

75. Panse, V.G. & Johnson, A.W. Maturation of eukaryotic ribosomes: acquisition of functionality. Trends Biochem Sci 35, 260-266 (2010).

76. Saito, K. et al. Functional analysis of a novel glioma antigen, EFTUD1. Neuro Oncol 16, 1618-1629 (2014).

77. Chaker-Margot, M., Hunziker, M., Barandun, J., Dill, B.D. & Klinge, S. Stage-specific assembly events of the 6-MDa small-subunit processome initiate eukaryotic ribosome biogenesis. Nat Struct Mol Biol 22, 920-923 (2015).

78. Perez-Fernandez, J., Martin-Marcos, P. & Dosil, M. Elucidation of the assembly events required for the recruitment of Utp20, Imp4 and Bms1 onto nascent pre-ribosomes. Nucleic Acids Res 39, 8105-8121 (2011).

79. Zhang, L., Wu, C., Cai, G., Chen, S. & Ye, K. Stepwise and dynamic assembly of the earliest precursors of small ribosomal subunits in yeast. Genes Dev 30, 718-732 (2016).

80. Sloan, K.E., Bohnsack, M.T., Schneider, C. & Watkins, N.J. The roles of SSU processome components and surveillance factors in the initial processing of human ribosomal RNA. RNA 20, 540-550 (2014).

157

81. Hofmann, E.R. et al. Identification and characterization of GRIM-1, a cell-death- associated gene product. J Cell Sci 123, 2781-2791 (2010).

82. Awad, D. et al. Inhibiting eukaryotic ribosome biogenesis. BMC Biol 17, 46 (2019).

83. Gadal, O. et al. A nuclear AAA-type ATPase (Rix7p) is required for biogenesis and nuclear export of 60S ribosomal subunits. EMBO J 20, 3695-3704 (2001).

84. Ferreira-Cerca, S., Poll, G., Gleizes, P.E., Tschochner, H. & Milkereit, P. Roles of eukaryotic ribosomal proteins in maturation and transport of pre-18S rRNA and ribosome function. Mol Cell 20, 263-275 (2005).

85. Kruger, T., Zentgraf, H. & Scheer, U. Intranucleolar sites of ribosome biogenesis defined by the localization of early binding ribosomal proteins. J Cell Biol 177, 573-578 (2007).

86. Zhang, D. et al. Aggregation of Ribosomal Protein S6 at Nucleolus Is Cell Cycle- Controlled and Its Function in Pre-rRNA Processing Is Phosphorylation Dependent. J Cell Biochem 117, 1649-1657 (2016).

87. Weller, S., Gould, S.J. & Valle, D. Peroxisome biogenesis disorders. Annu Rev Genomics Hum Genet 4, 165-211 (2003).

88. Hasan, S., Platta, H. & Erdmann, R. Import of proteins into the peroxisomal matrix. Frontiers in Physiology 4 (2013).

89. Ogura, T. & Wilkinson, A.J. AAA+ superfamily ATPases: common structure--diverse function. Genes Cells 6, 575-597 (2001).

90. Deosaran, E. et al. NBR1 acts as an autophagy receptor for peroxisomes. J Cell Sci 126, 939-952 (2013).

91. Gonzalez, K.L. et al. A pex1 missense mutation improves peroxisome function in a subset of Arabidopsis pex6 mutants without restoring PEX5 recycling. Proc Natl Acad Sci U S A 115, E3163-E3172 (2018).

92. Seo, J.G., Lai, C.Y., Miceli, M.V. & Jazwinski, S.M. A novel role of peroxin PEX6: suppression of aging defects in mitochondria. Aging Cell 6, 405-413 (2007).

93. Weller, S. et al. Alternative splicing suggests extended function of PEX26 in peroxisome biogenesis. Am J Hum Genet 76, 987-1007 (2005).

94. Chu, B.B. et al. Cholesterol transport through lysosome-peroxisome membrane contacts. Cell 161, 291-306 (2015).

95. Maxfield, F.R. & Wustner, D. Intracellular cholesterol transport. J Clin Invest 110, 891- 898 (2002). 158

96. Park, H. et al. Peroxisome-derived lipids regulate adipose thermogenesis by mediating cold-induced mitochondrial fission. J Clin Invest 129, 694-711 (2019).

97. Dirkx, R. et al. Absence of peroxisomes in mouse hepatocytes causes mitochondrial and ER abnormalities. Hepatology 41, 868-878 (2005).

98. Islinger, M., Voelkl, A., Fahimi, H.D. & Schrader, M. The peroxisome: an update on mysteries 2.0. Histochem Cell Biol 150, 443-471 (2018).

