PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/194290

Please be advised that this information was generated on 2019-06-02 and may be subject to change. The role of -32 in the increased cardiovascular risk associated with chronic inflammatory diseases

Michelle Susan Martina Adriana Damen “Op de kaft staat jouw linkerhand omdat die de afgelopen jaren zo hard heeft The role of interleukin-32 in the increased gewerkt; in de sportschool, in het lab, proostend met vrienden. Vanaf jongs af aan heb jij waarschijnlijk dingen opgepakt met die linker hand om er naar te kijken en cardiovascular risk associated with chronic het te onderzoeken. De harde lijnen in de hand symboliseren je kracht, niet alleen inflammatory diseases fysiek maar ook mentaal. Je hebt veel overwonnen in de afgelopen jaren en toch grijp je steeds nieuwe kansen aan, dat vind ik knap. De houding van de hand is wat zachter. Gewoon omdat je stiekem een heel lief en zacht persoon bent. De positie van de vingers vormen samen met de “cellen/eiwitten/moleculen” op de achterkant een hart, voor het cardiovasculaire onderwerp van het proefschrift.”

Copyright Pleun Hemelaar

Proefschrift ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de rector magnificus prof. dr. J.H.J.M. van Krieken, volgens besluit van het college van decanen in het openbaar te verdedigen op

woensdag 22 augustus 2018

om 12.30 uur precies colofon door ISBN: 978-94-92801-46-3 © Michelle Damen, 2018 Michelle Susan Martina Adriana Damen cover ontwerp: Pleun Hemelaar geboren op 23 maart 1990 layout: proefschrift-aio.nl druk: proefschrift-aio.nl te Eindhoven Promotoren: Prof. dr. Leo A.B. Joosten Prof. dr. Niels P. Riksen

Copromotor: Dr. Calin Popa

Manuscript commissie: Prof. dr. Gerard A.P.J.M. Rongen (voorzitter) Prof. dr. H.A.H. (Karin) Kaasjager (Universiteit Utrecht) Prof. dr. Angela H.E.M. Maas

“Hard work beats talent, when talent doesn’t work hard” Tim Notke CONTENTS

Chapter 1 General Introduction, aim and outline of the thesis 9 Adapted from: Interleukin-32 in chronic inflammatory conditions is associated with a higher risk of cardiovascular diseases. Atherosclerosis. 2017;264:83-91

Chapter 2 IL-32 promoter SNP rs4786370 predisposes to modified 25 lipoprotein profiles in patients with rheumatoid arthritis Scientific Reports 2017;7:41629

Chapter 3 Genetic variant rs4786370 of IL-32 is associated with the 45 ex vivo production in PBMCs treated with αTNF therapy in rheumatoid arthritis patients Submitted

Chapter 4 Human IL-32 transgenic mice develop adipokine profiles 65 resembling those of obesity-induced metabolic changes Manuscript in preparation

Chapter 5 Interleukin-32 upregulates the expression of ABCA1 79 and ABCG1 resulting in reduced intracellular lipid concentrations in primary human hepatocytes Atherosclerosis 2018, in press

Chapter 6 HIV-infected individuals on successful antiretroviral 109 treatment show differences in cardiovascular risk profile when stratified for IL-32 promotor SNP rs4786370 Submitted

Chapter 7 Summary, General Discussion & Future Perspectives 135

Chapter 8 Nederlandse samenvatting 151

Dankwoord 156 List of publications Curriculum Vitae 162 1

CHAPTER 1 General introduction, aim and outline of the thesis

Adapted from “Interleukin-32 in chronic inflammatory conditions is associated with a higher risk of cardiovascular diseases”. Atherosclerosis. 2017;264:83-91

Michelle S.M.A. Damen Calin D. Popa Mihai G. Netea Charles A. Dinarello Leo A.B. Joosten

8 9 Chapter 1 – General introduction, aim and outline of the thesis

GENERAL INTRODUCTION Nowadays, cardiovascular diseases (CVD) are the leading cause of mortality around the globe, leading to about 18 million deaths per year5. In most cases, the Cardiovascular diseases underlying pathological process of CVD is atherosclerosis and for the most part of 1 Ancient Egyptians already considered the heart as the central organ of the human the last century atherosclerosis was considered to be a cholesterol storage disease. body with hieroglyphic images of the heart showing various vessels attached. It was However, in the last 30 years a new concept regarding the role of local chronic thought that Imhotep, a prominent high priest (ca. 3000-2500 BC) was the first to in the onset and development of atherosclerosis and its complications describe that the pulse is associated with the heart1. Later (ca. 1555 BC), cardiac has gained awareness. rhythm disturbances and heart failure were described. Nevertheless, the first to discover the pulmonary circulation with a link between the heart and arteries, was Inflammation in cardiovascular diseases doctor Claudius Galenus from Pergamon (130-210 AD)2. Additionally, only in 1628, Nowadays, it is known that CVD risk and mortality are increased in patients suffering William Harvey (1578-1657) described the closed circulation of blood and the role from chronic inflammatory diseases such as rheumatoid arthritis (RA), inflammatory of the heart in pumping blood around which is depicted in Figure 1. Eventually, bowel disease (IBD) or human immunodeficiency virus (HIV) infection6-8. Traditional Leonardo da Vinci (1452-1512) was the first to describe arteriosclerosis. He observed risk factors such as hypertension, smoking, obesity and inadequate lifestyle, however, effects of age on the anatomy of vessels, studied the flow of fluids in an experimental do not fully account for the increased CVD risk in these patients9, 10. An increased setup and described the idea of nutrition and exercise on arterial wall thickening3. burden of atherosclerosis has been indicated to be responsible for the observed All of these aspects still represent important topics of research within the field of higher CV risk. Subsequently, common pathophysiological pathways to be shared cardiovascular (CV) diseases today. by atherosclerosis and for example RA or HIV have been hypothesized to explain the association of these chronic inflammatory disorders with CVD11. Atherosclerosis is seen as an active chronic inflammatory disease, in which immune cells and systemic markers of inflammation such as pro-inflammatory , including interleukin (IL)-1β, IL-6, TNFα and immune complexes play an important role in blood vessel pathology. In addition, the recent CANTOS trial was able to show and validate for the first time that reducing inflammation by reducing IL-1β using (150mg) significantly reduced the risk for cardiovascular events12. Furthermore, many cell types that are involved in both early and late-stage atherosclerosis also play a role in RA or HIV, including endothelial cells, and , B- and T cells.

Endothelial dysfunction occurs at sites where the endothelial cell (EC) layer is damaged or activated by smoking, hypertension and dyslipidemia. When damage occurs, endothelial cells are activated and start to express adhesion molecules, such as intercellular adhesion molecule 1 (ICAM-1), selectins and vascular cell adhesion molecule 1 (VCAM-1) on their cell surface. Expression of adhesion molecules is induced upon inflammation by pro-inflammatory mediators such as IL-1β, TNFα. Figure 1. Cardiovascular models over the course of time4. (A) Erasistratus’ model (B) Galen’s model (C) Colombo’s model (D) Harvey’s model the closed circulation. These factors together with the secretion of chemoattractant mediators including IL-8 and chemoattractant -1 (MCP-1) determine cell recruitment of monocytes and other inflammatory cells into the vascular wall13. Monocytes will differentiate into macrophages, which become foam cells by uptake of oxidized LDL (oxLDL) Additionally, plaque macrophages release more pro-inflammatory mediators (IL-1β, IL-8, TNFα) which attract more immune cells to the plaque

10 11 Chapter 1 – General introduction, aim and outline of the thesis

including T-cells (Figure 2). CD4+ T cells lacking CD28 are known to invade unstable In general, cholesterol is an important lipid in vertebrates, which is crucial in for atherosclerotic plaque regions resulting in weakening of the fibrous cap14. In example cellular membranes and for synthesis of steroid hormones, which have addition, these types of CD4+ T cells can secrete Th1 associated-cytokines and main physiological functions. Since cholesterol, cholesterol esters and triglycerides 1 possibly contribute to atherosclerotic damage15, 16. Besides, CD4+ T cells can kill are only minimally soluble in water they are transported in the bloodstream within endothelial cells in vitro, contributing to endothelial cell injury in for example lipoproteins. Several types of lipoproteins exist with differences in density depending coronary plaques17. As mentioned above, IL-1β was shown to be highly important in on the specific composition with in order of low to high density; chylomicrons, very- this process of inflammation after showing that inhibition of IL-1β by a human low density lipoprotein (VLDL), intermediate-denisty lipoprotein (IDL; also known monoclonal antibody canakinumab decreased the risk for myocardial infarction18. as VLDL-remnant), , low-denisty lipoprotein (LDL) and high-density lipoprotein This supports the thought of lowering inflammation by itself next to controlling lipid (HDL). Some particles are known to contribute to atherosclerosis such as LDL, concentrations during the development of CVD. VLDL and IDL whereas HDL is thought to have anti-atherogenic properties19. Traditionally, the “atherogenic lipid profile” is characterized by increased LDL and VLDL, but decreased HDL. Plaque macrophages cannot engulf native, unmodified LDL. However, after oxidation of LDL, monocytes/macrophages start to accumulate cholesterol20. Oxidation of LDL leads to increased binding affinity to multiple receptors on macrophages such as CD36 and Scavenger receptor A (SRA), increasing uptake and foam cell formation. LDL particles can substantially vary in size and recent studies showed that mostly small dense LDL particles are associated with CVD21. In contrast, HDLc plasma concentrations have been negatively correlated with CVD making HDLc anti-atherogenic. However, HDL is also heterogeneous and current treatment options trying to increase HDLc concentrations did not show any beneficial effects on atherosclerosis22. This finding led to an important new idea that HDL function might be more relevant than the actual concentration when it comes to cholesterol efflux and anti-atherogenic properties.

Besides HDLc, ATP-binding cassette subfamily A/G member 1 (ABCA1, ABCG1) and apolipoprotein A-1 (apoA1) are also involved in reverse cholesterol efflux and are other anti-atherogenic mediators. ApoA1 is the main particle of HDL and ABCA1 and ABCG1 are important transporters in macrophages, facilitating cholesterol efflux. Deficient or dysfunctional apoA1, ABCA1 and/or ABCG1 in mice and human studies, resulted in increased atherosclerosis23, 24. Figure 2. Schematic illustration of the pathophysiology of atherogenesis.

These lipid concentrations and functions can however be affected during Lipids and cardiovascular diseases inflammation. There are indications that HDL can be modulated in a way that Next to these above-mentioned mechanisms, other important risk factors such as it loses its anti-atherogenic properties and becomes another pro- atherogenic cholesterol and lipid composition have been described as players in atherosclerosis. lipid25, 26. In low grade chronic inflammatory diseases including RA and HIV, these Interestingly, these factors can be modulated by inflammatory molecules, as it lipid concentrations are known to be affected and can vary enormously depending would happen during prolonged periods of systemic inflammation. on stage of the disease or type of treatment. Interestingly, it can be concluded that both inflammation as well as cholesterol itself can affect the cardiovascular disease risk profile. Inflammation can directly affect cholesterol concentrations

12 13 Chapter 1 – General introduction, aim and outline of the thesis

and can modulate cholesterol composition and shift it towards a pro-atherogenic profile, showing the importance of trying to interfere with inflammation in chronic inflammatory diseases. 1

The discovery of IL-32 In 1992 Dahl et al. discovered a molecule that was called transcript 4 (NK4). expression of NK4 was increased when peripheral blood mononuclear cells (PBMCs), especially NK or -stimulated T cells were exposed to a high dose of interleukin-2 (IL-2)27, 28. Moreover, it was hypothesized that NK4 could play a role in cell adhesion since it contains an Arg-Gly-Asp (RGD) sequence which is important for mediating cell adhesion for a variety of proteins29. In 2005 Kim et al. showed that recombinant NK4 induced TNFα production in Raw 264.6 macrophages showing classical pro-inflammatory properties leading to renaming the molecule to interleukin-32 (IL-32)30. The IL32 gene resides on 16 p13.3 and consists of eight small exons. Besides , IL-32 is expressed in several other species like bovine and swine, but not in mice. Moreover, of IL-32 mRNA results in at least nine different isoforms: IL32γ, IL32β, IL32α, IL-32δ, IL-32ζ, IL- Figure 3. Schematic illustration of IL-32 signaling. 32ε, IL-32θ, IL-32η and IL-32sm, of which not all functions are known yet31, 32. No receptor for IL-32 has been discovered so far, although binding of IL-32 to proteinase IL-32 in inflammation 3 and integrins, αVβ3 and αVβ6 have been described31, 33. Additionally, recombinant IL-32 has been reported to be involved in a wide range of chronic inflammatory IL-32γ, in the presence of functional αVβ3, induced in vitro endothelial cell tube diseases such as, RA, IBD, COPD and HIV37-41. Of these diseases, RA is often described formation showing a role for IL-32 in angiogenesis34. IL-32 protein can be induced as the standard prototype autoimmune disease with cytokines such as IL-6 and by IFN-y. Moreover, IL-32 can stimulate pro-inflammatory mediators by activation TNFα playing an important role in disease pathogenesis. Therapeutic blockade of of signalling pathways nuclear factor –kB (NF-kB) and p38 mitogen-activated these cytokines or their receptors showed to improve disease activity, however it protein (MAP) kinase30. The intra-cellular nucleotide-binding oligomerization did not cure the disease. Therefore, new mediators and pathways involved in the domain (NOD)1 and NOD2 ligands can also synergize with IL-32 for IL-6 and IL-1β disease are still under investigation and could lead to new therapeutic targets. production through a caspase 1-dependent mechanism35. In line with this, silencing Interestingly, a few years ago research started to focus on IL-32 in RA pathogenesis. of endogenous IL-32 induces attenuated IL-1β, IL-6, IL8 and TNFα production in A clear correlation between the IL-32 protein expression in synovial tissue biopsies endothelial cells and monocytes36 (Figure 3). In general, it can be concluded that IL- of RA patients and the disease activity has been demonstrated42. Besides, both 32 acts as a pro- and an endogenous intracellular regulator of synovial tissue macrophages as well as synovial fibroblasts are known to produce cytokine production. Although the knowledge on the functions of IL-32 has expanded TNFα, which is a potent inducer of IL-32 and vice versa. Moreover, when RA patients significantly in the last decade, many aspects regarding the exact mechanism of were treated with anti-TNF therapy, significant lower levels of IL-32 were detected action in health and diseases still remain unknown. in synovial tissue biopsies, showing a link between TNFα and IL-32 in RA patients43. As a proof of concept, recombinant human IL-32 was injected into the joints of mice, resulting in joint swelling, cartilage damage and influx of pro-inflammatory cells. In case of TNFα deficient mice, these effects were remarkably decreased. Furthermore, the pro-inflammatory effects of IL-32 in human samples of synovial tissue biopsies and monocytes were observed and mRNA expression of IL-32γ was upregulated in specimens from RA patients compared to these from healthy individuals44. This all

14 15 Chapter 1 – General introduction, aim and outline of the thesis

shows that IL-32 is involved in RA pathogenesis and its chronic inflammatory state. IL-32 expression was detected in arterial vessel walls, endothelial cells and Next to this clear role for IL-32 in RA pathogenesis and the link with TNFα, IL-32 macrophages48, 49. Moreover, stimulation of Akt by pro-inflammatory cytokines has also been described to be involved in another chronic inflammatory disease, could strongly induce expression of IL-32 in endothelial cells49. Macrophages are 1 HIV. HIV patients showed increased expression levels of IL-32 mRNA and protein highly present in atherosclerotic lesions and show a highly increased mRNA and compared to healthy individuals. In addition, upon silencing of endogenous IL-32, protein expression of IL-32 when stimulated with pro- inflammatory mediators HIV replication increased tremendously. This suggests that endogenous IL-32 is including, -gamma (IFNγ) and a Toll-like receptor (TLR)3 ligand Poly I:C50, 51. highly important in suppressing HIV and just like in RA, could be an interesting new Moreover, macrophages are key players in controlling cholesterol levels in blood therapeutic target. vessel walls. They take up cholesterol, can turn into foam cells, and participate in the reverse cholesterol transport (RCT) by expressing functional ATP-binding cassette IL-32 in cardiovascular diseases transporter A1 (ABCA1) and ATP-binding cassette transport G1 (ABCG1)52. Since Since inflammatory pathways have been associated with the pathogenesis of macrophages are able to express high levels of IL-32 and in the same time actively chronic inflammatory diseases like RA, IBD, HIV, it is important to study the role of regulate circulating cholesterol levels, one could hypothesize that IL-32 may be these newly described pathways and the corresponding cytokines involved in the somehow related to cholesterol concentrations. Possibly, the intra-cellular capacities context of CVD and atherosclerosis. of IL-32 can interfere with the regulation of cholesterol transporters or foam cell formation and therefore influence circulating cholesterol levels. Despite previous The pro-inflammatory cytokine TNFα is seen as an important mediator in these latter work describing the IL-32 protein and its various isoforms, knowledge about genetic diseases. TNFα by itself is known to be associated with CVD and atherosclerosis45, 46. mutations or the function of each isoform is scarce. Recently, it was described that Furthermore, it can promote inflammation via induction of other cytokines such as, a single nucleotide polymorphism (SNP) in the promoter region of IL32 seems to IL-1β, IL-6 and importantly IL-32. Interestingly, IL-32 is capable of inducing TNFα, affect the expression of the isoforms in PBMCs. Both RA and HIV patients with this creating a positive inflammatory loop, which in turn can cause a non-resolving SNP showed differences in IL-32 isoform expression and in RA patients higher IL- inflammatory response contributing to a pro-atherogenic environment. In this way 32 protein concentrations were observed. Intriguingly, it was demonstrated that IL-32 by itself and via induction of other pro-inflammatory cytokines (IL-1β, IL-6, IL- this promoter SNP is associated with higher levels of HDLc53. This mutation in the 8, TNFα) can maintain an inflammatory response. IL32 gene could therefore contribute to a lower risk for CVD in patients with chronic inflammatory diseases like RA and HIV in which IL-32 is known to play a role. From IL-1β it is known that it plays a role in atherogenesis by acting on various aspects involved in vascular inflammation. The endothelium is the primary target of IL-1β triggering systemic inflammation. Additionally, IL-1β sets off the production of AIM AND OUTLINE OF THIS THESIS

IL-6, inhibitors of fibrinolysis, prostaglandin E2 (PGE2) and adhesion molecules in the endothelium, all of them participating in pathological conditions47. Additionally, IL- IL-32 is a pro-inflammatory cytokine involved in various immunological processes. 1β is the master cytokine of innate immunity and affecting its expression to control The scope of thesis was to investigate the mechanisms and possible therapeutic atherogenesis has been an important topic of research. Interestingly, the CANTOS potential of IL-32 in inflammation and more specifically the role of IL-32 in the trial was able to show that controlling IL-1β by canakinumab (150mg) resulted in development of cardiovascular diseases in chronic inflammatory diseases (Fig. 4). decreased hsCRP and IL-6 levels without affecting lipid concentrations, lowering the risk for new myocardial infarctions12, 18. Besides the effects of canakinumab on In Chapter 2 of this thesis, we studied whether the IL-32 promoter SNP rs4786370 IL-1β as shown by the CANTOS trial, an unexpected role for IL-32 as driver of the has a functional effect in RA patients and individuals at risk for cardiovascular immune responses via IL-1β was found in another study. When IL-32 levels were diseases. In these subjects, we characterized the lipid profile focusing mostly on decreased by siRNA, the pro-coagulant, pro-inflammatory and cytokine effects of IL- HDL concentrations and stratified these to the different genotypes. 1β, like IL-1β- induced ICAM-1 production, were reduced remarkably36. Therefore, IL- 32 is hypothesized to also be an important player in atherosclerosis. In line with this,

16 17 Chapter 1 – General introduction, aim and outline of the thesis

In Chapter 3, we explored in more detail the role of the IL-32 promoter SNP in RA differences were observed in cholesterol transporter mRNA expression of ABCA1/ patients and their disease activity. By using disease activity scores (DAS28-CRP) of ABCG1 also dependent on the promoter SNP. Finally, we studied the expression of RA patients from the RA BIOTOP cohort we were able to correlate the genotypes calnexin mRNA dependent on the promoter SNP trying to elucidate the mechanism 1 to 6 months follow-up DAS28-CRP scores to anti-TNF treatment (etanercept or behind the IL-32 SNP dependent regulation of cholesterol transporter expression. adalimumab). We showed a possible predictive role for the IL-32 promoter SNP on ex vivo response to anti- TNF treatment in clinical responders to either etanercept In Chapter 7, we integrate the insights we have gained into the role of IL-32 in the or adalimumab. mechanism behind inflammation-induced cardiovascular disease, and speculate on the potential of IL-32 as a new therapeutic agent in this mechanism to try and In Chapter 4, an in vivo model of human IL32 transgenic mice were studied to explore improve treatment of patients suffering from a chronic inflammatory disease with the role of IL-32 on lipid metabolism and diabetes. Moreover, the effect of IL-32 an increased risk for CVD. Moreover, we propose a hypothesis of IL-32 being more presence in mice on their adipose tissue phenotype and possible inflammatory state like a novel transcription factor than a cytokine as it is described to this moment. was studied. We exposed IL32 transgenic mice to normal chow diet and characterized their IL-32-induced phenotype by measuring ex vivo cytokine production, circulating plasma cytokines, adipose tissue cell size and inflammatory markers. Interestingly, adipose tissue cell size was significantly increased in IL32 transgenic mice. In line with this, circulating concentrations of leptin were increased in transgenic mice compared to wild-type mice. Since all mice were only studied under normal chow conditions, these findings suggest IL-32 plays an important role in lipid metabolism. These observations indicated that it would be of high importance to explore the effect of a high fat diet regimen in IL-32tg mice.

