<<

DEGREE PROJECT IN BIOTECHNOLOGY, SECOND CYCLE, 30 CREDITS STOCKHOLM, SWEDEN 2020

A zebrafish-based system to study the impact of environmental factors in Inflammatory Bowel Disease (IBD)

MIKAELA WESTLING

KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF ENGINEERING SCIENCES IN CHEMISTRY, BIOTECHNOLOGY AND HEALTH M.Sc Thesis in Biotechnology Mikaela Westling ______

Abstract (English) Inflammatory Bowel Disease (IBD) is a chronic disorder that affects millions of people worldwide. Although the etiology behind the disease is yet unknown, current theories propose a complex interplay between genetic susceptibility, exposure to environmental factors and exacerbated immune responses. While important efforts have been made to link genetics and environmental factors to IBD pathogenesis, a major challenge remains to assign them a causative role. Particularly since most of the IBD-risk genetic polymorphisms are found in non-coding regions (NCRs) with unknown regulatory activity, and for the lack of knowledge about how environmental factors can modulate the function of these elements in ​ vivo. A main problem to address this challenge in IBD research is the lack of an appropriate ​ model system in vivo that allows for high-throughput experiments with combinations of ​ different IBD-risk factors, while keeping the in vivo context. In this work, we sought to ​ ​ overcome this issue by using a zebrafish reporter for a specific human IBD-risk NCR, in order to investigate the modulation of this element by two groups of common environmental factors: pollutants, such as PolyFluoroAlkyl Substances (PFASs); and diet, by activation of dietary sensors.

We found that the activity of the WT-NCR in zebrafish larvae was increased in the presence of PFAS, while the activation of the dietary sensor PPARδ decreased the activity. These data ​ ​ lead us to suggest that the function of PFAS can be counteracted by PPARδ activation. Therefore, we propose zebrafish as a suitable in vivo model in which we can screen for ​ potentially harmful or beneficial effects of environmental factors in the activity of human non-coding regions.

Keywords: IBD, non-coding regions; zebrafish, in vivo, transgenic dual reporters; enhancer ​ ​ ​ activity; polyfluoroalkyl substances; dietary sensors ​

- 1 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Abstract (Swedish) Inflammatorisk tarmsjukdom (IBD) är en kronisk störning som drabbar miljontals människor världen över. Även om etiologin bakom sjukdomen fortfarande är okänd, föreslår nuvarande teorier ett komplext samspel mellan genetisk mottaglighet, exponering av miljöfaktorer och förvärrat immunförsvar. Även om stora ansträngningar har gjorts för att koppla genetik och miljöfaktorer till IBD-patogenes, återstår en stor utmaning att tilldela dem en orsakande roll. Särskilt eftersom de flesta av IBD-riskgenetiska polymorfismer finns i icke-kodande regioner (NCR) med okänd reglerande aktivitet samt för bristen på kunskap om hur miljöfaktorer kan modulera funktionen hos dessa element in vivo. Ett huvudproblem för att möta denna ​ ​ utmaning i IBD-forskning är avsaknaden av ett lämpligt modellsystem in vivo som möjliggör ​ experiment med hög kapacitet och kombinationer av olika IBD-riskfaktorer in vivo. I detta ​ ​ arbete försökte vi få svar på denna fråga genom att använda en zebrafiskreporter för ett specifikt humant IBD-risk icke-kodande område. Detta möjliggjorde att vi kunde undersöka modulering av två gemensamma miljöfaktorer: föroreningar, såsom PolyFluoroAlkyl-ämnen (PFASs); och diet, genom aktivering av dietsensorer.

Vi fann att aktiviteten i WT-NCR hos zebrafisklarver ökade i närvaro av PFAS, medan aktiveringen av dietsensorn PPARδ minskade aktiviteten. Denna data leder till att vi antyder att funktionen för PFAS kan motverkas genom PPARδ-aktivering. Därför föreslår vi zebrafisk som en lämplig in vivo-modell, i vilken vi kan screena för potentiellt skadliga eller ​ ​ gynnsamma effekter av miljöfaktorer i mänskligt icke-kodande DNA.

- 2 - M.Sc Thesis in Biotechnology Mikaela Westling ______Table of Contents

Abstract (English) 1

Abstract (Swedish) 2

Introduction 4 Inflammatory Bowel Disease 4 Genetics in IBD 4 Environmental factors in IBD 5 PolyfluoroAlkyl Substances (PFASs) 5 Dietary sensors 6 Studying genetic-environment interactions in vivo 7 Zebrafish reporters for human non-coding regions 8

Materials and Methods 10 Material, equipment and software 10 Workflow 11 Chemicals 11 Maintenance of ZF and exposure to compounds 12 Live Imaging and fluorescence analysis 12 Macros 13 RNA extraction 13 qPCR 14 Statistical analysis 15 Ethical considerations 16

Results 17 PFAS exposures 17 PFAS exposures in a context of inflammation 19 Activation of dietary sensors 21 RT-qPCR 23

Discussion 25 Environmental pollutants 25 Activation of dietary sensors 26 Limitations 27 Future aspects 28

Conclusion 29

Acknowledgments 30

References 31

Appendix 36

- 3 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Introduction

Inflammatory Bowel Disease

Inflammatory Bowel Disease (IBD), including Crohn’s disease (CD) and (UC), is an immune-mediated disorder that causes damage in the gastrointestinal (GI) tract of humans and causes complications such as diarrhea, fatigue, bloody stools, abdominal pain, or the more severe complications such as cancer, anemia or arthritis [1, 2]. The disease is categorized according to the tissue that is affected, where the inflammation in CD mostly affects the whole digestive system, whereas in UC it is restricted to mucosal inflammation mostly in the colon [3]. Studies have shown that both the incidence and prevalence of IBD has increased during the past decades in western and industrialized countries, with over 2 million and 1.5 million affected in Europe and North America, respectively [4]. Interestingly, studies also show an increased prevalence and incidence of IBD in countries in part of the world which starts to become more industrialized during the 21st century, such as eastern Europe, Africa, South America and Asia, where the disease previously was uncommon [1, 2]. This has made IBD to be considered as a global disease, and questions arise whether there is a correlation between industrialization and IBD.

Although the etiology of IBD is yet unknown, it has been proposed that it lies in the complex interaction between genetics, , the integrity of the intestinal epithelial barrier and exposure to various amounts of environmental factors [1]. Several efforts have been made to understand how these factors play a role in IBD pathogenesis. However, the molecular mechanisms behind the disease are far from being fully understood. This is likely due to the lack of experimental settings that allow simultaneous combinations of different IBD-risk factors in vivo. In this work, we will focus on addressing how environmental factors ​ ​ can modulate IBD-risk genetic polymorphisms in vivo. ​ ​

Genetics in IBD

It is well known in humans and animal models that single nucleotide polymorphisms (SNPs) that generate mutations in specific -coding (such as IL10R) are causative for IBD, but they represent only a minor fraction of total IBD patients. In most IBD patients, genetic mutations do not explain disease manifestation. To date, Genome-Wide Association Studies (GWAS) have identified around 241 IBD-risk polymorphisms, where the majority of them are located in human NCRs [5, 6]. The human genome consists of 2% -coding regions, which makes the remaining 98% NCRs. Due to a lack of knowledge and difficulties

- 4 - M.Sc Thesis in Biotechnology Mikaela Westling ______in how to investigate NCRs, they have historically been viewed as junk in comparison to coding-regions [7]. However, it has been proven that around 80% of the regulatory regions located in NCRs can regulate gene expression in both close and distant protein-coding genes, working as enhancers or promoters which can contribute to the pathogenesis of the disease, including IBD [8, 9, 10]. These studies have greatly increased the interest in studying the function of NCRs and their role in disease lately.

Enhancers can be located very distant, up to several megabases, from the genes they regulate. It functions by binding to transcription factors (TFs), which in turn binds to transcription factor binding sites (TFBSs) at the genomic DNA [11]. Several of the SNPs identified by GWAS are predicted to be located in DNA regulatory elements (DRE) that are active in specific cell types, such as intestinal epithelial- and immune cells [10]. In order to functionally validate the impact of disease-risk SNPs in NCRs, a general pipeline denominated CAUSEL (Characterization of Alleles USing Editing of Loci) was established, where multiple fine-mapped SNPs are epigenetically profiled, and later examined by comparing expression traits between isogenic cell lines [12]. However, the lack of tools to identify the impact of SNPs in the function of NCRs in vivo, and the role they play in the ​ ​ disease progression remain as major challenges in the field.

