Characterizing the role of host lipid metabolism on microsporidia infection in Caenorhabditis elegans
by
JiHae Jeon
A thesis submitted in conformity with the requirements for the degree of Master's of Science
Graduate Department of Molecular Genetics University of Toronto
© Copyright by JiHae Jeon 2021
Characterizing the role of host lipid metabolism on microsporidia infection using the nematode Caenorhabditis elegans
JiHae Jeon
Master’s of Science
Department of Molecular Genetics University of Toronto
2021
Abstract
Microsporidia are fungal obligate intracellular pathogens that primarily manifest in immunocompromised individuals. Microsporidia have reduced metabolic capabilities and so they have strong host-dependence for many metabolic processes. However, despite their growing medical importance, microsporidia are poorly understood. I investigated the impact of host lipid metabolism on microsporidia infection using the nematode Caenorhabditis elegans and its natural microsporidian pathogen Nematocida parisii. A previous metabolomics screen performed on N. parisii infected C. elegans identified upregulation of lipid metabolites in three lipid biosynthesis pathways: phosphatidylcholines, acylcarnitines and ceramides. I also identified an increase in the level of lipid droplet associated lipase,
ATGL-1, upon infection. In addition, by screening 25 lipid mutant strains, animals defective in producing sphingosine showed resistance to infection whereas supplementing sphingosine increased susceptibility, suggesting sphingosine may be involved in promoting microsporidian growth. Together, my research of lipid regulation by microsporidia can help to determine new therapeutic targets for microsporidian infection.
ii Acknowledgment
First and foremost, I’d like to thank my supervisor, Dr. Aaron Reinke, for giving me the opportunity to do this project in his lab and his continual support and guidance throughout my master’s degree. He has been an amazing mentor and his commitment and enthusiasm to science are truly inspirational. I am truly thankful for the opportunity to work in his lab.
I would also like to thank Dr. Nick Burton for his work on lipid metabolomics and my committee members, Dr. Brent Derry, Dr. Greg Fairn, and Dr. Amy Caudy for their valuable insights and guidance on my project. I am thankful for their thoughtful questions that helped me shape my critical thinking skills.
I would like to thank the entire Reinke lab for all the support and assistance they have offered me through the development of my work. Their helpful guidance and input helped me to survive the day-to-day life of grad school. I am grateful for all the meaningful conversations and good times (the usual all-you-can-eat sushi) I have shared with all of them.
Thank you to all of my incredible friends, Nuri who cheered me up all the way from
Waterloo, and especially my boyfriend, Simon, who always provided me with emotional support and encouragement. Thank you for being one of my biggest supporters.
Last but not the least, I would like to thank my parents, my sister and my grandmother, without whom, none of this would’ve been possible. Thank you to my sister, Alice, who always surprised me with treats whenever I am down. Finally, to my parents and grandmother especially, I thank you most ardently for your continuous support of my aspirations. I am forever indebted to my parents for giving me the opportunities and experiences that have made me who I am.
iii Table of Contents
ACKNOWLEDGMENT ...... III
LIST OF TABLES ...... VII
LIST OF FIGURES ...... VIII
LIST OF ABBREVIATIONS ...... IX
CHAPTER 1: INTRODUCTION ...... 1
1.1 Microsporidia ...... 1 1.1.1 History ...... 1 1.1.2 Microsporidia genome reduction ...... 1 1.1.3 Developmental morphology and life cycle ...... 3 1.1.4 Microsporidia infections in humans ...... 5 1.1.5 Microsporidia infections in C. elegans and N. parisii ...... 5
1.2 Lipid Metabolism and host-pathogen interaction ...... 7 1.2.1 Lipid metabolism in C. elegans ...... 7 1.2.1.1 Phosphatidylcholine biosynthesis and function ...... 8 1.2.1.2 Sphingolipid biosynthesis and function ...... 9 1.2.1.3 Carnitine shuttle and function ...... 10 1.2.2 Lipid metabolism in microsporidia ...... 11 1.2.3 Importance of lipids in parasitic infection ...... 12
1.3 Thesis Rationale ...... 14
CHAPTER 2 MATERIALS AND METHODS ...... 16
2.1 Materials and Methods ...... 16 2.1.1 Nematode strains and microsporidia culture ...... 16 2.1.1.1 C. elegans strains ...... 16 2.1.1.2 N. parisii spore preps ...... 17 2.2 Fasting experiment ...... 18 2.3 Lipid staining ...... 18
iv 2.3.1 Oil red O staining ...... 18 2.4 Microsporidian infection experiment ...... 19 2.4.1 Optimizing infection conditions ...... 19 2.4.2 24h, 48h, and 72h infection with atgl-1::gfp ...... 19 2.4.3 Infection assay between atgl-1::gfp and atgl-1(P87S) ...... 20 2.4.4 Lipid mutants N. parisii infection ...... 20 2.4.5 Pulse-infection assay ...... 20 2.4.6 DY96 and FISH staining using PFA fix ...... 21 2.4.7 DY96 staining using acetone fix ...... 21 2.4.8 FISH staining using acetone fix ...... 21 2.5 Immunofluorescent Microscopy ...... 22 2.5.1 Live imaging and Quantification ...... 22 2.5.2 Oil red O Quantification ...... 22 2.5.3 Embryo count and pathogen load analysis by DY96 ...... 22 2.5.4 Pathogen load and sporoplasm analysis by FISH ...... 22 2.5.5 Brood count analysis and Synthetic effects score calculation ...... 23 2.6 Sphingosine Supplementation ...... 23 2.7 Lipid metabolite quantification ...... 24
CHAPTER 3 RESULTS ...... 24
3.1 Introduction ...... 24 3.2 Understanding changes in host lipid metabolism in response to microsporidia infection ...... 27 3.2.1 ATGL-1 is upregulated in response to fasting ...... 27 3.2.2 The level of ATGL-1 increases in a time-dependent manner upon microsporidian infection ...... 29 3.2.3 Lipid stores are reduced in response to N. parisii infection, but the function of ATGL-1 is unknown...... 31
3.3 Characterizing the impact of host lipid metabolism on microsporidia infection ...... 36 3.3.1 Various lipid metabolites are upregulated upon N. parisii infection .... 36 3.3.2 A screen of 25 lipid mutants identifies lipid metabolites involved in resistance or hypersusceptibility to microsporidian infection ...... 39 3.3.3 Sphingosine supplementation increases susceptibility of C. elegans to N. parisii infection ...... 46 3.3.4 Sphingolipid mutants do not exhibit invasion defects ...... 48
v CHAPTER 4 DISCUSSION AND CONCLUSIONS ...... 50
4.1 Discussion ...... 50 4.1.1 Summary of main findings ...... 50 4.1.2 Changes in LDs and LD associated protein, ATGL-1, in response to N. parisii infection ...... 50 4.1.3 Sphingosine as the potential factor in promoting microsporidian growth ...... 52
4.2 Future Directions ...... 56 4.2.1 Investigate changes in LDs size and structure during infection ...... 56 4.2.2 Explore interaction between host’s sphingosine lipid metabolism pathways and N. parisii ...... 56
REFERENCES ...... 58
vi List of Tables
Table 1: List of worm strains used in the study……………………………………………..16
Table 2: A list of lipid mutants……………………………………………………………...39
Table 3: Lipid mutants and corresponding ‘synthetic effects’ scores...……………...……..41
vii List of Figures
Figure 1: Analysis of sequenced microsporidia genomes……………………………..……2
Figure 2: Structure of spore and general life cycle of microsporidia. …………….……..…4
Figure 3: Diagram of typical lipid droplet in C. elegans.………………………….….….…7
Figure 4: General structure of phospholipid………………………………………..….……8
Figure 5: Biosynthesis pathway of phosphatidylcholine.……...………………….…..….…9
Figure 6: Biosynthesis pathway of sphingolipids…………………………………….……10
Figure 7: Acyl carnitine shuttle system.…… ………………………….…………….….…11
Figure 8: Lipid metabolism pathways in microsporidians….……………………...…....…13
Figure 9: Quantification of ATGL-1::GFP level in fed and fasted fixed young adult stage atgl-1::gfp worms….…………….…………….…………….…………….………………28
Figure 10: Quantification of ATGL-1::GFP level in atgl-1::gfp at 24, 48, and 72 h post infection. ….…………….…………….…………….…………….…………….…………30
Figure 11: Changes in abundance of fat stores in response to N. parisii infection using Oil
Red O lipid dye. ….…………….…………….…………….…………….………….….…32
Figure 12: N. parisii infection in wildtype and ATGL-1 mutant animals. ….………….…34
Figure 13: Lipid metabolites identified from LC/MS profiling.….……………....….….…37
Figure 14: Quantification of embryos and pathogen load………………………………….43
Figure 15: N. parisii infection in wildtype and 25 lipid mutant animals. ….…….……..…44
Figure 16: Sphingosine supplementation reduces host fitness in response to N. parisii infection….…….…….…….…….…….…….…….…….…….…….…….…….……...….46
Figure 17: Sphingolipid mutant animals do not exhibit invasion defect upon N. parisii infection.….…….…….…….…….…….…….…….…….…….…….…….…….…...……49
viii List of Abbreviations
ASAH Acylsphingosine amidohydrolase
ASM Acid sphingomyelin phosphodiesterase
ATGL-1 Adipose Triglyceride Lipase 1
BME β-mercaptoethanol
C. elegans Caenorhabditis elegans
C1P Ceramide-1-phosphate
CDP-Choline Cytidine diphosphate choline
CDP-DAG Cytidine diphosphate diacylglycerol
CERK-1 Ceramide kinase 1
CGC Caenorhabditis Genetics Centre
CPT Carnitine O-palmitoyltransferase
CTP Cytidine triphosphate
DAPI 4′,6-diamidino-2-phenylindole
DY96 Direct Yellow 96
EDTA Ethylenediaminetetraacetic acid
ELO fatty acid elongase
FAT Fatty acid desaturase
FISH Fluorescent in situ hybridization
GFP Green fluorescent protein
HYL Homolog of yeast longevity gene
IPR Intracellular pathogen response
ix LAGR Longevity-assurance gene related
LD Lipid droplet
MAPK Mitogen-activated protein kinase
MRWB Modified Ruvkun's Witches Brew
NGM Nematode Growth Medium
N. parisii Nematocida parisii
NR Nile Red
PA Phosphatidic acid
PBS Phosphate-buffered saline
PC Phosphatidylcholine
PE Phosphatidylethanolamine
PEMT Phosphatidylethanolamine N-
methyltransferase
PFA Paraformaldehyde
PI Phosphatidylinositol
PIPES 1,4-Piperazinediethanesulfonic acid
PS Phosphatidylserine
PUFA Polyunsaturated fatty acids
RFP Red fluorescent protein
S1P Sphingosine-1-phosphate
SCF Skp, Cullin, F-box containing complex
SDS Sodium dodecyl sulfate
SPHK Sphingosine kinase
x SPTL Serine palmitoyl transferase
TTM Toxin-regulated targets of MAPK
*All images in “Chapter 1:Introduction” were created with BioRender.com
xi Chapter 1: Introduction
1.1 Microsporidia 1.1.1 History
Microsporidia are a group of fungal-related eukaryotes which comprise obligate intracellular parasites that infect a wide range of hosts, including humans. The first microsporidian infection was discovered in the 1800s by Louis Pasteur when he identified a disease known as pébrine or “pepper disease” that had been plaguing majority of silkworm populations in southern Europe. He then described microsporidian species Nosema bombycis to be the causal agent of this disease in silkworms. Since then about 150 years of microsporidian research characterized over 1200 individual species from approximately 150 genera of microsporidia. These pathogens are widely distributed in the environment and can infect a wide variety of hosts including vertebrate and invertebrates (Keeling and Fast 2002). They most commonly infect arthropods where they have been studied for their use as biological control agents for pests such as grasshoppers, corn borer and mosquitoes (Bjornson and Oi 2014). Moreover, historically, microsporidia had a substantial impact on global food chain where in nature, they have been linked to colony collapse disorder in United States and Europe in which a majority population of worker bees disappear leading to sudden colony deaths of pollinators such as honeybees. As well, microsporidia directly infect many wild and farmed fish and crustacean populations that are destined for human consumption (Stentiford et al. 2016). Despite their medical and agricultural importance, very little is known about microsporidia biology and treatment is limited, thus more research is needed to understand these pathogens to develop more efficient treatments.
1.1.2 Microsporidia genome reduction
As a consequence of the strict obligate intracellular parasitic lifecycle of microsporidia, the genomes of microsporidia are severely reduced and thus they are heavily dependent on their hosts. Their genome size ranges from only 2.3 Mbp, one of the smallest eukaryotic genomes, to 51.3 Mbp depending on species in question (Wadi and Reinke 2020). The first complete microsporidian genome belonging to Encephalitozoon cuniculi revealed valuable
1 information about the nature of the genome. The genome of E. cuniculi is 2.9 Mbp in size and encodes fewer than 2000 protein coding genes, supporting the idea that microsporidia have experienced substantial gene loss (Keeling and Fast 2002).
Figure 1: Analysis of sequenced microsporidia genomes (A) Phylogenetic tree of whole- genome sequenced microsporidia and related species. Illustrations to the right of each species represent host(s) that each species are reported to infect. (B) Protein content and (C) genome size for each species are compared with the largest at 51.3Mb, encoding approximately 4,200 proteins (Edhazardia aedis) and the smallest at 2.3Mb (Encephalitozoon species), encoding approximately 1,800 proteins. (D) Several morphological characteristics and key eukaryotic processes shared by each species. Entire figure reproduced from journal article by Wadi and Reinke (Wadi and Reinke 2020).
Genes that encode proteins involved in essential cellular functions such as DNA replication, transcription and translation were retained but as microsporidia are reliant on their hosts for many intracellular resources, they have lost numerous genes encoding complete metabolic or regulatory pathways which includes biosynthesis of primary metabolites such as amino acids, nucleotides and lipids and have limited capacity to generate energy from adenosine triphosphate (ATP), fatty acids oxidation, and tricarboxylic acids cycle. Meanwhile, genes encoding various transporters to import metabolites and nutrients from their hosts are overrepresented in their genome (Keeling and Slamovits 2004). The genome analysis of
2 microsporidia showed that substantial gene loss of many important biochemical pathways resulted in severely reduced metabolic potential, yet they are able to proliferate and thrive within host cells. Consequently, the hosts become energetically impoverished as they constantly generate metabolites (nucleotides, amino acids, and lipids) that are instead used by the microsporidia for energy (Cuomo et al. 2012). The mechanism behind how microsporidia are increasing biosynthetic output of host and stealing these factors remain to be further explored.
1.1.3 Developmental morphology and life cycle A typical life cycle of microsporidia is divided into three phases: the infective or environmental phase, the proliferative phase and the sporogonic phase. The first phase: infective/environmental phase, is the only extracellular phase of microsporidia where they exist as a spore form (Cali, Becnel, and Takvorian 2017). The spore is a single, highly organized cell which range in size from 1�m to 40 �m and are mostly in ovoid shape. The spore contains a sporoplasm, which contains the cytoplasm and nucleus of the spore and is the infectious material. (Keeling and Fast 2002). Spores are shed into the environment from the infected hosts which can be transmitted to the new hosts.