99. Jakobsson, A., Westerberg, R. & Jacobsson, A. Fatty acid elongases in mammals: their regulation and roles in metabolism. Prog Lipid Res 45, 237-249 (2006).

100. Novikoff, P.M. & Novikoff, A.B. Peroxisomes in absorptive cells of mammalian small intestine. J Cell Biol 53, 532-560 (1972).

101. Zaar, K., Volkl, A. & Fahimi, H.D. Association of isolated bovine kidney cortex peroxisomes with endoplasmic reticulum. Biochim Biophys Acta 897, 135-142 (1987).

102. Grabenbauer, M., Satzler, K., Baumgart, E. & Fahimi, H.D. Three-dimensional ultrastructural analysis of peroxisomes in HepG2 cells. Absence of peroxisomal reticulum but evidence of close spatial association with the endoplasmic reticulum. Cell Biochem Biophys 32 Spring, 37-49 (2000).

103. Fahimi, H.D., Baumgart, E. & Volkl, A. Ultrastructural aspects of the biogenesis of peroxisomes in rat liver. Biochimie 75, 201-208 (1993).

104. Kovacs, W.J. et al. Peroxisome deficiency causes a complex phenotype because of hepatic SREBP/Insig dysregulation associated with endoplasmic reticulum stress. J Biol Chem 284, 7232-7245 (2009).

105. Faust, P.L. & Kovacs, W.J. Cholesterol biosynthesis and ER stress in peroxisome deficiency. Biochimie 98, 75-85 (2014).

106. Raman, N., Weir, E. & Muller, S. The AAA ATPase MDN1 Acts as a SUMO-Targeted Regulator in Mammalian Pre-ribosome Remodeling. Mol Cell 64, 607-615 (2016).

107. Kallstrom, G., Hedges, J. & Johnson, A. The putative GTPases Nog1p and Lsg1p are required for 60S ribosomal subunit biogenesis and are localized to the nucleus and cytoplasm, respectively. Mol Cell Biol 23, 4344-4355 (2003).

108. Saveanu, C. et al. Sequential protein association with nascent 60S ribosomal particles. Mol Cell Biol 23, 4449-4460 (2003).

109. Steffen, K.K. et al. Ribosome deficiency protects against ER stress in Saccharomyces cerevisiae. Genetics 191, 107-118 (2012). 159

110. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21 (2013).

111. Love, M.I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550 (2014).

112. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102, 15545- 15550 (2005).

113. Mootha, V.K. et al. PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet 34, 267-273 (2003).

114. Szklarczyk, D. et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47, D607-D613 (2019).

115. Huang, D.W., Sherman, B.T. & Lempicki, R.A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature Protocols 4, 44-57 (2009).

116. Huang da, W., Sherman, B.T. & Lempicki, R.A. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 37, 1- 13 (2009).

117. Schmittgen, T.D. & Livak, K.J. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 3, 1101-1108 (2008).

118. Martins, T. et al. TGFbeta/Activin signalling is required for ribosome biogenesis and cell growth in Drosophila salivary glands. Open Biol 7 (2017).

160

6.7 Figures

A PTS1 Protein receptor Oenocyte specific Pex RNAi Pex5

Ubiquitination complex 5-day 0RU 200RU Pex2 Pex12 feeding Pex14 Pex10 Pex13 Pex5 Pex5 Oenocyte Recycling extraction

PTS1 Pex1 Pex6 PolyA-tailed RNA isolation Peroxisome matrix RNA-seq and data analysis

B C gRNA WT Tet-PEX5C11A Dox AAVS1 safe 0 0.1 1 0 0.1 1 HA-L HA-R (μg/ml) harbor locus FLAG

HA-L Tet-On(R)3G hPGKTRE3GS Pex5C11A-FLAG PuroR HA-R β-actin

Tet-PEX5C11A

t r o Import Pex14

D GFP-PTS1 PEX14 DAPI Ep Peroxisomal Import F Peroxisomal Number

m i 100 150

e p<0.001 p=0.89

o s

80 C

i /

Dox- x

o e

r 100 e

60 a

a u

m 40

P r

h 20

e t

Dox+ i

# Pex14 puncta per cell per puncta Pex14 #

% Cells with Cells % importwith PTS1 s

l 0 0 l

e Dox- Dox+ Dox- Dox+ T A 5 A C W 1 x 1 1 C11A Tete -PEX51 C11A

Tet-PEX5C P C %

WT Pex5C11A WT G H non-targeting siRNA +Dox / -Dox PEX2 PEX12 +Pex5 RNAi PEX10 PEX14 PEX13 PEX5 PEX1 Treatment C11A for 3 days PEX5 PEX6 PEX5 Cell lysates and RNA Peroxisome isolation matrix