Environmental factors in IBD

Although genetics play an important role in the pathogenesis of IBD, not all individuals who possess IBD-risk mutations in the genome develop the disease. Current theories propose that genetically predisposed individuals with chronic exposure to environmental triggers may result in an excessive inflammatory response, thus causing IBD [13]. Although genetic variations among individuals play a key role (by e.g increasing the production of proinflammatory cytokines), environmental triggers seem to possess a major influence in IBD pathogenesis [14]. Thus, common risk factors include urban pollutants, high consumption of fat and carbohydrates, smoking and stress among others.

PolyfluoroAlkyl Substances (PFASs) ​ Over the past few decades, the presence in the environment of a group of man-made organic pollutants, named Polyfluoroalkyl substances (PFASs), has dramatically increased. PFASs ​ consist of carbon chains bound to instead of hydrogen, with various functional groups attached at the end of the chain. The most commonly used and studied pollutants from this group are perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA) and perfluorohexane sulfonic acid (PFHxS) [15] (see supplementary fig. A1 in appendix for

- 5 - M.Sc Thesis in Biotechnology Mikaela Westling ______chemical structures). The strong bond between carbon and fluorine allows for water- and oil repelling attributes, making the compounds useful for packaging and commercial household products among others [16]. Thus, the pollutants are extremely persistent and widely spread throughout large parts of the world, resulting in contamination of as well as accumulation in living organisms [15, 16]. PFASs are absorbed in the GI tract of humans and distributed mainly to the and the plasma.

Since the pollutants are very stable, they are not metabolized, but excreted through the urine and feces. The estimated half-life of PFASs has been proven to be as long as between 3.5 and 8.5 years [17]. PFASs are reported to correlate with human adverse effects such as immunosuppressive responses, neurological disorders and an increased risk of cancer [16, 18]. Previous studies have also shown that increased levels of PFOS can lead to a modulation of the local immune system, resulting in pro-inflammatory effects leading to intestinal damage [19]. Both PFOS and PFOA have been associated with UC, however, the underlying molecular mechanism is not known [16, 18]. These findings highlight these pollutants as an important IBD-risk factor to further investigate.

Dietary sensors Another set of environmental factors that are in constant exposure in the GI tract are dietary-derived metabolites that activate dietary sensors. Most of the dietary sensors are classified as Nuclear Receptors (NRs), a well-studied superfamily of structurally related known to be ligand-activated transcription factors [20]. These receptors regulate a great number of pathways in the cells, including , immune response, cell proliferation and development, to only mention a fraction. Hence, looking at NRs activation by dietary-derived ligands that are common in westernized countries, such as cholesterol, lipid and carbohydrates, is of great importance for IBD research [21].

One of the NRs examined is Liver X Receptor (LXR), which in previous studies has been proven to enhance pro-inflammatory T-cells when repressed in mice, and plays an important role in regulating intestinal homeostasis. When mice were exposed to LXR agonists, the expression of inflammatory cytokines such as Tumor Necrosis Factor Alpha (TNFα) was strongly suppressed, making it a great candidate to further investigate [22, 23]. Another modulator that has been proven to protect against intestinal inflammation and colitis is Retinoic Acid (RA) and its activation of Retinoic Acid Receptor (RAR) [24]. RA is derived in the body from vitamin A, an important dietary compound found to be rich in eggs, fish,

- 6 - M.Sc Thesis in Biotechnology Mikaela Westling ______carrots and sweet potato amongst others [25]. The metabolite has been proven to be crucial for preservation in the intestinal barrier by shaping intestinal immune cell development and thereby the composition of B- and T cells in order to maintain homeostasis [26]. Interestingly, it has been reported that patients with IBD which had deficient levels of vitamin A were subjected to a higher likelihood of surgery and hospitalizations than those who retained normal levels [27]. Along with RAR, Proliferator-Activated Receptor (PPAR) also regulates the expression of tight junction proteins and regulates mucus secretion [28]. Previous studies have shown that PPARδ can play different roles depending on whether it is bound to a ligand or not. It can form complexes with other TFs in its unbound form, regulating gene transcription and suppressing proinflammatory cytokines [29]. In contrast, when bound to an agonist, it has been reported that inflammation was alleviated [30].

All of these nuclear receptors mentioned above (LXR, RAR, PPARs) have been validated in preclinical mouse models used for finding novel therapeutics in IBD, making them appropriate targets for investigating IBD-risk genetic factors [28].

Studying genetic-environment interactions in vivo ​ To date, most studies performed for investigating the complex interaction between the genetics and environmental factors in IBD are based on patient cohorts, where it is known that the patients have been exposed to an amount of various environmental factors. That allows for the association between exposures, specific genetic polymorphisms and disease [31]. Literature studies have revealed that genetic susceptibility does not cause IBD by itself, though it is an interplay between many different environmental factors, which are referred to as the exposome [32]. However, most of the functional studies only investigate one environmental factor at a time, lacking the real-world scenario where multiple genetic variations interplay with the exposome. Although in vitro cell-line reporter assays allow ​ ​ high-throughput screenings for combinations of environmental factors, they lack the in vivo ​ physiological context, as the studies are limited to specific isolated cell types and tissues [12]. Therefore, other in vivo approaches are needed to study both attributes simultaneously. ​ ​ Zebrafish (Danio rerio) is an excellent model organism for in vivo studies, due to its ​ ​ ​ ​ well-mapped genome and genetic similarity regarding intestinal- and immune function to the human genome. Furthermore, the transparency of embryos and larvae enables studies of fluorescence under a microscope without interrupting the system. In addition, the larvae have

- 7 - M.Sc Thesis in Biotechnology Mikaela Westling ______a very small size and high fecundity, allowing for growth in 96-well plates as well as breeding of around 200 embryos per mating. This results in reduced experimental costs and the allowance of high throughput screenings, while remaining the physiological context. Another good reason why zebrafish (ZF) can be used as an appropriate model for human behavior is the similarity in terms of the pharmacology of many drugs used for humans [33].

To date, there are available chemical models of ZF used for inducing intestinal inflammation similar to the human pathogenesis of IBD. Among them is the deleterious agent TriNitroBenzene Sulfonic acid (TNBS) used in this thesis work, which has proven to induce important pro-inflammatory cytokines and impair the intestinal homeostasis [34]. All of these attributes together announce ZF as a suitable model organism for investigation of how IBD-risk pollutants can modulate intestinal inflammation of IBD and to further develop drug screening for disease modifiers.

Zebrafish reporters for human non-coding regions

Functional analysis of human NCRs appears to remain a big challenge due to the lack of effective high-throughput models. However, it has been demonstrated that human NCRs can faithfully drive specific expression patterns in zebrafish and recapitulate the gene expression of human tissues [35]. In the host laboratory, a ZF dual reporter system was established to study the enhancer activity of IBD-risk human NCRs based on the work of Bhatia and colleagues [36]. Optimized plasmids for integration in the zebrafish genome [37] carrying human NCRs controlling the expression of fluorescent proteins were generated and injected in zebrafish embryos, in order to generate stable transgenic lines. In the established dual reporter system, the wild-type (WT) NCR drives expression of Green Fluorescent Protein (GFP), whereas the mutated IBD-risk variant drives expression of mCherry protein. Given that both vectors were injected into the same individuals, it is possible to verify the differences in the gene expression pattern between the wild type NCR and the respective IBD-risk variant simultaneously [36].

Among the IBD-risk SNPs located in human NCRs, we will focus on the SNP named rs17085007, described as IBD-risk in a GWAS work from a Japanese cohort [38]. This SNP is found on chromosome 13q12.13, where the nucleotide T (WT) has been substituted for C (IBD-risk). It has been shown that UC patients who possess the IBD-risk variant of rs17085007 suffer a higher risk of relapse to the disease [39]. Regarding the function, this region may contain regulatory sequences that can alter the gene expression of protein-coding genes [40]. In addition, it has been reported that the NCR containing the WT sequence of

- 8 - M.Sc Thesis in Biotechnology Mikaela Westling ______rs17085007 can drive expression in the intestine of zebrafish larvae [10]. The host laboratory has generated and established the zebrafish dual reporter line Tg(rs17085007_DR), which shows the enhancer activity for the NCR containing this SNP, but no differences between the expression patterns have been observed (see fig 1). Therefore, additional factors may influence the expression of this NCR, and potentially trigger differences in the expression pattern between the WT and the IBD-risk variants.