Two different models of microsporidia invasion of a host cell exist: active invasion and invasion synapse. Upon appropriate environmental triggers, spores germinate at which the characteristic structure called the polar tube is extruded out of the spore. The polar tube is a specialized invasion apparatus that is attached to an anchoring disk and is coiled within the spore. During the active invasion, the polar tube pierces and injects infectious sporoplasm directly into the cytoplasm of host cells (Franzen 2004). The second invasion method involves a formation of microenvironment called invasion synapse. Upon extrusion, the polar tube proteins interacts with host cell membrane to create an invagination that allows sporoplasm to travel down into the host cells (Han, Takvorian, and Weiss 2020). Once inside the host cell, the parasite enters phase two: proliferative phase, and is now referred to as meronts, a cell wall deficient form of microsporidia. The meronts repeatedly perform replication through binary or multiple fission to form multinucleate plasmodial. When the cell membrane of meronts thickens, they enter the last phase: sporogonic phase, and are referred to as sporonts. During the last phase, the sporonts are committed to spore
3 production. The sporonts divide to generate sporoblasts which undergo morphogenesis to re- generate spores. After the division, extrusion apparatus first develops which is followed by development of endospore layer and decrease in cell size. Once mature spores are formed, they are released for further disease transmission (Keeling and Fast 2002).
A
B
Figure 2: Structure of spore and general life cycle of microsporidia. (A) Diagram of the internal structure of a microsporidian spore. The spore coat consists of: an outer electron- dense exospore, an electron lucent endospore and plasma membrane. The spore contains one or two nuclei and the polar tube which is attached to the anchoring disk near the anterior end of the spore. The number of polar tube coils varies depending on species. Near the anterior and posterior region, the polar tube is surrounded by a pile of membrane sacs, polaroplast and the posterior vacuole, respectively. (B) Typical lifecycle of microsporidia. 1. Infective phase: Germination of the ingested mature spore which resulting in host cells invasion. The sporoplasm is injected into the cytoplasm of host cell. 2. Proliferative phase: Once inside, the sporoplasm develops into its proliferative forms, meronts, and further multiplies. 3.
4 Sporogonic phase: Cell membrane of meronts thickens to produce sporonts which develops into mature spores. Mature spores exit the cells and are released into the environment. 1.1.4 Microsporidia infections in humans As of current, microsporidia are considered priority pathogens by National Institutes of Health and with approximately 15 human infecting microsporidia species which comprise of four different genera: Encephalitozoon, Enterocytozoon, Nosema and Pleistophora. The first human case of microsporidia infection was reported in 1985 in an AIDS patient. The causal parasite was Enterocytozoon bieneusi and the infected patient suffered from chronic diarrhea. Since then, E. bieneusi has proven to be the most common human infecting microsporidia species followed by Encephalitozoon species (van Gool and Dankert 1995).
Microsporidia most commonly infect gastrointestinal tract particularly in immuno- compromised individuals such as HIV/AIDS patients, children, organ transplant recipients and the elderly. Once infected, microsporidia generally cause keratoconjunctivitis and chronic diarrhea that may result in dehydration, malnutrition, and malabsorption. Severe microsporidian infection may result in disseminated disease, infecting multiple other organs (Microsporidiosis 2013). Although they received wide interest as most common causal agent of chronic diarrhea and keratitis in HIV/AIDS-infected patients, recently microsporidia have been detected at increasing frequency in immunocompetent and healthy travelers with self- limiting symptoms (Weber et al. 1994). Despite the wide prevalence of microsporidia and increased incidences in immunocompetent individuals, no prophylactic anti-parasitic agents are identified. Moreover, the currently available treatments for microsporidia: albendazole and fumagillin, have numerous limitations such as ineffectiveness of albendazole against E. bieneusi infection and fumagillin being an unapproved FDA drug in addition to being not readily available (Weiss 2020). 1.1.5 Microsporidia infections in C. elegans and N. parisii The free-living nematode Caenorhabditis elegans has been developed as a biological model system to study host-pathogen interactions. They are readily found in the environment such as rotting fruits but until recently, no natural pathogens of C. elegans were isolated and only very few studies were reported on microsporidian infections in nematodes. Several microsporidia species were isolated from wild-caught Caenorhabditis nematodes that belong
5 to the genera: Nematocida, Enteropsectra and Pancytospora. Interestingly, phylogenetic analysis shows that several species from Enteropsectra and Pancytospora genus belong to clade containing human-infecting microsporidian species unlike the nematode-specific Nematocida genus (G. Zhang et al. 2016). These species are widely distributed around the world and specifically Nematocida species were found to undergo similar life cycle as other microsporidian species.
Along with collection of diverse natural microsporidian species, C. elegans provides a valuable whole-animal system to study microsporidian infection with many advantages such as transparent body, genetic tractability and similar morphological features of intestinal cells shared with mammals including humans. Nematocida parisii, also known as nematode killer from Paris, is first found and most common natural microsporidian pathogen isolated from wild C. elegans near Paris, France. N. parisii specifically cause lethal intestinal infection and its genome size of 4.1 Mbp is substantially reduced with only 2,661 predicted genes (Troemel 2016). Although its life cycle follows similar pattern as other microsporidian species, they show distinct two-phase cellular exit strategies. These parasites facilitate non- lytic cellular exit pathway where actin redistribution followed by hijacking of host exocytosis machinery allows them to exit the host intestinal cells without disturbing the cell integrity to minimize damage for increased efficiency in parasite transmission (Estes, Szumowski, and Troemel 2011).
C. elegans’ response to N. parisii is also intriguing as canonical defense pathways against bacterial and fungal infections which includes the conserved p38 MAPK pathway, are not important against N. parisii. Intriguingly, transcriptional response to N. parisii infection is similar to response against another intracellular pathogen, Orsay virus, but not to other extracellular pathogens and was later named intracellular pathogen response (IPR). IPR response includes ubiquitin-mediated response where SCF ubiquitin ligase components are up-regulated (Reddy et al. 2017). Substrates ubiquitylated by SCF ligases are degraded by the proteasome or by autophagy where during infection, individual pathogen, microsporidia cells and viral particles may be targeted by autophagy or proteasome (Bakowski et al. 2014). Nonetheless, further studies are necessary to understand the distinct immune response involved in microsporidian infections in C. elegans.
6 1.2 Lipid Metabolism and host-pathogen interaction 1.2.1 Lipid metabolism in C. elegans C. elegans break down lipids in food they consume like bacteria through lipid metabolism pathways to obtain nutrients and metabolites. Over the last decade, C. elegans has become an emerging animal model to understand lipid metabolism and related metabolic diseases due to an ease in visualization of lipid storage in whole animal level and majority of lipid genes with orthologs that are conserved in humans (Zhang et al. 2013). Biochemical analysis of C. elegans fat content revealed that similar to mammals, its composition comprised of triacylglycerol, phospholipids, a wide range of saturated, monounsaturated and polyunsaturated fatty acids (PUFAs), and sphingolipids. However, C. elegans lack dedicated adipocytes and instead store fat in droplets in their intestinal and hypodermal cells which can be visualized using various lipophilic dyes in their transparent body (Ashrafi 2007).
Lipid droplets (LDs) are evolutionarily conserved across almost all organisms with many important functions such as regulation of fat storage and mobilization. They are characterized by a neutral lipid core that is surrounded by a monolayer phospholipid membrane and proteins. Several LD-associated proteins have been identified in C. elegans which include the adipose triglyceride lipase (ATGL) ortholog ATGL-1 that is involved in fat metabolism and regulation of number and size of lipid droplets (Mak 2012). Although there are about 11 lipid metabolic pathways deduced from the C. elegans genome sequence, As described below, I will be focusing on three specific pathways: phospholipid/ phosphatidylcholine biosynthesis, sphingolipid and ceramide synthesis and carnitine shuttle (Ashrafi 2007).
Figure 3: Diagram of typical lipid droplet in C. elegans. Lipid droplets consist of a phospholipid monolayer that surrounds a neutral lipid core. Adipose triglyceride lipase,
7 ATGL-1, is one of several lipid-droplet localized proteins that accumulates on lipid droplets during fasting. The diameter of lipid droplets varies between 1-1.5µm in C. elegans.
1.2.1.1 Phosphatidylcholine biosynthesis and function The overall structure of phospholipids constitutes of glycerol, fatty acids, phosphoric acids and one additional polar group which can be substituted by choline, ethanolamine, serine or inositol that make up phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine or phosphatidylinositol respectively. Due to the amphipathic nature of phospholipids, they have an important structural role as major component of cell membranes and also a metabolic role in regulating communications between extra- and intracellular space (Da Costa and Ito 2003). Phospholipids make up bulk of cell membranes of C. elegans along with sphingolipids and cholesterol. From the four major categories of phospholipids, phosphatidylcholine (PC) and phosphatidylethanolamine (PE) are the most abundant phospholipids. Moreover, PC is composed of mostly polyunsaturated long chain of fatty acids (PUFA) compared to PE and other lipids such as triacylglycerol storage lipids (Satouchi et al. 1993). In this study, biosynthesis and function of PC will be reviewed as it appeared to be highly upregulated upon microsporidian infection in C. elegans.
Figure 4: General structure of phospholipid. Phospholipids are class of lipids with a hydrophilic ‘head’ and hydrophobic ‘tails’. The hydrophilic ‘head’ consists of a negatively charged phosphate and glycerol group. The phosphate can be attached to different molecules including choline, ethanolamine, serine, inositol and glycerol. The hydrophobic ‘tails’ are uncharged non-polar fatty acid chains which can vary in carbon group length and saturation degree.
8 There are two major biosynthesis pathways of phosphatidylcholine: CDP-DAG pathway and the Kennedy pathway. Two pathways diverge after the initial generation of phosphatidic acid (PA). In CDP-DAG pathway, PA is converted to CDP-DAG which can be used to synthesize phosphatidylinositol, phosphatidylglycerol and phosphatidylserine by the addition of inositol, glycerol-3-phosphate and serine group (Watts and Ristow 2017). The serine group of PS can be exchanged with ethanolamine and form PE which can undergo three sequential methylation catalyzed by PEMT to generate PC (van der Veen et al. 2017). The second pathway: Kennedy pathway, involves generating PC from choline that are transported into the cells via the transporters. Choline is first phosphorylated followed by the activation to form CDP-choline which then reacts with DAG (converted from PA) to form PC (Brendza et al. 2007). In mutants with low level of PC, the worms display larger LDs size, increased lipid stores and slowed reproduction rate, demonstrating important role of PC in C. elegans during development (Walker et al. 2011).
Figure 5: Biosynthesis pathway of phosphatidylcholine. Phosphatidylcholine (PC) can be synthesized by two different pathways: CDP-DAG followed by phosphatidylethanolamine methyl transferase (PEMT) pathway and Kennedy pathway. In the CDP-DAG pathway, phosphatidyl inositol (PI), phosphatidyl glycerol (PG), and phosphatidyl serine (PS) are synthesized. PS can be converted back to PE which can subsequently converted to PC by PEMT (enzyme labelled in red) via three successive methylation reactions. In Kennedy pathway, choline that enters the cell is phosphorylated to phosphocholine which is converted to CDP-choline. CDP-choline is then transferred to diacylglycerols to form PC.
1.2.1.2 Sphingolipid biosynthesis and function Sphingolipids are another group of important membrane lipid. Moreover, in C. elegans, sphingolipids are required for maintaining proper polarity of intestinal cells and lumen
9 formation during first larval stage as mutants that are incapable of synthesizing sphingolipid exhibited loss of continuous gut lumen due to a loss of polarity in intestinal cells (H. Zhang et al. 2011). Sphingolipids contain a backbone of a long-chain aliphatic amine called sphingoid base. Biosynthesis of sphingolipids begin with sphingoid base reacting with serine via serine palmitoyl transferase (SPTL) to form sphinganine. Sphinganine undergoes a further reaction with fatty acyl coenzymeA to produce ceramide via an enzyme called ceramide synthase. Ceramides can be modified to form more complex sphingolipids such as sphingomyelin by the addition of PC groups via sphingomyelin synthase. Additionally, ceramides can be broken down via acid ceramidase to generate sphingosine which can be phosphorylated into sphingosine-1-phosphate (S1P) (Watts and Ristow 2017). Several sphingolipid species have opposing roles in various cellular processes where for example, ceramide and sphingosine act as tumor-suppressor lipids whereas their phosphorylated derivatives, ceramide-1-phosphate and S1P act as tumor-promoting lipids (Pralhada Rao et al. 2013).
Figure 6: Biosynthesis pathway of sphingolipids. Synthesis of different sphingolipids begins with formation of sphinganine catalyzed by serine palmitoyl transferase (SPTL), followed by production of ceramide, sphingomyelin, and sphingosine. Ceramide and sphingosine can be phosphorylated to produce ceramide-1-phosphate and sphingosine-1- phosphate. The enzymes are labelled in red.
1.2.1.3 Carnitine shuttle and function Carnitine is characterized as a hydrophilic quaternary amine that function as a shuttle to import long fatty acids into mitochondria for �-oxidation. Carnitine can be synthesized in the body or taken in from the food. Blast analysis showed that in C. elegans, there are orthologs of carnitine biosynthesis genes that are conserved in humans (Strijbis, Vaz, and Distel 2010). When energy is needed by the cells, acyl-CoAs are transported to mitochondria for subsequent �-oxidation. Because mitochondrial membrane is impermeable
10 to both acyl-CoAs and fatty acids, carnitine binds to and import acyl residues from amino acids metabolism into the mitochondria.
Carnitine shuttle system begins with carnitine forming a high-energy ester bond with long chain fatty acids via carnitine palmitoyl transferase (CPT) which is located in the outer mitochondrial membrane to form acylcarnitines. Different acylcarnitine species are produced depending on the type of acyl residues carnitine conjugates with. Upon conjugation, acylcarnitines are transported across the inner mitochondrial membrane via carnitine acylcarnitine translocase (CACT). Inside mitochondria, another class of CPT removes carnitine from acylcarnitines to re-generate acyl-CoAs which then undergo �- oxidation. Once acyl residues are removed, carnitine can return to the cytoplasm for another cycle (Longo, Frigeni, and Pasquali 2016).
Figure 7: Acyl carnitine shuttle system. Fatty acids and acyl-CoA are transferred across the impermeable mitochondrial membrane via carnitine. Once acyl-CoA is transferred into the cytoplasm, carnitine is returned for another cycle. The enzymes involved in the shuttle system are labelled in red.
1.2.2 Lipid metabolism in microsporidia Although complete understanding of lipid metabolism in microsporidia remains to be further studied, current analysis of published microsporidian genomes shows a reduced genome resulting from an early bottleneck event that caused substantial gene loss yet keeping essential genes involved in exploiting host metabolites such as those encoding cell surface transporters. Across different microsporidian species, some enzymes that are widely
11 conserved are involved in biosynthesis of structural components of plasma membrane such as various phospholipids including phosphatidylinositol, phosphatidylethanolamine and phosphatidylcholine and sphingolipids. However, many of the lost genes are involved in de novo biosynthesis of basic biological building blocks which includes various lipids and so these molecules are either directly stolen from the host or indirectly acquired by hijacking host lipid metabolism pathways (Nakjang et al. 2013). The genes that are lost or maintained depends on species in question. For example, E. bieneusi lost genes involved in fatty acid metabolism in recent evolution (Akiyoshi et al. 2009) whereas in another human infecting species, Vittaforma cornea, most of these genes are detected (Mittleider et al. 2002).