In Chapter 5, we tried to unravel part of the mechanism behind the effect of IL- 32 on lipoprotein metabolism/ regulation of HDLc concentrations. We studied the expression of the three main isoforms of IL-32, IL-32α, IL- 32β and IL-32γ and cholesterol transporters ABCA1 and ABCG1 in primary liver cells and the HepG2 liver carcinogenic cell line. Furthermore, overexpression and silencing of IL-32 was performed to explore the effects on ABCA1 and ABCG1 expression. In both cell types, IL-32γ significantly correlated with ABCA1, ABCG1 and LXRα expression. Additionally, intracellular lipid content was decreased in HepG2 cells overexpressing IL-32 isoforms.

In Chapter 6, we studied the role of IL-32 and the possibility of IL-32 and the IL- 32 promoter SNP to affect inflammatory responses and lipoprotein metabolism in PBMCs from HIV-infected versus healthy individuals. IL-32 isoform mRNA expression was different in HIV patients compared to healthy individuals (in)dependent of the promoter SNP. HIV-infected showed more IL-32β mRNA expression in PBMCs after poly I:C stimulation compared to healthy individuals. Moreover, IL-32γ mRNA expression was higher in HIV-infected when bearing the TT genotype whereas healthy individuals showed more IL-32γ when bearing the CC genotype. Moreover,

18 19 Chapter 1 – General introduction, aim and outline of the thesis

REFERENCES 1 1. Willerson JT, Teaff R. Egyptian contributions to cardiovascular medicine. Tex Heart Inst J. 1996;23(3):191- 200. 2. Sternbach GL, Varon J, Fromm RE, Sicuro M, Baskett PJ. Galen and the origins of artificial ventilation, the arteries and the pulse. Resuscitation. 2001;49(2):119-22. 3. ElMaghawry M, Zanatta A, Zampieri F. The discovery of pulmonary circulation: From Imhotep to William Harvey. Glob Cardiol Sci Pract. 2014;2014(2):103-16. 4. Aird WC. Discovery of the cardiovascular system: from Galen to William Harvey. J Thromb Haemost. 2011;9 Suppl 1:118-29. 5. centre WM. Cardiovascular diseases (CVDs) Fact sheet. UPDATED 2017 ed. WHO webpage: WHO; 2017. 6. Teague H, Mehta NN. The Link Between Inflammatory Disorders and Coronary Heart Disease: a Look at Recent Studies and Novel Drugs in Development. Curr Atheroscler Rep. 2016;18(1):3. 7. Rungoe C, Nyboe Andersen N, Jess T. Inflammatory bowel disease and risk of coronary heart disease. Trends Cardiovasc Med. 2015;25(8):699-704. 8. Prasad M, Hermann J, Gabriel SE, Weyand CM, Mulvagh S, Mankad R, et al. Cardiorheumatology: cardiac involvement in systemic rheumatic disease. Nat Rev Cardiol. 2015;12(3):168-76. 9. Mirjafari H, Al-Husain A, Bruce IN. Cardiovascular risk factors in inflammatory arthritis. Curr Opin Lipidol. 2011;22(4):296-301. 10. Amaya-Amaya J, Sarmiento-Monroy JC, Mantilla RD, Pineda-Tamayo R, Rojas-Villarraga A, Anaya JM. Novel risk factors for cardiovascular disease in rheumatoid arthritis. Immunol Res. 2013;56(2-3):267-86. 11. Kelesidis T, Currier JS. Dyslipidemia and cardiovascular risk in human immunodeficiency virus . Endocrinol Metab Clin North Am. 2014;43(3):665-84. 12. Weber C, von Hundelshausen P. CANTOS Trial Validates the Inflammatory Pathogenesis of Atherosclerosis: Setting the Stage for a New Chapter in Therapeutic Targeting. Circ Res. 2017;121(10):1119-21. 13. Golia E, Limongelli G, Natale F, Fimiani F, Maddaloni V, Pariggiano I, et al. Inflammation and cardiovascular disease: from pathogenesis to therapeutic target. Curr Atheroscler Rep. 2014;16(9):435. 14. Liuzzo G, Goronzy JJ, Yang H, Kopecky SL, Holmes DR, Frye RL, et al. Monoclonal T-cell proliferation and plaque instability in acute coronary syndromes. Circulation. 2000;101(25):2883-8. 15. Liuzzo G, Kopecky SL, Frye RL, O’Fallon WM, Maseri A, Goronzy JJ, et al. Perturbation of the T-cell repertoire in patients with unstable angina. Circulation. 1999;100(21):2135-9. 16. Dumitriu IE. The life (and death) of CD4+ CD28(null) T cells in inflammatory diseases. Immunology. 2015;146(2):185-93. 17. Nakajima T, Schulte S, Warrington KJ, Kopecky SL, Frye RL, Goronzy JJ, et al. T-cell-mediated lysis of endothelial cells in acute coronary syndromes. Circulation. 2002;105(5):570-5. 18. Ridker PM, Everett BM, Thuren T, MacFadyen JG, Chang WH, Ballantyne C, et al. Antiinflammatory Therapy with Canakinumab for Atherosclerotic Disease. N Engl J Med. 2017;377(12):1119-31. 19. Carmena R, Duriez P, Fruchart JC. Atherogenic lipoprotein particles in atherosclerosis. Circulation. 2004;109(23 Suppl 1):III2-7. 20. Boullier A, Bird DA, Chang MK, Dennis EA, Friedman P, Gillotte-Taylor K, et al. Scavenger receptors, oxidized LDL, and atherosclerosis. Ann Ny Acad Sci. 2001;947:214-23. 21. Hoogeveen RC, Gaubatz JW, Sun WS, Dodge RC, Crosby JR, Jiang J, et al. Small Dense Low-Density Lipoprotein-Cholesterol Concentrations Predict Risk for Coronary Heart Disease The Atherosclerosis Risk in Communities ( ARIC) Study. Arterioscl Throm Vas. 2014;34(5):1069-77. 22. Barter PJ, Caulfield M, Eriksson M, Grundy SM, Kastelein JJP, Komajda M, et al. Effects of torcetrapib in patients at high risk for coronary events. New Engl J Med. 2007;357(21):2109-22. 23. Aiello RJ, Brees D, Bourassa PA, Royer L, Lindsey S, Coskran T, et al. Increased atherosclerosis in hyperlipidemic mice with inactivation of ABCA1 in macrophages. Arterioscl Throm Vas. 2002;22(4):630-7. Figure 4. Schematic overview of thesis.

20 21 Chapter 1 – General introduction, aim and outline of the thesis

24. Rached F, Santos RD, Camont L, Miname MH, Lhomme M, Dauteuille C, et al. Defective functionality of 47. Ridker PM, Rifai N, Pfeffer M, Sacks F, Lepage S, Braunwald E. Elevation of - HDL particles in familial apoA-I deficiency: relevance of alterations in HDL lipidome and proteome. J alpha and increased risk of recurrent coronary events after myocardial infarction. Circulation. Lipid Res. 2014;55(12):2509-20. 2000;101(18):2149-53. 25. Wang Y, Ji L, Jiang R, Zheng L, Liu D. Oxidized high-density lipoprotein induces the proliferation and 48. Dinarello CA. Blocking IL-1 in systemic inflammation. J Exp Med. 2005;201(9):1355-9. 1 migration of vascular smooth muscle cells by promoting the production of ROS. J Atheroscler Thromb. 49. Heinhuis B, Popa CD, van Tits BL, Kim SH, Zeeuwen PL, van den Berg WB, et al. Towards a role of 2014;21(3):204-16. interleukin-32 in atherosclerosis. Cytokine. 2013;64(1):433-40. 26. Van Lenten BJ, Hama SY, de Beer FC, Stafforini DM, McIntyre TM, Prescott SM, et al. Anti- inflammatory 50. Kobayashi H, Lin PC. Molecular characterization of IL-32 in human endothelial cells. Cytokine. HDL becomes pro-inflammatory during the acute phase response. Loss of protective effect ofHDL 2009;46(3):351-8. against LDL oxidation in aortic wall cell cocultures. J Clin Invest. 1995;96(6):2758-67. 51. Ahmad U, Ali R, Lebastchi AH, Qin L, Lo SF, Yakimov AO, et al. IFN-gamma primes intact human coronary 27. Panelli MC, Wang E, Phan G, Puhlmann M, Miller L, Ohnmacht GA, et al. Gene-expression profiling arteries and cultured coronary smooth muscle cells to double-stranded RNA- and self-RNA-induced of the response of peripheral blood mononuclear cells and melanoma metastases to systemic IL-2 inflammatory responses by upregulating TLR3 and melanoma differentiation-associated gene 5.J administration. Genome Biol. 2002;3(7):RESEARCH0035. Immunol. 2010;185(2):1283-94. 28. Dahl CA, Schall RP, He HL, Cairns JS. Identification of a novel gene expressed in activated natural killer 52. Roelofs MF, Joosten LA, Abdollahi-Roodsaz S, van Lieshout AW, Sprong T, van den Hoogen FH, et al. The cells and T cells. J Immunol. 1992;148(2):597-603. expression of toll-like receptors 3 and 7 in rheumatoid arthritis synovium is increased and costimulation 29. Ruoslahti E, Pierschbacher MD. Arg-Gly-Asp: a versatile cell recognition signal. Cell. 1986;44(4):517-8. of toll-like receptors 3, 4, and 7/8 results in synergistic cytokine production by dendritic cells. Arthritis 30. Kim SH, Han SY, Azam T, Yoon DY, Dinarello CA. Interleukin-32: a cytokine and inducer of TNFalpha. Rheum. 2005;52(8):2313-22. Immunity. 2005;22(1):131-42. 53. Rosenson RS, Brewer HB, Jr., Ansell BJ, Barter P, Chapman MJ, Heinecke JW, et al. Dysfunctional HDL and 31. Dinarello CA, Kim SH. IL-32, a novel cytokine with a possible role in disease. Ann Rheum Dis. 2006;65 atherosclerotic cardiovascular disease. Nat Rev Cardiol. 2016;13(1):48-60. Suppl 3:iii61-4. 54. Damen MS, Agca R, Holewijn S, de Graaf J, Dos Santos JC, van Riel PL, et al. IL-32 promoter SNP rs4786370 32. Hong JT, Son DJ, Lee CK, Yoon DY, Lee DH, Park MH. Interleukin 32, inflammation and . Pharmacol predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis. Sci Rep. 2017;7:41629. Ther. 2017;174:127-37. 33. Heinhuis B, Koenders MI, van den Berg WB, Netea MG, Dinarello CA, Joosten LA. Interleukin 32 (IL- 34. 32) contains a typical alpha-helix bundle structure that resembles focal adhesion targeting region of focal adhesion kinase-1. J Biol Chem. 2012;287(8):5733-43. 35. Nold-Petry CA, Rudloff I, Baumer Y, Ruvo M, Marasco D, Botti P, et al. IL-32 promotes angiogenesis. J Immunol. 2014;192(2):589-602. 36. Netea MG, Azam T, Ferwerda G, Girardin SE, Walsh M, Park JS, et al. IL-32 synergizes with nucleotide oligomerization domain (NOD) 1 and NOD2 ligands for IL-1beta and IL-6 production through a caspase 1- dependent mechanism. Proc Natl Acad Sci U S A. 2005;102(45):16309-14. 37. Nold-Petry CA, Nold MF, Zepp JA, Kim SH, Voelkel NF, Dinarello CA. IL-32-dependent effects of IL- 1beta on endothelial cell functions. Proc Natl Acad Sci U S A. 2009;106(10):3883-8. 38. Kim S. Interleukin-32 in inflammatory autoimmune diseases. Immune Netw. 2014;14(3):123-7. 39. Calabrese F, Baraldo S, Bazzan E, Lunardi F, Rea F, Maestrelli P, et al. IL-32, a novel pro-inflammatory cytokine in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2008;178(9):894-901. 40. Smith AJ, Toledo CM, Wietgrefe SW, Duan L, Schacker TW, Reilly CS, et al. The immunosuppressive role of IL-32 in lymphatic tissue during HIV-1 infection. J Immunol. 2011;186(11):6576-84. 41. Du Y, Wang W, Yang W, He B. Interleukin-32, not reduced by salmeterol/fluticasone propionate in smokers with chronic obstructive pulmonary disease. Chin Med J (Engl). 2014;127(9):1613-8. 42. Gasiuniene E, Lavinskiene S, Sakalauskas R, Sitkauskiene B. Levels of IL-32 in Serum, Induced Sputum Supernatant, and Bronchial Lavage Fluid of Patients with Chronic Obstructive Pulmonary Disease. COPD. 2016;13(5):569-75. 43. Joosten LA, Netea MG, Kim SH, Yoon DY, Oppers-Walgreen B, Radstake TR, et al. IL-32, a pro- inflammatory cytokine in rheumatoid arthritis. Proc Natl Acad Sci U S A. 2006;103(9):3298-303. 44. Heinhuis B, Koenders MI, van Riel PL, van de Loo FA, Dinarello CA, Netea MG, et al. Tumour necrosis factor alpha-driven IL-32 expression in rheumatoid arthritis synovial tissue amplifies an inflammatory cascade. Ann Rheum Dis. 2011;70(4):660-7. 45. Choi JD, Bae SY, Hong JW, Azam T, Dinarello CA, Her E, et al. Identification of the most active interleukin-32 isoform. Immunology. 2009;126(4):535-42. 46. Ohta H, Wada H, Niwa T, Kirii H, Iwamoto N, Fujii H, et al. Disruption of tumor necrosis factor-alpha gene diminishes the development of atherosclerosis in ApoE-deficient mice. Atherosclerosis. 2005;180(1):11-7.

22 23 2 CHAPTER 2 IL-32 promoter SNP rs4786370 predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis

Scientific Reports 2017;7:41629

*Michelle S.M.A. Damen* Rabia Agca* Suzanne Holewijn Jacqueline de Graaf Jéssica C. Dos Santos Piet L. van Riel Jaap Fransen Marieke J.H. Coenen Mike T. Nurmohamed Mihai G. Netea Charles A. Dinarello Leo A.B. Joosten** Bas Heinhuis** Calin D. Pop**

*Authors share first authorship, **Authors share senior authorship

24 25 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

ABSTRACT INTRODUCTION

Patients with rheumatoid arthritis (RA) are at higher risk of developing In patients with rheumatoid arthritis (RA), cardiovascular disease (CVD) represents cardiovascular diseases (CVD). Interleukin (IL)-32 has previously been shown to the leading cause of death1,2. Various studies have demonstrated that besides be involved in the pathogenesis of RA and might be linked to the development of behavioral risk factors and dyslipidemia, inflammation also plays a crucial role in atherosclerosis. However, the exact mechanism linking IL-32 to CVD still needs to the increased risk for CVD3. Additionally, inflammatory pathways in RA but also be elucidated. The influence of a functional genetic variant of IL-32 on lipid profiles other chronic inflammatory diseases, including psoriasis, have been proposed to 2 and CVD risk was therefore studied in whole blood from individuals from the NBS accelerate atherogenesis contributing to the increased CVD risk 4-6. These patients cohort and RA patients from 2 independent cohorts. Lipid profiles were matched are continuously exposed to varying levels of inflammatory mediators (e.g. to the specific IL-32 genotypes. Allelic distribution was similar in all three groups. cytokines) that may alter traditional CVD risk factors, including the lipid pattern, Interestingly, significantly higher levels of high density lipoprotein cholesterol both at the concentration and composition level7,8. Normally, a pro-atherogenic (HDLc) were observed in individuals from the NBS cohort and RA patients from the lipid profile consists of an increased total cholesterol (TC), low-density lipoprotein Nijmegen cohort homozygous for the C allele (p=0.0141 and p=0.0314 respectively). cholesterol (LDLc), triglycerides (TG), and a decreased high-density lipoprotein In contrast, the CC-genotype was associated with elevated low density lipoprotein cholesterol (HDLc). However, in RA patients the lipid profile varies throughout cholesterol (LDLc) and total cholesterol (TC) in individuals at higher risk for CVD different stages of the disease9,10. Particularly during active disease, these patients (plaque positive) (p=0.0396; p=0.0363 respectively). Our study shows a functional have low TC and LDLc levels, while their CVD risk is increased. Hence, due to the effect of a promoter single-nucleotide polymorphism (SNP) in IL32 on lipid profiles changeability of inflammatory activity and anti-inflammatory medication, the in RA patients and individuals, suggesting a possible protective role of this SNP individual lipid profiles may frequently fluctuate during the course of disease against CVD. making it hard to draw conclusions about the impact of these changes on CVD risk11. Of all lipids, HDLc is the most susceptible to inflammatory changes in terms of both concentration as well as composition12,13 14. In line with this, it was previously shown that HDLc becomes less anti-atherogenic or even pro-atherogenic in RA patients with an increased inflammatory status11.

Recently, IL-32 has been demonstrated to be an important key modulator of inflammation in RA15. In a previous study from our group, IL-32 was found to be highly expressed in synovial tissues from patients with moderate and severe rheumatoid arthritis and it was strongly correlated with the severity of joint inflammation. IL-32 can be induced by TNF and can on its own further potentiate TNF expression16,17. Given this fact and the well-known roles of TNF in both RA and atherosclerosis, IL-32 was recently proposed to contribute to the development of atherosclerotic plaques18-20. In 2009 Dinarello et al described IL-32 as a critical regulator of endothelial cell function, possibly promoting atherosclerosis by potentiating IL-1β-induced ICAM-1 and by producing pro-inflammatory cytokines in these cells21. This pro-atherogenic role of IL-32 was further supported by a recent report, which showed enhanced IL-32 expression in atherosclerotic arterial vessel walls22. Additionally, IL-32 was found to be expressed by macrophages, with highly increased expression after stimulation of these cells with pro-inflammatory components that were previously appointed to be involved in atherosclerosis (e.g. toll-like receptor (TLR) 3 agonist Poly I:C

26 27 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

and interferon-gamma (IFNγ))23-25. Moreover, macrophages are known to play an Male:#114 RA1: Female:#155; Male:#137 and RA2: Female:#232; Male:#121). Only important role in controlling cholesterol levels in blood vessel walls as they engulf within the RA2 cohort there was a larger difference in number of females to males, cholesterol, become foam cells, and take part in reverse-cholesterol transport nevertheless in none of the cohorts gender affected the HDLc (data not shown). (RCT)26,27. Knowing that IL-32 expression can be highly induced in these cells upon Moreover, individuals carrying one C-allele show a small increase in HDLc levels. inflammation, one can argue a role for IL-32 in cholesterol metabolism. Despite Concluding, the C-allele for the IL-32 promoter SNP is linked to an increase in HDLc the suggested role of IL-32 in inflammation, CVD and disease progression in RA, levels independent of gender or having RA. studies investigating the IL-32 protein function and IL-32 gene polymorphisms with 2 respect to these outcomes in RA are scarce22. However, our group recently showed A that a single-nucleotide polymorphism (SNP) in the promoter region of the IL-32 gene seemed to be associated with lower basal expression of IL-32β in peripheral CCCCC blood mononuclear cells (PBMCs) of RA patients28. The same data showed lower pro-inflammatory cytokine production in PBMCs after stimulation with various 2 28 compounds in patients bearing the CC genotype . Additionally, various studies, 2 prter 2p including genome-wide association studies (GWAS) analyses, show effects of gene polymorphisms in for example TNF, CD247 and anti-cyclic citrullinated peptide B (anti-CCP) RA and CVD18,29-31. The present study therefore aims to investigate the etpe ree functional implications of a SNP in the IL-32 gene on the lipid profile and CVD risk in RA patients. CC C

RESULTS Genotype distribution of an IL32 promoter SNP is comparable between individuals from the NN cohort and RA patients The genotype distribution of the IL32 promoter SNP (rs4786370) (Fig. 1A) was compared between three different cohorts. One cohort consisted of individuals 2 r etpe from the NBS NIMA cohort (NN). The other two cohorts consisted of RA patients, one group of patients treated at the Radboudumc Nijmegen (RA1) and the other at the Reade Clinic in Amsterdam (RA2). No significant differences in genotype frequencies were observed between the three cohorts (Figure 1B). Fig. 1. The IL-32 promoter SNP. A. Location of the IL32 promoter SNP within the IL32 region on . B. Genotype frequencies of the IL32 rs4786370 promoter SNP in individuals from The IL32 promoter SNP affects HDLc levels in both individuals from the the NN cohort (NN; CC:19.2%, CT:45.7%, TT:35%), RA patients from the Radboudumc Nijmegen NN cohort and RA patients (RA1; CC:16.1%, CT:51%, TT:32.9%) and RA patients from the Reade clinic Amsterdam (RA2; CC:23%, CT:47.1%, TT:29.9%). Chi-square analysis (IBM SPSS Statistics v.22) showed no significant differences We studied the concentration of HDLc in each cohort because of its importance in in genotype distribution between the cohorts. development of cardiovascular disease. As shown in figure 2, in all three cohorts individuals having the CC genotype show higher levels of HDLc compared to either individuals with the CT and TT genotype with a significant increase in HDLc within the NN and RA1 cohort (p=0.0141 and p=0.0314 respectively). Further analysis was performed. focussing on gender differences within each cohort (NN: Female:#120;

28 29 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

HDL Cholesterol A B HDL Cholesterol

HDL Cholesterol

y

y

mmol/L 2

mmol/L

% of genotype frequenc mmol/L

Plaques% of genotype frequenc Plaques Plaques LDL Cholesterol Plaques Total Cholesterol C D LDL Cholesterol Total Cholesterol NN RA1 RA2 rs4786370 genotype