Figure 1. The zebrafish dual reporter for the non-coding region containing the SNP rs17085007. (A) Schematics of the generated constructs. The WT-NCR controls expression of GFP, while the IBD-risk mutant-NCR controls the expression of mCherry. (B) Pictures taken from the intestinal part of the ZF larvae at 5 ​ dpf, representing gene expression of WT-GFP, mutant-mCherry, merged from both GFP and mCherry as well as a brightfield picture. Modified from Villablanca lab, unpublished

The question we seek to answer in this thesis is therefore whether environmental factors ​ can modulate the expression pattern of the NCR. Specifically, we seek to analyze whether PFAS and activation of dietary sensors can modulate the expression of the NCR containing the SNP rs17085007 in zebrafish.

- 9 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Materials and Methods

Material, equipment and software All the reagents for larvae exposures, RNA extraction as well as kits, laboratory apparatus and software used in the experiments are listed below in table 1 and 2, with their corresponding concentration and manufacturer.

Table 1. All the reagents and solutions with corresponding concentrations and respective manufacturer used in ​ ​ experiments are shown below

Table 2. All the kits, laboratory apparatus and softwares and respective manufacturer used in experiments are ​ ​ shown below

- 10 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Workflow A schematic representation of how the experiment was set up is shown below in figure 2.

Figure 2. Schematic representation of experimental pipeline. Embryos were screened upon 3 days post ​ fertilization (dpf) under a light microscope to ensure high growth rate and normal phenotype when eventually hatched to larvae. On day 3, the larvae were transferred to a 24 well-plate, distributing 10 larvae in each well and exposed to the tested compounds for 48 hours. At 5 dpf, two larvae were fixed in TRIzol for RNA extraction and posterior qRT-PCR assays. The remaining larvae were mounted in low-gelling point agarose and imaged in a fluorescence microscope. Images acquired were used for posterior analyses using ImageJ software

Chemicals The pollutants PFOS, PFAS and PFHxS were kindly provided by Dr. Emma Wincent (Institute of Environmental Medicine, Karolinska Institutet). These compounds were diluted in pure DMSO at a concentration of 200 µM, and stored at -20°C. For exposures, compounds ​ ​ were pre-diluted at a concentration of 2 µM in E3 (containing 5.0 mM NaCl, 0.33 mM CaCl2, ​ ​ ​ ​ 0.33 mM MgSO and 0.2 mM HEPES), and then finally diluted to 0.2 µM when exposed to ​4 ​ ​ the larvae. All dietary sensor agonists were stored at a concentration of 1 mM (using DMSO as diluent) at -20°C, and later diluted in E3 to a final concentration of 1 µM used for exposure ​ ​ to the larvae. All treatment groups contained a final concentration of 0.1 % DMSO, which is a common polar aprotic solvent used in pharmaceuticals due to its anti-inflammatory properties among others [41]. Control larvae were incubated in E3 medium containing 0.1% of DMSO.

- 11 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Maintenance of ZF and exposure to compounds The zebrafish dual-reporter for the human non-coding region containing the SNP rs17085007, henceforth referred to as Tg(rs17085007_DR), was generated in the host ​ ​ laboratory and kept in the Karolinska Institutet Zebrafish Core Facility. Embryos from this line were obtained at day 0 and kept in an incubator with a temperature of 28 ± 1 °C. In order ​ to achieve the highest possible growth rate, the dead or unfertilized embryos were discarded under the microscope constantly during the growth, likewise the E3 medium was exchanged every 24 h. Depending on the number of embryos achieved, they were kept at small or large petri dishes with 50 or 200 eggs/plate respectively. At 48 hours post-fertilization (hpf), the embryos were screened under a fluorescence microscope for double-positive fluorescence, meaning both GFP and mCherry. All the other embryos were discarded. At 72 hpf, positive larvae were transferred into a 24-well plate, with 10 larvae per well, and then treated with 1 mL of the 10 different conditions tested. The conditions varied depending on the amount of larvae obtained as well as the outcome of previous experiments. However, the tested conditions during all experiments were compiled and consisted of DMSO, PFOS, PFOA, PFHxS, TNBS, TNBS + PFOS, TNBS + PFOA, TNBS + PFHxS, RAR agonist, LXR agonist, PPARα agonist and PPARδ agonist. Due to the low water of many of the compounds, DMSO is a common solution added to experiments in order to speed up the administration process. The control contained an E3 medium with a final concentration of 0.2 % DMSO, a proper concentration based on previous studies [41]. The final concentrations of the compounds for the exposure can be found in table 1. All larvae were exposed for 48 h in an incubator with the temperature 28 ± 1°C, including an exchanged medium after 24 h. ​ ​

Live Imaging and fluorescence analysis After 5 dpf, the larvae were washed 3x with E3 medium and anesthetized with tricaine (MS-222, 0.04% in E3). Two larvae were kept in 200 µL of Trizol for RNA extraction, while ​ ​ the remaining larvae were mounted on a petri dish using 1 % of low gelling point agarose, to image them under the microscope. Larvae with a deviating phenotype, such as bent tails were discarded. The same applied to mortality. Notably, the larvae needed to be aligned with clear ​ visibility for the whole intestine, which was the area of which pictures were to be taken. Pictures were taken by a SMZ25 Research Stereo Microscope (Nikon), equipped with TRITC and GFP filters to image red and green fluorescence, respectively. The settings applied when taking pictures were 1 second with a gain of 2.0x for GFP, and 2 seconds and 2.0x gain for mCherry.

- 12 - M.Sc Thesis in Biotechnology Mikaela Westling ______

In order to analyze the fluorescence given by the non-coding dual reporter in which site and how much the NCR drives expression, the fluorescence areas and the intensities for the WT (GFP) and mutant (mCherry) variants of the non-coding reporter were examined simultaneously. The analysis was achieved by first examining the fluorescence of the entire intestine (referred to as whole intestine analysis), and further investigated by dividing the ​ ​ intestine in three different areas: Anterior intestine, Mid intestine and Posterior intestine (henceforth referred as sectioned intestine analysis). Admittedly, it is of highest interest to ​ ​ determine in which area of the intestine where most alterations occur [42].

Macros To make the analysis where one measures the area as well as the intensity of the fluorescence of the pictures taken in the fluorescence microscope, the software ImageJ was used. Different macros were designed to make crops of the raw images in the intestine area, and for the measurements of fluorescence area-intensity where a threshold was used for (see fig. 3 and appendix for scripts). Results were saved in an excel table for statistical analyses.

Figure 3. Diagram showing the threshold was set from the selected image, in order to measure positive ​ fluorescence area and intensity in the intestine. The same threshold was used for all selected images in each experiment

RNA extraction In order to isolate RNA from the tissue of ZFE, the TRIzol reagent procedure was utilized. The TRIzol reagent was kept stored at 4°C, and is able to maintain the integrity of RNA ​

- 13 - M.Sc Thesis in Biotechnology Mikaela Westling ______while dissolving the other cell components as well as RNases properly in one hour. The ZFE were fixed in TRIzol and stored in -80°C until the procedure could be started. Along with the TRIzol reagent, the solution was pulled through sterile syringes with a 23G needle and a 27G needle, 15x and 5x respectively. Furthermore, RNA extractions steps were followed according to manufacturer's instructions, with small modifications for the centrifugation steps where the settings were applied at 15,000 g instead of 10,000 g. The bench, pipettes and gloves were treated with RNase away to avoid RNA degradation and contamination. The RNA pellet was resuspended in 20 µL of RNase-free water followed by an incubation in a ​ heat block for 15 min at 60°C. ​

To be sure that the samples contained solely RNA and not DNA, a DNA-free kit was utilized ​ for the removement of DNA from contaminating the RNA preparations, containing rDNase I, 10x DNase I buffer and DNase inactivation reagent. 2 µL of 10x DNase I buffer was added to ​ the samples, followed by the addition of 0.5 µL rDNase I. The samples were briefly mixed ​ ​ and spun before incubated for 30 min in 37°C to allow the degradation of remaining DNA. ​ Importantly, the removal of rDNase I was of high priority since cDNA conversion was the next step. For this, 2 µL of DNase inactivation reagent was resuspended and vortexed ​ multiple times while incubated for 2 min at room T. The samples were then centrifuged at 10,000 x g for 2 min followed by the transfer of 20 µL of the RNA solution to a new ​ autoclaved Eppendorf tube. Finally, RNA quantification was determined using Nanodrop 1000 Spectrophotometer by adding 2 µL to the device which measured the purity ratio as ​ ​ well as sample concentration. qPCR In order to quantify the gene expression, quantitative Polymerase Chain Reaction (qPCR) was performed, analyzing three different genes of interest; GFP, mCherry and elongation ​ ​ ​ factor 1 α (ef1a) [43]. To enable normalization of the relative gene expression, the ​ ​ housekeeping gene ef1a was used, as previously reported to be reliable for ZF tissues [44]. ​ Previous studies have shown that ef1a has been identified as a part of the mitotic apparatus ​ and thus interact with microtubules. Its canonical function is to deliver aa-tRNA to the ribosome and a defect would consequently confer to a deleterious evolution [45].