Moreover, microsporidian genome analysis revealed enzymes involved in triacylglycerol (TAG) are conserved across several species of microsporidia (Nakjang et al. 2013) whereas enzymes involved in fatty acid biosynthesis such as fatty acid synthase complex are lacking, further supporting microsporidia utilizing cell surface-located transporters to uptake host- derived fatty acids. Interestingly, some microsporidian species such as E. cuniculi and N. parisii, encode fatty acid elongases and desaturases but no fatty acid synthase (Campbell et al. 2013). Similarly, it is suggested that cholesterol found in spore membrane and other sterols may also have host-origin since genes encoding components of isoprenoid biosynthesis pathway are missing in several lineages of microsporidia (Katinka et al. 2001). Full understanding of how microsporidia synthesize and metabolize lipids are lacking, however current analysis of microsporidian genome suggests possible link between microsporidia and host lipid metabolism during infection. 1.2.3 Importance of lipids in parasitic infection Lipids are main form of energy storage in cells with many functions including signaling, trafficking and cellular membrane organization and so they play key roles in host-pathogen interactions. Lipids are involved not only during the entry process of pathogens into the cells but also throughout the intracellular lifecycle of the pathogens as they replicate and grow in the host. Lipids have been linked with regulating immune responses during host-pathogen interactions, but host lipids and metabolism pathways are also recruited and hijacked by the pathogens to infect, survive, and multiply in the host. When pathogens infect the host, before they enter the host cells, they are first recognized by the clustering of lipids at the
12 docking site (Wenk 2006). Pathogens may modify host cell membrane composition where for instance, pathogen use their own phospholipases or trigger the host cell phospholipases to break down host cell membrane to acquire fatty acids for energy or activate various host signaling pathways (van der Meer-Janssen et al. 2010). Contrastingly, lipid signaling in the host cells can also orchestrate complex series of events that leads to the immune response of the host to eliminate pathogenic microorganisms.
Figure 8: Lipid metabolism pathways in microsporidians. Comparative analysis of published genomes of 11 microsporidian species revealed several key enzymes involved in biosynthesis of lipids including biosynthesis of some major phospholipids such as phosphatidylinositol, phosphatidylethanolamine, and phosphatidylcholine, sphingolipids and triacylglycerol. Enzymes involved in the synthesis are labelled in red. Red arrows indicate
13 the lipid molecules that are transported to the cell membrane. Blue arrows indicate molecules taken in from the environment. ABC: ATP-binding cassette transporters; CTL: choline-transporter-like transporter; MFS: Major Facilitator Superfamily.
In C. elegans, different lipid molecules and pathways have shown to play a role in immune response during infections. For instance, C. elegans modulate fatty acid synthesis pathways to induce antimicrobial peptides as part of its innate immune response upon fungal infection (K.-Z. Lee et al. 2010). In addition, polyunsaturated fatty acids (PUFAs) can modulate innate immunity in response to bacterial infection where gamma-linolenic acid and stearidonic acid are required for basal innate immunity such as regulating the expression of a number of immune specific genes and maintaining p38 MAP kinase pathway that protects animals from infections (Nandakumar and Tan 2008). Monounsaturated fatty acid oleate have also shown to be a requirement for innate immune activation and resistance to bacterial infection (Anderson et al. 2019).
Lipids are attractive targets for many pathogens to enable their survival and replication and thus many obligate intracellular parasites including microsporidia that have reduced metabolic capabilities, regulate or co-opt host lipid metabolism for nutrients and protection. For example, microsporidium, Tubulinosema ratisbonensis, hijacks host phosphatidic acid which is a critical limiting factor needed for its proliferation in Drosophila flies (Franchet et al. 2019). These lipid metabolites can then become potential targets to limit the pathogens’ growth and used to understand how the regulation of lipids may be used therapeutically to treat microsporidia infection. However, much is still unknown about the impact of lipid metabolism during microsporidian infection. 1.3 Thesis Rationale Over 16 million people are dying worldwide every year from infectious diseases caused by globally spread pathogenic microbes. Microsporidia are fungal-related obligate intracellular pathogens that are currently a major threat to agricultural industry and human health. They are associated with wide range of clinical syndromes in immunocompromised patients but cases of microsporidian infection in immunocompetent individuals are becoming more commonly recognized. Despite their prevalence and medical significance, these pathogens are poorly understood with limited and ineffective treatments. As an obligate intracellular
14 pathogen that have undergone severe genome reduction, they consequently evolved extreme host dependence for many metabolic processes including lipids. In this work, I utilize established C. elegans/N. parisii pathosystem to identify key host cellular process and metabolites exploited during microsporidian infection to enhance the understanding of microsporidia-host interactions and determine potential novel pharmacological targets in the parasite and the host.
First, to understand the changes in host lipid metabolism upon microsporidia infection, I investigated the changes in the C. elegans ortholog of the mammalian lipolytic enzyme, ATGL-1, in response to N. parisii infection. I identified initial increase in ATGL-1 protein level around the proliferative phase of microsporidia followed by substantial significant increase near mature spore formation, suggesting a fat catabolism during infection. Additionally, common lipid dye, oil Red O, was used to quantitatively measure lipid abundance. By comparing uninfected and infected animals throughout infection, I identified a decrease in fat stores upon infection. Despite, no significant changes in the pathogen load between strains with ATGL-1 overexpression and reduction of ATGL-1 function, this work revealed the role of ATGL-1 and LDs during microsporidian infection and enhanced understanding of how the lipids are regulated upon infection.
Second, to examine the impact of host lipid metabolism on microsporidia infection, I screened a panel of 31 strains of C. elegans that are defective in the production of phosphatidylcholines, acylcarnitines or sphingolipids which have been reported to be significantly upregulated upon infection in previous lipid metabolite screen. Through this mutant screening, I identified 12 mutants that exhibited change in host’s susceptibility to infection which revealed included mutants with defective enzymes involved in sphingolipid biosynthesis. Furthermore, sphingosine supplementation assay identified sphingosine as a potential limiting factor for the microsporidia to compensate for their genome reduction. Collectively, this work revealed C. elegans lipid genes that are involved in host’s resistance to microsporidia, and that are modulated by the parasite for survival. Together, these experiments highlighted the importance of host lipid metabolism to intracellular microsporidia replicative capacity and growth and uncovered novel lipid metabolites that may be used therapeutically to treat microsporidia infection.
15 Chapter 2 Materials and Methods
2.1 Materials and Methods 2.1.1 Nematode strains and microsporidia culture
2.1.1.1 C. elegans strains
All C. elegans strains were maintained at 20°C on Nematode Growth Medium (NGM) plates seeded with Escherichia coli OP50, as previously described (Brenner 1974). All strains were derived from the Bristol N2 wild-type animals and obtained from the Caenorhabditis Genetics Centre (CGC). Details of strains used are listed in Table 1.
Table 1: List of worm strains used in the study Strain name Text name Genotype N2 Wild-type Wild-type VS20 atgl-1::gfp hjIs67 [atgl-1p::atgl-1::GFP + mec- 7::RFP] VC20458 atgl-1(P87S) atgl-1 (gk176565) [P87S] III RB782 asah-2 asah-2(ok564) II RB1487 asm-3 asm-3(ok1744) IV RB1203 cerk-1 T10B11.2(ok1252) I VC20496 cpt-3 Whole-genome sequenced strain (Million Mutation Project strain) VC40798 cpt-4 Whole-genome sequenced strain (Million Mutation Project strain) VC40360 cpt-6 Whole-genome sequenced strain (Million Mutation Project strain) BX14 elo-1 elo-1(wa7) IV RB910 elo-3 elo-3(ok777) IV RB1264 elo-4 elo-4(ok1346) III VC410 elo-5 elo-5(gk208) IV
16 VC425 elo-6 elo-6(gk233) IV RB1046 elo-9 elo-9(ok993) II RB2135 F33D4.4 F33D4.4(ok2843) IV BX24 fat-1 fat-1(wa9) IV BX26 fat-2 fat-2 (wa17) IV BX30 fat-3 fat-3(wa22) IV BX17 fat-4 fat-4(wa14) IV BX107 fat-5 fat-5(tm420) V BX160 fat-7 fat-5 fat-7(wa36) fat-5(tm420) V BX110 fat-6; fat-5 fat-6(tm331) IV; fat-5(tm420) V BX106 fat-6 fat-6(tm331) IV BX156 fat-6; fat-7 fat-6(tm331) IV; fat-7(wa36) V BX153 fat-7 fat-7(wa36) V RB1036 hyl-1 hyl-1(ok976) IV RB1498 hyl-2 hyl-2(ok1766) X VC507 lagr-1 lagr-1(gk263) I VC916 sphk-1 sphk-1(ok1097) II RB1465 sptl-1 C23H3.4(ok1693) II VC2358 sptl-2 sptl-2(ok2753) V RB1579 sptl-3 sptl-3(ok1927) V RB1685 ttm-5 ttm-5(ok2095) I
2.1.1.2 N. parisii spore preps N. parisii spore was prepared using the previously established methods (Troemel et al. 2008) and all incubation steps were done at 20°C. After N. parisii infected animals are harvested, they are grinded with silicon beads then filtered through 5�m Milipore filter attached to 10 mL syringe. The filtered lysate should be devoid of any C. elegans debris. N. parisii spore concentration was quantified by staining with direct yellow 96 (DY96) and counting with a hemocytometer (Cell-Vu). N. parisii spore filtrates were frozen and stored at -80 °C in aliquot of 50�L, 100�L, and 200�L for later infection tests. The aliquots were thawed right before use.
17 2.2 Fasting experiment
N2 wild-type L1 larvae were grown on OP50 plated NGM plates at 20°C for 72 hours and the gravid adults were treated with bleaching solution (1M NaOH, 1.2% Hypochlorite) to leave only the embryos which are then washed and rotated in the M9 buffer (0.3% KH2PO4,
0.6% Na2HPO4, 0.5% NaCl, 1 mM MgSO4) overnight to hatch. Synchronized L1 larvae worms were plated on NGM plate seeded with E. coli OP50 for 3 days to the 1-day young adult stage. These worms were split into two groups: fed and fasted. The worms in fed group were plated on NGM plate seeded with E. coli OP50 and the worms in fasted group were plated on NGM plate without bacterial food. After 6 hours, the worms were washed with M9 buffer and mounted on slides for further analysis.
2.3 Lipid staining 2.3.1 Oil red O staining Oil red O staining was performed as previously described (Soukas et al. 2009) but omitting the freeze-thaw cycles. Synchronized L1 larvae were grown on OP50 seeded NGM plates for 72 h and 200-300 1-day young adult worms were harvested and washed twice with PBS and resuspended in 120 �L of PBS. 120 �L of 2× MRWB buffer (160 mM KCl, 40 mM
NaCl, 14 mM Na2EGTA, 1 mM Spermidine HCl, 0.4 mM Spermine, 30 mM Na PIPES at pH 7.4, 0.2% BME) containing 2% paraformaldehyde was added to the worms containing solution. The worms were washed once again in PBS then resuspended in 60% isopropanol to dehydrate for 15 minutes. After removing isopropanol, 250 �L of 60% oil red O stain was added and the worms were incubated overnight at room temperature. oil red O stain was removed, and stained worms were mounted on glass slides with 20 �L Vectashield containing DAPI. 60% oil red O stain was prepared by freshly diluting stock oil red O solution (0.5 g of oil red O powder/100 mL of 100% isopropanol solution and equilibrated for several days) with 40% water and sit at room temperature for at least 10 minutes. oil red O solution was filtered using 0.25 �m filters right before use.
18 2.4 Microsporidian infection experiment All strains used for infection were synchronized at L1 larvae stage as described above prior to the infection. All infections are conducted at 20 °C using the prepared N. parisii spore filtrate.
2.4.1 Optimizing infection conditions Four different concentrations of N. parisii spore were tested using N2 wildtype worms. The concentrations tested were: 1×106 spore/1000 L1s, 2×106 spore/1000 L1s, 3×106 spore/1000 L1s and 4×106 spore/1000 L1s. N2 wildtype worms were bleach synchronized at L1 larvae stage and infected for 72 h. The L1 larvae were plated on NGM plates seeded with OP50 and N. parisii spores and incubated at 20 °C. The infection plates were prepared as described above on 3.5 cm NGM plate. After either 48 h or 72 h, the worms were harvested, and stained with DY96 stain. About 20 worms per replicate were examined by fluorescence microscopy. The optimized N. parisii concentrations were used for subsequent infection assay using lipid mutants.
2.4.2 24h, 48h, and 72h infection with atgl-1::gfp Three different incubation periods were tested using atgl-1::gfp worms. Synchronized atgl- 1::gfp L1 larvae were divided into three groups where the worms were either infected for 24, 48, or 72 h. The L1 larvae in 24 and 48 h infected groups were first grown on NGM plates seeded with OP50 for 48 and 24 h, respectively. The worms were then washed in M9 and plated on new NGM plates seeded with OP50 and N. parisii spores (1×106 spore/1000 L1s) to be incubated at 20 °C to the 1-day young adult stage. Synchronized L1 larvae in 72 h infected group were plated on NGM plates seeded with OP50 and N. parisii spores (1×106 spore/1000 L1s) for 72 hr to the 1-day young adult stage. As a control for each incubation period, L1 larvae were grown on NGM plates seeded with E. coli OP50 without N. parisii spore and for 24 and 48 h infected group, the uninfected control groups were also washed in M9 and plated onto new NGM plate. 3.5 cm unseeded NGM plates were used for all the infection and prior to the infection, N. parisii spore in 250 �L 10X E. coli OP50 solution is plated onto the plate and swished around to thoroughly cover the entire plate. For the
19 control, only 250 �L 10X E. coli OP50 is seeded onto the plate. After 72 h from the initial plating of synchronized L1 larvae, 20 worms were examined by fluorescence microscopy for GFP intensity.
2.4.3 Infection assay with atgl-1::gfp and atgl-1(P87S) Synchronized atgl-1::gfp and atgl-1(P87S) L1 larvae were grown on NGM plates seeded with E. coli OP50 and N. parisii spores (1×106 spore/1000 L1s) for 48 h to L4 larvae stage. For control, L1 larvae were grown on NGM plate seeded with OP50 alone for 48 h. The plates were prepared as described above. After 48 h, the worms were harvested and washed in M9 +0.1% tween-20 to be fixed and stained with DY96 and FISH for imaging.