RNA-seq and data analysis Fig. 1 Transcriptomic analysis for peroxisomal stress response in Drosophila oenocytes and human cells. A Schematic diagram showing the key genes involved in peroxisomal import, as 161 well as oenocyte specific RNA-seq analysis. RNAi targets genes marked in red. B Schematic diagram for generation of CRISPR Knock-in HEK293 cells expressing PEX5C11A. C Western blot shows inducible PEX5C11A construct after treatment of Doxycycline. D Immunostaining showing GPF-PTS1 importing activity in wild type cells (Dox-) and PEX5C11A expressing cells (Dox+). Scale bar represents 10μm. E Quantification of D on % cells with PTS1 import activity in Dox- versus Dox+ cells. F Quantification in D showing the number of PEX14 positive peroxisomes in Dox- versus Dox+ cells. G – H Schematic diagram showing RNA-seq analysis using human cells.

A C

B Pex12 yw Oenocyte>RNAi 22 69 4 Pex12Pex12 248 Pex5 Pex1 1 3 ywyw 254 49 2 11 5 337 2 0 226 6 Pex1Pex1 1055 43 7 16 40 6 8 Pex5Pex5 1056 23 10 12 75 9 11 0 500 1000 1500 2 Number of DEGs Up-regulated genes yw Pex1 Pex12 Pex5 log2FC CG10383 Mito CG18815 ER COX7A scu Golgi D Pex5 RNAi E CG9512 CG10383 Pero GO:0006749 Glutathione Pex12 RNAi spidey Others GO:0006508 Proteolysis Pex1 RNAi Cyp6g1 Mito GO:0055114 Oxidation-reduction 10% ER 5% CG7530 Golgi GO:0042593 Glucose homeostasis p47 Pero 5% GO:0019216 Lipid metabolism CG18003Others 2.5% GO:0000502 Proteasome bgm GO:0005840 Ribosome 75.5% GO:0006412 Translation

0 10 20 30 40 50 -log10(P-value)

F G Pex5 RNAi Pex12 RNAi Pex12 yw GO:0000462 Maturation of SSU-rRNA Pex1 RNAi 15 55 Pex5 Pex1 GO:0006541 Glutamine metabolic process 24 2 4 GO:0005730 Nucleolus 328 4 1 494 5 GO:0006952 Defense response 10 22 10 7 GO:0000281 Mitotic cytokinesis 88

Down-regulated genes 0 2 4 6 8 10 -log10(P-value) Fig. 2 Pex1, Pex12 knock-down show distinct transcriptional pattern from Pex5. A Hierarchical clustering heatmap analysis, plotting log2 fold change (+RU / -RU) in yw, Pex1 RNAi, Pex12 RNAi, Pex5 RNAi oenocyte samples. B Number of differentially expressed genes (DEGs) across all oenocyte samples. C Venn diagram analysis showing genes commonly up regulated by yw, Pex1 RNAi, Pex12 RNAi and Pex5 RNAi (marked by red). D Gene Ontology (GO) term analysis on commonly induced genes from C. E Organelle localization of proteins produced commonly induced genes. F Venn diagram analysis showing genes commonly down regulated by yw, Pex1 RNAi, Pex12 RNAi, and Pex5 RNAi (marked by red). G Gene Ontology (GO) term analysis on commonly induced genes from F.

162

NAME NES FDR Type of RNAi Up or down regulated Ribosome biogenesis in eukaryotes 2.44 <0.0001 Pex5 Down Ribosome 1.79 0.021 Pex5 Down DNA replication 1.68 0.057 Pex5 Down Protein processing in endoplasmic reticulum -1.77 0.052 Pex5 Up Ribosome biogenesis in eukaryotes 2.89 <0.0001 Pex12 Down DNA replication 2.11 <0.0001 Pex12 Down RNA polymerase 1.91 0.001 Pex12 Down RNA transport 1.85 0.002 Pex12 Down Spliceosome 1.78 0.005 Pex12 Down RNA degradation 1.67 0.002 Pex12 Down Proteasome -1.90 0.01 Pex12 Up Protein processing in endoplasmic reticulum -1.46 0.188 Pex12 Up TOLL and IMD signaling pathway 1.74 0.041 Pex1 Down DNA replication 1.70 0.034 Pex1 Down Ribosome -2.41 <0.0001 Pex1 Up Proteasome -1.84 0.012 Pex1 Up Protein processing in endoplasmic reticulum -1.54 0.190 Pex1 Up Citrate cycle (TCA-cycle) -1.43 0.295 Pex1 Up