Before performing qPCR, the RNA samples needed to be converted to cDNA. For this, 8 µL ​ ​ of the RNA solution together with 8 µL of RNase-free water was mixed in new PCR-strip ​ ​ tubes, and 4 µL of the iScript reverse transcriptase supermix (BioRad) was pipetted into the ​ ​ solution. Samples were briefly vortexed and spun, and then placed in T100TM Thermal Cycler ​

- 14 - M.Sc Thesis in Biotechnology Mikaela Westling ______(BioRad) to run the program presented below in figure 4. After the run, the samples were ​ diluted with 20 µl nucleus free water (NFH2O), thus making a total of 40 µl of cDNA and ​ ​ stored at -20°C upon use. ​ ​

Figure 4. Program used to run cDNA conversion ​

The samples containing cDNA were amplified and quantified in duplicates by preparing a mix for each corresponding gene with 5 µl Master Mix (2x iTaq Universal SYBR Green ​ Supermix), 0.5 µl primer mix (containing both forward- and reverse primers diluted in

NFH2O to a final concentration of 5 µM each, see table 3) and 2.5 µl NFH2O. Furthermore, 2 ​ ​ ​ ​ µl of the mix was pipetted onto 384-well PCR plates together with 2 µl cDNA before ​ ​ entering the program (see fig. 5).

Figure 5. Program used for qPCR analysis ​ ​

Table 3. DNA sequence for the gene-specific primers used for qPCR, including the housekeeping gene ef1a as ​ ​ ​ well as GFP and mCherry ​ ​ ​ Gene Forward primer 5’-3’ Reverse primer 3’-5’

ef1a ACCTACCCTCCTCTTGGTCG GGAACGGTGTGATTGAGGGAA

GFP CTACCCCGACCACATGAAGC TCCTCCTTGAAGTCGATGCC

mCherry CACGAGTTCGAGATCGAGGG CCAGTAGTCGGGGATGTCGG

Statistical analysis For qPCR analysis the 2−ΔΔCt method was used, presenting the variations in the level of gene expression [46]. These results were plotted using “log fold change” values, thus visualizing ​2 whether the genes of interest were upregulated (over 0) or downregulated (below 0), respect to untreated controls.

- 15 - M.Sc Thesis in Biotechnology Mikaela Westling ______

All statistical analyses of the data, as well as plots, have been created using the software GraphPad Prism 6. The settings applied for normally distributed data obtained from the whole intestine was a parametric one-way ANOVA, with a complementary Dunnett's multiple comparisons test to determine whether the results were significant. The settings applied for the data collected from the post-mid-ant intestine was two-way ANOVA complemented with the same multiple comparisons. Results were considered statistically significant when p-values were equal to or lower than 0.05.

Ethical considerations The project is based on experiments on zebrafish, which has been taken into considerations by following the EU directive 2010/63/EU. In article 10, there are specific directions on how to treat larvae in order to fulfill animal welfare. One important note is that zebrafish embryos are not considered as animals before the stage of independent feeding, meaning 5 days post fertilization (5dpf). After that, a special permit is required [47]. Since this project also utilizes adult fish for breeding, the ethical permit N5756/17 (granted to Eduardo Villablanca) has been used to maintain adult strains.

- 16 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Results The dual reporter Tg(rs17085007_DR) was generated in the host laboratory (Morales R., ​ unpublished) in order to compare the expression between the WT and the IBD-risk variant of the NCR containing the studied SNP, by analyzing GFP and mCherry expression, respectively. However, we found no significant differences in the expression pattern when comparing GFP and mCherry expression (see fig. 1), suggesting that the mutant NCR has a comparable activity to the WT NCR at steady-state conditions. We then hypothesized that an additional challenge, such as exposure to environmental factors, may modulate the enhancer activity of this NCR. In order to check for potential regulators of this enhancer, we exposed larvae to a set of perfluoroalkyl substances (PFAS) members and to dietary sensors agonists.

PFAS exposures To test whether PFASs cause an alteration in the enhancer activity of the WT- and IBD-risk variant of the human NCR studied, we exposed Tg(rs17085007_DR) larvae to single doses of ​ PFOS, PFOA and PFHxS for 48 hours, renewing media with exposures after 24 hours. PFOS and PFOA were tested in four different experiments whereas PFHxS was tested in 2 different experiments. Both GFP and mCherry expression patterns were analyzed in the intestine as a whole, and also after dividing the intestine in anterior-mid-posterior sections. Here, representative images of the control- and PFOS conditions are shown (see fig. 6 and 7 e). The data for GFP showed an increased fluorescence area for the larvae treated with all three tested PFASs compared to the control treatment (see fig. 6 a). Analysis of the IBD-risk variant of the NCR represented by mCherry showed no significant changes in the expression area (see fig. 7), suggesting that the mutant NCR is impaired in increasing its activity upon pollutant exposure. Sectioned analysis showed that the increase in the GFP fluorescence area occurred in the mid intestine, while expression in the anterior and posterior intestine is comparable to the untreated control (see fig. 6 c). We did not observe changes in the fluorescence intensity of GFP and mCherry after PFAS exposures (see fig. 6 b, d and 7 b, d). These results indicate that the WT variant of the inserted human NCR is responding to PFASs in zebrafish, by means of increasing the expression area of GFP.

- 17 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Figure 6. WT-GFP fluorescence analyses from Tg(rs17085007_DR) larvae after PFAS exposures. GFP area 2 ​ was measured in squared millimeters (mm )​ , while fluorescence intensity was measured in arbitrary ​ fluorescence units. Top graphs show the analysis of the GFP area (A) and intensity (C) from the whole intestine ​ ​ after treatments. Each dot represents one larva, and one-way ANOVA analyses were performed. In the bottom graphs, partial analysis of GFP area (B) and intensity (D) from intestinal sections are shown. Bars represent ​ ​ mean ± standard deviation in each condition, and two-way ANOVA analyses were performed. In all graphs, N = 4 independent experiments for Control, PFOS and PFOA (25, 31 and 28 larvae respectively), while N = 2 * ** *** independent experiments for PFHxS (14 larvae in total). p < 0.05; p < 0.01; p​ < 0.001. (E) Representative ​ ​ ​ ​ pictures of Control (up) and PFOS-treated larva (down) are shown. Dashed yellow lines delimit intestinal regions (anterior, mid and posterior). Images were taken using 6x magnification

- 18 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Figure 7. Mutant-mCherry fluorescence analyses from Tg(rs17085007_DR) larvae after PFAS exposures. ​ ​ 2 mCherry area was measured in squared millimeters (mm )​ , while fluorescence intensity was measured in ​ arbitrary fluorescence units. Top graphs show the analysis of the mCherry area (A) and intensity (C) from the ​ ​ whole intestine after treatments. Each dot represents one larva, and one-way ANOVA analyses were performed. In the bottom graphs, partial analysis of mCherry area (B) and intensity (D) from intestinal sections are shown. ​ ​ Bars represent mean ± standard deviation in each condition, and two-way ANOVA analyses were performed. In all graphs, N = 4 independent experiments for Control, PFOS and PFOA (25, 31 and 28 larvae respectively), while N = 2 independent experiments for PFHxS (14 larvae in total). * p < 0.05. (E) Representative pictures of ​ ​ Control (up) and PFOS-treated larva (down) are shown. Dashed yellow lines delimit intestinal regions (anterior, mid and posterior). Images were taken using 6x magnification

PFAS exposures in a context of inflammation To further investigate the influence of PFAS on NCRs in the context of inflammation, the addition of an inflammatory stimulus was included in one pilot experiment. We co-exposed Tg(rs17085007_DR) larvae with PFASs and TNBS for 48h and we analyzed changes in the expression pattern of GFP (WT NCR) and mCherry (IBD-risk NCR), in where a representative image is shown for the conditions between PFAS and TNBS + PFAS (see fig. 8 and 9 e). TNBS treatment did not give significant differences when compared to untreated controls, although a trend of reduced intestine area in both GFP and mCherry was observed (see fig. 8 a,c and 9 a,c). The whole intestine for GFP area fluorescence showed that PFOS- and PFOA-induced GFP area was reduced when larvae were co-exposed with TNBS. Interestingly, we observed that co-exposure of PFASs with TNBS resulted in a decreased GFP area in the mid intestine, when compared to PFASs treatments alone (see fig. 8 c). As described with previous results, no significant outcome was obtained for GFP intensity (see

- 19 - M.Sc Thesis in Biotechnology Mikaela Westling ______fig. 8 d). In addition, no differences were observed in mCherry fluorescence in all the conditions tested (see fig. 9). These results suggest a downregulation in the enhancer activity of the WT-NCR when exposed to PFAS under inflammatory conditions.