2.4.4 Lipid mutants N. parisii infection Initial population of lipid mutants and N2 were grown on NGM plates seeded with E. coli OP50 for 3 days and were bleached synchronized at the first larval stage (L1). Synchronized wildtype, AWR17, and lipid mutant strains L1 larvae were divided into two groups where the worms were either infected for 48 or 72 h with 1×106 spore/1000 L1s and 2×106 spore/1000 L1s, respectively. To infect, L1 larvae were plated on unseeded NGM plates with N. parisii spore in 250 �L 10X E. coli OP50 solution and swished around to thoroughly cover the entire plate. As control, worms were plated on NGM plates with just 250 �L 10X E. coli OP50. After each time point, worms were harvested and fixed for further analysis.
2.4.5 Pulse-infection assay
Synchronized L1s were prepared from bleached population of worms grown on NGM plates seeded with E.coli OP50 for 3 days. In Eppendorf tube, for 400 �L total, 6000 L1s, 15×106 spores, 4 �L 10X OP50 and M9 were added to make the rest of the volume. The solutions were plated on 6 cm unseeded NGM plates and incubated at 21 °C for 3h. After 3h, worms were washed off with 1mL M9 + 0.1% Tween-20 four times. After removing supernatant, 700 �l acetone was added and incubated for 10 minutes. Then worms were FISH stained as described below.
20 2.4.6 DY96 and FISH staining using PFA fix Worms were washed off infection plates using M9 + 0.1% Tween-20 and fixed with 500 �L 4% paraformaldehyde (PFA diluted in 1×PBS/0.1%tween-20) for 30 minutes. Then worms were washed twice with 1 mL PBS + 0.1% Tween-20 and once with 1 mL hyb buffer (900 mM NaCl, 20 mM Tris [pH 8.0], 0.01% SDS in dH2O). Supernatant was then removed, 100 �L hyb buffer containing 5 ng/ul FISH probe (stocks: 1ug/ul, stored at -20 °C) was added and the worms were incubated at about 46 °C overnight. Probe used in the study was MicroB with Cal Fluor Red 610 attached (CTCTCGGCACTCCTTCCTG). MicroB probe was designed against small subunit ribosomal sequence specific to N. parisii and purified by Biosearch Technologies, Inc. After hybridization, samples were washed once with wash buffer (10 � 0.5M EDTA/1 mL hyb buffer) containing the chitin-binding dye direct yellow 96 (DY96) at 20 �g/mL. Worms were incubated in 500 �L wash buffer at 46 °C for 30 – 60 minutes. Afterwards, buffer was removed, and worms were mounted on glass slides using 20 �L Vectashield containing DAPI for imaging.
2.4.7 DY96 staining using acetone fix After 72 h infection, worms were harvested and washed using M9 + 0.1% Tween-20 until supernatant was clear and fixed with 700 �L acetone for 10 minutes. Fixed worms were washed twice with 1 mL PBS + 0.1% Tween-20 then incubated in 500 �L of DY96 staining solution (1X PBS, 0.1% Tween-20, 2.5 mg/mL DY96 (from 5 mg/ml stock in H2O), 1% SDS) for 30 minutes at 20 °C. Afterwards, supernatant was removed and stained worms were mounted on glass slides using 20 �L Vectashield containing DAPI for imaging.
2.4.8 FISH staining using acetone fix FISH staining was performed on worms in 48 h infected group. Worms were washed off plate using M9 + 0.1% Tween-20 and fixed with 1 mL acetone for 15 minutes. After fixation, worm were FISH stained following same FISH protocol and probe (MicroB) as described above for FISH staining using PFA fixed. Fixed worms were then mounted on glass slides with 20 �L Vectashield containing DAPI for analysis.
21 2.5 Fluorescence Microscopy 2.5.1 Live imaging and Quantification atgl-1::gfp animals were paralyzed in 1mM Levamisole for 2 minutes and then transferred onto a freshly prepared 3% agarose pad on glass slide. Imaging was performed on 1-day young adult stage worms using Zeiss AxioImager fluorescence microscope. Using ImageJ software, the level of GFP was quantified for each animal and normalized by worm size (net fluorescence divided by its total area).
2.5.2 Oil red O Quantification Images of oil red O stained worms were taken using a color camera equipped microscope in bright-field. ImageJ software was used to quantify the level of excess red intensity against transparent body. The color image was separated into RGB channel and the green channel was used for quantification as oil red O has been reported to absorb light at 510 nm (Ramírez-Zacarías, Castro-Muñozledo, and Kuri-Harcuch 1992). A minimum of 9 worms were quantified per each sample and the experiments were repeated 5 times. (Escorcia et al. 2018)
2.5.3 Embryo count and pathogen load analysis by DY96 DY96 stains microsporidia spore and embryos in worms and Zeiss AxioImager fluorescence microscope with 5X and 10X objective was used to image stained worms under same exposure times. Images taken at 5X magnification were used for brood count and about 50 worms per sample were quantified for the number of embryos. To measure pathogen load, ImageJ software was used to set to remove background fluorescence intensity from the embryos and only quantify the fluorescence intensity of microsporidia spore. All the images were set to the same threshold value and normalized to the nematode body area.
2.5.4 Pathogen load and sporoplasm analysis by FISH FISH stained worms were imaged using Zeiss AxioImager fluorescence microscope with a 10X objective. Same exposure times was used for all samples within a single experiment. Images were analyzed using ImageJ software where the level of FISH probe intensity was
22 quantified and normalized to the nematode body area to determine the amount of pathogen contained within the worm. For sporoplasm quantification, images from pulse-infection were analyzed using ImageJ where under 40X magnification, the number of sporoplasms present in each nematode’s body were quantified. A minimum of 20 worms were analyzed per sample.
2.5.5 Brood count analysis and Synthetic effects score calculation To analyze the effects of infection on brood size of animals, ‘Synthetic effect’ score (�) was calculated for each mutant (Mani et al. 2008). Synthetic effect score:
mutant infected mutant uninfected control infected − × mutant uninfected control uninfected control uninfected is the deviation of the observed fitness (brood size) of the lipid mutant strains infected with N. parisii from the expected fitness. Negative ɛ values represent an infected strain with lower brood size than expected, and positive ɛ values represent an infected strain with higher brood size than expected. Threshold of 0.12 was applied and mutants that passed ∣ɛ∣ ≥ 0.12 threshold were further analyzed for pathogen load.
2.6 Sphingosine Supplementation Commercial C17 branched-chain sphingosine (branched d17:1) (Hannich et al. 2017) were purchased from Avanti Polar Lipids. For preparation of sphingosine supplementation plates, 0.01 M of stock sphingosine dissolved in dimethyl sulfoxide (DMSO) was diluted to 50 �M to make 3.5 cm plates. The control DMSO (0.5%) plates were prepared by adding 20 �L of DMSO per 4 mL of NGM solution. The infection was performed as described above where L1 larvae were plated on unseeded NGM plates, sphingosine supplementation plates, and DMSO plates with N. parisii spore in 250 �L 10X E. coli OP50 solution. After each time point, worms were harvested and fixed for further analysis using DY96 and FISH staining.
23 2.7 Lipid metabolite quantification Synchronized L1 N2 animals were prepared by sodium hypochlorite treatment of mixed population of worms. Infections were carried out by adding 10,000 worms, 5 million N. parisii (ERTm1) spores, and 1 ml 10X OP50 per plate onto 13 10 CM NGM plates. For uninfected animals, 2.5K animals and 1 ml 10X OP50 were added per plate onto 10 10CM NGM plates. Animals were incubated at 21°C for 72 hours. For adult samples, worms from 1 plate of each condition were washed into 1.5 mL tubes and then washed twice with 1ml M9. To remove embryos and L1s, worms were allowed to settle for 1 minute, supernatant was removed, and 1 mL M9 was added. This was done a total of 4 times. Samples were then frozen in liquid nitrogen.
To prepare embryos, the remining infected and uninfected plates were treated with sodium hypochlorite in 15 mL conical tubes and washed 3X with 10 mL M9 and resuspended in 5 mL M9. To remove carcasses, tubes were left to settle for 4 minutes, supernatant removed, and washed twice with 10mL M9. Embryos were then added to a 10CM NGM plate and incubated for 3 hours at room temperature. Embryos were washed off plates with M9 into 1.5 mL tubes and froze in liquid nitrogen. Experiment was performed in triplicate.
Liquid chromatography separation with mass spectrometry detection (LC–MS) of intact lipid species was performed as described previously (Burton et al. 2018).
Chapter 3 Results
All experiments were performed by JiHae Jeon except for the metabolomic experiment in figure 13 which was performed by Nicholas O. Burton and Aaron W. Reinke
3.1 Introduction
With the advances in biomedical science, incidence of infectious disease was considerably reduced and appeared to be replaced by chronic diseases, yet with the emergence of
24 HIV/AIDS pandemic in 1980s, human population was again confronted by the re-emerging infectious diseases. Along with the increase in HIV/AIDS patients, overall population of immunocompromised patients is increasing due to an epidemiological transition caused by improvements in immunosuppressive treatments and increased life-expectancy (Bloom and Cadarette 2019). Such increase resulted in rising morbidity and mortality rate associated with invasive fungal infections. Infectious disease continues to be the leading global health problem, causing over 13 million deaths every year (Cohen 2000). Of different infectious disease, those caused by pathogenic eukaryotes including fungi are particularly more difficult to treat as they have complex life cycles with several different forms. Invasive fungal infections cause significant health problems in immunocompromised patients as most of these patients are undergoing immunosuppressive therapy and thus are unable to carry out with currently available treatments for fungal infections since it often involves antifungal therapy combined with reducing immunosuppression (Badiee and Hashemizadeh 2014).
Although microsporidia were first discovered in 1857 as infectious agent of silkworms, they were only infrequently identified in humans. However, with HIV/AIDS pandemic emerging around 1980s, they became recognized as an opportunistic pathogen of causing diarrhea and systemic disease in HIV/AIDS patients. Prevalence rates for microsporidiosis in HIV- positive patients range from 2 to 50% depending on geography and diagnostic techniques (Weber, Deplazes, and Schwartz 2000) and from 1.3 to 22% in non-HIV infected people. However, the actual rate may be much higher as microsporidiosis have non-specific or absent clinical signs and they are often ignored in different diagnostic tests (Didier 2005). Microsporidiosis in humans and animals are treated with albendazole and fumagillin but as these drugs are only effective against specific human infecting species with neurotoxic side effects, discovery of novel therapeutic interventions has become an active field of research (Rex and Stevens 2015).
Analysis focused on understanding host-pathogen interactions in regard to metabolism is imperative when studying infections by obligate intracellular parasites such as microsporidia. As an obligate intracellular parasite, these pathogens constantly interact with the host cells and manipulate the host metabolism, changing the level of available
25 metabolites to adjust to their needs. Their severely reduced genome size also reflects their extreme host reliance as they reduce their own metabolic pathways and compete with the host cells to uptake host metabolites (Xu et al. 2010). Thus, studying interplay between host metabolism and microsporidia provides an attractive approach to discover new therapeutic strategies. Host lipid metabolites are utilized by intracellular pathogens and interestingly microsporidia have also reduced their lipid metabolism.
In this study, I propose to use recently established C. elegans-N. parisii model system to study host-pathogen interactions involving lipid metabolism (Troemel 2016). C. elegans has been established as a model organism to study fat metabolism due to many of energy homeostasis genes and pathways shared among C. elegans and mammals including humans. In C. elegans, the majority of fat is stored as lipid droplets (LD) in intestinal epithelial cells, an organelle with a lipid core surrounded by phospholipid membrane (Vrablik et al. 2015). Recent study of LDs showed that host LDs are involved in maintaining energy homeostasis as well as interacting with the intracellular parasites at different steps of their life cycle. Several LD-associated proteins have been characterized including the adipose triglyceride lipase, ATGL (Herker and Ott 2012). C. elegans homologue, ATGL-1, accumulate in LDs during fasting and is required for the life span extension in response to dietary restriction (Zaarur et al. 2019). As N. parisii destructively replicate within the nematode’s intestinal tissue, the host’s ability to absorb nutrients is impaired and the host becomes energetically impoverished (Troemel et al. 2008). Since N. parisii infected C. elegans resemble the nutritionally deprived fasted nematodes, the role of ATGL-1 and LDs during microsporidian infection was investigated.
Moreover, a metabolomics was performed on wildtype C. elegans infected with N. parisii which identified various lipid metabolites that were significantly regulated. The upregulated lipids were in three main categories of lipid biosynthesis pathways: phosphatidylcholine, acylcarnitine and ceramide. These lipids have previously been reported to have involvement with the host during various parasitic infections. PC, a type of phospholipids, have been reported as being an essential mediator of malaria parasite survival during infection. Plasmodium parasites uptake PC from the host cells which allow different parasite proteins to be correctly localized to the parasitophorous vacuole membrane. Upon invasion of human
26 hepatocytes, these parasites envelop themselves and grow within a parasitophorous vacuole membrane, and hence without proper localization of parasite proteins, they are unable to establish infections in the host (Itoe et al. 2014). Another opportunistic pathogen, Pseudomonas aeruginoa, can detect and modulate host-derived sphingosine involved in host immune function which suggests their ability to alter and hijack host immune response for their survival (LaBauve and Wargo 2014). Moreover, acylcarnitine represents abundant energy sources for the intracellular pathogens as it can be metabolized into fatty acids and used in �-oxidation (Meadows, Willsey, and Wargo 2018).
Here, I report the changes in host lipid metabolism in response to microsporidia infection. Using a transgenic strain overexpressing ATGL-1, I identified changes in the ATGL-1 protein level induced upon the re-generation of infectious stage of N. parisii. Further investigation of LDs with the use of molecular lipophilic dyes revealed decrease in lipid stores in response to infection. Altogether, I uncovered potential interplay between the LD hydrolysis and microsporidia infection. As well, I performed N. parisii infection assay with a panel of 31 lipid mutant strains that are defective in the production of different types of metabolites in the identified three lipid metabolic pathways to identify genes that may be involved in the host resistance or parasite proliferation. The infection assay along with lipid supplementation assay suggests potential role of sphingosine in promoting microsporidian growth.
3.2 Understanding changes in host lipid metabolism in response to microsporidia infection
3.2.1 ATGL-1 is upregulated in response to fasting To confirm the role of ATGL-1 on dietary restriction, I used the VS20 strain that overexpresses ATGL-1 protein expression and contains chromosome-integrated transgenic array hjIs67 [atgl-1p::atgl-1::GFP+mec-7p::RFP] which has a green fluorescent protein fused to ATGL-1. In this study, I will be referring to VS20 strain as atgl-1::gfp. A previous study that investigated key genes involved in fasting-induced lipolysis, demonstrated that
27 ATGL-1::GFP colocalized with Nile red stained LDs and hence ATGL-1 localizes at the surface of LDs (J. H. Lee et al. 2014). To analyze the expression of ATGL-1, GFP signal was quantified as a proxy for the level of ATGL-1 expression and associated lipolytic activity. I performed feeding assay as described above and compared changes in the overall gfp intensity of fed and fasted worms. As expected from previous studies, fasting of atgl- 1::gfp for 6 h increased ATGL-1 protein level as demonstrated by an increase in brightness (Figure 9) and higher gfp intensity of fasted worms (Figure 9). The evident increase in the level of ATGL-1::GFP intensity upon fasting suggests a potential role of ATGL-1 lipase in mediating stored LDs hydrolysis in fasted worms to increase energy production.