Fig. 2. HDL cholesterol levels. A HDL cholesterol levels stratified for the IL-32 promoter SNP

(rs4786370) genotype in individuals from the NN cohort (TT#82, CT:#107, CC:#45, NN cohort) and mmol/L mmol/L RA patients from two different cohorts (RA1; TT:#96, CT:#149, CC:#47 and RA2; TT:#104, CT:#164, CC:#80). A significant induction in HDL cholesterol was observed in both individuals from theNN mmol/L mmol/L cohort and RA patients of the RA1 cohort carrying the C allele (NN; p=0.0141,RA1; p=0.0314 and RA2; p=0.8450). Values are expressed as means ± SEM. p-values are calculated using Mann-Whitney U-test, Plaques Plaques *p < 0.05 Graphpad prism v5.03. rs4786370 genotype rs4786370 genotype Plaques Plaques rs4786370 genotype rs4786370 genotype Fig. 3 Lipoproteins and presence of plaques. A-D. Lipoproteins stratified for IL-32 promoter SNP HDLc levels are affected by the IL32 promoter SNP independent of the rs4786370 and the presence or absence of plaques in individuals from the NN cohort. Genotype presence of plaques frequency, HDL cholesterol, LDL cholesterol and total cholesterol concentrations were determined Within the NN-cohort we were able to stratify the genotype frequency and in these individuals. Number of individuals CC-:25, CC+:20, CT-:61, CT+:46, TT-:37, TT+:45. Chi-square analysis (IBM SPSS Statistics v.22) showed no significant differences in genotype distribution between cholesterol levels for the presence or absence of plaques detected by ultrasound the groups. Values are expressed as means ± SEM. p-values are calculated using Mann-Whitney U- (Fig. 3A-D). The percentage of individuals with the CC genotype seemed to be test, *p < 0.05 Graphpad prism v5.03. higher in the plaque negative group compared to the plaque positive group; 20% vs. 18% respectively although this did not reach statistical significance (Fig.3A). Additionally, individuals carrying the CC or CT genotype showed higher HDLc levels HDLc is linked to the IL32 promoter SNP and the prevalence of CVD than individuals with the TT genotype independent of having plaques. Nevertheless, events in RA patients individuals with the CT genotype and having a plaque showed a decrease in HDLc CVD is a common problem in RA and the composition of cholesterol levels plays a (CT- versus CT +: p<0.0004) (Fig.3B). crucial role herein. To determine if the IL-32 promoter SNP was involved in both these parameters, genotypes and cholesterol levels were determined in RA patients (RA1 Both LDLc and TC were not affected by the SNP in IL32 itself. However, higher and RA2 cohort) with versus without a history in CVD (Fig.4A-D). No differences were levels of both LDLc and TC were observed in individuals who were found positive for observed in genotype distribution between the two groups (Fig.4A). Even so, HDLc plaques (Fig. 3C,D). levels were lower in individuals with a history of CVD, reaching lowest concentrations in individuals with the TT genotype (TT- versus TT+ p=0.0293)(Fig.4B). In contrast, patients carrying the CC genotype showed the highest levels of HDLc as was also

30 31 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

observed within the individuals from the NN cohort (Fig 3B). Besides, lower levels DISCUSSION of LDLc and TC were observed in individuals with a history of CVD (Fig. 4C,D). This was completely opposite from what was found in individuals from the NN cohort The major novel findings of this study are that a promoter SNP in the IL32 gene as shown in figure 3C,D. Overall, patients with a history of CVD and carrying the TT is equally distributed in individuals from the NN cohort as in RA patients, but at genotype had lowest HDLc, LDLc and TC (Fig.4B-D). the same time that this SNP causes an increase in HDLc in both groups. This effect doesn’t seem to be influenced by disease activity, the use of medication, the A B 2 HDL Cholesterol presence of plaques or a history of CVD (data not shown). In contrast, other lipid concentrations like LDLc and TC are not affected by the SNP but are linked with y HDL Cholesterol plaques or a history of CVD. This might further expand the knowledge about the

y underlying mechanisms for CVD in RA patients32-34.

mmol/L Although chronic inflammation is known to affect cholesterol concentration in the body resulting in lower HDLc and LDLc levels, some studies have not been able to

mmol/L % of genotype frequenc show the benefit of increasing HDLc34-38.

CVD% of genotype frequenc PREV CVD PREV Nevertheless, HDLc does play a role in lowering the risk for CVD, especially in CVD PREV 39,40 LDL Cholesterol CVD PREV Total Cholesterol individuals with chronic inflammatory diseases like RA . In RA there seems to be a C D tight relationship between the levels of HDLc, LDLc and disease activity, with lower LDL Cholesterol Total Cholesterol levels during periods of active disease41-43. Even though this might speak against an increased risk for CVD, it is more the ratio between the levels of these different lipids (and probably their composition/function) that matters when assessing CVD risk. The fact that HDLc is decreased to a greater extent than the TC in these patients, results in a higher atherogenic index (the ratio between TC and HDLc) and therefore mmol/L mmol/L an increased CVD risk37,44. In addition, HDLc is less capable of exercising its anti- 8 mmol/L mmol/L atherogenic functions if inflammation levels are still uncontrolled . Conversely, when RA patients receive standard treatment including DMARDs and biological CVD PREV CVD PREV agents like anti-TNF therapy, cholesterol levels might increase, which correlates 44,45 CVD PREV CVD PREV with the level of suppression of their disease activity . In our study, no differences were observed in disease activity between genotypes concluding that the variation in HDLc was not caused by difference in disease activity (Supplementary Fig. Fig. 4 Lipoproteins and CVD events. A-D. Lipoproteins stratified for IL-32 promoter SNP rs4786370 and the presence or absence of an previously described CVD. Allele frequency, HDL Cholesterol, S1A-B). Nevertheless, CVD is still the number one cause of death in RA, showing the LDL Cholesterol and Total Cholesterol concentrations were determined in patients with RA from importance of exploring how cholesterol metabolism and regulation is affected in the RA2 cohort (Reade center Amsterdam). Number of individuals: CC-:63(37.5%), CC+: 17(21.1%), these patients. CT-:145(30.2%), CT+:18(, TT-:90 , TT+: 14. Chi-square analysis (IBM SPSS Statistics v.22) showed no significant differences in allelic distribution between the cohorts. Values are expressed as means ± SEM. p-values are calculated using Mann-Whitney U-test, *p < 0.05 Graphpad prism v5.03. Genetic factors can also play a role in the development of the accelerated atherogenesis observed in RA patients as was described previously. Some studies describe that several gene polymorphisms for example in TNFα and NFkB1 have been found associated with an increased risk of CVD or subclinical atherosclerosis in RA18,46,47. Moreover, GWAS studies show many different gene polymorphisms

32 33 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

related to RA, also finding new polymorphisms which were not yet described When taken together, we suggest a novel role for a genetic polymorphism in the in other studies like polymorphisms in IL6ST, IRF5 and PTPN229,48. In at least one promoter region of IL-32 causes an increase in HDLc levels in individuals from the study which performed GWAS analyses, IL-32 was present in the data which could NN cohort and RA patients. This might have functional effects leading to a lower suggest a role for IL-32 in the disease49. In contrast, other studies investigating gene prevalence of carotid artery plaques. This has not been investigated in the present polymorphisms in IL-6 for example, were not able to link these to an increased risk of study, but should be taken into account in future investigations. Our results lead subclinical atherosclerosis or CVD in RA when it was hypothesized that these could to the assumption that IL-32 is a previously unrecognized cytokine involved in be associated50,51. the process of inflammation and cholesterol metabolism. However, currently we 2 are still investigating the exact mechanism behind the role of IL-32 in cholesterol IL-32 has previously been described to play a role in the pathogenesis of RA and metabolism regulation. additionally it has been suggested that IL-32 plays a role in atherosclerosis, presumably contributing to the increased CVD risk in this population. As mentioned earlier, recent data showed that a SNP in the promoter region of the IL-32 gene seemed METHODS to be associated with lower basal expression of IL-32β and lower pro-inflammatory cytokine production in PBMCs of RA patients bearing the CC-genotype28. Therefore, Patient cohorts we hypothesized that there is also a genetic polymorphism in the promoter region From the total number of participants from the NBS NIMA study, only a subgroup of IL-32 that could affect and possibly lower the CVD risk in RA patients. To our of 234 participants, from whom the IL32 promoter SNP was determined, were knowledge, this is the first study demonstrating that a functional SNP in IL-32 is included in this study. The NBS is a population-based survey as described before53. linked to an increase in HDLc in RA patients and individuals from the NN cohort, Participants aged 50–70 yr were asked to visit the hospital to perform six non- suggesting IL-32 by itself can affect cholesterol metabolism. Individuals carrying the invasive measurements of atherosclerosis (NIMA) including pulse wave velocity CT genotype already showed higher HDLc levels compared to individuals bearing (PWV), augmentation index (Aix), intima-media thickness (IMT), plaque thickness, the TT genotype, while having the CC genotype results in even higher HDLc. This ankle-brachial index (ABI) at rest and after exercise. Additionally, fasting venous suggests that having one C-allele might already affect IL-32 expression in a way that blood samples were collected. All participants filled out a questionnaire about their results in higher HDL cholesterol. The increased HDLc concentrations in individuals previous history of vascular disease, medication use, smoking habits, and exercise bearing the C-allele might be due to lower levels of TNFα, which is known to suppress habits. Prevalent CVD was defined as a reported myocardial infarction (MI), transient cholesterol synthesis52. Another possible explanation could be that expression of a ischemic attack (TIA), stroke (CVA), peripheral arterial disease (PAD), coronary specific IL-32 isoform affects intracellular pathways resulting in for example higher artery bypass or angioplasty, or treated angina. The Medical Ethics Committee of cholesterol efflux or increased level of apoA-I (the main protein component of the Radboud university medical center, Nijmegen (nr.2010-397), The Netherlands HDLc). The SNP however only seems to affect HDLc since no observation was made approved the study protocol, and all participants provided written informed in individuals from the NN cohort or RA patients linking the SNP to lower or higher consent54. Experiments were conducted according to the principles expressed in the LDLc or TC. In our study, RA patients with a history of CVD had lower LDLc and TC Declaration of Helsinki. compared to patients without a history of CVD and also lower levels than individuals from the NN cohort. This might be explained by the fact that these RA patients Patients with RA who fulfilled the 1987 ACR criteria and/or the 2010 ACR/EULAR probably received statins after going through an event. classification criteria for RA were recruited from the Radboudumc in Nijmegen, The Netherlands. These patients underwent a screening program of their CVD risk factors Our study has several limitations. Most of them are related to the difficulties of between July 2011 and August 2012. Disease-related parameters, lipid profile and comparing the three heterogeneous cohorts. Various measurements like blood history of cardiovascular events were registered (Table 1). In addition, the Nijmegen pressure or underlying disease (like diabetes) were not notified for all patients or inception cohort database was checked for patients who had already been screened individuals within the cohorts making it impossible to correct for all these factors. previously. The Nijmegen inception cohort is a prospective study that started in 1985 which includes regular visits for disease related parameters and blood samples

34 35 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

in patients with RA. Eventually, 297 patients have been identified as participants DNA isolation and Taqman genotyping of both the inception cohort and the CVD screening program. These patients were Blood was obtained from 297 RA patients from the Radboudumc (RA1), 353 RA included in this study. The stored blood samples of the inception cohort were used patients of the Jan van Breemen Institute (Reade)(RA2) and 234 individuals from the for the determination of the SNP in the IL-32 gene. NBS NIMA study (NN). Genomic DNA was extracted from leukocytes in peripheral venous blood as previously described55. The samples were quantified and evaluated The CARdiovascular research and RhEumatoid arthritis (CARRÉ) study is an ongoing for purity (260/280-nm ratio) with a NanoDrop ND-1000 spectrophotometer (Thermo prospective cohort investigating cardiovascular (CV) disease and CV risk factors in 353 Scientific). The genetic variant in the IL32 promoter (rs4786370) polymorphism was 2 patients with RA (CCMO P 01.0408L, METC 0105). In 2000, a random sample was drawn determined using the TaqMan SNP assay C_27972515_20, (Thermofisher, Foster of patients registered at the Jan van Breemen Institute (now Reade) in Amsterdam, City, CA, USA). The TaqMan qPCR assays were performed on the AB StepOnePlus The Netherlands. Patients were eligible if they fulfilled the 1987 American College of polymerase chain reaction system (Applied Biosystems). Negative controls were Rheumatology (ACR) classification criteria, were diagnosed with RA between 1989 included in the assay. No duplicates were used. and 2001 and were aged 50 to 75 years. Patient enrollment was between 2001 and 2002 with follow up visits in 2004-2005 and 2010-2011. CV endpoints were defined as IMT measurements a verified medical history of coronary, cerebral or peripheral arterial disease (Table 1). Carotid IMT was determined using an AU5 Ultrasound machine (Esaote Biomedica, Genova, Italy) with a MHz linear-array transducer. Longitudinal images of the distal- Cohort NBS NIMA (NN) Radboudumc (RA1) Reade (RA2) most 10 mm of both the far wall and the near wall of both common carotid arteries N 234 297 353 (CCA) were obtained in the optimal projection. Actual measurement of the IMT Age, years 61 ± 6 60 ± 12 63 ± 8 was performed off-line by the sonographer at the time of the examination, using Female, no (%) 120 (51.34) 155 (52.2) 232 (65.7) semi-automatic edge-detection software (MAth®Std Version 2.0, Metris, Argenteuil, Disease duration, years n.a. 9 (3-17) 7 (4 – 10) France). Reproducibility of our IMT measurements as investigated by the method Rheumatoid factor positive, no (%) n.a. 202 (68.2) 256 (72.5) of Bland and Altman had been reported before: the mean (±S.E.M.) difference for Anti-citrullinated protein antibodies positive, no. (%) n.a. 160 (53.9) 187 (54.5) repeated measurements by the same observer was 0.003 ± 0.007 mm56,57. History of CVD, no (%) 57 (24.4) 63 (22.6) 51 (14.4) DAS28 n.a. 2.97 (1.18) 3.90 (1.35) Statistical analysis Diabetes, no (%) 14 (6.0) 18 (6.1) 17 (4.8) Normality was tested using the D’Agostino and Pearson omnibus normality test. Systolic blood pressure, mmHg 130 ± 17 132 ± 17 142 ± 20 Continuous variables are presented as mean and standard deviation (SD) or as Diastolic blood pressure, mmHg 78 ± 10 77 ± 10 81 ± 9 median followed by the interquartile range (IQR). Total cholesterol, mmol/L 5.92 ± 1.11 5.2 ± 1.13 5.77 ± 1.12 HDL cholesterol, mmol/L 1.39 ± 0.40 1.3 ± 0.34 1.46 ± 0.48 Categorical variables are presented as number followed by percentage. The LDL cholesterol, mmol/L 3.89 ± 0.98 3.1 ± 1.07 3.69 ± 1.03 differences between allele frequency and lipid concentrations measured in RA Triglycerides, mmol/L 1.29 (0.91-1.81) 1.53 (1.09-2.14) 1.32 ( 3.04 – 4.42) patients and individuals from the NN cohort were analyzed using the Mann-Whitney test. Chi-square test was used to test for differences between categorical variables. Table 1. Baseline characteristics of the cohorts. Values are presented as mean ± SD, median (IQR) or A p-value less than 0.05 was considered statistically significant (*p<0.05 and numbers (percentages). **p<0.01). Data was analyzed using GraphPad Prism v5.0. CVD: cardiovascular disease, DAS28: disease activity score 28, HDL: high density lipoprotein; LDL: low-density lipoprotein.

36 37 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

ACKNOWLEDGEMENTS REFERENCES

This research was supported by grants from the Dutch Arthritis Foundation (Nr.13- 1. de Campos, O. A. et al. Assessment of cardiovascular risk in patients with rheumatoid arthritis using the 03-302) and the Nijmegen Institute for Infection, Inflammation and Immunity (N4i), SCORE risk index. Revista brasileira de reumatologia 56, 138-144, doi:10.1016/j.rbre.2015.09.005 (2016). 2. Lopez-Mejias, R. et al. Cardiovascular risk assessment in patients with rheumatoid arthritis: The the Netherlands. relevance of clinical, genetic and serological markers. Autoimmunity reviews 15, 1013-1030, doi:10.1016/j. autrev.2016.07.026 (2016). Authors’ contributions 3. Ross, R. Atherosclerosis--an inflammatory disease. The New England journal of medicine 340, 115-126, 2 MSMAD and RA, MTN, BH, SH, MC and CP contributed to acquisition, analysis and doi:10.1056/NEJM199901143400207 (1999). 4. Coumbe, A. G., Pritzker, M. R. & Duprez, D. A. Cardiovascular risk and psoriasis: beyond the traditional interpretation of the data. CP, LJ, MTN, RA, MGN, CD, JdS, PvR, JdG and JF assisted risk factors. The American journal of medicine 127, 12-18, doi:10.1016/j.amjmed.2013.08.013 (2014). in study design and patient inclusion. MSMAD , RA and CP drafted the manuscript. 5. John, H., Toms, T. E. & Kitas, G. D. Rheumatoid arthritis: is it a coronary heart disease equivalent? All authors critically reviewed the manuscript. All authors read and approved the 6. Current opinion in cardiology 26, 327-333, doi:10.1097/HCO.0b013e32834703b5 (2011). final manuscript. 7. Solomon, D. H. et al. Explaining the cardiovascular risk associated with rheumatoid arthritis: traditional risk factors versus markers of rheumatoid arthritis severity. Annals of the rheumatic diseases 69, 1920- 1925, doi:10.1136/ard.2009.122226 (2010). Competing interests 8. McMahon, M. et al. Proinflammatory high-density lipoprotein as a biomarker for atherosclerosis in All authors declare that they have no conflict of interest patients with systemic lupus erythematosus and rheumatoid arthritis. Arthritis and rheumatism 54, 2541- 2549, doi:10.1002/art.21976 (2006). 9. Popa, C. et al. Anti-inflammatory therapy with tumour necrosis factor alpha inhibitors improves high- density lipoprotein cholesterol antioxidative capacity in rheumatoid arthritis patients. Annals of the rheumatic diseases 68, 868-872, doi:10.1136/ard.2008.092171 (2009). 10. Park, Y. B. et al. Lipid profiles in untreated patients with rheumatoid arthritis.The Journal of rheumatology 26, 1701-1704 (1999). 11. van Halm, V. P. et al. Lipids and inflammation: serial measurements of the lipid profile of blood donors who later developed rheumatoid arthritis. Annals of the rheumatic diseases 66, 184-188, doi:10.1136/ ard.2006.051672 (2007). 12. Popa, C. D., Arts, E., Fransen, J. & van Riel, P. L. Atherogenic index and high-density lipoprotein cholesterol as cardiovascular risk determinants in rheumatoid arthritis: the impact of therapy with biologicals. Mediators of inflammation 2012, 785946, doi:10.1155/2012/785946 (2012). 13. Choi, H. K. & Seeger, J. D. Lipid profiles among US elderly with untreated rheumatoid arthritis--the Third National Health and Nutrition Examination Survey. The Journal of rheumatology 32, 2311-2316 (2005). 14. Georgiadis, A. N. et al. Atherogenic lipid profile is a feature characteristic of patients with early rheumatoid arthritis: effect of early treatment--a prospective, controlled study. Arthritis research & therapy 8, R82, doi:10.1186/ar1952 (2006). 15. Chung, C. P. et al. Lipoprotein subclasses determined by nuclear magnetic resonance spectroscopy and coronary atherosclerosis in patients with rheumatoid arthritis. The Journal of rheumatology 37, 1633- 1638, doi:10.3899/jrheum.090639 (2010). 16. Heinhuis, B. et al. Tumour necrosis factor alpha-driven IL-32 expression in rheumatoid arthritis synovial tissue amplifies an inflammatory cascade. Annals of the rheumatic diseases 70, 660-667, doi:10.1136/ ard.2010.139196 (2011). 17. Joosten, L. A. et al. IL-32, a proinflammatory cytokine in rheumatoid arthritis.Proceedings of the National Academy of Sciences of the United States of America 103, 3298-3303, doi:10.1073/pnas.0511233103 (2006). 18. Kim, S. H., Han, S. Y., Azam, T., Yoon, D. Y. & Dinarello, C. A. Interleukin-32: a cytokine and inducer of TNFalpha. Immunity 22, 131-142, doi:10.1016/j.immuni.2004.12.003 (2005). 19. Rodriguez-Rodriguez, L. et al. TNFA -308 (rs1800629) polymorphism is associated with a higher risk of cardiovascular disease in patients with rheumatoid arthritis. Atherosclerosis 216, 125-130, doi:10.1016/j. atherosclerosis.2010.10.052 (2011). 20.