Figure 8. WT-GFP fluorescence analyses from Tg(rs17085007_DR) larvae after PFAS exposures and with ​ ​ 2 co-exposure to TNBS. GFP area was measured in squared millimeters (mm )​ , while fluorescence intensity was ​ measured in arbitrary fluorescence units. Top graphs show the analysis of the GFP area (A) and intensity (C) ​ ​ from the whole intestine after treatments. Each dot represents one larva, and one-way ANOVA analyses were performed. In the bottom graphs, partial analysis of GFP area (B) and intensity (D) from intestinal sections are ​ ​ shown. Bars represent mean ± standard deviation in each condition, and two-way ANOVA analyses were performed. In all graphs, N = 1 independent experiment for Control, TNBS, PFOS, TNBS + PFOS, PFOA, TNBS + PFOA, PFHxS and TNBS + PFHxS (4, 6, 7, 6, 5, 6, 6 and 5 larvae respectively). * p < 0.05; ** p < 0.01; ​ ​ **** p​ < 0.0001. (E) Representative pictures of PFOA (up) and TNBS + PFOA-treated larva (down) are shown. ​ Dashed yellow lines delimit intestinal regions (anterior, mid and posterior). Images were taken using 6x magnification

- 20 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Figure 9. Mutant-mCherry fluorescence analyses from Tg(rs17085007_DR) larvae after PFAS exposures and ​ ​ 2 with co-exposure to TNBS. mCherry area was measured in squared millimeters (mm )​ , while fluorescence ​ intensity was measured in arbitrary fluorescence units. Top graphs show the analysis of the mCherry area (A) ​ and intensity (C) from the whole intestine after treatments. Each dot represents one larva, and one-way ANOVA ​ analyses were performed. In the bottom graphs, partial analysis of mCherry area (B) and intensity (D) from ​ ​ intestinal sections are shown. Bars represent mean ± standard deviation in each condition, and two-way ANOVA analyses were performed. In all graphs, N = 1 independent experiment for Control, TNBS, PFOS, TNBS + PFOS, PFOA, TNBS + PFOA, PFHxS and TNBS + PFHxS (4-7 larvae respectively). (E) ​ Representative pictures of PFOA (up) and TNBS + PFOA-treated larva (down) are shown. Dashed yellow lines delimit intestinal regions (anterior, mid and posterior). Images were taken using 6x magnification

Activation of dietary sensors ​ ​ We then sought to investigate whether activation of dietary sensors has an impact in enhancer activity of the human NCR cloned in the Tg(rs17085007_DR) line. As in previous ​ experiments, we exposed the larvae to dietary sensor agonists for 48 hours. LXR agonist (GW3965), RAR agonist (retinoic acid), PPARα agonist (WY14643) and PPARδ agonist (GW501516) were tested in three individual experiments, while whole and partial analyses were carried out for each fluorescent protein. A representative image shows a comparison between control and PFOS conditions (see fig. 10 and 11 e). In the whole intestine analysis, PPARδ agonist gave a significantly decreased GFP area when compared to the control. Even though RAR agonists did not show significant results, the results appeared to be decreased as well (see fig. 10 a). In the partial analysis for GFP, both RAR agonist and PPARδ agonist gave a significantly decreased GFP area in the mid intestine (see fig. 10 c). Regarding the

- 21 - M.Sc Thesis in Biotechnology Mikaela Westling ______intensity analysis, the only difference observed was an increased GFP intensity in the anterior intestine after LXR agonist exposure (see fig. 10 c). As previously, no alteration in the expression pattern was observed for the mCherry fluorescence (see fig. 11). These results imply a downregulation in the enhancer activity of the WT-NCR in the mid intestine when exposed to PPARδ and RAR agonists.

Figure 10. WT-GFP fluorescence analyses from Tg(rs17085007_DR) larvae after dietary sensors exposures. ​ 2 GFP area was measured in squared millimeters (mm )​ , while fluorescence intensity was measured in arbitrary ​ fluorescence units. Top graphs show the analysis of the GFP area (A) and intensity (C) from the whole intestine ​ ​ after treatments. Each dot represents one larva, and one-way ANOVA analyses were performed. In the bottom graphs, partial analysis of GFP area (B) and intensity (D) from intestinal sections are shown. Bars represent ​ ​ mean ± standard deviation in each condition, and two-way ANOVA analyses were performed. In all graphs, N = 4 independent experiment for Control (26 larvae), while N = 3 individual experiments for LXR agonist (GW3965), RAR agonist (retinoic acid), PPARα agonist (WY14643) and PPARδ agonist (GW501516) (22, 17, * ** **** 23 and 21 larvae respectively). p < 0.05; p < 0.01; p​ < 0.0001. (E) Representative pictures of control (up) ​ ​ ​ ​ and PPARδ agonist (GW501516)-treated larva (down) are shown. Dashed yellow lines delimit intestinal regions (anterior, mid and posterior). Images were taken using 6x magnification

- 22 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Figure 11. Mutant-mCherry fluorescence analyses from Tg(rs17085007_DR) larvae after dietary sensors ​ ​ 2 exposures. mCherry area was measured in squared millimeters (mm )​ , while fluorescence intensity was ​ measured in arbitrary fluorescence units. Top graphs show the analysis of the mCherry area (A) and intensity ​ (C) from the whole intestine after treatments. Each dot represents one larva, and one-way ANOVA analyses were performed. In the bottom graphs, partial analysis of mCherry area (B) and intensity (D) from intestinal ​ sections are shown. Bars represent mean ± standard deviation in each condition, and two-way ANOVA analyses were performed. In all graphs, N = 4 independent experiment for Control (26 larvae), while N = 3 individual experiments for LXR agonist (GW3965), RAR agonist (retinoic acid), PPARα agonist (WY14643) and PPARδ agonist (GW501516) ( 22, 17, 23 and 21 larvae respectively). (E) Representative pictures of control (up) and ​ PPARδ agonist (GW501516)-treated larva (down) are shown. Dashed yellow lines delimit intestinal regions (anterior, mid and posterior). Images were taken using 6x magnification

RT-qPCR In order to validate the findings observed in the fluorescence expression pattern of the human NCR cloned in the Tg(rs17085007_DR) line when exposed to all the tested compounds for 48 ​ hours, RT-qPCR analyses were performed. Here, we wanted to examine whether the mRNA expression level of the fluorescence protein mRNA correlated with their protein levels measured by area-intensity. The results show that the relative expression of the mRNA for GFP is decreased after PFOA and PFHxS treatments, whereas the expression from mCherry is widely spread (see fig. 12 a,b). Regarding NRs, relative gene expressions of GFP and mCherry show no differences between the treatments performed, and when compared to controls (see fig. 12 e, f).

- 23 - M.Sc Thesis in Biotechnology Mikaela Westling ______In general, the results are widely spread and do not correlate with the previous results obtained from the fluorescence analysis of live embryos-larvae.