A
B
100
80
60
40
20 ATGL-1::GFP level ATGL-1::GFP
0 Fed Fasted
28 Figure 9: Quantification of ATGL-1::GFP level in fed and fasted fixed young adult stage atgl-1::gfp worms Strain overexpressing ATGL-1 that contains chromosome- integrated transgenic array hjIs67 [atgl-1p::atgl-1::GFP + mec-7p::RFP] were split into control (Fed) and fasted groups. (A) Representative images of fixed atgl-1::gfp worms under feeding and fasting conditions visualized by fluorescence microscopy at the young adult stage. Images taken at 10X magnification. White lines indicate the boundaries of worm bodies. (B) Quantification of the GFP fluorescence level shown in A in fed and fasted groups (20 randomly selected worms per group). Elevated level of ATGL-1::GFP protein in fasted group compared to the control. Each experiment was repeated once. Error bars represent standard error of the mean.
3.2.2 The level of ATGL-1 increases in a time-dependent manner upon microsporidian infection
Since microsporidia are intracellular pathogens with significantly reduced metabolic potentials, they evolved strong host reliance for metabolites. N. parisii undergoes destructive replication and growth in C. elegans intestinal tissue which causes the worms to become nutritionally deprived similar to when the worms are fasted (Cuomo et al. 2012). Therefore, to investigate whether the fasting induced LD hydrolysis is induced in response to microsporidia infection, I performed N. parisii infection assay with atgl-1::gfp strain. The worms were incubated with N. parisii for different length of time to reflect three phases of microsporidia life cycle. atgl-1::gfp worms were infected for either 24, 48 or 72 h where most pathogens are either at the initial (sporoplasm and meronts), proliferative (meront) or sporogonic (spore) stage respectively. To analyze parasite burden in each worm, the level of ATGL-1::GFP intensity normalized to the body area of worm was quantified for 20 worms in each sample. Initially, worms in infected and control groups were fixed, prior to quantification. Unfortunately, ATGL-1::GFP appeared fairly ubiquitously throughout the whole-worm rather than exhibiting intestinal LDs specific localization. Thus, I was unable to correlate the changes in ATGL-1 level with N. parisii infection.
To minimize the invasiveness of fixative protocol and associated permeability issues, alternatively, I performed the same infection assay, but the worms were imaged live for further quantification. When visualized with fluorescence microscope, the ATGL-1::GFP appeared to localize specifically in the intestine of the worms and absent in other parts
29 (Figure 10). Interestingly, upon infection, the level of ATGL-1::GFP intensity is similar between not infected and infected worms during the initial infective stages when the microsporidia exist either as sporoplasm or meront in the intestinal cells. Upon longer exposure to these pathogens, worms have a slight but insignificant increase in ATGL- 1::GFP at 48 h post infection. Highest and significant increase in level of ATGL-1::GFP was exhibited at 72 h post infection which correlates to the sporogonic stage where microsporidia are re-generating infectious spore. Thus, these data suggest that either the host, C. elegans, or parasite, N. parisii, is utilizing the host stored fat storage via ATGL-1 mediated LD hydrolysis during the microsporidia infective stage.
A
30 n = 3 24 h uninf. ctrl24 h infection48 h uninf. ctrl48 h infection72 h uninf. ctrl72 h infection
B 100
80
60 *
40 p > 0.13 p > 0.05
ATGL-1::GFP level ATGL-1::GFP 20
0 24 h 24 h 48 h 48 h 72 h 72 h uninf. ctrl infection uninf. ctrl infection uninf. ctrl infection
Figure 10: Quantification of ATGL-1::GFP level in atgl-1::gfp at 24, 48, and 72 h post infection. Strain overexpressing ATGL-1 that contains chromosome-integrated transgenic array hjIs67 [atgl-1p::atgl-1::GFP + mec-7p::RFP] has GFP fused protein localized lipid droplets. Here, atgl-1::gfp worms were either not infected or infected with a low dose of N. parisii (1 million spores) for 24, 48 or 72 hours. (A) Representative images of live atgl- 1::gfp worms under different infection duration conditions visualized by fluorescence microscopy at the young adult stage. Lipid droplets are shown in green. Images taken at 10X magnification. White lines indicate the boundaries of worm bodies. (B) Quantification of the GFP fluorescence intensity shown in A normalized to uninfected control group (20 randomly selected worms per group). Elevated level of ATGL-1::GFP protein upon 72 hours infection compared to the initial 24 hours of infection. Each experiment was repeated three times. All error bars represent standard error of the mean over three independent experiments. p values were calculated by unpaired two-tailed t- test. * p < 0.05.
3.2.3 Lipid stores are reduced in response to N. parisii infection, but the role of ATGL-1 in infection is unknown.
ATGL-1 is a type of lipolytic enzyme that is involved in the hydrolysis of lipid reserves in C. elegans. To quantify changes in lipid abundance in response to N. parisii infection. I utilized a common lipid dye, Oil Red O. In addition to ATGL-1 overexpression mutant animals, I used ATGL-1 reduction-of-function mutant animals that were previously described to accumulate significantly more lipids than the wildtype (Zaarur et al. 2019, 1).
31 Partial loss-of-function atgl-1 mutant were obtained from the Million Mutation Project (Thompson et al. 2013) that has a substitution of a proline by a serine in the catalytic domain of ATGL, thus named atgl-1 (P87S). Upon fasting, this mutant did not deplete of its lipid reserves, demonstrating its reduced catalytic activity (Zaarur et al. 2019, 1). Therefore, atgl- 1 (P87S) and atgl-1::gfp were used to study the role of ATGL-1 and LDs in response to N. parisii infection.
To measure the changes in the lipid stores, the worms were infected with the optimized low dose (1×106 spores) of N. parisii and were fixed and stained with Oil Red O dye after 72 h. In uninfected condition, as expected, mutant animals overexpressing ATGL-1 appear to have reduced lipid stores compared to the wildtype animals whereas mutants with reduced- function ATGL-1 have increased lipid stores (Figure 11). Upon infection, I observed a depletion of lipid stores in all animals of wildtype, atgl-1::gfp and atgl-1 (P87S), suggesting that the lipid stores in animals are utilized during infection either by the host or microsporidia.
A
32 B Uninfected Infected
150 ✱✱✱ p > 0.4 ✱✱✱✱
100
50 % ORO intensity (% to control) 0 N2 atgl-1::gfp atgl-1(P87S)
Figure 11: Changes in abundance of fat stores in response to N. parisii infection using Oil Red O lipid dye. (A) Wildtype (N2), atgl-1::gfp, atgl-1 (P87S) worms were either not infected or infected with N. parisii and stained with Oil Red O. (B) Lipid stores of wildtype, atgl-1::gfp, and atgl-1 (P87S) were measured and normalized to wildtype uninfected control by ImageJ. p values were calculated by unpaired two-tailed t- test. Each dots represents each biological replicates. *p < 0.05.
Therefore, to test whether ATGL-1 benefits the host (C. elegans) or the parasite (N. parisii), I performed the infection assay on wildtype and atgl-1 mutants to compare the changes in fitness and parasite burden in response to infection. Initial infection was performed with low dose spores which appeared to over-infect the animals at 72h, rendering it difficult to look at the changes in fitness and parasite burden. Thus, a lower dose of 0.5 million spores were used instead to infect these animals and low dose of 1 million spores were used only for 48 h timepoint. Number of embryos produced by animal was used as a proxy for estimating the fitness of animals. When animals were grown in normal condition, atgl-1 (P87S) produced significantly fewer number of embryos compared to the wildtype animals whereas atgl- 1::gfp produced similar number of embryos. When these animals were infected for 72 h, no significant changes in the fitness were observed in all three strains (Figure 12A). The animals appeared to produce similar number of embryos even in presence of N. parisii. Both DY96 and FISH stain were used to look at the pathogen load at 72 h timepoint, and similar to the fitness data, no significant changes were observed between the three strains upon
33 infection. Both the number of spores and meronts appear to be not significantly changed regardless of whether the animals have fully functional ATGL-1 or not (Figure 12B&C). Similarly, when I measured the pathogen load in worms stained with FISH probes after 48 h, the changes were insignificant (Figure 12D). Although, atgl-1::gfp appear to have enhanced susceptibility due to increased pathogen load than atgl-1 (P87S) because the changes are not significant, it is difficult to conclude whether overexpressing ATGL-1 enhanced susceptibility to N. parisii infection.
Live imaging of atgl-1::gfp showed increased in the level of ATGL-1::GFP at 72 h of infection corresponding to the reduced lipid stores as shown with oil red O staining, suggesting that ATGL-1 may play a role during microsporidia infection via hydrolyzing lipid reserves in C. elegans. However, both atgl-1::gfp and atgl-1 (P87S) do not appear to have enhanced susceptibility or resistance in response to infection when compared to wildtype and thus further study is needed to fully understand the role of ATGL-1 during infection.
A Uninfected Infected ns ns 30
ns 20
10 Embryos/worm
0 N2 atgl-1::gfp atgl-1(P87S)
34 B C ns ns ns ns 250 200 ns ns
200 150
150 100 100 50 50 Pathogen load (% to control) Pathogen load (% to control) 0 0 N2 atgl-1::gfp atgl-1(P87S) N2 atgl-1::gfp atgl-1(P87S)
D ns
250 ns ns
200
150
100
50 Pathogen load (% to control) 0 N2 atgl-1::gfp atgl-1(P87S)
Figure 12: N. parisii infection in wildtype and ATGL-1 mutant animals. Wildtype and ATGL-1 mutant animals (atgl-1::gfp and atgl-1 (P87S)) were either not infected or infected with N. parisii for 48 and 72 h. (A) Number of embryos per animal was quantified in wildtype, atgl-1::gfp, and atgl-1 (P87S) animals stained with DY96 after 72 h. Each dot represents embryo counts of single worm from three biological replicates. (B) and (C) Parasite burden was measured in wildtype, atgl-1::gfp, and atgl-1 (P87S) animals stained with DY96 (B) or FISH probes (C) after 72 h. (D) Parasite burden was measured in wildtype, atgl-1::gfp, and atgl-1 (P87S) animals stained with FISH probes after 48 h. The values for A - D were normalized to wildtype control group. Each dot represents the mean of each biological replicates. p values were calculated by unpaired two-tailed t- test. *p < 0.05.
35 3.3 Characterizing the impact of host lipid metabolism on microsporidia infection 3.3.1 Various lipid metabolites are upregulated upon N. parisii infection Metabolomic screen was performed in collaboration with Dr. Nick Burton from University of Cambridge. Dr. Burton compared the level of lipid metabolites in uninfected and N. parisii-infected 1-day young adult stage C. elegans and embryos from parents’ groups that either were or were not infected with N. parisii.
In total, 1451 different ionized lipids were reliably detected by mass spectrometry. In the sample of embryos, he identified 62 metabolites with more than two-fold change with a significant p value (p < 0.01). He discovered that none of these metabolites overlapped with those that exhibited differential expression in response to osmotic stress (Burton et al. 2018). Interestingly, a previous paper demonstrated that pals genes are highly induced by microsporidia infection and function as transcriptional regulator of intracellular pathogen response (IPR) genes. Of pals gene, pals-22, which is part of a clade in pals gene family is not induced by intracellular infection and acts as a repressor of IPR gene expression. In their study, investigation of interaction between pals-22 and different characterized stress response pathways suggested that IPR genes provide a novel pathway to mediate protection against microsporidia and is distinct from previously characterized stress response pathways (Reddy et al. 2017). Similarly, non-overlapping lipid metabolites between infection and osmotic stress further provides support a distinct role of IRP genes.
Dr. Nick Burton observed a greater number of metabolites that were differently expressed in infected adult C. elegans populations when compared to uninfected populations. 169 out of 1451 metabolites exhibited more than two-fold change with significant p value (p < 0.01).
36 A
37 B
Figure 13: Lipid metabolites identified from LC/MS profiling. (A) and (B) A heat map showing list of lipid metabolites that are upregulated (A) and downregulated (B) by more than two times Log2-fold change with significant p-value (p < 0.01). The lipid metabolites were compared between the infected and not infected adults and embryos from infected and not infected parents. In each table, the first column lists the lipid ions, the second and third column shows the average fold change in adult animals and the embryos respectively. The metabolites are divided into three categories. (A) The first, second and third table represents phosphatidylcholine, acyl-carnitine and ceramide respectively. (B) The first and second table represent phosphatidylcholine and ceramide respectively.
Out of these metabolites, those that belong to phosphatidylcholines category exhibited more than 1000-fold change which are likely the metabolites generated by microsporidia. The
38 upregulated lipid metabolites were in three main categories of lipid biosynthesis pathways which are biosynthesis of phosphatidylcholine, acylcarnitine and ceramide.
3.3.2 A screen of 25 lipid mutants identifies lipid metabolites involved in resistance or hypersusceptibility to microsporidian infection To test the function of the lipids involved in biosynthesis of phosphatidylcholine, acylcarnitine, and ceramide, Dr. Nick Burton identified a panel of 31 C. elegans lipid mutant strains that are defective in the production of different types of lipids (Table 2) which have previously been reported to be involved in the host-pathogen interaction. I hypothesized that some of the lipids in the three lipid biosynthesis categories are either regulated by the host C. elegans and/or co-opted by the parasite, N. parisii, to become resistant or susceptible to the microsporidia infection respectively. Therefore, in order to determine which lipid mutants exhibit enhanced resistance or susceptibility upon infection with microsporidia, I performed a microsporidia infection assay with N. parisii.
Table 2: A list of lipid mutants in each of three categories: phosphatidylcholine, acyl- carnitine, and ceramide. Lipid mutants are listed under each of the lipid categories in their gene name. The mutants labelled in red were not included in the experiments.
Phosphatidylcholine Acyl-Carnitine Ceramide fat-1 cpt-3 asah-2 fat-2 cpt-4 asm-3 fat-3 cpt-6 cerk-1 fat-4 F33D4.4 fat-5 hyl-1 fat-6 hyl-2 fat-6;fat-7 lagr-1 fat-7 sphk-1 fat-5 fat-7 sptl-1 fat-5;fat-6 sptl-2
39 elo-1 sptl-3 elo-3 ttm-5 elo-4 elo-5 elo-6 elo-9
Prior to infecting lipid mutants, I optimized concentration of N. parisii to use for the experiments. Appropriate concentrations of N. parisii spores and infection duration were determined so as to not over-infect the worms that it becomes difficult to observe the resistance phenotype as well as to accurately illustrate how the microsporidia growth changes in increasing degrees of infection. Two different infection concentrations: 1 million spores for low dose and 2 million spores for high dose, were determined. To compare resistance phenotype, AWR17 strain was used which is known to resist N. parisii infection as it prevents the pathogens from initial invasion of the host.