38 39 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

21. Rus, H. G., Niculescu, F. & Vlaicu, R. Tumor necrosis factor-alpha in human arterial wall with atherosclerosis. 41. Bou Malham, S. & Goldberg, A. C. Cardiovascular risk reduction: the future of cholesterol lowering drugs. Current Atherosclerosis 89, 247-254 (1991). opinion in pharmacology 27, 62-69, doi:10.1016/j.coph.2016.01.007 (2016). 22. Popa, C., Netea, M. G., van Riel, P. L., van der Meer, J. W. & Stalenhoef, A. F. The role of TNF-alpha in 42. Maradit-Kremers, H. et al. Increased unrecognized coronary heart disease and sudden deaths in rheumatoid chronic inflammatory conditions, intermediary metabolism, and cardiovascular risk. Journal of lipid arthritis: a population-based cohort study. Arthritis and rheumatism 52, 402-411, doi:10.1002/art.20853 (2005). research 48, 751-762, doi:10.1194/jlr.R600021-JLR200 (2007). 43. Solomon, D. H. et al. Patterns of cardiovascular risk in rheumatoid arthritis. Annals of the rheumatic diseases 65, 23. Nold-Petry, C. A. et al. IL-32-dependent effects of IL-1beta on endothelial cell functions. Proceedings 1608-1612, doi:10.1136/ard.2005.050377 (2006). of the National Academy of Sciences of the United States of America 106, 3883-3888, doi:10.1073/ 44. Lazarevic, M. B. et al. Dyslipoproteinemia in the course of active rheumatoid arthritis. Seminars in arthritis and pnas.0813334106 (2009). rheumatism 22, 172-178 (1992). 24. Heinhuis, B. et al. Towards a role of interleukin-32 in atherosclerosis. Cytokine 64, 433-440, doi:10.1016/j. 45. Boers, M. et al. Influence of glucocorticoids and disease activity on total and high density lipoprotein cholesterol 2 cyto.2013.05.002 (2013). in patients with rheumatoid arthritis. Annals of the rheumatic diseases 62, 842-845 (2003). 25. Ahmad, U. et al. IFN-gamma primes intact human coronary arteries and cultured coronary smooth muscle 46. Choy, E. & Sattar, N. Interpreting lipid levels in the context of high-grade inflammatory states with a focus on cells to double-stranded RNA- and self-RNA-induced inflammatory responses by upregulating TLR3 rheumatoid arthritis: a challenge to conventional cardiovascular risk actions. Annals of the rheumatic diseases 68, and melanoma differentiation-associated gene 5. Journal of immunology 185, 1283-1294, doi:10.4049/ 460-469, doi:10.1136/ard.2008.101964 (2009). jimmunol.0902283 (2010). 47. Schimmel, E. K. & Yazici, Y. Increased lipid levels but unchanged atherogenic index in rheumatoid arthritis patients 26. Buttice, G., Miller, J., Wang, L. & Smith, B. D. Interferon-gamma induces major histocompatibility treated with biologic disease modifying antirheumatic drugs: published experience. Clinical and experimental class II transactivator (CIITA), which mediates collagen repression and major histocompatibility class rheumatology 27, 446-451 (2009). II activation by human aortic smooth muscle cells. Circulation research 98, 472-479, doi:10.1161/01. 48. Dahlqvist, S. R., Engstrand, S., Berglin, E. & Johnson, O. Conversion towards an atherogenic lipid profile in RES.0000204725.46332.97 (2006). rheumatoid arthritis patients during long-term infliximab therapy. Scandinavian journal of rheumatology 35, 107- 27. Roelofs, M. F. et al. The expression of toll-like receptors 3 and 7 in rheumatoid arthritis synovium is 111, doi:10.1080/03009740500474578 (2006). increased and costimulation of toll-like receptors 3, 4, and 7/8 results in synergistic cytokine production 49. Lopez-Mejias, R. et al. NFKB1-94ATTG ins/del polymorphism (rs28362491) is associated with by dendritic cells. Arthritis and rheumatism 52, 2313-2322, doi:10.1002/art.21278 (2005). cardiovascular disease in patients with rheumatoid arthritis. Atherosclerosis 224, 426-429, doi:10.1016/j. 28. Li, J., Hsu, H. C. & Mountz, J. D. Managing macrophages in rheumatoid arthritis by reform or removal. atherosclerosis.2012.06.008 (2012). 29. Current rheumatology reports 14, 445-454, doi:10.1007/s11926-012-0272-4 (2012). 50. Palomino-Morales, R. et al. A1298C polymorphism in the MTHFR gene predisposes to cardiovascular risk in 30. Libby, P., Ridker, P. M. & Hansson, G. K. Progress and challenges in translating the biology of rheumatoid arthritis. Arthritis research & therapy 12, R71, doi:10.1186/ar2989 (2010). atherosclerosis. Nature 473, 317-325, doi:10.1038/nature10146 (2011). 51. Okada, Y. et al. Meta-analysis identifies nine new loci associated with rheumatoid arthritis in the Japanese 31. Damen, M. et al. SAT0025 Shift in Genetic Composition of an IL-32 Promoter Polymorphism Resuls in a population. Nature genetics 44, 511-516, doi:10.1038/ng.2231 (2012). Higher Cytokine Production in RA Patients. Annals of the rheumatic diseases 74, 657-658, doi:10.1136/ 52. Xiao, X. et al. Genome-wide association studies and gene expression profiles of rheumatoid arthritis: An analysis. annrheumdis-2015-eular.5540 (2015). Bone & joint research 5, 314-319, doi:10.1302/2046-3758.57.2000502 (2016). 32. Stahl, E. A. et al. Genome-wide association study meta-analysis identifies seven new rheumatoid arthritis 53. Lopez-Mejias, R. et al. Influence of elevated-CRP level-related polymorphisms in non-rheumatic Caucasians risk loci. Nature genetics 42, 508-514, doi:10.1038/ng.582 (2010). on the risk of subclinical atherosclerosis and cardiovascular disease in rheumatoid arthritis. Scientific reports 6, 33. Suzuki, A. et al. Functional haplotypes of PADI4, encoding citrullinating enzyme peptidylarginine deiminase 31979, doi:10.1038/srep31979 (2016). 4, are associated with rheumatoid arthritis. Nature genetics 34, 395-402, doi:10.1038/ng1206 (2003). 54. Lopez-Mejias, R. et al. Lack of association between IL6 single nucleotide polymorphisms and cardiovascular 34. Teruel, M. et al. Association of CD247 polymorphisms with rheumatoid arthritis: a replication study and a disease in Spanish patients with rheumatoid arthritis. Atherosclerosis 219, 655-658, doi:10.1016/j. meta-analysis. PloS one 8, e68295, doi:10.1371/journal.pone.0068295 (2013). atherosclerosis.2011.07.124 (2011). 35. Gordon, T., Castelli, W. P., Hjortland, M. C., Kannel, W. B. & Dawber, T. R. High density lipoprotein as a 55. Ma, A. Z., Zhang, Q. & Song, Z. Y. TNFa alter cholesterol metabolism in human macrophages via PKC-theta- protective factor against coronary heart disease. The Framingham Study. The American journal of dependent pathway. BMC biochemistry 14, 20, doi:10.1186/1471-2091-14-20 (2013). medicine 62, 707-714 (1977). 56. Hoogendoorn, E. H. et al. Thyroid function and prevalence of anti-thyroperoxidase antibodies in a population 36. Arca, M. et al. Usefulness of atherogenic dyslipidemia for predicting cardiovascular risk in patients with with borderline sufficient iodine intake: influences of age and sex. Clinical chemistry 52, 104-111, doi:10.1373/ angiographically defined coronary artery disease. The American journal of cardiology 100, 1511- 1516, clinchem.2005.055194 (2006). doi:10.1016/j.amjcard.2007.06.049 (2007). 57. Holewijn, S., den Heijer, M., Swinkels, D. W., Stalenhoef, A. F. & de Graaf, J. The metabolic syndrome and its traits 37. Mahdy Ali, K., Wonnerth, A., Huber, K. & Wojta, J. Cardiovascular disease risk reduction by raising HDL as risk factors for subclinical atherosclerosis. The Journal of clinical endocrinology and metabolism 94, 2893-2899, cholesterol--current therapies and future opportunities. British journal of pharmacology 167, 1177-1194, doi:10.1210/jc.2009-0084 (2009). doi:10.1111/j.1476-5381.2012.02081.x (2012). 58. Miller, S. A., Dykes, D. D. & Polesky, H. F. A simple salting out procedure for extracting DNA from human 38. Lakatos, J. & Harsagyi, A. Serum total, HDL, LDL cholesterol, and triglyceride levels in patients with nucleated cells. Nucleic acids research 16, 1215 (1988). rheumatoid arthritis. Clinical biochemistry 21, 93-96 (1988). 59. ter Avest, E., Holewijn, S., Stalenhoef, A. F. & de Graaf, J. Variation in non-invasive measurements of vascular 39. Hurt-Camejo, E. et al. Elevated levels of small, low-density lipoprotein with high affinity for arterial function in healthy volunteers during daytime. Clinical science 108, 425-431, doi:10.1042/CS20040300 (2005). matrix components in patients with rheumatoid arthritis: possible contribution of phospholipase A2 to 60. ter Avest, E., Holewijn, S., Bredie, S. J., Stalenhoef, A. F. & de Graaf, J. Remnant particles are the major determinant this atherogenic profile. Arthritis and rheumatism 44, 2761-2767 (2001). of an increased intima media thickness in patients with familial combined hyperlipidemia (FCH). Atherosclerosis 40. Lee, Y. H., Choi, S. J., Ji, J. D., Seo, H. S. & Song, G. G. Lipoprotein(a) and lipids in relation to inflammation 191, 220-226, doi:10.1016/j.atherosclerosis.2006.03.025 (2007). in rheumatoid arthritis. Clinical rheumatology 19, 324-325 (2000).

40 41 Chapter 2 – IL-32 SNP and modified lipoproteins in RA

SUPPLEMENTAL FIGURE

IL-32 promoter SNP rs4786370 predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis

Supplementary Fig. S1A Supplementary Fig. S1B 2

RA1 cohort RA2 cohort

e

DAS28 score DAS28 DAS28 scor DAS28

rs4786370 genotype rs4786370 genotype

Supplementary Fig. S1. DAS28-scores split on the IL-32 promoter SNP genotype within the two RA-patients cohorts. A. represents DAS28-scores of RA patients from the RA1 cohort. B. Represents DAS28-scores of RA patients in the RA2 cohort.

42 43 CHAPTER 3 Genetic variant rs4786370 of 3 IL-32 is associated with the ex vivo cytokine production of anti- TNF treated PBMCs isolated from rheumatoid arthritis patients

Submitted

Michelle S.M.A. Damen1 Kiki Schraa1 Lieke Tweehuysen2 Alfons A. den Broeder2 Mihai G. Netea1,3 Calin D. Popa2,4 Leo A.B. Joosten1,*

44 45 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

ABSTRACT INTRODUCTION

Objective: Still about 60% of RA patients do not achieve good response with Rheumatoid arthritis (RA) is a common autoimmune and chronic inflammatory biological disease-modifying anti-rheumatic drugs bDMARD treatment (including disease affecting about 1% of the population1, 2. RA is characterized by persistent TNF inhibitors, TNFi’s). Previously, a strong link between TNFα and interleukin joint inflammation, progressive disability and ongoing systemic inflammation, which (IL)-32 has been reported in RA. We hypothesize that a promoter single nucleotide can lead to joint deformity and low quality of life. Moreover, these characteristics polymorphism (SNP) rs4786370 in IL-32 can affect clinical responsiveness to TNFi’s are also able to increase the risk for atherosclerosis and cardiovascular disease in RA patients, potentially serving as new biomarker in RA. (CVD), which is the main cause of death in these patients3-5. The clinical course of RA varies tremendously between patients from spontaneous remission, to mild joint Methods: Peripheral mononuclear cells (PBMCs) from RA patients and healthy symptoms to severe bone destruction. Early and aggressive treatment has been 3 individuals were stimulated with RPMI or recombinant human (rh)TNFα to study shown to improve the outcome. The introduction of biological disease-modifying mRNA and protein expression of pro-inflammatory cytokines. Moreover, “ex vivo anti- rheumatic drugs (bDMARDs) in combination with “treat-to-target” treatment response” and clinical response to anti-TNFα therapy (etanercept, adalimumab) strategy significantly improved the disease outcomes in RA. These bDMARDs measured by disease activity scores (DAS28), of RA patients were measured and however still only achieve good response in about 40% of RA patients. The were stratified for the IL-32 SNP. variation in clinical response to bDMARDs could be explained by variations in drug concentration and pharmacokinetics, which in turn are influenced by age, gender, Results: Stimulation of PBMCs from RA patients showed higher IL-32 protein renal or liver function6. Alternatively, the genetic background may also play a role production and a tendency towards higher IL-32β/IL-32γ mRNA expression and the interplay with the other factors could conduct towards specific profiles compared to healthy individuals. Patients bearing the CC genotype showed higher and increase the chance of achieving a good clinical response, suggesting a niche IL-32 protein expression and produced more cytokines. The DAS28 did not depend personalized medicine. on the presence of the promoter SNP, however, the “ex vivo” cytokine response did have a different pattern in clinical responders depending on the genotype. The pathogenesis of RA still remains partly unknown but results in a chronic inflammatory state. The initial phase involves the activation of T and B cells and the Conclusion: IL-32 mRNA and protein production was higher in RA patients compared induction of pro-inflammatory cytokines such as IL-6, IL- 1β and TNFα7-9. TNFα is to healthy individuals, with a trend towards higher concentrations in patients clearly of high importance in the pathogenesis of RA, which is shown by the fact that bearing the CC genotype. Furthermore, IL-1 beta production in the CC-genotype TNFi’s can effectively reduce the chronic inflammation in RA10,11. Moreover, TNFα might predict clinical response to etanercept or adalimumab. These data indicate is also capable of inducing other pro-inflammatory mediators, such as that IL-32 could play a role in predicting the response to treatment in RA. and cytokines IL-6, IL-1β and IL-32, all found to be important in RA11,12. Studies of the last decade have shown that the cytokine interleukin-32 (IL-32) by itself is a strong inducer of TNFα and the expression levels of IL-32 in synovial biopsies correlated with inflammation severity in RA13,14. Moreover, overexpression of IL-32γ in human synovial fibroblasts followed by stimulation of TLR2/NOD2, showed a potent induction of TNFα mRNA15. In contrast, when IL-32 was suppressed, TNFα production was decreased in human monocytes, all showing the important pro-inflammatory properties of IL- 32 and its close relation with TNFα15,16. Despite knowing the interaction between these two cytokines and the importance in RA, research on the specific role of IL-32 in RA remains scarce. Our group recently showed a role for a promoter single-nucleotide polymorphism (SNP) in IL-32 that seemed to be associated with cytokine production, IL-32 isoform expression and high-density

46 47 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

cholesterol (HDLc) levels in RA patients17,18. The present study therefore aims to The Netherlands) and collecting the buffy-coat enriched layer. Cells were washed investigate the possible predictive implications of this SNP in the IL-32 gene on the twice in cold PBS and concentrations were adjusted to 5 × 106 cells/ml in RPMI-1640, severity of the disease and the clinical response to TNFI’s (including adalimumab or supplemented 2 mM l- glutamine, 1 mM pyruvate and 50 μg/ml gentamicin (GIBCO etanercept) in RA patients. Invitrogen, Carlsbad, CA). Mononuclear cells (5 × 105) in a 100-μl volume were added to round-bottom 96-well plates (Greiner, Nurnberg, Germany) and incubated with either 100 μl of culture medium (negative control), Poly I:C (50uL Invivogen), or MATERIAL AND METHODS recombinant human TNFα (rhTNFα 10ng R&D Systems). After 24 hour incubation at 37°C, the supernatants were stored at - 20°C until further use. Besides these Patient cohort experiments, additionally 5 × 105 PBMCs were pre-incubated in round bottom Blood samples were obtained from patients included in the prospective longitudinal 96-well plates for one hour at 37 °C with therapeutic in vivo concentrations of 3 prediction cohort study BIO-TOP [Biologic Individual Optimized Treatment Outcome adalimumab or etanercept. Taking into account the different half-life times, dosing Prediction] and isolated immune cells were used for various assays. RA patients >18 and treatment intervals, and therapeutic concentration ranges of the anti-TNFα, years, treated in the Sint Maartenskliniek (Nijmegen, the Netherlands) who were the same concentration of 5 µg/mL was added for both anti-TNFα bDMARDS (4-6). going to start with (or switch to) a biological disease-modifying anti-rheumatic drug Human IgG was used as negative control. Thereafter, cells were stimulated with (bDMARD) were included in this study. The local ethical committee (CMO region either Pam3Cys (a TLR2 agonist) or heat killed Candida albicans (ATCC MYA-3573 Arnhem-Nijmegen, NL47946.091.14) was responsible for approval of the BIO-TOP (UC 820)). After 24 hours, supernatants were stored at -20 °C until assayed. study and a detailed description is available in the Dutch trial register (NTR4647 clinical trial, 17-jun-2014, NTR). Additionally, blood samples of healthy individuals Cytokine measurements were used. Various cytokines were determined in supernatant after stimulation of PBMCs with Written informed consent was received from all donors. Experiments with human recombinant human TNFα (10 ng/ml) (R&D Systems), recombinant human IL-1β blood were performed in accordance with the Declaration of Helsinki. (1ng/mL R&D Systems) or Poly I:C (TLR3 agonist)(50 µg/ml) (Invivogen) for 24hours, by commercially available ELISA kits according to manufacturer’s instructions. DNA isolation and taqman genotyping Concentrations of human IL-1β, IL-1Ra (R&D Systems, Inc., Minneapolis, MN, Whole blood obtained from 329 RA patients in the BIOTOP study was used to perform USA) and IL-6, IL-8, IL-10 (Sanquin Reagents, Amsterdam, The Netherlands) were genomic DNA extraction. Genomic DNA was isolated from whole blood using measured. IL-32 production was measured in cell lysates (Triton X 100 0.5%) of the Qiagen (Valencia, CA, USA) isolation kit and following the standard protocol. PBMCs stimulated with various ligands, using the commercially available ELISA kit The samples were quantified and evaluated for purity (260/280-nm ratio) with a (R&D Systems, Inc, Minneapolis, MN, USA). In brief, Maxisorp plates (Nunc) were NanoDrop ND-1000 spectrophotometer (Thermo Scientific). Using the Taqman SNP coated with AF3040 (R&D Systems) diluted in Phosphate Buffered Saline (PBS) assay C_27972515_20 (Thermofisher, Foster City, CA, USA), each genetic variant in at a concentration of 0.4 αg/ml and incubated overnight at room temperature. the IL32 promoter (rs4786370) polymorphism was determined. The TaqMan qPCR Plates were blocked with PBS containing 1 % BSA (Sigma-Aldrich) for 1 hour at assays were performed on the AB StepOnePlus polymerase chain reaction system room temperature. The standard curve was prepared by diluting recombinant IL-32 (Applied Biosystems). Negative controls were included in the assay. No duplicates ranging from 5000 pg/ml until 39.06 pg/ml in PBS containing 5 % BSA. After a 2hour were used. incubation with the cell lysates, detection antibody was added (BAF3040, R&D Systems), 0.1 µg/ml in PBS with 5 % BSA for 1hour. Streptavidin (R&D Systems) was Isolation and ex-vivo stimulation of peripheral blood mononuclear cells added for 30 minutes at room temperature after which substrate buffer was used to (PBMCs) develop a color reaction that was measured by a plate reader. At baseline (before start bDMARD), venous blood was collected into three 10 mL EDTA tubes after informed consent. Within 24 hours, PBMCs were isolated by density gradient centrifugation using Ficoll-Paque PLUS (GE Healthcare, Zeist,

48 49 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

A B Clinical assessments Disease activity was measured with the 28-joints disease activity score using C-reactive protein (DAS28-CRP) during outpatient clinical visits performed in usual care after 3 and 6 months (±1 month). Primary outcome was the DAS28-CRP based European League Against Rheumatism (EULAR) response criteria (good versus moderate/no response) at month 6.

Statistical analysis Normality was tested using the D’Agostino normality test. Continuous variables C D are presented as mean and standard deviation (SD). The differences IL-32 mRNA 3 expression, IL-32 protein concentrations and cytokine concentrations were analyzed using the Mann-Whitney test. A p-value less than 0.05 was considered statistically significant (*p<0.05 and **p<0.01). Data was analyzed using GraphPad Prism v5.3.

RESULTS

IL-32 isoform mRNA expression in RA patients versus healthy subjects PBMCs of RA patients (n=22) and those of healthy individuals (n=7) were isolated Fig 1. mRNA and expression of IL-32 isoforms IL-32β, IL-32γ, in RA patients and healthy individuals and IL-32β and IL-32γ mRNA expression were determined. First, IL-32β and IL-32γ in (un)stimulated PBMCs also separated on IL-32 promoter SNP genotypes. A). mRNA expression of isoforms expression was determined in unstimulated PBMCs, independent of the IL-32β in PBMCs from healthy individuals versus RA patients in unstimulated PBMCs (n=7 healthy individuals; n=22 RA patients). B). mRNA expression of IL-32γ in PBMCs from RA patients and healthy IL-32 promoter SNP genotype (Fig. 1A-B). Although no significant differences were individuals in unstimulated PBMCs (n=7 healthy individuals; n=22 RA patients). C). mRNA expression observed, there is a clear trend towards more IL-32β and IL-32γ mRNA expression of IL-32β in unstimulated PBMCs or PBMCs stimulated with rhTNFα from RA patients versus healthy in the RA patients compared to the healthy individuals (Fig. 1A-B). Thereafter we individuals with the CC- versus TT-genotype for the IL-32 promoter SNP (n=13 TT; n=9 CC Ra patients). examined whether the IL-32 promoter SNP genotype (T/C) would influence this D). mRNA expression of IL-32γ in unstimulated PBMCs or PBMCs stimulated with rhTNFα from RA patients versus healthy individuals with the CC- versus TT-genotype for the IL-32 promoter SNP (n=13 latter result (Figure 1C-D). Unstimulated PBMCs of RA patients with the TT-genotype TT; n=9 CC RA patients). showed a tendency towards higher IL-32β mRNA expression compared to healthy individuals (Fig.1C). RA patients bearing the CC-genotype did not show differences on IL-32β mRNA expression neither in unstimulated nor in the case PBMCs have been IL-32 protein expression is highest in RA patients bearing the CC- stimulated with recombinant human TNFα (rhTNFα) (Fig. 1C). Exploring specifically genotype IL-32γ, a trend towards higher IL-32γ mRNA expression in unstimulated PBMCs of Even though there were no significant differences observed in IL-32β and IL- patients bearing the CC genotype could be observed. When stimulated with rhTNFα, 32γ isoforms mRNA expression between RA patients and healthy individuals, RA patients bearing the TT-genotype seemed to express more IL-32γ compared to independent of IL-32 promoter SNP, the endogenous IL-32 protein expression was the unstimulated condition and the expression of IL-32γ in the group bearing the CC- investigated. IL-32 protein concentration was determined in the two groups as a genotype. However, none of these observations were statistical significant (Fig. 1D). whole as well as stratified for IL-32 promoter SNP (Fig.2A-B). Figure 2A shows a significantly higher IL-32 protein expression in RA patients compared to healthy individuals in unstimulated PBMCs. Of note, the production of IL-32 protein was significantly higher especially in the RA patients with the CC-genotype (Fig. 2B), whereas in the group bearing the TT-genotype it did not reach statistical significance.