Figure 12. Relative gene expression of GFP and mCherry after exposure to all the compounds tested in this ​ project. All the larvae have been exposed to the compounds for 48 hours and compared and normalized to a housekeeping gene (ef1a) as well as the control condition, represented as log fold change. Each dot represents ​ ​ ​2 2 ZF larvae from one separate experiment. (A) Relative gene expression of GFP obtained from exposure to PFOS, PFOA and PFHxS ([PFAS] = 200 nM; control = 1 mL E3). (B) Relative gene expression of mCherry ​ obtained from exposure to the same conditions. (C) Relative gene expression of GFP obtained from exposure to ​ all PFASs as well as TNBS + PFASs ([PFAS] = 200 nM; [TNBS] = 70 µg/mL; control = 1 mL E3). (D) ​ ​ ​ Relative gene expression of mCherry obtained from exposure to the same conditions. (E) Relative gene ​ expression of GFP obtained from exposure to LXR agonist, RAR agonist, PPARα agonist and PPARδ agonist ([agonists] = 200 nM; control = 1 mL E3). (F) Relative gene expression of mCherry obtained from exposure to ​ the same conditions. (G) A representative picture visualizing the performance of qPCR at 5 dpf of the larvae, ​ after 48 hours of treatment with compounds

- 24 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Discussion Although GWAS have revealed several IBD-risk polymorphisms located in NCRs, their functions need to be further investigated in order to better understand their contribution to IBD pathogenesis [1]. Moreover, the impact of environmental factors in the activity of the identified IBD-risk polymorphisms still constitutes a major challenge, since current tools are limited to provide insights regarding disease-risk genetic-environment interactions in vivo. In ​ ​ this work, we sought to discover potential environmental modulators of the enhancer activity of an IBD-risk NCR, by using zebrafish reporters. The results found in this work imply that the enhancer activity of a human NCR can be modulated by environmental factors, a finding that has never been demonstrated in vivo until now. ​ ​

Environmental pollutants Previous results from the host laboratory demonstrated that the human NCR variant containing either the WT or the IBD-risk variant of the SNP rs17085007 drives gene expression in the ZF intestine under unchallenged conditions (see fig. 1) [36]. Hence, we stimulated the Tg(rs170085007_DR) reporter larvae and thereby investigated if ​ environmental factors could modulate the enhancer activity of the NCR in the intestine in ​ vivo. For this purpose, Tg(rs17085007_DR) larvae were exposed to PFASs and dietary sensor ​ ​ ​ ​ agonists, both of which have been proven to contribute to the pathogenesis of IBD in different levels [14]. Interestingly, the increased intestinal GFP area obtained from the WT-NCR after PFASs exposures suggests the WT-NCR to be a PFAS responsive element, while the IBD-risk NCR is not. Here, one hypothesis may be that the WT-NCR promotes the expression of protective genes against PFAS, in order to block tissue damage and potentially avoid intestinal inflammation. Since the IBD-risk NCR does not show the same pattern, it might be that this SNP actually does not enhance protection against intestinal inflammation, resulting in the initiation of IBD.

In order to further investigate how PFASs modulate the gene expression of Tg(rs17085007_DR), we wanted to analyze if the fluorescence pattern driven by PFASs ​ exposure is altered under inflammatory conditions. This was done by co-exposing the larvae to PFASs and TNBS, a chemical substance known to induce intestinal inflammation in ZF [34], and then comparing them with exposure to PFAS only. Notably, the results showed a significant decrease in the WT-NCR expression, represented by GFP fluorescence area, when treated with both TNBS and each of all three PFASs in the mid intestine. Since no significant

- 25 - M.Sc Thesis in Biotechnology Mikaela Westling ______results were found for the mutant-NCR, which showed no differences in the fluorescence area nor intensity when co-exposed to PFASs and TNBS, we hypothesize that the IBD-risk mutation in the NCR generates an impaired response to PFASs, that may contribute to IBD pathogenesis.

Activation of dietary sensors Furthermore, the impact in the expression of the NCR after activation of dietary sensors for diet-derived metabolites common in westernized countries was investigated. Among the dietary sensors tested, we found that PPARδ activation negatively regulates the expression of the WT-NCR, by means of reduced GFP expression area, while no differences were observed in the mutant-mCherry NCR. To confirm that all the agonists tested were functional, RT-qPCR can be done for target genes of the tested NRs that are known to be activated in zebrafish. This will exclude the possibility of a negative result as a consequence of a non-functional agonist. Regarding the observations after PPARδ agonist exposures, it has been reported previously that PPARδ activation ought to induce inflammation when bound to an agonist [30]. Besides, in the pilot experiment of co-exposures of PFASs with TNBS (see fig. 8), a similar pattern to PPARδ activation in the WT-NCR was observed when larvae were exposed to TNBS, known to induce intestinal inflammation [34]. These points suggest that the decreased levels of GFP in the WT-NCR are due to an increased intestinal inflammation in the larvae after PPARδ agonist treatment. However, since intestinal inflammation can trigger changes in the zebrafish intestine architecture [48, 49], further experiments need to be done to address if the phenotype observed is actually due to a direct action of the dietary sensor in the NCR activity or an indirect effect caused by potential intestinal inflammation.

Lastly, the mRNA expression data obtained by RT-qPCR showed deviating results compared to the protein levels measured by fluorescence. RT-qPCR is a robust technique for quantification of the mRNA levels due to its simplicity to set up and perform the analysis. Also, the quantification of small amounts of cDNA compose qPCR a favorable technique to utilize for this purpose [50]. However, since cDNA from the whole larvae was collected for analysis, it does not show what is happening specifically in the intestine. There might be a risk that the expression arising from the intestine is masked by the expression from other tissues as well. To solve this, the intestine can be dissected from the carcass and analyzed by qPCR separately, thus allowing to compare mRNA and protein expression levels in a more precise fashion. Another technique that could be utilized for this purpose is in situ ​ hybridization, in which the spatial information is captured. That allows us to visualize the

- 26 - M.Sc Thesis in Biotechnology Mikaela Westling ______precise place from where the expression is coming from, and analyze tissue-specific changes in the mRNA expression pattern of fluorescent proteins.

However, another possible explanation for these results may be that the relationship between the mRNA and the protein levels does not always correlate positively, meaning that the mRNA expression might decrease with time while the protein expression is still highly expressed [51]. Therefore, even though the mRNA and protein expression may differ, it does not indicate that the phenotypes visualized by fluorescence are incorrect. In order to validate the protein levels, other quantifications can be performed by using antibody-based detection techniques, such as ELISA and Western blot from whole larvae or from dissected intestines.

Limitations Although the Zebrafish organism contains many attributes as an in vivo model to study the ​ ​ combination of genetics and environmental factors in disease, including IBD, while preserving the physiological context, there are still limitations when using this model. Notably, the immune system of ZF differs from the human at the developmental stage used in this study, significantly due to the late emergence of the adaptive immune system, which is fully functional after 4-6 weeks post-fertilization [33]. Consequently, the ZF larvae survive only on the innate immune system during this period of time. This may be a caveat when trying to translate the obtained results to humans. However, since the observed phenotypes come from intestinal cells, where the tissue composition is comparable to humans [52], the results presented in this work may have the potential to be translated to humans. Alternatively, another model that could have been used for these experiments is the mouse, where IBD is widely studied. However, the experimental settings would have taken more time, more research funds [4], and they would go against the 3R principles, which seek to minimize the use of animals and promote the use of alternative models, such as the zebrafish larvae. Working with mice would have required highly invasive methods that have a negative impact on animal welfare, since they do not possess the attribute of being transparent, and then the biological processes cannot be followed in real-time without intervention. Thus, the choice landed in using ZF.

An additional limitation in the dual reporter setting is that the fluorescence generated by the mCherry protein is lower than for GFP. This is a common problem for all red fluorescent proteins due to reduced phototoxicity and lower fluorescence backgrounds [53]. This suggests that the alterations in the fluorescence obtained from GFP are not visible in mCherry due to the lower appearance of fluorescence. Nevertheless, this could be solved by the

- 27 - M.Sc Thesis in Biotechnology Mikaela Westling ______generation of a new reporter for the studied NCR, in which the vector containing the WT-NCR instead carries the mCherry protein, while the SNP carries the GFP. If the results appear to be reproducible, it would further improve the robustness of the model and one can conclude that the changes are not due to the fluorescence differences coming from the FP.