I screened 25 out of 31 lipid mutant strains for infection assay and the remaining 6 strains were eliminated due to slower growth and development and difficulty in culturing (Table 2). To determine the impact of infection on the animal’s fitness, I quantified the number of embryos in DY96 stained animals (Figure 14A, 15A). Because many of the lipid mutants may have reduced fitness on their own, I calculated ‘Synthetic effects’ scores (ɛ) to determine whether the changes in fitness of mutants upon N. parisii infection were additive or synergistic (Table 3). Conservative threshold of 0.12 was used and I identified 12 mutant strains which passed the threshold: fat-1, fat-5, elo-1, elo-3, elo-4, elo-9, cpt-3, cpt-4, hyl-1, lagr-1, asah-2, and sphk-1. From these, 10 strains were assigned a negative ɛ score (fat-1, fat-5, elo-1, elo-3, elo-4, elo-9, cpt-3, hyl-1, lagr-1, sphk-1), indicating that these mutants produced fewer embryos than expected upon infection. Remaining 2 strains were given a positive ɛ score (cpt-4, and asah-2), suggesting that higher number of embryos were scored for these mutants upon infection compared to the expected counts. To further characterize these mutants, I analyzed the parasite burden from DY96 stained spore (Figure 14A, 15B) and FISH stained meronts (Figure 14B, 15C).
40 Table 3: Lipid mutants and corresponding ‘synthetic effects’ scores. ‘Synthetic effects’ scores were calculated as described in the methods and those that pass the threshold of ∣ℇ∣ ≥ 0.12 are labelled in red. p values were calculated by unpaired two-tailed t- test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
Gene Product Strain Text name ℇ Fatty acid elongase BX14 elo-1 -0.2349 RB910 elo-3 -0.3864 RB1264 elo-4 -0.1654 VC425 elo-6 -0.0915 RB1046 elo-9 -0.2268 Fatty acid desaturase BX24 fat-1 -0.1225 BX30 fat-3 -0.0271 BX17 fat-4 -0.0456 BX107 fat-5 -0.3531 BX160 fat-7; fat-5 -0.0301 BX110 fat-6; fat-5 0.1045 BX106 fat-6 0.0226 Carnitine palmitoyl transferase VC20496 cpt-3 -0.1879 VC40798 cpt-4 0.5045 Acylsphingosine amidohydrolase RB782 asah-2 0.1342 (ceramidase) Acid sphingomyelinase RB1487 asm-3 0.0188 Ceramide kinase RB1203 cerk-1 0.0165 Sphingolipid Δ4 desaturase RB2135 F33D4.4 -0.0451 RB1685 ttm-5 -0.0955 Ceramide synthase RB1036 hyl-1 -0.1715 RB1498 hyl-2 -0.0681 VC507 lagr-1 -0.1299 Sphingosine kinase VC916 sphk-1 -0.1343 Serine palmitoyltransferase RB1465 sptl-1 -0.0728 RB1579 sptl-3 -0.0751
41 I first examined the parasite burden of these mutants using DY96 (Figure 15B) and I identified two stains, fat-7; fat-5 and cpt-4, to exhibit drastic and significant decrease in parasite load which suggest that these mutants may have enhanced resistance to N. parisii infection. Contrastingly, elo-1 mutant animals demonstrated slight yet significant increased parasite burden, implying that they may be more susceptible to N. parisii compared to wild- type. Unfortunately, examining spore stage of N. parisii in DY96 stained animals proved to be challenging because as shown in Figure 15B, the quantified parasite burdens of remaining mutants were highly variable between the three biological replicates, rendering it difficult to predict the effects of corresponding mutations on the infectivity of the pathogen. One reason behind the inconsistent results could be due to the chitin-binding property of DY96 stain. DY96 binds to the chitin that is present in both the eggshell of C. elegans’ embryo and spore wall of N. parisii and thus when quantifying parasite burden of DY96 stained worms, it is challenging to separate the unwanted fluorescence signal of the embryos from the signal of N. parisii spore.
An alternative approach was used to examine the pathogen burden using FISH probe, particularly at the meront stage 48 h post infection. Initially, both DY96 and FISH staining were performed at 72 h time point. However, FISH protocol involved incubation of fixed worms at 46C which ruptured the 1-day adult stage worms, particularly the mutants, as lipid-deficient mutants were more sensitive to rupture upon heat. Therefore, only DY96 was used for 72 h old worms and FISH probe was used on 48 h time point worms. When quantifying pathogen load using FISH probe (Figure 15C), wild-type worms infected at a higher concentration of N. parisii (4 million spores) was used as a positive control for comparison of increased susceptibility. From FISH-stained worms, several mutants such fat- 1, fat-5, and cpt-3 were identified to have increased fluorescence intensity corresponding to higher pathogen load. Furthermore, asah-2 mutant animals showed significantly lower fluorescence intensity compared to the wild-type animals. All these mutant animals showed significant changes in the fluorescence intensity and passed the threshold of 0.12 for their synthetic effect scores.
42 A
B
43 Figure 14: Quantification of embryos and pathogen load. (A) A chitin-binding dye, Direct Yellow 96, was used to label chitin-containing C. elegans embryos and spore coat of N. parisii spores. The image shows the wildtype animals fixed and stained 72 hours post infection. The solid lines point to the embryos and the dotted lines point to the spores. Bar ⎯ 145 �M. B) Fluorescent in situ hybridization probe used to stain parasite rRNA to analyze N. parisii load. The image shows the wildtype animals fixed and stained 48 hours post infection. The solid lines point to the meronts. Bar ⎯ 107 �M.
To determine which genes of lipid biosynthesis pathways are involved in either the resistance or susceptibility of C. elegans against N. parisii infection, I examined the impact of infection on both the brood size and pathogen burden of the lipid mutants. Out of the potential candidate mutants, two strains: hyl-2 and asah-2 appeared to be phenotypically similar to wildtype when not infected but upon infection, asah-2 demonstrated resistance whereas hyl-2 exhibited increased susceptibility. Interestingly, two mutants are defective for enzymes involved in sphingolipid biosynthesis pathway. ASAH-2 is a ceramidase which breaks down ceramide to produce sphingosine and HYL-2 is a ceramide synthase that synthesizes ceramide from sphingosine. Therefore, this analysis revealed that sphingosine is the common metabolite between these two mutants. The opposite phenotype of asah-2 to and hyl-2 suggests that sphingosine may be promoting microsporidia growth and when the animals are not able to produce sphingosine, animals have less pathogen and confer resistance to infection. Given that the differences in the level of sphingosine conferred strongest phenotype, I further focused on exploring the role of sphingosine in microsporidia infection.
A Uninfected Infected
200
150
100
50 Embryos / worm (% to control) 0
N2 fat-1 fat-3 fat-4 fat-5 fat-6 elo-1 elo-3 elo-4 elo-6 elo-9 cpt-3 cpt-4 hyl-1 hyl-2 sptl-1 sptl-3 ttm-5 asah-2 asm-3 cerk-1 lagr-1 sphk-1 AWR17 F33D4.4 fat-6;fat-5 fat-7;fat-5
44
B ✱✱ ✱ ✱✱ ✱ ✱ ✱✱ ✱✱ 250 ✱ ✱ 200 ✱✱✱✱
150
100
50 Pathogen load (% to control) 0 N2 fat-1 fat-3 fat-4 fat-5 fat-6 elo-1 elo-3 elo-4 elo-6 elo-9 cpt-3 cpt-4 hyl-1 hyl-2 sptl-1 sptl-3 ttm-5 asah-2 asm-3 cerk-1 lagr-1 sphk-1 AWR17 F33D4.4 fat-6;fat-5fat-7;fat-5
C
✱ ✱ ✱✱ ✱ ✱✱ ✱ 1000 ✱ ✱✱✱✱ 800 ✱✱✱✱
600
400
200 Pathogen load (% to control) 0 N2 fat-1 fat-4 fat-5 fat-6 elo-1 elo-3 elo-4 elo-9 cpt-3 cpt-4 hyl-1 hyl-2 sptl-1 ttm-5 N2-4M asah-2 lagr-1 sphk-1 AWR17 F33D4.4 fat-6;fat-5fat-7 fat-5
Figure 15: N. parisii infection in wildtype and 25 lipid mutant animals. A panel of C. elegans lipid mutants and wildtype control was either not infected or infected with a high or low dose of N. parisii for 72 h or 48 h respectively. Each color box represents three different lipid biosynthesis pathways: blue (phosphatidylcholine), orange (acylcarnitine), and purple (ceramide). (A) Quantification of embryos per animal in wildtype, AWR17 and mutant strains under uninfected and high dose infected conditions at young adult stage (72 h). The brood size was normalized to the average brood size of the control. ‘Synthetic effect’ score was calculated for each mutants and mutants that passed ∣ℇ∣ ≥ 0.12 have been identified as potential candidate strains and further analyzed for parasite burden (B-C). (B) Quantification of the parasite burden during infective phase of microsporidia (spore stage) in wildtype, AWR17 and mutant strains that are infected for 72 h and stained with DY96. Spore counts were normalized to the average spore counts of the control. (C) Quantification of the parasite burden during proliferative phase of microsporidia (meronts stage) in wildtype, AWR17 and mutant strains infected for 48 h and stained with FISH probes. Meronts counts were normalized to
45 the average meronts counts of the control. Wildtype infected at 4 million is used as positive control for mutants with increased parasite burden. Each dot represents the mean of each biological replicates. p values were calculated by unpaired two-tailed t- test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
3.3.3 Sphingosine supplementation increases susceptibility of C. elegans to N. parisii infection
From the initial lipid mutant’s infection assay, sphingosine was determined to potentially play a role during microsporidian infection. To further analyze how the presence of sphingosine affects the susceptibility or resistance of the animals, I performed sphingosine supplementation assay where the wildtype and asah-2 mutant animals were plated on sphingosine supplemented plates during infection. Since asah-2 mutant animals are defective in producing sphingosine, feeding the animals with sphingosine can help to determine whether the supplementation can rescue the susceptible phenotype of the mutants.
A Uninfected Infected
ns ✱ ns ns 30
20
10 Embryos / worm
0 +DMSO +SPH +DMSO +SPH
N2 asah-2
46 B ✱ ✱✱✱
250 ns ns
200
150
100
50 Pathogen load (to control) 0 +DMSO +SPH +DMSO +SPH
N2 asah-2
C ✱ ✱✱ ✱ 500 ns ✱✱ ns ✱
400
300
200
100 Pathogen load (to control) 0 +DMSO +SPH +DMSO +SPH
N2 asah-2
Figure 16: Sphingosine supplementation reduces host fitness in response to N. parisii infection. N. parisii infection assay of wildtype and asah-2 animals grown on control plate (NGM and DMSO) or sphingosine supplemented plate. The animals were either not infected or infected with a high or low dose of N. parisii for 72 hours or 48 hours respectively. (A) Quantification of embryos per animal in wildtype and asah-2 animals stained with DY96. Sphingosine supplementation causes greater reduction in the fitness of animals in response to infection. Each dot represents embryo count of single worm of three biological replicates. (B) and (C) Quantification of the parasite burden in wildtype and asah-2 animals stained with DY96 or FISH after 72 h (B) and 48 h (C), respectively. Animals grown on sphingosine supplemented plate show enhanced susceptibility to N. parisii infection. Each dot represents the mean of each biological replicates. A-C: The values are normalized to the
47 wildtype control group. All error bars represent standard error of the mean over three repeated independent experiments. p values were calculated by unpaired two-tailed t- test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p< 0.0001.
Previous paper showed that sphingoid bases that are produced by C. elegans are structurally different from those of other animals because worms are able to produce both iso- and anteiso-branched species. Further analysis showed that while organisms like S. cerevisiae predominantly produce sphinganine (d18:0), mammals and worms have the ability to desaturate the sphinganines. The most predominantly found species in C. elegans is C17 branched-chain sphingosine (branched d17:1) (Hannich et al. 2017). Branched-chain sphingosine is not commercially available, so d17:1 sphingosine was used to prepare the sphingosine supplemented plates for the infection.
The wildtype and asah-2 mutant animals were grown and infected on NGM, DMSO, and sphingosine supplemented plates for 48 h and 72 h to analyze the changes in the fitness and parasite burden of the animals in response to infection. As shown in Figure 16A, both the wildtype and asah-2 mutant animals show greater significant reduction in the number of embryos when they were grown on sphingosine supplemented plates. Similarly, the parasite burden quantification from both the DY96-stained (Figure 16B) and FISH-stained (Figure 16C) animals, demonstrate a significant increase in the amount of parasite present in the animal’s body when they are supplemented with sphingosine. asah-2 mutant animals previously showed that they are more resistance to infection than the wildtype and so the data from sphingosine supplementation assay suggests that feeding sphingosine to the mutant animals can reverse the enhanced resistance phenotype. As well, the wildtype animals with properly functioning sphingosine producing enzymes also show significantly increased susceptibility, suggesting that the presence of extra sphingosine can also change the phenotypes of wildtype animals upon infection. However, further study would be necessary to fully understand how the sphingosine is promoting microsporidian growth.
3.3.4 Sphingolipid mutants do not exhibit invasion defects Animals can become resistant to parasitic infection because they are able to clear the infection once they are infected, but the animals could also prevent the initial invasion of the
48 pathogen from entering the host cells that is required to propagate the infection. To analyze whether sphingosine impacts the initial invasion or the proliferation of the parasites, pulse- infection assay was performed on the wild-type, asah-2 and hyl-2 mutant animals. hyl-2 animals have a defective ceramide synthase that synthesize ceramide from sphingosine and so contribute to the mutant animals with higher amount of sphingosine.
Pulse-infection assay involves pulse-inoculating the animals for a short period of time (e.g. 1-3 h) with very high dose and after fixation, the number of invasion events are quantified. I infected the animals for 3 h and the number of sporoplasms in worm’s body were used as the number of invasion events. Upon quantification, I observed that the mutant and wildtype animals showed similar number of sporoplasms (Figure 17) with no significant differences, demonstrating that the resistance or susceptibility phenotypes observed in these mutants are not likely to be the result of invasion defect.
ns
ns ns 10
8
6
4 # of Sporoplasm 2
0 N2 asah-2 hyl-2
Figure 17: Sphingolipid mutant animals do not exhibit invasion defect upon N. parisii infection. Wildtype, asah-2 and hyl-2 animals were infected with N. parisii spores for 3h and immediately fixed. The number of invasion events (counted as number of sporoplasms) were quantified. Data are from three independent experiments and 20 animals were measured per experiment. p values were calculated by unpaired two-tailed t- test.
49 Chapter 4 Discussion and Conclusions 4.1 Discussion 4.1.1 Summary of main findings
C. elegans with its natural microsporidian pathogen N. parisii (Troemel 2016) provides a genetically tractable model system to study the impact of host lipid metabolism on microsporidia infection. Although data from infection of ATGL-1 mutant animals were inconclusive, data from atgl-1::gfp infection and oil red O staining experiments suggest that throughout the infection, ATGL-1 protein level increases and the lipid reserves are depleted.
The impact of the host lipid metabolism during parasitic infections have been studied with several pathogens where lipids represent a rich resource for energy and important component of various cellular signaling pathways. However, interplay between host lipid metabolism and microsporidia is not widely studied and how the host lipid metabolism is perturbed during microsporidia infection in C. elegans is currently unknown. In this study, I proposed to determine lipid metabolites that are essential either for pathogen survival and replication or for host’s survival. Infection of lipid mutants demonstrated a possible role of sphingosine in promoting microsporidian growth which was further supported by sphingosine supplementation experiments. Additionally, the pulse-infection experiments of sphingosine mutant animals showed that the parasite is able to invade the host cells initially, but the level of sphingosine may impact the proliferation of the parasite, resulting in the enhanced resistance phenotypes of mutants with reduced sphingosine level.