Table 1 GSEA pathway analysis results under Pex5, Pex12 and Pex1 RNAi. NES: normalized enrichment score. “Up or down regulated” indicates whether genes within that pathway are induced or reduced under RNAi treatment. 163

A B C 3 FDR = 0.05 FDR = 0.18 3 FDR = 0.19 NES = -1.77 NES = -1.46 NES = -1.54 All genes All genes 3 All genes ER pathway ER pathway 2 ER pathway 2

2 Density 1 1 1

0 0 0 0 1 2 3 0 1 2 3 4 0 1 2 3 Fold change (Pex5RNAi / Ctrl) Fold change (Pex12RNAi / Ctrl) Fold change (Pex1RNAi / Ctrl) dros-ER-mapping D 7 Genotypes Pex1RNAi 6 Pex12RNAi Pex5RNAi 5

FC FC (RNAi/ctrl) 3

1 Fold change (KD/ctrl) 0 CG301561 2 5 Sec61α1 2 5 Sec611 2 5β CG30611 2 5 1Ubc72 5 1Sar12 5 1 p472 5 Hsp70Ba1 2 5 Hsp70Bb1 2 5 x 1 x x x x 1 x x 1 x x 1 x x 1 x x 1 x x 1 x x 1 x e x e e x1 e e x e e x e e x e e x e e x e e x e e x e P e P p e p p e p p e p p e p p e p p e p p e p p e p - P - - p - - p - - p - - p - - p - - p - - p - - p - 6 - 6 a - a ta - ta 1 - 1 7 - 7 r1 - r1 7 - 7 a - a b - b 5 6 5 h a h e ta e 6 1 6 c 7 c a r1 a 4 7 4 B a B B b B 1 5 1 lp h lp b e b 0 6 0 b c b a p 4 p 0 B 0 0 B 0 0 1 0 a lp a 1 b 1 3 0 3 U b U S S p 7 0 7 7 0 7 3 0 3 1 a 1 6 1 6 G 3 G U S p 7 p p 7 p G 3 G 6 1 6 c 6 c C G C s p s s p s G c 6 c e c e C H s H H s H C C e c e S e S H H FigureC 3 SPeroxisomale S S stress induced endoplasmic reticulum changes in oenocytes. A-C S Density plot showing the fold change (Pex RNAi / Control) of ER pathway genes in oenocytes. D Plotting of fold change in selected genes involved in ERAD pathway across Pex1 RNAi, Pex12 RNAi, and Pex5 RNAi samples.

164

A B C C11A specific GO terms (UP) PEX5KD specific GO terms (UP) C11A PEX5KD GO:0005777 Peroxisome GO:0006955 Immune Response GO:0042632 Cholesterol homeostasis GO:0005634 Nucleus 4397 26 354 GO:0005615 Extracellular space GO:0051252 RNA metabolism GO:0008202 Steroid metabolism GO:0051171 Nitrogen metabolism GO:0019752 Carboxylic acid metabolism Up-regulated genes GO:0043167 Ion Binding GO:0006629 Lipid metabolism

0 10 20 30 0 5 10 15 20 -log10(P-value) -log10(P-value)

D E F C11A specific GO terms (Down) PEX5KD specific GO terms (Down) C11A PEX5KD

GO:0005739 Mitochondrion GO:0030055 Cell-substrate junction

2996 6 233 GO:0006412 Translation GO:0001933 Negative regulation of protein phosphorylation GO:0042254 Ribosome biogenesis GO:0043409 Negative regulation of MAPK cascade GO:0022613 Ribonucleoprotein complex biogenesis Down-regulated genes

0 10 20 30 40 50 0 2 4 6 8 -log10(P-value) -log10(P-value)

H I G Oxidative Phosphorylation MAPK signaling 0 0.6 -0.1 0.5 NAME NES FDR RNAi Regulation 0.4 -0.2 0.3 -0.3 Spliceosome 1.85 0.003 PEX5 Down 0.2 -0.4 0.1 -0.5 0 -0.6

Cholesterol Metabolism -1.45 0.019(P-value) PEX5 UP Enrichment Score Enrichment

PEX5&LIPOPROTEIN Up in Ctrl Up in C11A Up in Ctrl Up in C11A N J Regulation of lipolysis in adipocytes K HIPPO signaling APOB 0 0 APOA2 -0.2 -0.2 -0.4 PLTP -0.4 APOC2 -0.6 -0.6