50 51 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

A B A pro-inflammatory status of RA PBMCs bearing the CC-genotype After observing differences in IL-32 mRNA isoforms expression and IL-32 protein expression between RA patients and healthy individuals, we examined the potential role of the CC-genotype in RA patients. We were interested to determine whether the genotypes in IL-32 promoter SNP would be associated with a different inflammatory response. In order to explore this, we investigated the capacity of ex- vivo cytokine production by PBMCs after exposure with various proinflammatory stimuli. Figures 3A-3C clearly showed that PBMCs isolated from RA patients bearing the CC-genotype produce more IL-6 and IL-8, compared to TT genotype. Fig 2. Protein expression of total IL-32 in unstimulated PBMCs from healthy individuals and RA patients also stratified for IL32 promoter SNP. A) A statistical significant difference was found in IL-32 protein expression between these groups (p<0,0044) using the Mann-Whitney test in Graphpad Prism 3 Depending on the stimuli, the enhanced production of IL-8 reached statistical v5.03 (n=8 TT healthy individuals and n=25 TT RA patients versus n= 8 CC healthy individuals and n=14 significance after PBMCs have been stimulated with rhIL-1β (Fig. 3B). In line with CC RA patients). B) IL-32 protein concentrations were stratified for the IL-32 promoter SNP genotypes these results, PBMCs bearing the CC genotype produce more IL-1Ra, although this TT and CC. A significantly higher concentration of IL-32 protein was found in the RA patients carrying was not statistical significant. the CC-genotype compared to healthy individuals with the same genotype (p<0.0105).

DAS28-CRP is not affected by IL-32 promoter SNP genotypes A B After observing that RA patients, particularly those bearing the CC-genotype of the IL-32 SNP promoter, produce more pro-inflammatory cytokines, we thought to investigate whether this could have an impact on disease activity of these RA patients. To quantify disease activity, we used well-established DAS28-CRP score. As shown in Figure 4A, no differences were observed in DAS28-CRP scores between the two genotypes for the IL-32 promoter SNP at baseline. In the subgroups of patients who started treatment with respectively adalimumab and etanercept. IL-32 genotype did not influence the DAS28-CRP at three as well as six months afterwards (Fig. 4B-C). C

IL-32 genotype could help to predict the response on etanercept and adalimumab in RA In order to provide a possible explanation on the fact that the IL-32 promoter SNP did not influence the response of RA patients to therapy with adalimumab or etanercept in the first six months, we further thought to assess whether the genotype could affect the ex vivo response of PBMCs to various stimuli in the presence of etanercept or adalimumab. In addition, we also wanted to address the question whether patients who eventually show a good clinical response to these Fig 3. Cytokine measurements in PBMCs from RA patients and stratified for the IL-32 promoter SNP bDMARDs have a different inflammatory response at baseline dependent on the after stimulation with either Poly I:C and rhIL-1β or rhTNFα. A) IL-6 protein concentration in PBMCs of IL-32 genotype. More specifically, if RA patients with either the CC or TT genotype RA patients with the TT- versus CC-genotype for the promoter SNP, at basal level and after stimulation B) IL-8 protein concentration in PBMCs of RA patients with the TT- versus the CC-genotype for the would respond better to the specific bDMARD by showing a different cytokine IL-32 promoter SNP, at basal level and after stimulation C) IL- 1Ra protein concentration in PBMCs of response (Fig. 5A-D and Supplementary Fig S1 and Supplementary tables T1, T2). RA patients with the TT- versus CC-genotype for the IL-32 promoter SNP at basal level as well as after IL-1β production by PBMCs tends to decrease after Candida albicans stimulation in stimulation. Statistical significance was shown by * when p<0.05, GraphpadPrism V5.03. (n=27 TT RA patients and n= 15 CC RA patients).

52 53 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

A B the presence of either adalimumab or etanercept (Fig. 5A). When the IL-1β cytokine production was stratified for clinical responders versus non-responders to either etanercept or adalimumab (Fig. 5B and Suppl. Fig. S1), higher IL-1 production in the clinical responders was noted, though not reaching statistical significance. We further divided these groups according to their genotype for the IL-32 promoter SNP. Figure 5C/D and supplementary Fig. S1 indicate that for the clinical responders to either etanercept or adalimumab, RA patients bearing the CC genotype seem to produce more IL-1β after ex vivo stimulation of PBMCs, even reaching a statistical significant difference in the case of etanercept treatment. In contrast, no such effect was observed for RA patients bearing the TT genotype. To try and explain this effect C D 3 in more detail we calculated the percentage of ex vivo responders to etanercept and adalimumab. This data however does not show to be able to predict a possible effect of the IL-32 promoter SNP on the clinical responsiveness of RA patients to anti-TNFα therapy (Supplementary Tables T1-T2).

A B

Fig 5. Percentage of IL-1β production corrected for baseline IL-1β production induced by Candida IgG stimulation of PBMCs from RA patients in the presence of anti-TNFα treatment (etanercept). A)IL-1β cytokine production after ex vivo stimulation of PBMCs from RA patients with additional etanercept or adalimumab for 24h independent of clinical treatment (n=121). B) IL-1β cytokine production after ex vivo stimulation of PBMCs from RA patients with additional etanercept, stratified for RA patients that clinically responded or not to etanercept (n= 26 non-responders; n=19 responders). C) IL-1β cytokine production after ex vivo stimulation of PBMCs from RA patients with additional etanercept, stratified C for clinical responders to etanercept treatment and IL-32 promoter SNP (n=4 CC; n=15 TT RA patients). D) IL-1β cytokine production after ex vivo stimulation of PBMCs from RA patients with additional etanercept, stratified for clinical non-responders to etanercept and IL-32 promoter SNP genotype (n=9 CC and n=17 TT RA patients). Statistical significance was shown by * when p<0.05, GraphpadPrism V5.03.

DISCUSSION

Fig 4. Disease activity scores (DAS) 28 in RA patients stratified on IL-32 promoter SNP used therapy (etanercept or adalimumab). A) DAS 28CRP separated on IL-32 promoter SNP genotype and In this study, we showed that mRNA expression levels of IL-32 isoforms IL-32β and independent of therapy (n = 37 CC vs 85 TT). B) DAS28CRP separated on IL-32 promoter SNP genotype IL-32γ tended to be higher in PBMCs isolated from RA patients as compared to and Etanercept therapy (n =13 CC vs 32 TT). C) DAS28CRP separated on IL-32 promoter SNP genotype healthy individuals. Additionally, intracellular IL-32 protein production was higher and Adalimumab therapy (n = 4 CC vs 13 TT). in RA patients, especially in those bearing the CC-genotype. These patients also tended to produce more pro-inflammatory cytokines in-vitro. However, this fact did not translate into a higher disease activity, as assessed by DAS28-CRP scores.

54 55 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

Finally, ex-vivo PBMCs stimulations in the presence of the intended medication term outcomes. Unfortunately, this study showed that the promoter SNP in IL-32 could potentially identify the patients who would become responders according to was not such a marker. Even though it has functional effects on cytokine production clinical criteria. The determination of IL-32 promoter SNP in this context proved to by PBMCs or on the HDLc concentration, no effect was seen on the level of disease have crucial importance. activity, as measured by DAS28-CRP. These results are somehow in contrast with the study of Gui et al., which observed correlations between the IL-32 levels and IL-32 mRNA isoforms have been studied so far in many different cell types and disease activity23. The number of patients enrolled and the different design might diseases including chronic obstructive pulmonary disease (COPD), various types of explain this discrepancy. Alternatively, IL-32 might differently impact the various cancer and RA14,19-21. Nevertheless, the exact expression pattern of the isoforms in components of DAS28 score, for instance the number of tender joints or the VAS. specific cells or tissues and the function of each isoform still need to be elucidated21. It is known that the structure of the two isoforms IL-32β and IL-32γ is quite similar Single nucleotide polymorphisms in other cytokines such as TNFα, IL-1β and IL-6 3 and that IL-32γ mRNA can be spliced into IL-32β22. Both isoforms seem to be involved have been studied in possible association to response to treatment in RA patients25-28. in pro-inflammatory processes, including the induction of other pro-inflammatory Since IL-32 is strongly linked to TNFα expression and can induce IL-6 and IL-1β, it cytokines and chemokines like IL-8, IL-6, IL-1β and TNFα. In our study, we found a was worth investigating whether this SNP also had an effect on clinical response higher IL-32 protein production from PBMCs of RA patients together with a slightly to treatment. No differences were observed when looking at the clinical response higher IL-32 mRNA expression in the same patients. This is in line with previous data independent of treatment or anti-TNFα (etanercept or adalimumab) treatment in pointing to a higher IL-32 mRNA expression in fibroblast-like synoviocytes (FLS) particular, when separated on the promoter SNP for IL-32. of RA patients and also in the PBMCs22,23. Interestingly, here we show that only in patients bearing the CC-genotype, the increase in plasma IL-32 concentrations Besides the clinical response, the ex vivo response of PBMCs stimulated with reaches statistical significance. This could also be due to the low number of patients additional etanercept or adalimumab and Candida albicans or Pam3Cys was studied included for this analysis. Alternatively, this might suggest that the studied SNP in by looking at IL-1β cytokine production after stimulation. A previous study by Popa IL-32 promoter has functional consequences. Finally, a difference between isoforms C et al. could not detect differences in IL-1β cytokine production after anti-TNFα expression has been observed within the patients, which varied from the expression treatment29. Interestingly, our data however showed that PBMCs of RA patients pattern seen in healthy individuals. bearing the CC genotype produced more IL-1β after ex vivo stimulation only within the group of RA patients that clinically responded to either etanercept or possibly In a previous study, we indicated that the promoter SNP is associated with different adalimumab treatment. No such differences were observed either for the clinical levels of high-density lipoprotein cholesterol (HDLc) concentrations in RA patients, non-responders or patients bearing the TT genotype. This was in line with data from namely higher HDLc concentrations17. These results already suggested that this SNP another study by Kayakabe K et al. which showed that IL-1β could serve as a possible might be functional. In the present report, we extended our observations at the level predictive measurement for response to anti-TNFα treatments30. of cytokine production and immune responses. PBMCs from RA patients showed a difference in cytokine production when the patients were separated according to These results suggest that even though the IL-32 promoter SNP does not have an their genotype. Our study is the first to show that and therefore confirms the fact effect on DAS28 scores, there might be a role for the genetic polymorphism in the that the promoter SNP serves a functional effect. promoter region of IL-32 in the prediction of clinical response to anti-TNFα treatment in RA patients also taking into account the observed influence on the production Given the contribution of IL-32 and TNF to the inflammatory cascade in RA, we further of pro-inflammatory cytokines in these patients. Given the previously indicated investigated whether the SNP in IL-32 promoter region impacts disease activity in relation between the promoter SNP and HDLc levels in RA patients and the new these patients. Absolute levels of disease activity are set to define the disease state findings of the functional effect of the promoter SNP on cytokine production and of the patients and over time can be used to define improvement or response to response to treatment, these data show a possible additional role of IL-32 and its treatment24. Finding a marker that could predict the latter would therefore be useful, promoter SNP in RA. Our results therefore lead to an area of research in which the although current treat to target treatment strategies already have very good long effects of IL- 32 and the promoter SNP in RA have to be studied relevant to response

56 57 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

to treatment and cardiovascular diseases within these patients. In line with this, we ACKNOWLEDGEMENTS are currently performing studies to further elucidate the mechanisms behind the role of IL-32 in cholesterol metabolism regulation. We would like to thank the rheumatologists in the Sint Maartenskliniek for participation in patient recruitment and all the patients who were willing to Some limitations could be envisaged in our study. One of the most important one participate. might be related to the low number of patients per genotype (TT vs CC) for the IL-32 promoter SNP, especially in the analysis of the data concerning only patients who Funding received etanercept or adalimumab. This is caused by the multiple stratifications This research was supported by grants from the Dutch Arthritis Foundation e.g. genotypes, clinical responses; medications/treatments that were performed. (Reumafonds)(Nr.13-03-302) and the Nijmegen Institute for Infection, Inflammation This could interfere with statistical analysis and power in case a small difference in and Immunity (N4i), the Netherlands. MGN was funded by a Spinoza Grant of 3 response to treatment would have been observed and sample size is small. Moreover, the Netherlands Organization for Scientific Research, and a Competitiveness differences in cytokine production of TNFα were not measured directly and because Operational Program Grant of the Romanian Ministry of European Funds (FUSE). of multiple freeze-thaw cycles, this was no longer possible. Therefore, this study lacks a very important cytokine measurement within this group of patients. Authors’ contribution Authors MSMAD, KS, LT, AAdB, MGN, CAD and LABJ contributed to the design of the In conclusion, the present study shows that IL-32 mRNA and protein production was study and acquisition of data. MSMAD, KS, CAD and LABJ additionally contributed higher in RA patients compared to healthy individuals. Moreover, this study is the to the analysis and interpretation of data. All authors furthermore contributed to first to show a slightly higher IL-32 expression in patients bearing the CC-genotype drafting and critically revising the manuscript to create an approved version for for the IL-32 promoter SNP (rs4786370). Additionally, the promoter SNP tended submission to the journal. to be associated with an increased expression of pro-inflammatory cytokines IL-6 and IL-8 produced by PBMCs of RA patients. These findings add to the previously described functional effect of the IL-32 promoter SNP on HDLc concentrations within RA patients. Nevertheless, we were unable to show an association between the promoter SNP and disease activity and clinical response to adalimumab and etanercept.

However, most interestingly, we were able to show a link between the promoter SNP and the ex vivo induced cytokine production of IL-1β in RA patients that clinically responded to etanercept (or adalimumab).

Therefore, we suggest that the exploration of the role of IL-32 in RA should be continued in future research, focussing more specifically on treatment response, inflammation induced cardiovascular disease burden and cholesterol metabolism abnormalities.

58 59 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

REFERENCES

1. Selmi C, Generali E, Massarotti M, et al. New treatments for inflammatory rheumatic disease. Immunol 29. Heinhuis B, Koenders MI, van de Loo FA, et al. Inflammation-dependent secretion and splicing of Res IL- 32{gamma} in rheumatoid arthritis. Proc Natl Acad Sci U S A 2011;108(12):4962-7. doi: 10.1073/ 2. 2014;60(2-3):277-88. doi: 10.1007/s12026-014-8565-5 pnas.1016005108 3. Harris ED, Jr. Rheumatoid arthritis. Pathophysiology and implications for therapy. N Engl J Med 30. Gui M, Zhang H, Zhong K, et al. Clinical significance of interleukin-32 expression in patients with 4. 1990;322(18):1277-89. doi: 10.1056/NEJM199005033221805 rheumatoid arthritis. Asian Pac J Allergy Immunol 2013;31(1):73-8. 5. Gabriel SE. Heart disease and rheumatoid arthritis: understanding the risks. Ann Rheum Dis 2010;69 31. van Gestel AM, Haagsma CJ, van Riel PL. Validation of rheumatoid arthritis improvement criteria Suppl 1:i61-64. doi: 10.1136/ard.2009.119404 that include simplified joint counts. Arthritis Rheum 1998;41(10):1845-50. doi: 10.1002/1529- 6. Firestein GS. Evolving concepts of rheumatoid arthritis. Nature 2003;423(6937):356-61. doi: 10.1038/ 0131(199810)41:10<1845::AID-ART17>3.0.CO;2-K nature01661 32. Miceli-Richard C, Comets E, Verstuyft C, et al. A single tumour necrosis factor haplotype influences 7. Dhawan SS, Quyyumi AA. Rheumatoid arthritis and cardiovascular disease. Curr Atheroscler Rep the response to adalimumab in rheumatoid arthritis. Ann Rheum Dis 2008;67(4):478-84. doi: 10.1136/ 3 8. 2008;10(2):128-33. ard.2007.074104 9. Daien CI, Morel J. Predictive factors of response to biological disease modifying antirheumatic drugs: 33. Padyukov L, Lampa J, Heimburger M, et al. Genetic markers for the efficacy of tumour necrosis factor towards personalized medicine. Mediators Inflamm2014;2014:386148. doi: 10.1155/2014/386148 blocking therapy in rheumatoid arthritis. Ann Rheum Dis 2003;62(6):526-9. 10. Mateen S, Zafar A, Moin S, et al. Understanding the role of cytokines in the pathogenesis of rheumatoid 34. Davila-Fajardo CL, Marquez A, Pascual-Salcedo D, et al. Confirmation of -174G/C interleukin-6 gene arthritis. Clin Chim Acta 2016;455:161-71. doi: 10.1016/j.cca.2016.02.010 promoter polymorphism as a genetic marker predicting antitumor necrosis factor treatment outcome. 11. Feldmann M, Brennan FM, Maini RN. Role of cytokines in rheumatoid arthritis. Annu Rev Immunol Pharmacogenet Genomics 2014;24(1):1-5. doi: 10.1097/FPC.0000000000000013 12. 1996;14:397-440. doi: 10.1146/annurev.immunol.14.1.397 35. Harrison P, Pointon JJ, Chapman K, et al. Interleukin-1 promoter region polymorphism role in rheumatoid 13. McInnes IB, Schett G. Cytokines in the pathogenesis of rheumatoid arthritis. Nat Rev Immunol arthritis: a meta-analysis of IL-1B-511A/G variant reveals association with rheumatoid arthritis. 14. 2007;7(6):429-42. doi: 10.1038/nri2094 Rheumatology (Oxford) 2008;47(12):1768-70. doi: 10.1093/rheumatology/ken374 15. Taylor PC, Feldmann M. Anti-TNF biologic agents: still the therapy of choice for rheumatoid arthritis. Nat 36. Popa C, Barrera P, Joosten LA, et al. Cytokine production from stimulated whole blood cultures in Rev Rheumatol 2009;5(10):578-82. doi: 10.1038/nrrheum.2009.181 rheumatoid arthritis patients treated with various TNF blocking agents. Eur Cytokine Netw 2009;20(2):88- 16. Brennan FM, McInnes IB. Evidence that cytokines play a role in rheumatoid arthritis. J Clin Invest 93. doi: 10.1684/ecn.2009.0150 17. 2008;118(11):3537-45. doi: 10.1172/JCI36389 37. Kayakabe K, Kuroiwa T, Sakurai N, et al. Interleukin-1beta measurement in stimulated whole blood 18. Brennan FM, Chantry D, Jackson A, et al. Inhibitory effect of TNF alpha antibodies on synovial cell cultures is useful to predict response to anti-TNF therapies in rheumatoid arthritis. Rheumatology interleukin-1 production in rheumatoid arthritis. Lancet 1989;2(8657):244-7. (Oxford) 2012;51(9):1639-43. doi: 10.1093/rheumatology/kes094 19. Kim SH, Han SY, Azam T, et al. Interleukin-32: a cytokine and inducer of TNFalpha. Immunity 20. 2005;22(1):131-42. doi: 10.1016/j.immuni.2004.12.003 21. Joosten LA, Netea MG, Kim SH, et al. IL-32, a proinflammatory cytokine in rheumatoid arthritis. Proc Natl Acad Sci U S A 2006;103(9):3298-303. doi: 10.1073/pnas.0511233103 22. Heinhuis B, Koenders MI, van de Loo FA, et al. IL-32gamma and Streptococcus pyogenes cell wall fragments synergise for IL-1-dependent destructive arthritis via upregulation of TLR-2 and NOD2. Ann Rheum Dis 2010;69(10):1866-72. doi: 10.1136/ard.2009.127399 23. Hong J, Bae S, Kang Y, et al. Suppressing IL-32 in monocytes impairs the induction of the proinflammatory cytokines TNFalpha and IL-1beta. Cytokine 2010;49(2):171-6. doi: 10.1016/j.cyto.2009.10.003 24. Damen MS, Agca R, Holewijn S, et al. IL-32 promoter SNP rs4786370 predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis. Sci Rep 2017;7:41629. doi: 10.1038/srep41629 25. Damen M, Heinhuis B, Tweehuysen L, et al. SAT0025 Shift in Genetic Composition of an IL-32 Promoter Polymorphism Resuls in a Higher Cytokine Production in RA Patients. Annals of the Rheumatic Diseases 2015;74(Suppl 2):657-58. doi: 10.1136/annrheumdis-2015-eular.5540 26. Li D, Chen D, Zhang X, et al. c-Jun N-terminal kinase and Akt signalling pathways regulating tumour necrosis factor-alpha-induced interleukin-32 expression in human lung fibroblasts: implications in airway inflammation. Immunology 2015;144(2):282-90. doi: 10.1111/imm.12374 27. Sorrentino C, Di Carlo E. Expression of IL-32 in human is related to the histotype and metastatic phenotype. Am J Respir Crit Care Med 2009;180(8):769-79. doi: 10.1164/rccm.200903- 0400OC 28. Hong JT, Son DJ, Lee CK, et al. Interleukin 32, inflammation and cancer. Pharmacol Ther 2017;174:127-37. doi: 10.1016/j.pharmthera.2017.02.025

60 61 Chapter 3 – Genetic variant in IL-32 and ex vivo cytokine production

Supplementary Figure/tables Supplementary Table T1

A B Clinical Responders to Etanercept CC TT Candida Etan Candida Etan IL-1β 84,40 84,88 Percentage red (non-responders) 25,00 26,67 Percentage green (responders) 75,00 73,33 Pam3Cys Etan Pam3Cys Etan IL-1β 163,57 97,26 Percentage red (non-responders) 100 40,00 3 Percentage green (responders) 0,00 53,33

C Table T1. Overview of the percentage of IL-1β production corrected for Candida IgG or Pam3Cys IgG (as 100%) ex vivo stimulation in PBMCs of RA patients clinically responding to etanercept treatment and stratified for the IL-32 promoter SNP (n=4 CC versus n=15).