A third problem which arose when using Tg(rs17085007_DR) was the high variability that ​ ​ could be seen in the IBD-risk NCR. Beyond the lower fluorescence emitting from mCherry, the variability also constitutes a problem, making it hard to fully rely on these results. The expression patterns could be affected by several factors, such as the random integration site in the genome of the vectors containing the NCRs, as well as the number of copies that are integrated in the genome. This means that some ZF may contain single or multiple copies of the vectors containing the NCRs in the genome, resulting in higher variability in the fluorescence [54]. To address this problem, the transgenic line needs to be further crossed and tested to follow Mendelian inheritance, by counting the number of positive transgenic fish when crossing a homozygous individual with a WT (~50% of positive embryos will indicate Mendelian segregation). This will allow us to validate the current results in mCherry on a later generation which does not contain potential duplications of the plasmids injected.

Future aspects To date, many of the available systems have the possibility to investigate one IBD-risk environmental factor at a time, thus, missing the effect of the interaction of multiple factors, which resembles a real-world situation. In daily life, one is not only exposed to a single environmental factor, but a mixture of several of them simultaneously. If we manage to overcome the limitations with the ZF system proposed in this work, which allows the analysis of multiple genetic and environmental factors at the same time, we will have a very robust model with the ability to identify mixtures of environmental factors that can represent a risk factor, mimicking the complexity of the exposome. In addition, we will be able to decipher a function of different NCRs. The establishment of this dual reporter system in ZF in vivo will ​ have the potential to identify which compounds may modulate the function of IBD-risk NCRs, but also it opens the possibility to test different NCRs linked to other diseases, making it an excellent system to study mechanisms and potential therapeutic targets for human diseases.

- 28 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Conclusion The results of this work are summarized in fig. 13. In brief, we found that PFASs increase the activity of the NCR containing the WT variant of the SNP rs17085007, while the effect is lost in the presence of TNBS. In contrast, PPARδ activation inhibits the enhancer activity of the WT-NCR. These results suggest that the WT-NCR acts as a response element to PFASs, which function is lost in the context of inflammation.

For this study, we demonstrated that zebrafish is a suitable model to study genetic-environment risk factor interactions in the context of IBD. Specifically, we were able to screen for environmental factors that modulate the enhancer activity of human NCRs, which constitutes a novel and undescribed approach to further understand the interaction of IBD-risk factors in vivo. ​ ​

Figure 13. Proposed roles of the WT-NCR when exposed to PFASs and PPARδ ​ ​

- 29 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Acknowledgments I would like to express my deepest gratitude to my supervisor Associate Professor Eduardo Villablanca, who has not only provided me with the opportunity to perform my thesis at his laboratory with fantastic colleagues, but also for sharing his enthusiastic encouragement and enormous expertise in the area of immunology and research science. During these 20 weeks, I have learned more than I could have ever expected. I also want to acknowledge my supervisor Postdoctoral Rodrigo Morales, with whom I have had daily contact with. Not only has he guided me throughout the whole project, but he has also provided a huge amount of knowledge and patience when needed. Booking the qPCR machine and microscope, screening embryos, mounting larvae, taking the pictures, taught me how to make the analysis and interpret the results to only mention a fraction. He has also improved me in my thesis writing and presentation technique, which I will forever be very grateful for. In addition, I want to thank my supervisor based at KTH, Professor Torbjörn Gräslund, for giving helpful advice regarding the thesis structure and project plan during this period of time.

Main supervisor: Professor Torbjörn Gräslund PhD. Medical Protein Technology, School of Engineering Sciences in ​ Chemistry, Biotechnology and Health

Co-supervisors: Associate Professor Eduardo J. Villablanca, Immunology and Allergy division, Department of Medicine Postdoctoral Rodrigo Morales, Immunology and Allergy division, Department of Medicine

- 30 - M.Sc Thesis in Biotechnology Mikaela Westling ______

References

[1] Ng SC, Shi HY, Hamidi N, Underwood FE, Tang W, Benchimol EI, et al. Worldwide incidence and prevalence of inflammatory bowel disease in the 21st century: a systematic review of population-based studies. The Lancet. 2017 Dec 23;390(10114):2769–78.

[2] Kaplan GG. The global burden of IBD: from 2015 to 2025. Nat Rev Gastroenterol Hepatol. 2015 Dec;12(12):720–7.

[3] Wehkamp J, Götz M, Herrlinger K, Steurer W, Stange EF. Inflammatory Bowel Disease. Dtsch Arztebl Int. 2016 Feb 5;113(5):72–82.

[4] Loftus EV et al. of Inflammatory Bowel Disease. Gastroenterol Clin N Am. 2002. ​ 31(1):1–20.

[5] Verstockt B, Smith KG, Lee JC. Genome-wide association studies in Crohn’s disease: Past, present and future. Clin Transl Immunology. 2018;7(1):e1001.

[6] Huang H, Fang M, Jostins L, Umićević Mirkov M, Boucher G, Anderson CA, et al. Fine-mapping inflammatory bowel disease loci to single-variant resolution. Nature. 2017 Jul;547(7662):173–8.

[7] Zhen Y, Andolfatto P. Methods to Detect Selection on Noncoding DNA. Methods Mol Biol. 2012;856:141–59.

[8] Mirza AH, Kaur S, Brorsson CA, Pociot F. Effects of GWAS-Associated Genetic Variants on lncRNAs within IBD and T1D Candidate Loci. PLOS ONE. 2014 Aug;9(8):e105723.

[9] Dunham I, Kundaje A, Aldred SF, Collins PJ, Davis CA, Doyle F, et al. An integrated encyclopedia of DNA elements in the human genome. Nature. 2012 Sep;489(7414):57–74.

[10] Mokry M, Middendorp S, Wiegerinck CL, Witte M, Teunissen H, Meddens CA, et al. Many Inflammatory Bowel Disease Risk Loci Include Regions That Regulate Gene Expression in Immune Cells and the Intestinal Epithelium. Gastroenterology. 2014 Apr;146(4):1040–7.

[11] Meddens CA, van der List ACJ, Nieuwenhuis EES, Mokry M. Non-coding DNA in IBD: from sequence variation in DNA regulatory elements to novel therapeutic potential. Gut. 2019;68(5):928–41.

[12] Spisák S, Lawrenson K, Fu Y, Csabai I, Cottman RT, Seo J-H, et al. CAUSEL: an epigenome- and genome-editing pipeline for establishing function of noncoding GWAS variants. Nature Medicine. 2015 Nov;21(11):1357–63.

[13] Baumgart DC, Carding SR. Inflammatory bowel disease: cause and immunobiology. The Lancet. 2007 May 12;369(9573):1627–40.

- 31 - M.Sc Thesis in Biotechnology Mikaela Westling ______

[14] Fakhoury M, Negrulj R, Mooranian A, Al-Salami H. Inflammatory bowel disease: clinical aspects and treatments. J Inflamm Res. 2014 Jun 23;7:113–20.

[15] Knutsen HK, Alexander J, Barregård L, Bignami M, Brüschweiler B, Ceccatelli S, et al. Risk to human health related to the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food. EFSA Journal. 2018;16(12):e05194

[16] Lee YJ. Potential health effects of emerging environmental contaminants perfluoroalkyl compounds. Yeungnam Univ J Med. 2018 Dec 31;35(2):156–64.

[17] Keil DE. Immunotoxicity of Perfluoroalkylated Compounds. In: DeWitt JC, editor. Toxicological Effects of Perfluoroalkyl and Polyfluoroalkyl Substances. Cham: Springer International Publishing; 2015;239–48.

[18] Xu Y, Li Y, Scott K, Lindh CH, Jakobsson K, Fletcher T, et al. Inflammatory bowel disease and biomarkers of gut inflammation and permeability in a community with high exposure to perfluoroalkyl substances through drinking water. Environmental Research. 2020 Feb 1;181:108923.

[19] Suo C, Fan Z, Zhou L, Qiu J. Perfluorooctane sulfonate affects intestinal immunity against bacterial infection. Sci Rep. 2017 12;7(1):5166.

[20] Mazaira GI, Zgajnar NR, Lotufo CM, Daneri-Becerra C, Sivils JC, Soto OB, et al. The Field: A Historical Overview and Future Challenges. Nuclear Receptor Research. 2018 Jul 26;5.

[21] Weikum ER, Liu X, Ortlund EA. The nuclear receptor superfamily: A structural perspective. 2018 Nov 27(11):1876-92

[22] Jakobsson T, Vedin L-L, Hassan T, Venteclef N, Greco D, D’Amato M, et al. The oxysterol receptor LXRβ protects against DSS- and TNBS-induced colitis in mice. Mucosal Immunology. 2014 Nov;7(6):1416–28.