APOE Score Enrichment APOC1 * Up in Ctrl Up in C11A APOA4 Up in Ctrl Up in C11A APOA1 Fold Change (KD/Ctrl) Change Fold ** L Spliceosome M Cholesterol metabolism APOA5 **** 0.6 0 APOC3 **** 0.4 -0.2 0 2 4 6 0.2 -0.4

Fold Change (RNAi / Ctrl) 0 -0.6 Enrichment Score Enrichment

Up in Ctrl Up in PEX5KD Up in Ctrl Up in PEX5KD

Figure 4: Peroxisome dysfunction up-regulates inflammation and cholesterol pathways in human cells. A Venn diagram analysis showing commonly up-regulated genes by PEX5C11A induction and PEX5 RNAi. B-C Gene ontology analysis on up-regulated genes only in PEX5C11A or only in PEX5 RNAi cells. D Venn diagram analysis showing commonly down-regulated genes by PEX5C11A induction and PEX5 RNAi. E-F Gene ontology analysis on down-regulated genes only in PEX5C11A or only in PEX5 RNAi cells. G List of enriched pathways from GSEA analysis on PEX5 RNAi cells. H-K GSEA enrichment profiles on PEX5C11A: Oxidative phosphorylation, MAPK signaling, regulation of lipolysis in adipocytes, HIPPO signaling. L-M GSEA enrichment profiles on PEX5 RNAi: spliceosome, cholesterol metabolism. N Selected genes from cholesterol metabolism pathways on PEX5 RNAi cells. Y-axis represents fold change (RNAi / Control).

165

PEX5 C11A - DOWN NAME NES FDR Ribosome 2.35 <0.0001 Oxidative phosphorylation 1.97 0.024 Parkinson disease 1.72 0.013 Valine and isoleucine degradation 1.67 0.024

PEX5 C11A – UP (Top 20 pathways) NAME NES FDR Signaling pathways regulating pluripotency of stem cells -2.06 <0.0001 Spnocerebellar ataxia -2.00 <0.0001 MAPK signaling pathway -1.94 0.001 MicroRNAs in cancer -1.89 0.004 Osteoclast differentiation -1.88 0.003 Neuroactive ligand-receptor interaction -1.87 0.004 Regulation of lipolysis in adipocytes -1.85 0.004 HIPPO signaling pathway -1.84 0.004 Type II diabetes mellitus -1.83 0.004 WNT signaling pathway -1.83 0.004 Acute myeloid leukemia -1.82 0.003 Breast cancer -1.80 0.004 Cushing syndrome -1.80 0.004 B cell receptor signaling pathway -1.78 0.005 CAMP signaling pathway -1.78 0.004 Aldosterone-regulated sodium reabsorption -1.78 0.004 CGMP-PKG signaling pathway -1.77 0.004 Proteoglycans in cancer -1.76 0.005 Olfactory transduction -1.76 0.004 TH1 and TH2 cell differentiation -1.75 0.005

Table 2 Selected list of PEX5C11A regulated pathways

166

A Ribosome Biogenesis Gene Set B Ribosome Gene Set

8 3 Genotypes Genotypes Pex12RNAi-Dros C11A-hm 6 Pex5RNAi-Dros Pex5RNAi-Dros 2 PEX5RNAi-hm

Density 4

1 2

0 0 0.5 1 1.5 2 2.5 0.5 1 Fold change (RNAi / Ctrl) Fold change (RNAi / Ctrl) C 90S pre-ribosome components Nucleolus Gene / Gene pre-rRNA 5’ 3’ Fold change in Pex5 KD (fly) < 0.7 UTP-A complex Pol I 18S 5.8S 25S Fold change in Pex12 KD (fly) < 0.7 WDR43/CG30349 UTP8 TCOF1 Fold change in PEX5 KD (hm) < 0.7 UTP4 / l(3)72Dn UTP15 / CG3071 U3 snoRNP complex rRNA modification SNORD3A/3C/3B-1/3B-2 Pseudouridylation HEATR1 / UTP9 l(2)k09022 2’-O-methylation DKC1 / Nop60B NHP2 / NHP2 WDR75 FBL / Fib NOP56 / nop56 GAR1 NOP10 / CG7637 NOP58 / nop5 SNU13 / hoip Box H/ACA snoRNPs UTP-B complex Box C/D snoRNPs UTP18 TBL3 90S pre-ribosome WDE36 WDR3