4.1.2 Changes in LDs and LD associated protein, ATGL-1, in response to N. parisii infection
This work defines changes in host lipid metabolism that occur in response to microsporidia infection. Lipid droplets (LD) have been shown to be the target of various intracellular pathogens upon infection. To determine how LDs are utilized upon microsporidian
50 infection, a LD localized protein, ATGL-1 (lipase) that catabolizes intestinal LDs, was investigated. Previous studies suggested that ATGL-1 accumulate in LDs of fasted animals and since N. parisii infected C. elegans resemble the nutritionally deprived fasted nematodes, among several LD-associated proteins, ATGL-1 was examined in this study. To validate whether ATGL-1 is upregulated upon fasting, atgl-1::gfp worms were starved for 6 hours and ATGL-1::GFP level was quantified which showed that the fasted worms appeared brighter compared to the fed worms (Figure 9) corresponding to higher quantified fluorescence intensity as shown in previous studies. To analyze ATGL-1 expression upon infection, ATGL-1 overexpressing GFP transgenic strain and partial loss-of-function mutant animals were infected and assessed throughout parasite’s lifecycles. Initially, atgl-1::gfp strains were infected with three different incubation periods where each period corresponded to a specific life stage of the parasite. Intriguingly, ATGL-1 expression appeared to be initially induced around the proliferative phase (48 h) of microsporidia and was highest near mature spore formation (72 h), suggesting ATGL-1 may be mediating fat catabolism during infection. Additionally, with oil red O staining of atgl-1::gfp animals after 72 h infection, I observed a decrease in the lipid reserves corresponding to increase in level of ATGL-1 but the changes had insignificant p-value. The experiment should be repeated to confirm whether there is a significant lipid reserves reduction upon infection. The similar phenotypes were observed for the wildtype and atgl-1 (P87S) animals with partial loss-of function mutation of ATGL-1 where both animals showed depletion of lipid reserves in response to infection, but the reduction was only significant for atgl-1 (P87S). The data shows that the reduction of lipid stores is a common response also shared by wildtype animals in addition to ATGL-1 overexpressing animals, suggesting that during microsporidian infection, there is a change in host’s lipid metabolism.
Results from atgl-1::gfp infection assay suggest that ATGL-1 may be involved in the interplay between C. elegans lipid metabolism and N. parisii infection, but whether the lipase is utilized by the host or parasite is inconclusive. The lipase may be used by the host to catabolize lipid reserves in LDs for energy to defend against infection, but the parasite may also exploit host’s LDs for the same energy to grow and replicate. Therefore, I performed infection assay of atgl-1::gfp and atgl-1(P87S) but unfortunately, no significant
51 differences were observed in the fitness or susceptibility of mutants in response to infection. This could be due to performing infection assay with insufficient number of spores since the data from fitness measurement showed that the animals, including the wildtype control, produced similar number of embryos in both not infected and infected conditions, suggesting that concentration of spores used were insufficient to induce changes in the fitness. To study whether ATGL-1 protein enhances susceptibility or resistance in response to microsporidia infection, one could repeat the infection assay but optimize the spore concentration to ensure sufficient concentration was used to effectively observe the changes in the fitness and parasite burden.
Results from both atgl-1::gfp infection assay and oil red O staining suggest that LD- associated lipase, ATGL-1, is hydrolyzing lipid stores at the spore-formation stage of the parasite resulting in decreased lipid reserves. However, further study is needed to determine whether ATGL-1 is involved in providing enhanced resistance or susceptibility against microsporidia pathogens possibly through lipid catabolism.
4.1.3 Sphingosine as the potential factor in promoting microsporidian growth
This work describes an effort to characterize the impact of host lipid metabolism on microsporidia infection. By screening a library of 31 different C. elegans lipid mutant strains defective for proteins involved in biosynthesis of phosphatidylcholine, acylcarnitine and sphingolipids that collectively cover 3 major classes of lipid metabolites that were upregulated upon microsporidia infected C. elegans, 12 candidate mutants were identified. I established an infection assay where mutants are infected with N. parisii and microsporidia growth is measured using probes specific for different developmental stages of the pathogen. The infection experiment was performed at different microsporidia concentrations throughout the lifecycle of N. parisii to observe how the microsporidia growth and host response change over time and at increasing degrees of infection. Any change in the host fitness was assessed by measuring the number of embryos and the infection load in the animal. Because lipid mutant animals often present defects in development, it complicates
52 the interpretation of the results whether changes in susceptibility is due to synergistic interaction of mutation and infection or due to animals with initial reduced fitness. Therefore, ‘synthetic effects’ scores were calculated for fitness values that measures the deviation of observed lipid mutant’s phenotype from the expected phenotypes. I used the threshold of 0.12 and a total of 12 strains were selected. Intriguingly, mutant animals that are part of the sphingolipid pathway, specifically around the production of sphingosine, showed an interesting opposing phenotype where animals that were defective in producing sphingosine (asah-2) showed enhanced resistance and animals that were defective in enzymes that depletes sphingosine (hyl-2) showed enhanced susceptibility. In addition to the opposing phenotype of asah-2 and hyl-2, sphingolipid desaturase mutants, F33D4.4 and ttm- 5, show that less spores are being formed when pathogen load was analyzed. Defective sphingolipid desaturase in animals lead to reduction in the sphingosine synthesis. The enhanced resistance of these animals could be due to the overall reduction in the sphingosine production similar to asah-2 mutant animals.
Apart from sphingolipid mutants, few non-sphingolipid mutants also demonstrated either enhanced resistance or susceptibility in response to N. parisii infection which includes elo-3 and fat-7;fat-5 involved in synthesis of phosphatidylcholine and cpt-3 involved in acylcarnitine shuttle system. ELO-3 is a fatty acid elongase and FAT-7 and FAT-5 are fatty acid desaturases. When infected, elo-3 mutant animals showed increased pathogen load and drastic reduction in the number of embryos produced whereas fat-7;fat-5 mutant animals showed significant reduction in the pathogen load. Previous studies showed that elo-3 is required for the expression of acyl-CoA dehydrogenases, acdh-1, (Watson et al. 2013) which is regulated by mdt-15. MDT-15 is involved in innate immunity and functions to induce resistance to P. aeruginosa infection (Pukkila-Worley et al. 2014). Moreover, fat-7 has also previously shown to affect pathogen susceptibility through the oleate production in which oleate was shown to be required for host resistance to P. aeruginosa infection (Anderson et al. 2019). Although the interaction between fatty acids metabolism and other parasitic infections have been previously demonstrated, further study is needed to understand a link between fatty acids metabolism and microsporidia infection. In addition, CPT-3 is a shuttle protein that transports long-chain fatty acids across the mitochondrial
53 membrane for �-oxidation (Watts and Ristow 2017). cpt-3 mutant animals showed increased susceptibility in response to N. parisii infection. Fatty acid �-oxidation is one of the major processes that C. elegans use to provide energy and so the increased susceptibility and reduced fitness of these animals could be due to the defective enzyme that is involved in energy production via fatty acids metabolism. Nonetheless, an involvement of carnitine transferase protein in microsporidian infection requires more detailed study.
The initial screen identified sphingosine to be a crucial limiting factor for microsporidia to compensate for genome reduction. Restricting the production of sphingosine limit the pathogen growth, conferring increased resistance to the animals. Although sphingosine is involved in production of bioactive molecules that regulate the immune system upon infection by the fungal pathogen Cryptococcus neoformans (Rollin-Pinheiro et al. 2016), the function of this lipid during microsporidian infection is unknown. To confirm that sphingosine supports microsporidia growth, I performed a sphingosine feeding assay where the animals were infected on sphingosine supplemented plates. If sphingosine is promoting the parasite growth, I expect that feeding sphingosine to asah-2 mutant animals should benefit the microsporidia and make animals more susceptible, reversing the enhanced resistance phenotype. The results from sphingosine supplementation experiments showed that both the wildtype and asah-2 mutant animals became more susceptible to infection when they were fed sphingosine, as expected. The animals produced significantly a smaller number of progeny and had higher pathogen load when grown on sphingosine supplemented plates. Interestingly, the presence of extra sphingosine also resulted in lower fitness of the wildtype animals, further suggesting the potential use of sphingosine by the parasite for their benefit
Nonetheless, the possibility that increased resistance of sphingosine mutants is due to the invasion defects preventing the parasite from initial invasion could not be excluded. Thus, I performed pulse-infection experiments to compare the number of invasion events. I observed no significant differences in the number of sporoplasms found in wildtype, asah-2 and hyl-2 animals. The results from pulse-infection and asah-2 48h infection experiments
54 suggest that N. parisii is able to initially infect the intestinal cells in asah-2 animals to the same extent as in the wildtype animals,
Additionally, from the metabolomic data that compared the lipid metabolites abundance between infected and not infected adult worms, I observed that in the infected animals, ceramide species containing 20 carbon appeared to be highly expressed with average fold- change of greater than 80. Previous study showed that although the long-chain bases can vary in their chemical structure, in nematodes, C. elegans, only d17:iso-sphinganine is produced from the condensation of serine and branched-chain fatty acid C15:iso. d17:iso- sphinganine is further modified to produce d17:0 iso-ceramide (Dall and Færgeman 2019; Hannich et al. 2017). On the contrary, studies of yeast lipidomics have identified the ability of Saccharomyces cerevisiae in producing sphingolipids containing 20 carbons where C20- ceramide is one of the most abundant dihydroceramide species found in yeast (Montefusco, Matmati, and Hannun 2014). Since C. elegans is unable to synthesize C20-sphingolipids, the presence of C20-ceramide in infected adult animals could have been synthesized by the microsporidia as it is a fungal organism similar to S. cerevisiae. However, due to the severely reduced genomes of microsporidian species, N. parisii has only retained partial sphingolipid biosynthesis pathway and does not encode a dihydroceramide desaturase or a ceramidase (Desjardins et al. 2015). Interestingly, many C17- ceramide species are down regulated upon infection. Although the metabolomic data does not distinguish the lipids made by C. elegans from those potentially made by microsporidia during infection, looking at the relative amount of the various lipid metabolites can help to determine the origin of the lipids. For instance, substantial increase in the level of metabolite that is poorly abundant might suggest that it is made by the microsporidia. Nonetheless, it can be suggested that microsporidia may be hijacking the C. elegans enzymes to make C20-ceramides. Alternatively, different enzymes may be present in microsporidia to generate C20- ceramides.
Results from metabolomics and infection of sphingosine mutant animals indicate an important role of sphingolipids during microsporidian infection where presence of sphingosine appear to promote microsporidian growth and ceramide species that are only
55 found in fungal species are highly abundant in the infected animals. To confirm whether the C20-ceramide species are produced by N. parisii, one can add labelled exogenous lipid to C. elegans to see whether the parasite can take up the sphingolipid from the host and perform microsporidia-specific modifications.
4.2 Future Directions 4.2.1 Investigate changes in LDs size and structure during infection
C. elegans depletes its lipid reserves upon microsporidia infection as shown with increased ATGL-1 and reduced oil red O intensity in 72 h infected animals. Whether the presence of additional ATGL-1 protein and increased lipids hydrolysis is beneficial to the host or the parasite needs to be further studied. However, the results from the infection experiment in animals with ATGL-1 overexpression and reduction of function mutation were inconclusive. One could repeat the assay with new optimized spore concentration that significantly lower fitness of wildtype animals and study how ATGL-1 mutant animals behave in response to same condition. In addition, to studying the overall level of lipid stores, one can study a role of LDs in host-pathogen interactions. Different lipid dyes, like Nile Red, can be used to stain and quantify size of LDs. For instance, during viral infection, such as hepatitis C, the virus interacts with host’s LDs leading to increase in both the number and size of LDs (J. Zhang, Lan, and Sanyal 2017). Similar infection assay can be performed where the infected animals will be stained with Nile Red to determine the changes in number and size of LDs and also whether the microsporidia appear to co-localize with LDs in the intestine.
4.2.2 Explore interaction between host’s sphingosine lipid metabolism pathways and N. parisii
With the metabolomics data showing upregulation of lipids in the infected animals, I hypothesized that host’s lipids may be exploited by the parasite or used by the host. Of different lipid metabolites present in C. elegans, sphingolipid pathways demonstrate a potential role in response to infection: mutant animals defective in sphingosine production
56 appear to have enhanced resistance which phenotype can be reversed by supplementation of d17:1 sphingosine. In addition to the sphingosine supplementation experiments, the role of sphingosine can be further investigated with the use of RNAi. Since completely abolishing the ability of C. elegans in synthesizing sphingosine can be lethal, RNAi can be used to silence the gene expression of enzymes involved in sphingosine biosynthesis such as asah-2 and F33D4.4. Both asah-2 and F33D4.4 have paralogs asah-1 and ttm-5, respectively and so RNAi can be used to silence the expression of its paralogs in these mutants. RNAi can inhibit the synthesis of sphingosine in these mutants and so infecting these animals can allow one to investigate whether the presence of sphingosine is completely essential for microsporidia growth.
Similar to sphingosine supplementation, a complementation experiment can be performed to see whether adding back the wildtype copy of the defective genes can reverse the phenotypes in sphingosine mutants. For instance, I predict that injecting wildtype copy of asah-2 gene into asah-2 mutant animals can reverse the enhanced resistance phenotype, thus making the animals more susceptible to infection, similar to feeding them with sphingosine. In addition to determining the necessity of sphingosine in infection, the localization of enzymes involved in sphingolipid pathways can be determined with the use of fosmid reporters carrying GFP. Enzymes like acid ceramidase, ASAH-2, and ceramide desaturase, F33D4.4, are responsible for producing sphingosines from either cleaving or desaturating ceramide, respectively. One can analyze the GFP localization of these proteins to examine their localization during infection. I predict that the proteins may be localizing near the parasites in order for the parasites to exploit the host’s sphingosine or even hijack the host’s sphingolipid synthesis pathways during invasion and replication within host cells. All these experiments can further help us to elucidate how microsporidia is interacting with sphingosine during the infection.
57 References
Akiyoshi, Donna E. et al. 2009. “Genomic Survey of the Non-Cultivatable Opportunistic Human Pathogen, Enterocytozoon Bieneusi.” PLOS Pathogens 5(1): e1000261.
Anderson, Sarah M. et al. 2019. “The Fatty Acid Oleate Is Required for Innate Immune Activation and Pathogen Defense in Caenorhabditis Elegans.” PLOS Pathogens 15(6): e1007893.
Ashrafi, Kaveh. 2007. “Obesity and the Regulation of Fat Metabolism.” WormBook : the online review of C. elegans biology 9: 1–20.
AU - Escorcia, Wilber, Dana L. AU - Ruter, James AU - Nhan, and Sean P. AU - Curran. 2018. “Quantification of Lipid Abundance and Evaluation of Lipid Distribution in Caenorhabditis Elegans by Nile Red and Oil Red O Staining.” JoVE (133): e57352.