Supplementary Table T2

Clinical responders to Adalimumab CC TT Candida Adal Candida Adal IL-1β 99,74 95,95 Percentage red (non-responders) 0 62,50 Percentage green (responders) 100 37,50 Figure S1. Production ofIL-1β induced by Candida IgG stimulation of PBMCs from RA patients in the Pam3Cys Adal Pam3Cys Adal presence of anti- TNFα treatment (adalimumab). A) IL-1β cytokine production after ex vivo stimulation of PBMCs from RA patients with additional adalimumab, stratified for RA patients that clinically IL-1β 59,37 91,05 responded or not to adalimumab (n=8 non- responders; n=9 responders). B) IL-1β cytokine production Percentage red (non-responders) 0 25,00 after ex vivo stimulation of PBMCs from RA patients with additional adalimumab, stratified for clinical Percentage green (responders) 100 62,50 responders to adalimumab treatment and IL-32 promoter SNP (n=1 CC; n=8 TT RA patients). C) IL-1β cytokine production after ex vivo stimulation of PBMCs from RA patients with additional adalimumab, Table T2. Overview of the percentage of IL-1β production corrected for Candida IgG or Pam3Cys IgG (as stratified for clinical non-responders to adalimumab and IL-32 promoter SNP genotype (n=3 CC and 100%) ex vivo stimulation in PBMCs of RA patients clinically responding to adalimumab treatment and n=5 TT RA patients). stratified for the IL-32 promoter SNP (n=1 CC versus n=8).

62 63 CHAPTER 4 Human IL-32 transgenic mice develop adipokine profiles resembling those of obesity- 4 induced metabolic changes

Manuscript in preparation

Michelle S.M.A. Damen Dov Ballak Zackary Sapinsley Xiyuan Bai Ed Chan Douglas R. Seals Charles A. Dinarello Calin D. Popa Leo A.B. Joosten

64 65 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

ABSTRACT INTRODUCTION

Low-grade inflammation has been suggested to be involved in the development of Over the years obesity has become a major problem worldwide with about 1.9 insulin resistance in obese subjects. The present study aims to provide additional billion adults aged 18years or older being overweight and 650 million being obese1. evidence strengthening the role of interleukin (IL)-32 in this key process. Using Being overweight is a strong indicator for the development of type II diabetes an IL-32 transgenic (IL-32tg) mice model, we observed that IL-32tg fed a normal (T2D) and other related complications such as coronary artery disease (CAD) and diet had a higher body weight, due to more white adipose tissue (WAT) displaying atherosclerosis, leading to an increased morbidity and mortality2. A state of chronic larger adipocytes histologically. These changes have metabolic consequences, with low-grade inflammation has been documented in obesity. In case of overnutrition, significant higher leptin levels and a trend towards hyperinsulinaemia and normal lipids will accumulate in adipocytes causing an expansion of the adipose tissue (AT) glycaemia, suggesting a certain degree of pre-diabetic insulin resistance state. In which will activate c-Jun N-terminal kinase (JNK) and nuclear factor-kappa B (NF- addition, adipocytes of IL-32tg mice were more prone to induce a pro-inflammatory κB) signalling pathways and might alter the production of various adipokines such inflammatory response locally, which would further contribute to the development as pro- inflammatory cytokines interleukin-6 (IL-6), tumor necrosis factor alpha of insulin resistance and type2 diabetes mellitus (T2D). In conclusion, the results of (TNFα), adiponectin and leptin3-6. Furthermore, chemokines are released, which our study provide for the first time evidence of a direct contribution of IL-32 to the results in infiltration of AT, maintaining the low- grade inflammatory pathophysiology of insulin resistance and T2D, rendering IL-32 as a novel therapeutic state7. This low-grade inflammation at the AT level may also affect other primary 4 target for this 21st century major health problem. metabolic tissues such as the liver and skeletal muscle, known for their important role in glucose homeostasis under insulin actions, contributing to a lower insulin sensitivity under normal glucose concentrations8,9.

Therefore, the inflammatory state in adipose tissue is suggested to be the most important link between increased adipose tissue mass and insulin resistance in obese patients. The precise mechanism linking inflammation/cytokines in adipose tissue to insulin resistance and the occurrence of T2D are currently still under investigation.

Previous studies indicated that TNFα produced by the adipocytes was able to induce insulin resistance in animal models, while blocking TNFα showed improved insulin resistance in the same settings2,10.

Nevertheless, similar interventions targeting inflammation in humans either didn’t pass clinical trials or showed mixed results, underlying the complexity and importance to further evaluate the link between inflammation and diabetes2,11. Interleukin (IL)- 32 is a recently described intracellular cytokine acting as an important regulator of TNF production and other inflammatory processes12-15. It plays a pathogenic role in various inflammatory diseases including rheumatoid arthritis, which in turn is likely to be associated with a higher prevalence of metabolic syndrome16-19. It has been recently suggested that IL-32 may play an important role in the pathogenesis of obese-associated insulin resistance. IL-32 expression was higher in visceral and subcutaneous AT from obese subjects and in monocytes exposed to adipocyte- conditioned media obtained from obese subjects as compared to lean volunteers20.

66 67 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

A link with other inflammatory cytokines such as TNF and IL-6 has been suggested Blood measurments as responsible for these effects. Nevertheless, overexpression of IL- 32γ, a more Tail vein blood (140-180μL) was collected at the beginning of the study and after potent isoforms, in a streptozotocin-induced type 1 diabetic mice model was able to 8weeks, in EDTA-coated tubes for later use. Circulating concentrations of leptin, contribute to initial islet β-cell injury21. adiponectin, insulin, glucose, cholesterol and free fatty acids (FFA) were measured using commercially available ELISA kits. The aim of the present study is to provide additional evidence strengthening the role of IL-32 in the development of obesity and insulin resistance and possible T2D. Using Histology a humanized IL-32 transgenic mice model, we searched for direct evidence of IL-32 Epididymal white adipose tissue (eWAT) were dissected, weighed, and parts were implication in obesity-associated insulin resistance and intermediary metabolism fixed in 4% formaldehyde until further processing. Morphometry of fat cells was homeostasis. assessed by the use of a digital image analysis method (KS-400 software; Zeiss Axiophoto microscope, 40x magnification).

MATERIAL AND METHODS ANIMALS Rna isolation and gene expression Total RNA was isolated from animal epididymal white adipose tissue (eWAT) 10 (12-16weeks old) Male β-actin IL32tg, obtained from dr. Xiyuan Bai and 10 or cultured cells using TRIzol Reagent (Invitrogen, Carlsbad, CA) following the 4 (12-16weeks old) male WT mice on a C57BL/6 background obtained from Jackson manufacturer’s instructions. RNA was reverse-transcribed (iScript cDNA Synthesis Laboratories were housed with 3-4 mice per cage in filter top cages with water and Kit; Bio-Rad Laboratories) and real-time PCR was performed using specific primers food ad libitum. IL-32tg mice were generated as previously described22. Briefly, the (see table 1) and Power Sybr Green PCR master mix (Applied Biosystems) using a open reading frame (ORF) of IL-32γ cDNA was transferred into pCAGGS expression the Step-one Real-Time PCR system (Applied Biosystems, Foster City, CA). Melt vector, which carried β- actin promoter that drives gene expression in all tissues. curve analysis was included to assure a single PCR product was formed. Values were Next, the complete sequence was injected into mouse zygotes of the C57BL/6 strain corrected using the housekeeping gene 36B4. to generate IL-32 transgenic mice. As control mice, wild-type littermates were included. They were all fed a regular low fat diet (Research Diet Services, D15030305) Statistical analysis for 16 weeks. The housing temperature was held at 23°C and a 12:12h light-dark Graphs were created with GraphPad 5.03 Prism software. All mRNA results are cycle was maintained. Bodyweight, blood triglyceride levels, glucose and insulin expressed as relative expression means ± SEM. ELISA data were analyzed by the were monitored at the end of the study. Additionally, food intake and bodyweight Mann-Whitney U-test, with p<0,05 as the minimum level of significance. were measured throughout the course of the study. Food intake was monitored on a per cage basis. Heart, liver, blood and epididymal white adipose tissue (eWAT) were dissected, weighed, and immediately frozen in liquid nitrogen. All animal RESULTS procedures were reviewed and approved by the National Jewish Health Institutional Animal Care and Use Committee (IACUC). Increased epididymal white adipose tissue weight and cell size in IL32tg mice independent of bodyweight Body weight from both IL-32tg and C57BL/6 WT mice Cytokine measurements were measured at the initiation of the study and with a repeated measurement Murine Leptin (10x), Adiponectin (2000x), CXCL-1, IL-1β, IL-6, TNFα, IL-10 and IL- every week till the end of the study at 18weeks. Figure 1A shows that the IL-32tg 1RA were determined by standard sandwich ELISA. ELISA kits for mice were used mice had a higher bodyweight from the initiation of the experiments, compared to according to the manufacturer instructions (R&D systems, Minneapolis, MN, USA). the C57BL/6 WT mice. Over time the difference in bodyweight did not change much between groups even though IL-32tg mice appear to eat slightly more compared to WT mice (Fig. 1B-C). In order to further explain the differences in body weight, the weight of various tissues was measured. IL-32tg mice had significantly heavier

68 69 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

A B eWAT than C57BL/6 WT mice (Fig 1D). No differences were observed between  "*-%(  !%*)(" -%( C57BL/6 WT mice and IL-32tg mice in the weight of the heart, spleen or liver,      respectively (Supplemental Fig. 1 A-C). When we further explored the histology of  

AT we discovered that adipocytes in eWAT were significantly enlarged in IL-32tg '&  -# &     mice compared to C57BL/6 WT mice (Fig, 1E-F). ' (*# -% *#  %( "  *- 

)(   A B "  %* ' -, #!"*'+ (*#% #&*#%

  !     

        -#   -# 

!    *  

!"  ! * # * C D )

&     (,.&%( &. )," ( -,   #&( '   



     *! &  #% , $)  & '& 4

# 

C D '') (  '' #&*$    "   &%(

  ,.  . ),     (   -

& ! !   "*  $ #!

&*  

   -#   -#  '' #    E F   *!   *!  $)&",-"+)&    &

E F & 

& &

'')  &

+) '')  ",-"   $)&

    -#   -#

Fig 2. Circulating levels of leptin, adiponectin, insulin, glucose, cholesterol, FFA, IL-32tg versus C57BL/6 WT mice. (A)Circulating leptin levels in IL-32tg mice versus C57BL/6 WT mice. (B) Circulating adiponectin levels in IL-32tg mice versus C57BL/6 WT mice. (C) Circulating insulin levels in IL-32tg mice versus C57BL/6 WT mice. (D) Circulating Glucose levels in IL-32tg mice versus C57BL/6 WT mice. (E) Circulating cholesterol levels in IL-32tg mice versus C57BL/6 WT mice. (F) Circulating FFA levels in IL- Fig 1. Various weight measurements in IL-32tg mice versus C57BL/6 WT mice. (A) Body weight per 32tg mice versus C57BL/6 WT mice. group. (B) Difference in weight gain over the course of time. (C) Food intake between groups. (D) Tissue weights of eWAT in IL- 32tg mice versus C57BL/6 WT mice. (E) eWAT cell size in C57BL/6 WT mice. (F) eWAT cell size in IL-32tg mice. (magnification scale 40x)

70 71 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

A B         Mice overexpressing human IL-32 develop functional changes    

 

  resembling metabolic syndrome  

     After observing that IL-32tg mice are heavier due to more eWAT and larger       adipocytes.we were interested to explore whether this may have pathophysiological   

   consequences. Therefore, we measured circulating levels of various compounds     mirroring the intermediary metabolism homeostasis. As mentioned before,       adipokines are key molecules involved in the metabolic processes at AT level. In our study, we indicated that the circulating leptin levels were significantly elevated C D     within the IL-32tg mice compared to the WT mice (p=0.003)(Fig. 2A). No significant    

 

 

differences have been observed in adiponectin, insulin and glucose levels, yet a trend      towards higher insulin concentrations in the IL-32tg has been depicted, suggesting   

   a certain degree of insulin resistance with compensatory hyperinsulinemia in these  

 

mice (Figure 2 B-D). Finally, circulating cholesterol and FFA have been measured but    4 showed no significant difference between the two groups (Fig. 2E-F), with a slight         trend towards higher cholesterol levels in IL-32tg mice. Fig 3. Cytokine measurements in eWAT of IL-32tg mice versus C57BL/6 WT mice. (A) IL-6 production in Human IL-32 transgenic mice have an altered profile of inflammatory eWAT tissue of IL-32tg mice versus C57BL/6 WT mice. (B) CXCL-1 production in eWAT tissue of IL-32tg mice versus C57BL/6 WT mice. (C) IL-1Ra production in eWAT tissue of IL-32tg mice versus C57BL/6 mediators in eWAT WT mice. Because the histological changes of eWAT in IL-32tg mice turned out to have metabolic consequences, we were interested to explore whether these changes are favouring the development of a low-grade inflammation at eWAT level, as observed levels and a trend towards hyperinsulinaemia with normal glycaemia, suggesting in obesity-induced T2D patients. Accordingly, spontaneous cytokine production of a decreased glucose tolerance characteristic to pre-diabetes stages. Finally, we eWAT has been assessed. There was an increased production of pro-inflammatory observed that eWAT of IL-32tg mice is able to elicit a stronger pro-inflammatory cytokine IL-6 (p=0.049) and CXCL-1 in unstimulated eWAT of IL-32tg response, allegedly contributing to the low-grade inflammation described in AT of mice compared to WT mice (Fig. 3A-B). T2D patients.

Production of other pro-inflammatory cytokines such as IL-1β and TNFα were Inflammation in AT has been indicated to play a role in obesity-associated metabolic almost undetectable. Additionally, anti-inflammatory cytokines IL-1Ra and IL-10 changes23. Various inflammatory cytokines have been studied24-26, however the were determined, showing no difference in IL-1RA between the two groups, but role of IL-32 in this field is relatively new. IL- 32 is known for its pro-inflammatory significantly lower levels of IL-10 in the IL-32tg mice (p=0.013) (Fig 3C-D). capacities in many different diseases27,28. One of the chronic inflammatory diseases in which IL-32 is known to play a role is rheumatoid arthritis (RA). Increased IL-32 concentrations were correlated to disease activity and a promoter SNP in IL-32 was DISCUSSION found to be associated with high-density lipoprotein cholesterol concentrations in RA patients29,30. Furthermore RA is associated with increased prevalence of In the present study, we show for the first time that eWAT of IL-32tg mice on regular metabolic syndrome which could suggest that IL-32 could play a role herein31,32. chow diet displays morphological and functional changes favouring the development of insulin resistance and T2D. The eWAT in IL-32tg is heavier compared to WT mice, Following up on the findings of Catalán V et al., we were able to show that IL-32tg most probably due to larger adipocytes in IL-32tg mice. This resulted in higher leptin mice were heavier than the WT controls, and this difference remained unchanged

72 73 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

throughout the study period despite a slight increase in food intake in the IL-32tg mice. This is in line with the observation that IL-32 concentrations were higher in To conclude, the present study shows for the first time that there is a direct obese patients and decreased after weight loss20. Interestingly, Lee et al. showed association between IL-32 and intermediary metabolism in mice, and that AT might that mice overexpressing IL-32β on a high-fat diet were protected against hepatic play an important role in this processes. Our data corroborated with the recent steatosis and inflammation33. In contrast, overexpression of IL-32γ, a more potent results of Catalan further underline the hypothesis that IL-32 could indeed be an isoforms, in a streptozotocin-induced type 1 diabetic mice model contributed to essential player in promoting obesity and obesity-induced comorbidities. initial islet β-cell injury21 Various isoforms of IL-32 seem to play different roles in inflammation and changes to metabolic processes.

The increased eWAT mass and adipocytes size were likely to have functional consequences in the IL-32tg mice. We found a significant increase in circulating leptin concentrations in mice which could be the beginning of a leptin resistant state. Previous studies have shown that leptin production regulates energy balance by inhibiting hunger and lowering appetite (REF). However, in obesity/metabolic syndrome, the increased adipose tissue disturbs adipokine regulation which results 4 in a low-grade inflammatory state. Expression of pro-inflammatory mediators such as leptin and IL-6 are upregulated whereas anti-inflammatory mediators including adiponectin are reduced34. Moreover, increased leptin production in obese individuals was linked to leptin resistance due to the inadequate response to lower for example food intake. Additionally, it has been suggested that leptin resistance is associated with impaired transport of leptin across the blood brain barrier (BBB) leading to accumulation of triglycerides in adipose tissue, liver and pancreas which results in decreased insulin sensitivity35. Pancreatic β-cell receptors may show a decreased responsiveness in the presence of chronically induced leptin levels which results in increased insulin secretion. Hyperinsulineamia in turn increases obesity leading to a further increased leptin production and creating a positive pro-diabetic feedback loop36.

Besides changes in leptin and insulin concentrations, changes in cytokine profile and inflammatory state are also linked to obesity and metabolic syndrome even though the exact role has not been clearly established23,37. IL-32 is known to be able to induce other pro-inflammatory cytokines including IL-6, IL-1β and TNFα. Our findings of increased IL-6 and CXCL-1 but decreased anti-inflammatory cytokine IL- 10 are therefore in line with previous data. Furthermore, Catalán V et al. showed that pro-inflammatory cytokines including IL-6 and TNFα were decreased when IL-32α was silenced, showing a role for IL-32 in the low-grade inflammatory status of obese individuals. Since IL-32 was found to be elevated in AT of obese patients, IL-32 seems to be part of the positive feedback loop maintaining the inflammatory status in AT of these patients.