UTP6 / CG7246 PWP2 / CG12325 FCF1 UTP14A/14C/ DROSHA CG12301 5’ UTP-C complex 3’ 18S 5.8S 25S Ribosomal proteins CSNK2B / Ste12DOR U3 CSNK2A1/2/3 complex NOL6 / Mat89Ba EMG1 RRP7A / CG9107 DEAD-box NAT10 / helicases BMS1/CG7728 l(1)G0020 MPP10 complex RCL1/Rtc1 MPHOSPH10 / CG13097 3’ 18S 5.8S 25S IMP3/CG4866 IMP4 U3 complex RPP38/30/40 POP4/5/7 pre-40S POP1 pre-60S RPP25/L 18S 5.8S 25S RMRP

GTPBP4

Nob1 5.8S 25S GNL3 GNL2 18S REXO1/2/5 /CG12877 XRN1/2 NVL 5.8S 25S MDN1 Pre-40S HRR25 RBM28 Pol III Export factors 5S RAN NXF1/2/2B/3/5 /Nxf3 XPO1 NXT1/2 / Nxt1 5.8S 25S NMD3 5S EIF6/eIF6 Nucleus Pre-60S

Cleavage RIOK1 / RIOK1 SBDS SPATA5 / CG5776 AK6 / AK6 RIOK2 Cytoplasm EFL1 / CG33158 LSG1 / Ns3

Mature 60S Mature 40S 18S 5.8S 25S 5S

167

Figure 5 Peroxisome dysfunction represses ribosomal genes in both flies and humans: A Density plot showing the fold change (Treatment / Control) of ribosome biogenesis pathway genes in oenocytes and PEX5 RNAi in human cells. B Density plot showing the fold change (Treatment / Control) of ribosome pathway genes in oenocytes and PEX5C11A in human cells. C Schematic diagram showing ribosome biogenesis pathway, the role of biogenesis factors and their fold change in oenocytes (Pex5, Pex12 RNAi) and human cells (PEX5 RNAi). Genes reduced the expression under treatment for more than 0.7-fold are noted on the diagram (red arrow indicates genes reduced under Pex5 RNAi in oenocytes, yellow indicates Pex12 RNAi, blue indicates PEX5 KD in human cells).

168

A D. melanogaster Mammalian B 5’ETS ITS1ITS2 3’ETS 5’ETS ITS1 ITS2 3’ETS 1 3 pre-rRNA (47S) 1 3 pre-rRNA (47S) 2 4 Probes: ETS ITS1 2 4 45S 45S 43S 30S 32S a (33S) d (23S) b 41S 26S 21S 32S 18S b 21S-C 12S c 18S-E e(12S) 18S-E 7S

18S 5.8S 2S 28Sa 28Sb 18S 5.8SS/L 28S

1.5 F C ETS D ITS1 E Uncleaved ETS/18S 18S G Uncleaved ETS/18S H 1.5 1.5 ns 2.5 18S 0.0015 0.015 0.009 ns 0.0049 0.0024 0.0001 1.0 2.0 1.0 0.0010 1.0 0.010 1.5 0.5 0.5 0.0005 0.5 1.0

Relative Expression Relative 0.005

Relative Expression Relative

Relative Expression Relative Relative Expression Relative

Relative Expression Relative 0.5 0.0000 0.0 0.0 0.0 Expression Relative U U R R 0.000 U U 0 0 0.0 R R U U 0 U U 0 0 R R 2 R R x x 0 0 0 0 0 o o x x 2 0 0 2 -D D o o 2 + -D D S S + T T S S E E 8 8 1 1 I J K L