[23] Herold M, Breuer J, Hucke S, Knolle P, Schwab N, Wiendl H, et al. Liver X receptor activation promotes differentiation of regulatory T cells. PLoS One. 2017 Sep 19;12(9).

[24] Hurst RJM, De Caul A, Little MC, Kagechika H, Else KJ. The Retinoic Acid Receptor Agonist Am80 Increases Mucosal Inflammation in an IL-6 Dependent Manner During Trichuris muris Infection. J Clin Immunol. 2013 Nov;33(8):1386–94.

[25] Ross AC. Diet in Vitamin A Research. Methods Mol Biol. 2010;652.

[26] Klepsch V, Moschen AR, Tilg H, Baier G, Hermann-Kleiter N. Nuclear Receptors Regulate Intestinal Inflammation in the Context of IBD. Front Immunol. 2019 May 14;10.

- 32 - M.Sc Thesis in Biotechnology Mikaela Westling ______[27] Ananthakrishnan AN, Bernstein CN, Iliopoulos D, Macpherson A, Neurath MF, Ali RAR, et al. Environmental triggers in IBD: a review of progress and evidence. Nature Reviews Gastroenterology & Hepatology. 2018 Jan;15(1):39–49.

[28] Kidani Y, Bensinger SJ. LXR and PPAR as integrators of lipid homeostasis and immunity. Immunol Rev. 2012 Sep;249(1):72–83.

[29] Jiang Y, Li Q, Jia M, Yan Z. PPARδ: A Potential Therapeutic Target for the Treatment of Metabolic Hypertension. Int J Hypertens. 2019;2019:7809216.

[30] Liu Y, Colby JK, Zuo X, Jaoude J, Wei D, Shureiqi I. The Role of PPAR-δ in Metabolism, Inflammation, and Cancer: Many Characters of a Critical Transcription Factor. Int J Mol Sci. 2018 Oct 26;19(11)

[31] Denson LA, Curran M, McGovern DPB, Koltun WA, Duerr RH, Kim SC, et al. Challenges in IBD Research: Precision Medicine. Inflamm Bowel Dis. 2019 16;25(Suppl 2):S31–9.

[32] Abegunde AT, Muhammad BH, Bhatti O, Ali T. Environmental risk factors for inflammatory bowel diseases: Evidence based literature review. World J Gastroenterol. 2016 Jul 21;22(27):6296–317.

[33] Diaz OE, Morales RA, Das S, Villablanca EJ. Experimental Models of Intestinal Inflammation: Lessons from Mouse and Zebrafish. In: Hedin C, Rioux JD, D’Amato M, editors. Molecular Genetics of Inflammatory Bowel Disease. Cham: Springer International Publishing; 2019:47–76

[34] Oehlers SH, Flores MV, Okuda KS, Hall CJ, Crosier KE, Crosier PS. A chemical enterocolitis model in zebrafish larvae that is dependent on microbiota and responsive to pharmacological agents. Dev Dyn. 2011 Jan;240(1):288–98.

[35] Fisher S, Grice E, Vinton R, Bessling S, Mccallion A. Fisher, S., Grice, E.A., Vinton, R.M., Bessling, S.L. & McCallion, A.S. Conservation of RET regulatory function from human to zebrafish without sequence similarity. Science 312, 276-279. Science (, NY). 2006 May 1;312:276–9.

[36] Bathia S, Gordon CT, Foster RG, et al. Functional Assessment of Disease-Associated Regulatory Variants In Vivo Using a Versatile Dual Colour Transgenesis Strategy in Zebrafish. PLoS Genetics. 2015 June 1;11(10)

[37] Kawakami K. Tol2: a versatile gene transfer vector in vertebrates. Genome Biol. 2007;8(Suppl 1):S7.

[38] Asano K, Matsushita T, Umeno J, Hosono N, Takahashi A, Kawaguchi T, et al. A genome-wide association study identifies three new susceptibility loci for ulcerative colitis in the Japanese population. Nat Genet. 2009 Dec;41(12):1325–9.

- 33 - M.Sc Thesis in Biotechnology Mikaela Westling ______[39] Asano K, Esaki M, Umeno J, Hirano A, Maehata Y, Moriyama T, et al. Contribution of susceptibility variants at FCGR2A and 13q12 to the risk of relapse among Japanese patients with ulcerative colitis. J Gastroenterol. 2015 Nov 1;50(11):1094–102.

[40] Kabakchiev B, Silverberg MS. Expression quantitative trait loci analysis identifies associations between and gene expression in human intestine. Gastroenterology. 2013 Jun;144(7):1488–96, 1496.e1-3.

[41] Christou M, Kavaliauskis A, Ropstad E, Fraser TWK. DMSO effects larval zebrafish (Danio rerio) behavior, with additive and interaction effects when combined with positive controls. Science of The Total Environment. 2020 Mar 20;709:134490.

[42] Schlomann BH, Wiles TJ, Wall ES, Guillemin K, Parthasarathy R. Bacterial Cohesion Predicts Spatial Distribution in the Larval Zebrafish Intestine. Biophysical Journal. 2018 Dec 4;115(11):2271–7.

[43] McCurley AT, Callard GV. Characterization of housekeeping genes in zebrafish: male-female differences and effects of tissue type, developmental stage and chemical treatment. BMC Molecular Biology. 2008 Nov 12;9(1):102.

[44] Tang R, Dodd A, Lai D, McNabb WC, Love DR. Validation of Zebrafish (Danio rerio) Reference Genes for Quantitative Real-time RT-PCR Normalization. Acta Biochim Biophys Sin (Shanghai). 2007 May;39(5):384–90.

[45] Sasikumar AN, Perez WB, Kinzy TG. The Many Roles of the Eukaryotic Elongation Factor 1 Complex. Wiley Interdiscip Rev RNA. 2012 Jul;3(4):543–55.

[46] Rao X, Huang X, Zhou Z, Lin X. An improvement of the 2ˆ(–delta delta CT) method for quantitative real-time polymerase chain reaction data analysis. Biostat Bioinforma Biomath. 2013 Aug;3(3):71–85.

[47] Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes Text with EEA relevance. OJ L 276 33–79.

[48] Fleming A, Jankowski J, Goldsmith P. In vivo analysis of gut function and disease changes in a zebrafish larvae model of inflammatory bowel disease: a feasibility study. Inflamm Bowel Dis. 2010 Jul;16(7):1162–72.

[49] Oehlers SH, Flores MV, Hall CJ, Swift S, Crosier KE, Crosier PS. The inflammatory bowel disease (IBD) susceptibility genes NOD1 and NOD2 have conserved anti-bacterial roles in zebrafish. Disease Models & Mechanisms. 2011 Nov 1;4(6):832–41.

[50] Smith CJ, Osborn AM. Advantages and limitations of quantitative PCR (Q-PCR)-based approaches in microbial ecology. FEMS Microbiol Ecol. 2009 Jan 1;67(1):6–20.

- 34 - M.Sc Thesis in Biotechnology Mikaela Westling ______[51] Maier T, Güell M, Serrano L. Correlation of mRNA and protein in complex biological samples. FEBS Letters. 2009 Dec 17;583(24):3966–73.

[52] Brugman S. The zebrafish as a model to study intestinal inflammation. Dev Comp Immunol. 2016;64:82–92.

[53] Siegel AP, Baird MA, Davidson MW, Day RN. Strengths and Weaknesses of Recently Engineered Red Fluorescent Proteins Evaluated in Live Cells Using Fluorescence Correlation Spectroscopy. Int J Mol Sci. 2013 Oct 14;14(10):20340–58.

[54] Ung CY, Guo F, Zhang X, Zhu Z, Zhu S. Mosaic Zebrafish Transgenesis for Functional Genomic Analysis of Candidate Cooperative Genes in Tumor Pathogenesis. J Vis Exp. 2015 Mar 31;(97)

- 35 - M.Sc Thesis in Biotechnology Mikaela Westling ______Appendix

Figure A1. Molecular structures of the environmental pollutants PFOS, PFOA and PFHxS ​ ​

Figure A2. Macro for cropping images ​

- 36 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Figure A3. Macro for whole intestine analysis ​

- 37 - M.Sc Thesis in Biotechnology Mikaela Westling ______

Figure A4. Macro for intestine section analysis ​

- 38 - www.kth.se