74 75 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

REFERENCES

1. WHO. Obesity and Overweight. WHO fact sheet. 2017. 22. Choi J, Bae S, Hong J, Ryoo S, Jhun H, Hong K, et al. Paradoxical effects of constitutive human IL- 2. Lontchi-Yimagou E, Sobngwi E, Matsha TE, Kengne AP. Diabetes mellitus and inflammation. Curr Diab 32{gamma} in transgenic mice during experimental colitis. Proc Natl Acad Sci U S A. 2010;107(49):21082- Rep. 2013;13(3):435-44. 6. 3. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest. 2006;116(7):1793- 801. 23. Tack CJ, Stienstra R, Joosten LA, Netea MG. Inflammation links excess fat to insulin resistance: the role of 4. Sharma M, Vikram NK, Misra A, Bhatt S, Tarique M, Parray HA, et al. Assessment of 11-beta the interleukin-1 family. Immunol Rev. 2012;249(1):239-52. hydroxysteroid dehydrogenase (11-betaHSD1) 4478T>G and tumor necrosis factor-alpha (TNF-alpha)- 24. Netea MG, Joosten LA, Lewis E, Jensen DR, Voshol PJ, Kullberg BJ, et al. Deficiency of interleukin-18 in 308G>A polymorphisms with obesity and insulin resistance in Asian Indians in North India. Mol Biol Rep. mice leads to hyperphagia, obesity and insulin resistance. Nat Med. 2006;12(6):650-6. 2013;40(11):6261-70. 25. Suganami T, Nishida J, Ogawa Y. A paracrine loop between adipocytes and macrophages aggravates 5. Coppack SW. Adipose tissue changes in obesity. Biochem Soc Trans. 2005;33(Pt 5):1049-52. inflammatory changes: role of free fatty acids and tumor necrosis factor alpha. Arterioscler Thromb Vasc 6. Kroder G, Bossenmaier B, Kellerer M, Capp E, Stoyanov B, Muhlhofer A, et al. Tumor necrosis factor- Biol. 2005;25(10):2062-8. alpha- and hyperglycemia-induced insulin resistance. Evidence for different mechanisms and different 26. Gao D, Madi M, Ding C, Fok M, Steele T, Ford C, et al. Interleukin-1beta mediates macrophage- effects on insulin signaling. J Clin Invest. 1996;97(6):1471-7. induced impairment of insulin signaling in human primary adipocytes. Am J Physiol Endocrinol Metab. 7. Sell H, Eckel J. Chemotactic cytokines, obesity and type 2 diabetes: in vivo and in vitro evidence for a 2014;307(3):E289-304. possible causal correlation? Proc Nutr Soc. 2009;68(4):378-84. 27. Damen M, Popa CD, Netea MG, Dinarello CA, Joosten LAB. Interleukin-32 in chronic inflammatory 8. Petersen KF, Dufour S, Savage DB, Bilz S, Solomon G, Yonemitsu S, et al. The role of skeletal muscle insulin conditions is associated with a higher risk of cardiovascular diseases. Atherosclerosis. 2017;264:83-91. resistance in the pathogenesis of the metabolic syndrome. Proc Natl Acad Sci U S A. 2007;104(31):12587- 28. Hong JT, Son DJ, Lee CK, Yoon DY, Lee DH, Park MH. Interleukin 32, inflammation and cancer. Pharmacol 4 94. Ther. 2017;174:127-37. 9. Muniyappa R, Sowers JR. Role of insulin resistance in endothelial dysfunction. Rev Endocr Metab Disord. 29. Joosten LA, Netea MG, Kim SH, Yoon DY, Oppers-Walgreen B, Radstake TR, et al. IL-32, a proinflammatory 2013;14(1):5-12. cytokine in rheumatoid arthritis. Proc Natl Acad Sci U S A. 2006;103(9):3298-303. 10. Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose Expression of Tumor-Necrosis-Factor-Alpha - 30. Damen MS, Agca R, Holewijn S, de Graaf J, Dos Santos JC, van Riel PL, et al. IL-32 promoter SNP rs4786370 Direct Role in Obesity-Linked Insulin Resistance. Science. 1993;259(5091):87-91. predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis. Sci Rep. 2017;7:41629. 11. Goldfine AB, Fonseca V, Shoelson SE. Therapeutic Approaches to Target Inflammation in Type 2 Diabetes. 31. Zhang J, Fu L, Shi J, Chen X, Li Y, Ma B, et al. The risk of metabolic syndrome in patients with rheumatoid Clin Chem. 2011;57(2):162-7. arthritis: a meta-analysis of observational studies. PLoS One. 2013;8(10):e78151. 12. Kim SH, Han SY, Azam T, Yoon DY, Dinarello CA. Interleukin-32: A cytokine and inducer of TNF alpha. 32. Ferraz-Amaro I, Gonzalez-Juanatey C, Lopez-Mejias R, Riancho-Zarrabeitia L, Gonzalez-Gay MA. Immunity. 2005;22(1):131-42. Metabolic syndrome in rheumatoid arthritis. Mediators Inflamm. 2013;2013:710928. 13. Joosten LA, Dinarello CA, Van den Berg WB. Il-32, a novel challenge in rheumatoid arthritis. Arthritis 33. Lee DH, Hong JE, Yun HM, Hwang CJ, Park JH, Han SB, et al. Interleukin-32beta ameliorates metabolic Rheum. 2006;54(9):S301-S2. disorder and liver damage in mice fed high-fat diet. Obesity (Silver Spring). 2015;23(3):615-22. 14. Zhou YQ, Zhu Y. Important Role of the IL-32 Inflammatory Network in the Host Response against Viral 34. Cornier MA, Dabelea D, Hernandez TL, Lindstrom RC, Steig AJ, Stob NR, et al. The metabolic syndrome. Infection. Viruses-Basel. 2015;7(6):3116-29. Endocr Rev. 2008;29(7):777-822. 15. Nold-Petry CA, Nold MF, Zepp JA, Kim SH, Voelkel NF, Dinarello CA. IL-32-dependent effects of IL-1 beta 35. Saito K, Tobe T, Yoda M, Nakano Y, Choi-Miura NH, Tomita M. Regulation of gelatin-binding protein 28 on endothelial cell functions. P Natl Acad Sci USA. 2009;106(10):3883-8. (GBP28) gene expression by C/EBP. Biol Pharm Bull. 1999;22(11):1158-62. 16. Heinhuis B, Koenders MI, van Riel PL, de Loo FAV, Dinarello CA, Netea MG, et al. Tumour necrosis factor 36. Thorand B, Zierer A, Baumert J, Meisinger C, Herder C, Koenig W. Associations between leptin and the alpha-driven IL-32 expression in rheumatoid arthritis synovial tissue amplifies an inflammatory cascade. leptin / adiponectin ratio and incident Type 2 diabetes in middle-aged men and women: results from the Ann Rheum Dis. 2011;70(4):660-7. MONICA / KORA Augsburg study 1984-2002. Diabet Med. 2010;27(9):1004-11. 17. Joosten LAB, Dinarello CA, Van den Berg WB. IL-32, a novel target in rheumatoid arthritis. Ann Rheum 37. Netea MG, Joosten LA. The NLRP1-IL18 Connection: A Stab in the Back of Obesity-Induced Inflammation. Dis. 2006;65:108-. Cell Metab. 2016;23(1):6-7. 18. Parra-Salcedo F, Contreras-Yanez I, Elias-Lopez D, Aguilar-Salinas CA, Pascual-Ramos V. Prevalence, incidence and characteristics of the metabolic syndrome (MetS) in a cohort of Mexican Mestizo early rheumatoid arthritis patients treated with conventional disease modifying anti-rheumatic drugs: the complex relationship between MetS and disease activity. Arthritis Res Ther. 2015;17. 19. Dao HH, Do QT, Sakamoto J. Increased frequency of metabolic syndrome among Vietnamese women with early rheumatoid arthritis: a cross-sectional study. Arthritis Res Ther. 2010;12(6). 20. Catalan V, Gomez-Ambrosi J, Rodriguez A, Ramirez B, Valenti V, Moncada R, et al. Increased Interleukin-32 Levels in Obesity Promote Adipose Tissue Inflammation and Extracellular Matrix Remodeling: Effect of Weight Loss. Diabetes. 2016;65(12):3636-48. 21. Jhun H, Choi J, Hong J, Lee S, Kwak A, Kim E, et al. IL-32gamma overexpression accelerates streptozotocin (STZ)-induced type 1 diabetes. Cytokine. 2014;69(1):1-5.

76 77 Chapter 4 – IL-32tg mice and obesity-induced metabolic changes

SUPPLEMENTARY FIGURES

A B               

  

       

 

   

C          

  4 



 

Supplementary Figure S1. Weight of several organs in wild-type mice versus IL32tg mice. A). Weight of the heart in wild-type mice versus IL32tg mice. B) Weight of the spleen in wild-type mice versus IL32tg mice. C) Weight of the liver in wild-type mice versus IL32tg mice.

78 79 CHAPTER 5 Interleukin-32 upregulates the expression of ABCA1 and ABCG1 resulting in reduced intracellular lipid concentrations in primary human hepatocytes 5

In Press

Michelle S.M.A. Damen1 Jéssica Cristina dos Santos1,2 Rob Hermsen3 J. Adam van der Vliet4 Mihai G. Netea1 Niels P. Riksen1 Charles A. Dinarello1,5 Leo A.B. Joosten1* Bas Heinhuis1*

80 81 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

ABSTRACT INTRODUCTION

Background and aims: The role of interleukin (IL-)32 in inflammatory conditions is Cardiovascular diseases (CVD) are currently the leading cause of death in developed well-established, however the mechanism behind its role in atherosclerosis remains countries with atherosclerosis as the most important contributor to the disease unexplained. Our group reported a promoter single nucleotide polymorphism in burden1,2. Atherosclerosis is characterized by inflammation and the accumulation of IL-32 associated with higher high-density lipoprotein (HDL) concentrations. We lipids in the vessel wall causing plaque formation3. Additionally, triggers such as, hypothesize that endogenous IL-32 in liver cells, a human monocytic cell line and smoking, hypertension, dyslipidemia and hyperglycemia are known to contribute to carotid plaque tissue, can affect atherosclerosis by regulating (HDL) cholesterol plaque formation3-5. Ongoing inflammation in the plaque further activates monocytes homeostasis via expression of cholesterol transporters/mediators. to differentiate into macrophages, which will take up lipids to generate foam cells. Moreover, many studies have indicated that pro-inflammatory cytokines including Methods: Human primary liver cells were stimulated with recombinant human (rh) TNFα, IL-1β, IL-6, IFNγ, contribute to the development of atherosclerosis6-8. TNFα and poly I:C to study the expression of IL-32 and mediators in cholesterol pathways. Additionally, IL-32 was overexpressed in HepG2 cells and overexpressed Recently, Heinhuis et al. suggested that the intracellular pro-inflammatory cytokine and silenced in THP-1 cells to study the direct effect of IL-32 on cholesterol interleukin (IL)-32 could play an important role in atherosclerosis9. IL-32 has been transporters expression and function. shown to play a role in inflammatory diseases with an increased risk for CVD, such as rheumatoid arthritis (RA) and human immunodeficiency virus (HIV)10-13. IL-32 can be Results: Stimulation of human primary liver cells resulted in induction IL-32α, spliced into various isoforms with IL-32α, IL-32β and IL-32γ being most intensively IL-32β and IL-32γ mRNA expression (p<0.01). A strong correlation between the studied and IL- 32γ being the most active isoform14. In atherosclerotic plaques, IL32 5 expression of IL-32γ and ABCA1, ABCG1, LXRa and apoA1 was observed (p<0.01) is expressed, and in macrophages IL32 overexpression increases the expression and intracellular lipid concentrations were reduced in the presence of endogenous of chemokine (C-C motif) ligand 2 (CCL2), soluble vascular cell adhesion molecule IL-32 (p<0.05). Finally, IL32γ and ABCA1 mRNA expression were upregulated in (sVCAM-1), matrix metalloproteinase 1 (MMP1), MMP9, and MMP13. IL-32 promotes carotid plaque tissue and when IL-32 was silenced in THP-1 cells, mRNA expression inflammation by induction of pro-inflammatory cytokines like TNFα, IL-6, IL-1β and of ABCA1 was strongly reduced. IL-815, 16. Furthermore, IL-32 seems to be a regulator of endothelial cell function were it enhances IL-1β-induced intracellular adhesion molecule 1 (ICAM-1)17. Conclusion: Regulation of IL-32 in human primary liver cells, HepG2- and THP-1-cells strongly influences the mRNA expression of ABCA1, ABCG1, LXRα and apoA1 and In contrast to these pro-atherogenic actions, IL32 could also have anti- affects intracellular lipid concentrations in the presence of endogenous IL-32. These atherosclerotic effects by increased HDL cholesterol18. Recently, a promoter data, for the first time show an important role for IL32 in cholesterol homeostasis. single nucleotide polymorphism (SNP) in IL32 was found to be associated with HDL cholesterol (HDLc) concentrations in both RA patients as well as individuals with an increased CVD risk, again suggesting a role for IL-32 in CVD18. Individuals homozygous for the C- allele showed higher HDLc concentrations compared to individuals being heterozygous or homozygous for the T- allele. HDLc is considered to be atheroprotective even though Mendelian randomization studies and drug trials showed there is no causality dependent on HDL cholesterol levels but rather atheroprotective HDL functions are more relevant metrics to analyze19,20. Therefore, IL32 could still have an atheroprotective role. The HDL metabolism starts in the liver, which is an important organ in HDL biosynthesis and a regulator of plasma HDL concentrations. Additionally, the small intestine and especially enterocytes are involved in biosynthesis of HDLc. Dietary lipids including, cholesterol esters and

82 83 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

triacylglycerols are first hydrolyzed in the interstitial lumen of the intestine after human serum and penicillin/streptomycin. This study (2014-1453) was reported and which the product are taken up by enterocytes21. These products are re-synthesized approved by our ethical committee of the Radboud University Medical Centre. by the enterocytes and packed into either chylomicrons or HDLc for secretion into the circulation. sMoreover, the biosynthesis of HDL involves synthesis and secretion The human liver cell line HepG2 was cultured in complete Dulbecco’s Modified of apolipoproteins (apoA-I and apoA-II) followed by acquisition of lipids and Eagle’s Medium with glutamax (Gibco) containing 10 % heat-inactivated Fetal Calfs generation of mature HDL22,23. Lipidation of nascent, discoidal apoA-I containing Serum (FCS), pyruvate (Gibco) and gentamycin (Gibco). Trypsin was used to detach HDL particles must occur to form mature HDL. A critical participant herein is the the adherent cells and the cell line was passaged twice a week in a 1:5 ratio. ATP-binding cassette A1 (ABCA1) expressed on liver cells and enterocytes. These nascent, discoidal apoA-I containing HDL particles are secreted via hepatic and The human monocytic cell line THP1 was used to study overexpression and silencing enterocyte ABCA1 and are matured in circulation via the lecithine-cholesterol- of IL-32 and its effect on cholesterol transporters. The cell line was cultured in acyltransferase (LCAT) function and formation of spherical HDL particles24. In the Roswell Park Memorial Institute (RPMI) 1640 medium containing heat-inactivated absence of ABCA1, extremely low levels of HDLc and apoA-I are observed which 10 % FCS, pyruvate and gentamycin. Cells were growing in suspension and passaged contribute to an increased risk for CVD. twice a week in fresh medium.

An important mechanism involved in regulation of excessive cholesterol is reverse Fresh carotid artery plaque tissue was kindly provided by the department of surgery cholesterol transport (RCT). During RCT, excessive cholesterol is transferred from from the Radboud University Medical Center and upon arrival was separated into 4 peripheral tissues and the arterial wall back to the liver for removal from the pieces to perform various analyses. Samples were then stored at -80°C for further use. body25,26. This cholesterol efflux is also mediated via ABCA1 which is induced by activation of liver X receptor alpha (LXRα)27-29. Alternatively, ABCG1 transports Quantitative PCR 5 excessive cholesterol outside the cells30,31. Upon return to the liver, HDLc can be Human primary liver cells were seeded ~200,000 cells per well in a 24 well plate in 0.5 taken up by scavenger receptor class B, type 1 (SR-B1) for degradation of HDL. ml Williams B medium containing 10 % human serum and penicillin/streptomycin. Another alternative pathway by which HDLc is metabolized and transported to the The next day, cells were stimulated with TNF (100 ng/ml) (R&D Systems) or Poly I:C liver is via the cholesteryl ester (CE) transfer protein (CETP). Knowing the effect of (50 µg/ml) (Invivogen) for 24 hours. Subsequently, medium was removed and stored IL-32 on HDLc concentrations and the importance of the role of the liver in HDLc and and 0.5 ml Tri-reagent (Sigma-Aldrich) was added per well. After lysing the cells cholesterol efflux, one can argue a role for IL-32 in HDLc metabolism and synthesis by incubating the cells with Tri-reagent for 30 minutes at room temperature, the in liver cells. However, despite previous studies on the role of IL-32 in cardiovascular solution containing lysed cells were was stored at - 20 °C until further processing. disease, studies investigating the exact mechanism behind how IL-32 influences HepG2 cells and THP-1 cells were lysed in TriZol reagent and also stored for later use. cholesterol homeostasis remain scarce. The present study aims to investigate the Carotid artery tissue was crushed in TriZol reagent using Magnalyser green beads precise role of IL-32 on HDLc homeostasis, focussing on cholesterol transporters (Roche). RNA was isolated as previously described33. After RNA isolation, mRNA involved in this process in human primary liver cells and the human monocytic cell was transformed into cDNA by applying an iScript kit (Bio-Rad) to transform mRNA line THP1. into cDNA. IL-32 primers were previously developed and other primers sequences were extracted from the Harvard Primerbank database34. Primers were produced MATERIAL AND METHODS by Biolegio (Nijmegen, The Netherlands) and a StepOnePlus qPCR system (Applied Biosystems) was used to analyze relative mRNA expression. Relative expression was Culturing primary human liver cells, HepG2 and THP-1 cell lines and calculated by normalizing for GAPDH and 2^-dCt method. carotid artery plaque tissue Patients undergoing liver surgery in our hospital donated human liver tissue. The Western blotting anonymized liver tissue was used to isolate primary human liver cells as previously Twenty-four hours after stimulation, primary liver cells were washed with PBS and reported32. Human liver cells were seeded in Williams B medium containing 10 % lysed with standard cell lysis buffer on ice for 30 minutes. Subsequently, cells were

84 85 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

scraped with a cell scraper and transferred into a safe-lock tube. were boiled was replaced by the same fresh medium and transfection medium containing IL- in Laemmli buffer under denaturing and reducing conditions. SDS-Page gels (12%) 32 plasmids (pCDNA3-IL32, pCDNA3-IL32, pCDNA3-IL32) or a control plasmid were prepared and marker plus proteins were loaded onto the gels. After running (pCDNA3-eGFP) was prepared. Transfection medium per transfection was prepared the gels, proteins were transferred onto nitrocellulose membranes with an iBlot as follow; 1) dissolve 0.5 g plasmid DNA in 25 l serum free RPMI-1640, 2) add 1.5 apparatus (Invitrogen). After transferring the proteins, blots were blocked with 5 % l Fugene HD (Promega) directly to the DNA mixture, 3) vortex 1-2 seconds and milk proteins in Tris-buffered Saline (TBS) with 0.1 % tween-20 (Invitrogen) for at incubate for 15 minutes at room temperature, 4) add 25 l dropwise per well while least 1 hour at room temperature. Next, blots were washed in TBS-T (TBS containing gently shaking the plate, 5) incubate for 24 or 48 hours at 37 °C and 5 % CO2. After the 0.1% tween-20) and incubated overnight on the roller mixer at room temperature incubation, RNA and protein samples were isolated for determining gene expression in TBS-T containing primary antibody against IL-32 (AF3040, R&D Systems), at and IL-32 protein expression as previously described. BODIPY flowcytometry assay a concentration of 0.2 µg/ml, or against ApoA1 (Acris Antibodies GmbH, Herford, Transfected HepG2 cells and THP-1 cells were used for the BODIPY FACS analysis. Germany), at a concentration of 1.0 µg/ml. The following day, blots were washed with In more detail, 1 million (HepG2) or 2,5x10^5 (THP-1) cells per well were used for the TBST and the IL-32 blots were incubated with rabbit-anti-mouse-HRP (Dako P0449, transfection after which cells were detached, spun down and resuspended in PBS 1:5000 in TBST) while the ApoA1 blots were incubated with rabbit-anti-mouse-HRP containing 4 % Formaldehyde (FA) for 15 minutes on ice. Thereafter, the cells were (Dako P0260, 1:5000 in TBST) and incubated for 1 hour at room temperature on spun down again and resuspended in PBS with 5 % BSA and BODIPY (1:500) for 45 the roller mixer. Next, blots were washed and incubated with ECL (GE Health care minutes on ice in the dark, after which the BODIPY signal was measured. Life Sciences) before scanning the blots. Finally, actin was detected on the blots by using anti-actin antibodies (Santa Cruz Biotechnology) and appropriate secondary Overexpression and silencing of IL-32 in THP1 cells antibody (Dako) followed by ECL incubation and scanning of the blots. THP-1 cells (15 x 106 cells/15mL) were differentiated into macrophages (75 cm2 – tissue culture flask; Corner) in RPMI-1640 including 10 % FCS, PMA (Sigma-Aldrich) 5 IL-32 ELISA at 10 ng/mL, β-mercaptoethanol (Sigma-Aldrich) at 50 µM and incubated for 48 Maxisorp plates (Nunc) were coated with AF3040 (R&D Systems) diluted in Phosphate hours at 37 °C and 5 % CO . 2.5 x 106 cells/800 µL were electroporated by using Amaxa Buffered Saline (PBS) at a concentration of 0.4 g/ml and incubated overnight at Nucleofactor technology (Lonza, Basel) according with the protocol described by35. room temperature. Next morning, plates were blocked with PBS containing 1 % For IL-32 knockdown, 1 µg of ON-TARGETplus SMARTpool siRNA per transfection BSA (Sigma-Aldrich) for 1 hour at room temperature. Standard curve was prepared was used or 1 µg of ON-TARGETplus SMARTpool control siRNA (Dharmacon Inc.), by diluting recombinant IL-32 ranging from 5000 pg/ml until 39.06 pg/ml in PBS sequences were describe by36. For IL-32 overexpression, 0.5 µg of pCDNA3 plasmid containing 5 % BSA. Standard curve was added followed by the supernatant samples expressing human IL-32γ of eGFP was used as a control. Transfected cells (3 x 105/100 from the primary liver cells. ELISA plates were incubated for 2 hours on a shaker. µL) were plated on flat-bottom 96-well plates (Costar) and 100 µL of transfection After the incubation, plates were washed and detection antibody was added medium was added. After 4 and 24 h the cell monolayers were collected by adding (BAF3040, R&D Systems), 0.1 µg/ml in PBS with 5 % BSA. Plates were incubated for 200 µL of TRIzol and stored at -80°C until mRNA extraction. Comparable studies 1 hour at room temperature on the shaker. Subsequently, plates were washed and were performed in order to determine transfection (eGFP) efficiency which was streptavidin (R&D Systems) was added and the plates were incubated for 30 minutes around 30%. at room temperature on the shaker. After the last incubation, plates were washed and substrate buffer was added and the color reaction was closely monitored until Foam cell formation of THP-1 cells and human derived percoll the reaction was terminated by adding stop solution. Finally, the plates were read by monocytes a plate-reader and IL-32 concentrations were calculated. THP-1 cells (15 x 106 cells/15mL) were differentiated into macrophages (75 cm2 – tissue culture flask; Corner) in RPMI-1640 including 10 % FCS, PMA (Sigma-Aldrich) Overexpression of IL-32 in HepG2 cell line at 10 ng/mL, β-mercaptoethanol (Sigma-Aldrich) at 50 µM and incubated for 48

Half million HepG2 cells were seeded per well in a 24-well plate in RPMI-1640 including hours at 37 °C and 5 % CO2. 5 x 10^5 cells per well were seeded in 24 wells plate

5 % FCS and incubated overnight at 37 °C and 5 % CO2. The following day, medium and starved for 4h in RPMI-1640 supplemented 2 mM l-glutamine, 1 mM pyruvate