RPS6 Merge RpS6 Merge p<0.0001 0.8 25 0.0005

ctrl 0.6 20

) Dox

- (

0.4 15 /total intensity /total

0.2 10

RNAi

Fly Oenocytes Fly

HEK293 cells HEK293 RPS6 5 Pex5 0.0

U U

R R (+) Dox 0 0 0

0 (%) cells cells/total rpS6 Nucleus 2 x x o o D D ) ) (- + (

Figure 6 Impaired ribosome biogenesis genes in human and flies: A-B Overview of D. melanogaster, and Mammalian rRNA biogenesis Pathways. rRNA biogenesis intermediates are generally conserved among vertebrates; 5.8S rRNA is cleaved into 2S and a short 5.8S, and 28S rRNAs are cleaved into 28Sa and 28Sb. Blue arrows indicate the locations of the primers for ETS. Green label indicates primer locations for ITS1. Red arrow indicates primers for 18S. C-F The relative amounts of the ETS, ITS1, unprocessed ETS/18S ratio, 18S were measured by qPCR using RNA isolated from control oenocytes or Pex5 RNAi oenocytes. Data are presented as the mean ± s.e.m (n = 2). G-H The relative amounts of the unprocessed ETS/18S ratio, 18S were measured by qPCR using RNA isolated from control cells (-DOX) or PEX5C11A (+DOX) cells. Data are presented as the mean ± s.e.m (n = 3). I-L Analysis of ribosome proteins (RPs) 169 nucleolar localization. I Depletion of Pex5 in oenocytes results in the accumulation of RpS6 nucleolus. Scale bar shows 3.3μm. J Quantification of RpS6 intensity in nucleolus normalized to total intensity within the cell. K Expressing PEXC11A (+Dox) results in the accumulation of RpS6 nucleolus in HEK293 cells (n=6). Scale bar shows 20μm. L Quantification of number of cells containing RpS6 signal in nucleolus normalized to total number of cells, in control (-Dox) and PEX5C11A expressing cells (+Dox) (n = 6).

170 Supplementary figure 1

Pex5_200 B

Pex5_0RU yw_200 Genotypes Pex12_200 C11A - DOX yw_0RU C11A PEX5KD C11A + Dox Pex12_0RU C11A_Ctrl PEX5_Ctrl PEX5RNAi-hm Pex1_200 PEX5control-hm

Pex1_0RU

C D Cluster 4, specifically up-regulated by RU feeding: Cluster 2, commonly induced by Pex5 and Pex12:

GO:0034976 response to ER stress GO:0048477 oogenesis GO:0005975 carbohydrate metabolic process GO:0031057 histone modification GO:0005978 glycogen biosynthetic process GO:0006351 transcription GO:0005811 lipid particle GO:0022008 neurogenesis 0 1 2 3 4 5

-log10(p value) 0 2 4 6 8 -log10(p value)

C D Cluster 9&11, commonly induced by Pex5 and Pex1: Cluster 10&12, commonly induced by Pex12 and Pex1:

GO:0006635 fatty acid beta-oxidation GO:0043248 proteasome assembly

GO:0000398 mRNA splicing GO:0006412 translation

GO:0006457 protein folding GO:0006120 mitochondrial electron transport GO:0005777 peroxisome GO:0005840 ribosome GO:0032543 mitochondrial translation

0 5 10 15 20 0 20 40 60

-log10(p value) -log10(p value)

E F

Mitochondria Cell cycle

171

Supplementary figure 1: A PCA analysis on gene expression from oenocyte samples: Pex1, Pex12, Pex5 and yw. 0RU represents the control group, 200RU represents the RNAi group. B PCA analysis on gene expression from human cell culture samples: PEX5C11A (-DOX) versus PEX5C11A (+DOX); control versus PEX5RNAi. C-D GO analysis for clusters identified in Figure 2A. E Oenocyte genes that are commonly repressed by all Pex knockdowns but not affected by RU feeding. F Genes that are commonly induced by PEX5 RNAi and PEX5C11A in human cell cultures.Supplementary figure2

A B

Up in Ctrl Up in Pex5RNAi Up in Ctrl Up in Pex12RNAi Up in Ctrl PEX5KD

C Ribosome (Dros) Ribosome (Hm) Ribosome (Hm) 0.8 0.5 0.6 0.6 0.3 0.4 0.4 0.2

0.2 0.2 Enrichment Score Enrichment Enrichment Score Enrichment 0 0 Score Enrichment 0

yw Pex5 Up in Ctrl Up in Pex5RNAi Up in Ctrl Up in PEX5 KD Up in Ctrl Up in C11A log2(FC) value

D All-gene-control-C11A E All-gene-control-PEX5 (hm) F ER pathway

G Pex5 (dros) H Pex12 (dros) I Pex1 (dros)

Supplementary figure 2: A GSEA enrichment profiles on ribosome biogenesis pathway in Pex5 RNAi, Pex12 RNAi of oenocyte samples and PEX5 knockdown in human cell. B Heatmap analysis plotting log2 (fold change) value in yw samples (200 RU/0RU) and Pex5 RNAi samples in oenocytes. C GESA enrichment profiles in ribosome pathway in Pex5 RNAi in oenocyte samples, PEX5 knockdown and PEX5C11A samples in human cells. D-E Density plot on fold 172 change of all genes in PEX5C11A group and PEX5 knockdown group. F Density plot on fold change of ER pathway genes in PEX5C11A group and PEX5 knockdown group. G-I GSEA enrichment profiles on protein processing in endoplasmic reticulum in Pex5 RNAi, Pex12 RNAi and Pex1 RNAi samples in oenocytes. 6.8 Appendix