Badiee, Parisa, and Zahra Hashemizadeh. 2014. “Opportunistic Invasive Fungal Infections: Diagnosis & Clinical Management.” The Indian journal of medical research 139(2): 195–204.
Bakowski, Malina A. et al. 2014. “Ubiquitin-Mediated Response to Microsporidia and Virus Infection in C. Elegans.” PLoS Pathogens 10(6): e1004200.
Bjornson, Susan, and David Oi. 2014. 1 Microsporidia Biological Control Agents and Pathogens of Beneficial Insects. 1st ed. https://doi.org/10.1002/9781118395264.ch25.
Bloom, David E, and Daniel Cadarette. 2019. “Infectious Disease Threats in the Twenty- First Century: Strengthening the Global Response.” Frontiers in immunology 10: 549–549.
Brendza, Katherine M. et al. 2007. “Phosphoethanolamine N-Methyltransferase (PMT-1) Catalyses the First Reaction of a New Pathway for Phosphocholine Biosynthesis in Caenorhabditis Elegans.” Biochemical Journal 404(3): 439–48.
Brenner, S. 1974. “The Genetics of Caenorhabditis Elegans.” Genetics 77(1): 71–94.
Burton, Nicholas et al. 2018. “Neurohormonal Signaling via a Sulfotransferase Antagonizes Insulin-like Signaling to Regulate a Caenorhabditis Elegans Stress Response.” Nature Communications 9.
Cali, Ann, James Becnel, and Peter Takvorian. 2017. “Microsporidia.” In Handbook of the Protists: Second Edition, , 1559–1618.
Campbell, Scott E. et al. 2013. “The Genome of Spraguea Lophii and the Basis of Host- Microsporidian Interactions.” PLOS Genetics 9(8): e1003676.
58 Cohen, Mitchell L. 2000. “Changing Patterns of Infectious Disease.” Nature 406(6797): 762–67.
Cuomo, Christina A et al. 2012. “Microsporidian Genome Analysis Reveals Evolutionary Strategies for Obligate Intracellular Growth.” Genome research 22(12): 2478–88.
Da Costa, T.H.M., and M.K. Ito. 2003. “PHOSPHOLIPIDS | Physiology.” In Encyclopedia of Food Sciences and Nutrition (Second Edition), ed. Benjamin Caballero. Oxford: Academic Press, 4523–31. http://www.sciencedirect.com/science/article/pii/B012227055X009172.
Dall, Kathrine B, and Nils J Færgeman. 2019. “Metabolic Regulation of Lifespan from a C. Elegans Perspective.” Genes & nutrition 14: 25–25.
Desjardins, Christopher A. et al. 2015. “Contrasting Host–Pathogen Interactions and Genome Evolution in Two Generalist and Specialist Microsporidian Pathogens of Mosquitoes.” Nature Communications 6(1): 7121.
Didier, Elizabeth S. 2005. “Microsporidiosis: An Emerging and Opportunistic Infection in Humans and Animals.” Acta Tropica 94(1): 61–76.
Estes, Kathleen A, Suzannah C Szumowski, and Emily R Troemel. 2011. “Non-Lytic, Actin-Based Exit of Intracellular Parasites from C. Elegans Intestinal Cells.” PLoS pathogens 7(9): e1002227–e1002227.
Franchet, Adrien, Sebastian Niehus, Gaëtan Caravello, and Dominique Ferrandon. 2019. “Phosphatidic Acid as a Limiting Host Metabolite for the Proliferation of the Microsporidium Tubulinosema Ratisbonensis in Drosophila Flies.” Nature Microbiology 4(4): 645–55.
Franzen, Caspar. 2004. “Microsporidia: How Can They Invade Other Cells?” Trends in Parasitology 20(6): 275–79. van Gool, Tom, and Jacob Dankert. 1995. “Human Microsporidiosis: Clinical, Diagnostic and Therapeutic Aspects of an Increasing Infection.” Clinical Microbiology and Infection 1(2): 75–85.
Han, Bing, Peter M Takvorian, and Louis M Weiss. 2020. “Invasion of Host Cells by Microsporidia.” Frontiers in microbiology 11: 172–172.
Hannich, J. Thomas et al. 2017. “Structure and Conserved Function of Iso-Branched Sphingoid Bases from the Nematode Caenorhabditis Elegans.” Chemical Science 8(5): 3676–86.
Herker, Eva, and Melanie Ott. 2012. “Emerging Role of Lipid Droplets in Host/Pathogen Interactions.” The Journal of biological chemistry 287(4): 2280–87.
59 Itoe, Maurice A et al. 2014. “Host Cell Phosphatidylcholine Is a Key Mediator of Malaria Parasite Survival during Liver Stage Infection.” Cell host & microbe 16(6): 778–86.
Katinka, Michaël et al. 2001. “Genome Sequence and Gene Compaction of the Eukaryote Parasite Encephalitozoon Cuniculi.” Nature 414: 450–53.
Keeling, Patrick J., and Naomi M. Fast. 2002. “Microsporidia: Biology and Evolution of Highly Reduced Intracellular Parasites.” Annual Review of Microbiology 56(1): 93– 116.
Keeling, Patrick J, and Claudio H Slamovits. 2004. “Simplicity and Complexity of Microsporidian Genomes.” Eukaryotic cell 3(6): 1363–69.
LaBauve, Annette E., and Matthew J. Wargo. 2014. “Detection of Host-Derived Sphingosine by Pseudomonas Aeruginosa Is Important for Survival in the Murine Lung.” PLOS Pathogens 10(1): e1003889.
Lee, Jung Hyun et al. 2014. “Lipid Droplet Protein LID-1 Mediates ATGL-1-Dependent Lipolysis during Fasting in Caenorhabditis Elegans.” Molecular and cellular biology 34(22): 4165–76.
Lee, Kwang-Zin et al. 2010. “The Fatty Acid Synthase Fasn-1 Acts Upstream of WNK and Ste20/GCK-VI Kinases to Modulate Antimicrobial Peptide Expression in C. Elegans Epidermis.” Virulence 1(3): 113–22.
Longo, Nicola, Marta Frigeni, and Marzia Pasquali. 2016. “Carnitine Transport and Fatty Acid Oxidation.” Biochimica et biophysica acta 1863(10): 2422–35.
Mahmud, Siraje Arif, Mohammed Adnan Qureshi, Madhab Sapkota, and Mark W. Pellegrino. 2020. “A Pathogen Branched-Chain Amino Acid Catabolic Pathway Subverts Host Survival by Impairing Energy Metabolism and the Mitochondrial UPR.” PLOS Pathogens 16(9): e1008918.
Mak, Ho Yi. 2012. “Lipid Droplets as Fat Storage Organelles in Caenorhabditis Elegans: Thematic Review Series: Lipid Droplet Synthesis and Metabolism: From Yeast to Man.” Journal of lipid research 53(1): 28–33.
Mani, Ramamurthy et al. 2008. “Defining Genetic Interaction.” Proceedings of the National Academy of Sciences 105(9): 3461.
Meadows, Jamie A, Graham G Willsey, and Matthew J Wargo. 2018. “Differential Requirements for Processing and Transport of Short-Chain versus Long-Chain O- Acylcarnitines in Pseudomonas Aeruginosa.” Microbiology (Reading, England) 164(4): 635–45. van der Meer-Janssen, Ynske P M, Josse van Galen, Joseph J Batenburg, and J Bernd Helms. 2010. “Lipids in Host-Pathogen Interactions: Pathogens Exploit the Complexity of the Host Cell Lipidome.” Progress in lipid research 49(1): 1–26.
60 “Microsporidiosis: (Last Updated November 6, 2013; Last Reviewed November 6, 2013).” 2013. The Pediatric Infectious Disease Journal 32. https://journals.lww.com/pidj/Fulltext/2013/11002/Microsporidiosis___Last_updated _November_6,_2013_.21.aspx.
Mittleider, Derek et al. 2002. “Sequence Survey of the Genome of the Opportunistic Microsporidian Pathogen, Vittaforma Corneae.” The Journal of eukaryotic microbiology 49: 393–401.
Montefusco, David J, Nabil Matmati, and Yusuf A Hannun. 2014. “The Yeast Sphingolipid Signaling Landscape.” Chemistry and physics of lipids 177: 26–40.
Nakjang, Sirintra et al. 2013. “Reduction and Expansion in Microsporidian Genome Evolution: New Insights from Comparative Genomics.” Genome Biology and Evolution 5(12): 2285–2303.
Nandakumar, Madhumitha, and Man-Wah Tan. 2008. “Gamma-Linolenic and Stearidonic Acids Are Required for Basal Immunity in Caenorhabditis Elegans through Their Effects on P38 MAP Kinase Activity.” PLoS genetics 4(11): e1000273–e1000273.
Pralhada Rao, Raghavendra et al. 2013. “Sphingolipid Metabolic Pathway: An Overview of Major Roles Played in Human Diseases” ed. Philip W. Wertz. Journal of Lipids 2013: 178910.
Pukkila-Worley, Read et al. 2014. “The Evolutionarily Conserved Mediator Subunit MDT- 15/MED15 Links Protective Innate Immune Responses and Xenobiotic Detoxification.” PLoS pathogens 10(5): e1004143–e1004143.
Ramírez-Zacarías, J. L., F. Castro-Muñozledo, and W. Kuri-Harcuch. 1992. “Quantitation of Adipose Conversion and Triglycerides by Staining Intracytoplasmic Lipids with Oil Red O.” Histochemistry 97(6): 493–97.
Reddy, Kirthi C. et al. 2017. “An Intracellular Pathogen Response Pathway Promotes Proteostasis in C. Elegans.” Current Biology 27(22): 3544-3553.e5.
Rex, John H., and David A. Stevens. 2015. “39 - Drugs Active against Fungi, Pneumocystis, and Microsporidia.” In Mandell, Douglas, and Bennett’s Principles and Practice of Infectious Diseases (Eighth Edition), eds. John E. Bennett, Raphael Dolin, and Martin J. Blaser. Philadelphia: Content Repository Only!, 479-494.e4. http://www.sciencedirect.com/science/article/pii/B9781455748013000394.
Rollin-Pinheiro, Rodrigo, Ashutosh Singh, Eliana Barreto-Bergter, and Maurizio Del Poeta. 2016. “Sphingolipids as Targets for Treatment of Fungal Infections.” Future medicinal chemistry 8(12): 1469–84.
Satouchi, Kiyoshi et al. 1993. “Phospholipids from the Free-Living NematodeCaenorhabditis Elegans.” Lipids 28(9): 837–40.
61 Soukas, Alexander A et al. 2009. “Rictor/TORC2 Regulates Fat Metabolism, Feeding, Growth, and Life Span in Caenorhabditis Elegans.” Genes & development 23(4): 496–511.
Stentiford, G D et al. 2016. “Microsporidia - Emergent Pathogens in the Global Food Chain.” Trends in parasitology 32(4): 336–48.
Strijbis, Karin, Frédéric Vaz, and Ben Distel. 2010. “Enzymology of the Carnitine Biosynthesis Pathway.” IUBMB life 62: 357–62.
Thompson, Owen et al. 2013. “The Million Mutation Project: A New Approach to Genetics in Caenorhabditis Elegans.” Genome research 23(10): 1749–62.
Troemel, Emily R et al. 2008. “Microsporidia Are Natural Intracellular Parasites of the Nematode Caenorhabditis Elegans.” PLOS Biology 6(12): e309.
Troemel, Emily R. 2016. “Host-Microsporidia Interactions in Caenorhabditis Elegans, a Model Nematode Host.” Microbiology Spectrum 4(5). https://www.asmscience.org/content/journal/microbiolspec/10.1128/microbiolspec.F UNK-0003-2016. van der Veen, Jelske N. et al. 2017. “The Critical Role of Phosphatidylcholine and Phosphatidylethanolamine Metabolism in Health and Disease.” Membrane Lipid Therapy: Drugs Targeting Biomembranes 1859(9, Part B): 1558–72.
Vrablik, Tracy L. et al. 2015. “Lipidomic and Proteomic Analysis of Caenorhabditis Elegans Lipid Droplets and Identification of ACS-4 as a Lipid Droplet-Associated Protein.” Biochimica et Biophysica Acta 1851(2015): 1337–45.
Wadi, Lina, and Aaron W. Reinke. 2020. “Evolution of Microsporidia: An Extremely Successful Group of Eukaryotic Intracellular Parasites.” PLOS Pathogens 16(2): e1008276.
Walker, Amy K et al. 2011. “A Conserved SREBP-1/Phosphatidylcholine Feedback Circuit Regulates Lipogenesis in Metazoans.” Cell 147(4): 840–52.
Watson, Emma et al. 2013. “Integration of Metabolic and Gene Regulatory Networks Modulates the C. Elegans Dietary Response.” Cell 153(1): 253–66.
Watts, Jennifer, and Michael Ristow. 2017. “Lipid and Carbohydrate Metabolism in Caenorhabditis Elegans.” Genetics 207: 413–46.
Weber, R, R T Bryan, D A Schwartz, and R L Owen. 1994. “Human Microsporidial Infections.” Clinical microbiology reviews 7(4): 426–61.
Weber, R., P. Deplazes, and D. Schwartz. 2000. “Diagnosis and Clinical Aspects of Human Microsporidiosis.” Contributions to microbiology 6: 166–92.
62 Weiss, Louis M. 2020. “109 - Microsporidiosis.” In Hunter’s Tropical Medicine and Emerging Infectious Diseases (Tenth Edition), eds. Edward T. Ryan et al. London: Content Repository Only!, 825–31. http://www.sciencedirect.com/science/article/pii/B9780323555128001095.
Wenk, Markus R. 2006. “Lipidomics of Host–Pathogen Interactions.” FEBS Letters 580(23): 5541–51.
Xu, Tao et al. 2010. “Revealing Parasite Influence in Metabolic Pathways in Apicomplexa Infected Patients.” BMC bioinformatics 11 Suppl 11(Suppl 11): S13–S13.
Zaarur, Nava et al. 2019. “ATGL-1 Mediates the Effect of Dietary Restriction and the Insulin/IGF-1 Signaling Pathway on Longevity in C. Elegans.” Molecular Metabolism 27(2019): 75–82.
Zhang, Gaotian et al. 2016. “A Large Collection of Novel Nematode-Infecting Microsporidia and Their Diverse Interactions with Caenorhabditis Elegans and Other Related Nematodes.” PLOS Pathogens 12(12): e1006093.
Zhang, Hongjie et al. 2011. “Apicobasal Domain Identities of Expanding Tubular Membranes Depend on Glycosphingolipid Biosynthesis.” Nature cell biology 13: 1189–1201.
Zhang, Jingshu, Yun Lan, and Sumana Sanyal. 2017. “Modulation of Lipid Droplet Metabolism-A Potential Target for Therapeutic Intervention in Flaviviridae Infections.” Frontiers in microbiology 8: 2286–2286.
Zhang, Yuru et al. 2013. “Comparative Genomics and Functional Study of Lipid Metabolic Genes in Caenorhabditis Elegans.” BMC genomics 14: 164.
63