Supplementary Table 1: Primer list

Oligonucleotides Sequence 5'-3' Species ITS1-F TTATTGAAGGAATTGATATATGCC Drosophila ITS1-R ATGAGCCGAGTGATCCAC Drosophila ETS-F GCTCCGCGGATAATAGGAAT Drosophila ETS-R ATATTTGCCTGCCACCAAAA Drosophila ETS-F-unprocessed cgagtgctatataaaaatggccg Drosophila ETS-R-unprocessed gcatataactactggcaggatc Drosophila 18S-f TGGTCTTGTACCGACGACAG Drosophila 18S-r GCTGCCTTCCTTAGATGTGG Drosophila ETS-F-unprocessed ctcgccgcgctctacctt Homo Sapiens ETS-R-unprocessed gcgcccgtcggcatgtattagctc Homo Sapiens 18S-f CTTTCGATGGTAGTCGCCGT Homo Sapiens 18S-r CCTTGGATGTGGTAGCCGTT Homo Sapiens

173

CHAPTER 7. CONCLUSIONS

The process of aging has held a longstanding fascination for the scientific community. In recent decades, aging research has been accelerated thanks to advancement in the realm of molecular and cellular biology. In this thesis, we utilized Drosophila oenocytes, a hepatocyte- like tissue, to decipher the molecular mechanism of aging and its interaction with oxidative stress, with surprising findings on the role of peroxisome in this process.

Firstly, our oenocyte translatomic analysis identified many conserved genes and pathways between Drosophila oenocytes and mammalian liver, highlighting the functional similarities between the two tissues. Genes involved in pathways, such as oxidative phosphorylation, ribosome, proteasome, peroxisome, innate immune response, MAPK cascade and lipid metabolism are similarly regulated between oenocytes and liver during aging. Thus, our translatome analysis provide important genomic resource for future dissection of oenocyte function and provide insights on the mechanism of liver aging.

Our studies are the first to establish the causal role of liver aging on cardiac arrhythmia.

Even though numerous evidence have suggested the close interaction between liver function and cardiovascular diseases, most of the studies are epidemiological and based on associations. We present genetic evidence showing that upd3 in Drosophila oenocytes can non-autonomously induce cardiac arrhythmia. Furthermore, we also show that oenocyte-specific knockdown of upd3 is sufficient to block aging induced cardiac arrhythmia. We show that the age-dependent induction of upd3 is due to impaired peroxisomal import and elevated JNK signaling in aged oenocytes. Furthermore, oenocyte-specific overexpression of Pex5 can block age-related upd3 induction and rescues cardiac arrhythmicity during aging. Our studies identify an important role 174 of hepatocyte-specific peroxisomal import in mediating non-autonomous regulation of cardiac aging.

Due to the important role of peroxisome in aging regulation and the fact that peroxisome’s activity is largely impaired during this process, we next sought to understand transcriptomic responses to peroxisome import defects. Our studies confirm previous findings that peroxisomes are highly dynamic, and they interact closely with other organelles, especially mitochondria and ER. We also identify that different peroxisome biogenesis factor may control transcriptomic responses differently. Our study identifies a novel phenomenon by which peroxisome dysfunction conservatively impair ribosome biogenesis in the nucleolus. The molecular mechanism of this interaction is yet unknown, possibly through damaging biogenesis factors localized on the nucleolus.

Concluding remarks and future perspectives

In summary, our studies provide insights in the fields of peroxisome, inflammaging and inter-tissue communications. Increasing evidence suggests that peroxisomal function is affected during aging, due to impaired peroxisomal import function. Our studies are the first to prove that peroxisomal dysfunction is the cause of age-related arrhythmia, and we demonstrate that hepatic peroxisome function could be targeted to prevent aging-related cardiovascular diseases. Impaired peroxisomal function has been implicated in other diseases, such as in neurodegenerative disorders, beta cell failure, and diabetes-related complications even cancers. The molecular mechanisms underlying the role of peroxisomal dysfunction in these diseases are not well understood.

Based on the essential role of peroxisomes in lipid and ROS metabolism, peroxisomal dysfunction could contribute to pathogenesis of age-related diseases through imbalanced ROS 175 homeostasis, membrane lipid remodeling, or generation of lipid signaling. It is yet unknown whether peroxisome dysfunction influences other organelles through these mechanisms. How does peroxisomes affect mitochondria function, ER function and nucleolus? Finally, whether and how impaired peroxisome in hepatocytes affects distant organs non-autonomously, through cytokine secretion or metabolic remodeling. Understanding these questions will hopefully shed lights on age-related pathologies and provide therapeutic targets.