86 87 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

A B C IL-32 mRNA IL-32 mRNA IL32 mRNA ) ) and 50 μg/ml gentamicin (GIBCO Invitrogen, Carlsbad, CA) after which oxidized ) 8 400 1500 * ** ** x1000 H x1000 H x1000 LDL was added (25ug/mL) for 24h at 37 °C and 5 % CO . Human pheripheral blood H 2 6 300 ** Ct Ct Ct PD PD PD -d -d -d 1000 (2 GA (2 mononuclear cells were isolated using density centrifugation over Ficoll-Paque. GA (2 GA r r r on on 4 200 on fo fo fo

ed ed ed essi essi essi 500 37 2 100 rrect rrect ex pr rrect ex pr Next, percoll isolation of monocytes was performed as previously described . In ex pr co co co ve ve ve ti ti ti la la brief, 150-200 x 10^6 PBMCs were layered on top of a hyper-osmotic Percoll solution 0 0 la 0

Re

Re d Re F F F med me and centrifuged for 15 minutes at 580 g. The interphase layer was isolated and cells TN TN med TN Poly I:C Poly I:C Poly I:C were washed with cold PBS. Cells were resuspended in RPMI+++ An extra purification D E F IL-32 protein step was added by adhering Percoll isolated monocytes to polystyrene flat bottom ly I:C WB quantification 800 Medium TNFa Po * * n plates (Corning, NY, USA) for 1h at 37 °C and 5 % CO2; subsequently a washing step on

1 IL-32b 25kDa 600 essi with warm PBS was performed to yield maximal purity. Once, percoll monocytes ] ex pr pg/ ml ve Actin 40kDa [ 400 ti Donor 2 were obtained they were differentiated to macrophages for 6 days in 10% human la orrected for Acti -3 c Re pool serum. On day 6 medium was removed and cells were starved for 4h similar to IL 200 IL-32b 25kDa 2 F the THP-1 cells before being fed oxidized LDL for 24h. After 24h supernatants were med TN 0 Poly I:C Actin 40kDa d F C me TN collected and cells were stored in TRIzol reagent for mRNA isolation. Donor Poly I: Extracellular

3 IL-32b 25kDa Statistics Actin 40kDa Statistical analysis were performed by using the Mann-Whitney U test, Spearman Donor Intracellular correlation test or One-way Anova including Kruskal-Wallis test and Dunn’s Multiple 5 comparison test. In each figure the applied statistical test is indicated. Fig. 1 Induction of IL-32 expression in human primary liver cells. (A) Stimulation with TNFα induced significantly IL-32α (n=7, Mann-Whitney U test, p=0.0070). (B)Significant differences between TNFα and Poly I:C induced IL-32β (n=7, Mann-Whitney U test, p=0.0023). (C) Poly I:C stimulation enhanced RESULTS the expression of IL-32γ both compared with medium control (n=7, Mann-Whitney U test, p=0.0111) or TNFα (n=7, Mann-Whitney U test, p=0.0070). (D) Induction of IL-32β protein by TNFα or Poly I:C in primary liver cells from 3 donors (actin as loading control). (E) Relative expression of L-32β protein Induction of IL-32 isoforms in human primary liver cells by TNFα or TLR3 expression corrected for Actin expression after stimulation (med, TNFα, Poly I:C). (F) IL-32 protein ligand poly I:C expression after TNFα and Poly I:C stimulation in culture supernatants of human primary liver cells Primary liver cells were stimulated with recombinant human TNFα (rhTNFα) or (n=5 (with replicates), Mann-Whitney U test, medium vs Poly I:C p=0.0176; TNFα vs Poly I:C p=0.0471). Poly I:C to study whether these cells were capable of expressing IL-32. Stimulation of cells with rhTNFα resulted in a slight upregulation of IL-32α mRNA expression Induction of IL-32 in human primary liver cells leads to enhanced (Fig. 1A). Moreover, the protein level of IL-32β was determined intracellularly and expression of important regulators of cholesterol homeostasis which was induced mostly by rhTNFα but also Poly I:C stimulation (Fig. 1D-E). Poly I:C correlate strongly with IL-32γ mRNA expression stimulation resulted in a slight increase in IL-32γ mRNA expression but almost no to Here, we explored whether TNFα- or TLR3-poly I:C-induced IL-32 expression in minor increase in either IL-32β or IL-32α respectively compared to medium control human primary liver cells modulates the expression of cholesterol transporters and (Fig. 1A-B-C). Furthermore, Poly I:C stimulation resulted in higher extracellular IL-32 regulators of cholesterol homeostasis. Furthermore, we studied whether there was protein levels (Fig. 1F). a correlation between mRNA expression of IL-32 isoforms and the components involved in the cholesterol homeostasis. As shown in figure 2A-D, stimulation of human primary liver cells with poly I:C resulted in increased expression of ABCA1, ABCG1, ApoA1 and LXRα mRNA. Besides mRNA levels, protein levels of ApoA1 were also studied and showed to be increased after stimulation with poly I:C, and possibly

88 89 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

A(A) (B)B (CC) (DD) after rhTNFα stimulation (Fig. 2E-F). As a positive control to study the function of ABCA1 mRNA ABCG1 mRNA ApoA1 mRNA LXRa mRNA ** ** * ** 2000 1500 3000 * 2000 the cells, we determined the rhTNFα- and poly I:C-induced IL-8 mRNA expression * * * x1000) x1000) x1000) H H H x1000) 1500 H Ct Ct Ct 1500 in the liver cells (supplementary Fig 1A). Additionally, a strong positive correlation Ct PD PD PD PD -d -d -d 1000 2000 -d (2 GA (2 (2 GA GA (2 GA r r r between the IL-32γ isoform and ABCA1, ABCG1, ApoA1 and LXRα expression was r on on on 1000 fo fo fo on 1000

fo d d ed ed essi essi essi te essi observed after the cells were stimulated with poly I:C (Fig.2F-I). Cells stimulated with 500 1000 te 500 500 rrec rrect rrect ex pr ex pr ex pr rrec ex pr co co co co ve ve rhTNFα also showed positive correlations to the same mediators (supplementary ve ve ti ti ti ti la la la 0 0 0 la 0

Re Re Fig 1B-E). These correlations were not observed when looking at IL-32α nor IL-32β Re Re F F F med med med med F TN TN TN TN expression and these mediators (Supplementary Table S1). Poly I:C Poly I:C Poly I:C Poly I:C E(E) F(F)

Endogenous expression of IL-32γ and ABCA1 mRNA in HepG2 cells ly I:C kDa ly I:C WB quantification kDa Med TNFa Po Med TNFa Po To further study the effect of IL-32 isoforms on the expression of cholesterol n on mediators, a human liver cell line was used (HepG2 cell line) (Fig. 3A-D). At first, 70 70 essi ex pr ve unstimulated HepG2 cells were studied to look at the basal mRNA expression levels ti la orrected for Acti

35 c of IL-32 and cholesterol mediators. Similar to the observation made in human 35 Re 25 25 F primary liver cells, also HepG2 cells showed expression of the three isoforms of IL- ApoA 1 med TN ApoA 1 15 Poly I:C 32 (IL-32α, IL-32β and IL-32γ) and ABCA1 (Fig. 3A-D). Interestingly, IL-32γ expression 15 was already upregulated after 4 hours of culture in only DMEM complete medium with 10% FCS hi, together with ABCA1 expression, while IL-32α didn’t show any 40 40

Acti n 5 expression and IL-32β only minor expression (Fig. 3A-B). As shown in figure 4A and Acti n

4B the relative expression IL-32α and IL-32β increased after 24h but resulted in G(G) (H)H (I)I J(J) an even more pronounced expression after 48h of culture. Within the same time, expression of both IL-32γ and ABCA1 decreased (Fig. 3C-D). Highly interesting, a 1 positive correlation between IL-32 isoforms and ABCA1 expression was again mostly 1 LXRα ABCG ApoA 1 observed for the IL-32γ isoform (Fig. 3E and Supplementary figure 2). ABCA

IL-32γ IL-32γ IL-32γ IL-32γ

Fig. 2 Induction of cholesterol transporters ABCA1/ABCG1, HDL particle ApoA1 and transcription factor LXRa in human primary liver cells and their correlation with IL-32γ after Poly I:C stimulation of cells for 24h. (A) Stimulation with Poly I:C significantly induced ABCA1 (n=7, Mann-Whitney U test, medium vs Poly I:C p=0.0041; TNFα vs Poly I:C p=0.011). (B) ABCG1 expression was significantly induced by Poly I:C (n=7, Mann-Whitney U test, medium vs Poly I:C p=0.0023; TNFα vs Poly I:C p=0.0262). (C) HDL particle ApoA1 was significantly induced by Poly I:C (n=7, Mann- Whitney U test, medium vs Poly I:C p=0.0260; TNFα vs Poly I:C p=0.0411). (D) Poly I:C stimulation significantly induced transcription factor LXRα (n=7, Mann-Whitney U test, medium vs Poly I:C p=0.0023; TNFα vs Poly I:C p=0.0262). (E) Induction of ApoA1 protein after Poly I:C stimulation in 2 different donors. (F) Relative expression of ApoA1 protein corrected for Actin expression in two donors. (G) Positive correlation between IL-32γ and ABCA1 (Spearman r=0.4727, p=ns). (H) Positive correlation between IL-32γ and ABCG1(Spearman r=0.9091, p=0.0003***). (I) Positive correlation between IL-32γ and LXRα(Spearman r=0.6726, p=0.0277*). (J) Positive correlation between IL- 32γ and ApoA1 (Spearman r=0.9455, p<0.0001***). All graphs showing relative mRNA expression adjusted for GAPDH expression. (n=5, data includes replicates).

90 91 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

(A) (B) A B IL-32 mRNA IL-32 mRNA Overexpression of IL-32α, IL-32β or IL-32γ results in strong induction of

) * 10 * 15.0 * cholesterol transporters and reduced intracellular lipid concentrations 12.5 x1000 x1000) H H 8 in HepG2 cells Ct Ct PD PD -d -d 10.0 Since we noted strong correlations between mRNA expression of IL-32 isoforms and (2 GA (2 GA 6 r r on on 7.5

fo cholesterol transporters, we were interested in the function of these cholesterol fo

4 d * ed essi essi te 5.0 transporters in the presence of IL-32. HepG2 cells were therefore transfected with rrec rrect ex pr ex pr 2 2.5 pCDNA3 constructs containing either eGFP (as a negative control), IL-32α, IL-32β or co co ve ve ti ti nd la la 0 0.0 IL-32γ. As shown in figure 4A, overexpression of the various isoforms of IL-32 resulted Re Re h h h h h h 4 4 24 48 24 48 in protein expression of that specific isoform after 48h of transfection. Moreover, Time Time overexpression of the IL-32α isoform resulted in a trend towards an increased expression in LXRα and ABCA1 (Fig. 4B-C). Additionally, overexpression of the IL- (CC ) (D) D IL-32 mRNA ABCA1 mRNA 32β and IL-32γ isoforms resulted in an increased expression of LXRα (non-significant for IL-32γ), ABCA1 and ABCG1 (Fig. 4B,C,D). However, important mediators such 80 * 200 * * as ApoA1(mRNA and protein) and SR-B1 (mRNA) were not affected in HepG2 cells x1000) H x1000 ) 60 H * Ct 150 Ct PD PD overexpressing the isoforms of IL-32 (Supplemental Fig. S3). Finally, we performed -d -d (2 GA (2 GA r r a boron-dipyrromethene (BODIPY) staining to study the intracellular lipid content on

40 on

fo 100 fo

ed ed essi essi of the transfected versus untransfected HepG2 cells. BODIPY fluorescence was 20 50 rrect rrect ex pr ex pr significantly lower in HepG2 cells transfected with the IL-32β isoforms and showed co co ve ve

ti 5 ti a trend towards lower expression in IL-32α and IL-32γ transfected compared to the la la 0 0

Re h h h Re h h h 4 4 24 48 24 48 untransfected control or eGFP transfected negative control (Fig.4E). The strongest Time Time reduction of BODIPY was observed after transfection of HepG2 cells with IL-32β.

E(E) Correlation

100 4 h 4 h 4 h 48 h 4 h 24 h 24 h 1 24 h 24 h 10 48 h 48 h

AB CA 48 h

1 Relative expression (2-dCtx1000) 110100 1000 IL-32

Fig. 3 Time-dependent induction of IL-32γ and cholesterol transporter ABCA1 in human HepG2 cells. A) Expression of IL-32α is significantly induced after 48h compared with 24h, while at 4h the expression of IL-32α was not detectable (4replicates, Mann-Whitney U test, p=0.0286). B) Expression of IL-32β was already detected at 4h and significantly enhanced after 24h and after 48h (4 replicates, Mann- Whitney U test, p=0.0286). C) Expression of IL-32γ was already high at 4h and decreased after 24h and 48h (4 replicates, Mann-Whitney U test, p=0.0286). D) Expression of the cholesterol transporter ABCA1 was high at 4h and significantly decreased after 24h and 48h (4 replicates, Mann- Whitney U test, p=0.0286). E) Correlation between IL-32γ and ABCA1 at different time points showed to be highly significant p<0.0001 and a Spearman r of 0.9580. All conditions were kept in DMEM complete medium with 10% Fetal Calf Serum heat-inactivated. Time course started after plating the cells.

92 93 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

A (A) Silencing of IL-32γ strongly reduces ABCA1 mRNA expression kDa pCDNA3 overexpression As described above, overexpression of the IL-32γ isoform in HepG2 liver cells 100 eGFP IL-32a IL-32b IL-32g 70 resulted in an increased expression and possible function of ABCA1. To study the role 55 of IL-32γ into more detail, we studied IL-32γ and ABCA1 expression in carotid artery 40 plaque tissue and performed knockdown experiments of IL-32γ. Due to the fact that 35 HepG2 cells did not survive the procedure to knockdown IL-32γ, we were prompted 25 IL-32 to use another cell line. We performed silencing and overexpression experiments in 15 a human monocytic cancer cell line (THP1 cell line). This, because ABCA1 expression and function is also very important for monocytes in the circulation, contributing 40 Actin to reverse cholesterol transport. Using THP1 cells, IL-32γ overexpression resulted 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h in high expression of IL-32γ as well as ABCA1 (Fig. 5A,D). When IL-32γ was silenced, B (B) (C)C IL-32γ and ABCA1 expression were completely diminished (Fig. 5B,E). This effect LXRa mRNA (24h) ABCA1 mRNA (24h) was not caused by the transfection method itself, since THP1 cells transfected with 15 * 15 * eGFP or spC negative controls did not result in these effects on ABCA1 expression x1000) H x1000) H Ct Ct PD PD -d 10 -d 10 * (Fig. 5A,B,D,E). The spIL-32 construct is also capable of silencing other isoforms of (2 GA (2 GA r r on on fo fo

* IL-32 as is shown by the silencing of IL-32β (Fig. 5. C). Besides, since we observed ed ed essi 5 essi 5 similar expression patterns of induced cholesterol transporters in THP-1 cells, we rrect rrect ex pr ex pr co co ve ve ti

ti performed overexpressing experiments in THP-1 cells to perform the BODIPY la 0 la 0 5

Re β γ Re β γ staining. Overexpressing IL-32γ in THP-1 cells resulted in a decrease of BODIPY IL32 IL32 eGFP IL32α eGFP IL32α IL32 IL32 pCDNA3 overexpression pCDNA3 overexpression fluorescence (Supplementary figure 4). Moreover, after observing the importance of IL-32γ on ABCA1 expression in liver cells and THP-1 cells, we were curious if this D (D) (E)E ABCG1 mRNA (24h) BODIPY effect is also present in plaque tissue in which ABCA1 expression can affect foamcell 15 10 y ti * formation. We were able to show an increased expression of IL-32γ and ABCA1 mRNA x1000) H ens 9 t

Ct ** in carotid artery plaque tissue (Fig. 5F-G). Finally, THP-1 derived macrophages and In PD

-d 10

ne 8 (2 GA

r human derived percoll monocytes (differentiated to macrophages) were loaded on sce fo 7 ed essi with oxidized LDL (25ug/mL; 50ug/mL respectively) for 24h to study the expression 5 uo re Fl rrect

ex pr 6

an of IL-32 isoforms and mediators involved in cholesterol metabolism in foam cells. co ve ti Me 5 la 0 l α β Re γ IL32 IL32 eGFP IL32α IL32β IL32 Contro eGFP IL32γ Interestingly, IL-32 isoform expression of IL-32α, IL-32β and IL-32γ seemed to be pCDNA3 overexpression pCDNA3 overexpression decreased or unchanged compared to normal macrophages (Supplementary figure 5). Nevertheless, cholesterol mediators ABCA1, ABCG1 and LXRα seemed to be induced, Fig. 4 Modulation of transcription factor LXRα and cholesterol transporters ABCA1/ABCG1 resulted in albeit not significantly (Supplementary figure 5). lower intracellular lipid content after IL-32 overexpression in human HepG2 cells. A) Overexpression of different IL-32 isoforms in human HepG2 liver cells at 24h and 48h. IL-32α (approx. 19 kDa), IL-32β (approx. 26kDa), IL-32γ (approx. 29kDa) B) Transcription factor LXRα was significantly induced after overexpression of IL-32β (5 replicates, Mann- Whitney U test, p=0.0317). C) Overexpression of IL-32α (p=0.0317), IL-32β (p=0.0159), IL-32γ (p=0.0159) significantly induced ABCA1 expression (5 replicates, Mann-Whitney U test). D) Overexpression of IL-32β and IL-32γ enhanced the expression of ABCG1 (5 replicates, Mann-Whitney U test, p=0.0079). E) Mean fluorescent intensity was significantly reduced after overexpression of IL-32β (4 replicates, Mann-Whitney U test, p=0.0286). Replicates are from 2 independent experiments with technical replicates within each experiment.

94 95 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

A(A) (B)B (CC) IL-32 overexpression IL-32 silencing IL-32β silencing DISCUSSION 5000 * 700 * ** 600 * x1000) H x1000 ) H 4000 In the present study, we aimed to investigate the effect of IL32 on proteins involved Ct Ct

PD 500 PD -d -d (2 GA (2 GA 3000 400 in cholesterol metabolism in liver cells and atherosclerotic plaques. We show for the r r on on fo fo

300

2000 ed ed

essi first time that both human primary liver cells and a HepG2 cell line express IL-32 essi 200 rrect rrect ex pr ex pr 1000 mRNA and protein and that human carotid artery tissue expresses IL-32γ mRNA.

co 100 co ve ve ti ti la la 0 0 Moreover, the three main IL-32 isoforms, IL-32γ, IL-32β and IL-32α can be induced Re Re h h h h h h either by recombinant human (rh)TNFα or a synthetic analogue of a double-stranded

WT 4-24 WT 4-24 spC 4-24 spIL-32 4-24 RNA virus (Poly I:C) stimulation in these liver cells. mRNA expression of components

pCDNA-eGFPpCDNA-IL32g 4-24 4-24 important in cholesterol metabolism such as, ABCA1, ABCG1, LXRα and apoA1 D(D) E(E) (F)F (GG) are correlated to expression levels of the IL-32γ isoforms in unstimulated human ABCA1 mRNA ABCA1 mRNA IL-32γ mRNA ABCA1 mRNA primary liver cells and HepG2 cells. Additionally, overexpression of IL-32β and IL- 3000 * 200 * ** 32γ resulted in induction of ABCA1 and ABCG1 in HepG2 cells and overexpression ** *

2500 x1000) H x1000) H x1000) H x1000 ) 150 ** H of IL-32γ induction of ABCA1 expression in THP-1 cells. By performing the BODIPY Ct Ct Ct PD PD PD Ct -d PD -d -d 2000 -d (2 GA (2 GA (2 GA (2 GA r r

r flowcytometry assay, we could show that in the presence of endogenous IL-32, r on fo on

on 100 fo

1500 ion d d fo d fo ed essi ess te essi essi te te 1000 intracellular lipid concentrations were decreased. Moreover, silencing of IL-32 in rrect ex pr rrec 50 ex pr rrec ex pr rrec co ex pr co ve 500 ve THP-1 cells caused a strong reduction of ABCA1 expression. Lastly, we also observed ti ti co co ve ve la la ti ti Re la Re la 0 0 increased mRNA expression of IL-32 and ABCA1 in human carotid artery tissue Re Re h h h h h h HC HC 5 PLQ PLQ obtained from carotid endarterectomy surgery. WT 4-24 WT 4-24 spC 4-24 spIL-32 4-24

pCDNA-eGFPpCDNA-IL32g 4-24 4-24 A recent study suggested the existence of an association between IL-32 and HDL

Fig. 5 IL-32 regulates cholesterol transporter ABCA1 in THP1 cells and is expressed in human carotid cholesterol, which can become relevant for patients with an increased risk to develop artery plaque tissue. (A) Human THP1 cells transfected with different plasmids (pCDNA3-eGFP or CVD such as RA patients. A single nucleotide polymorphism (SNP) in the promoter pCDNA3-IL32γ) or no plasmids (wt control) showed significant induction of IL-32γ when the cells region of IL32 was described to be correlated with higher HDLc concentrations in RA were transfected with pCDNA3-IL32γ (p<0.05* or p<0.01** (6 replicates, One-Way ANOVA, Kruskal patients, suggesting a possible role for IL-32 in determining CVD risk18. Moreover, test, Dunn’s Multiple Comparison Test)). (B) Silencing of IL-32γ showed significant reduction of IL-32γ 38 expression compared to the wt control or smartpool-control (6 replicates, One-Way ANOVA, Kruskal- this SNP possibly results in more IL-32 protein expression . This suggests, higher IL- Wallis test, Dunn’s Multiple Correction Test). (C) Silencing of IL-32β showed significant reduction of 32 protein concentrations are linked with higher HDLc concentrations. IL-32β expression compared to the wt control or smartpool-control (6 replicates, One-Way ANOVA, Kruskal-Wallis test, Dunn’s Multiple Correction Test).(D) Overexpression of IL-32γ iduced significant Next to IL-32 mRNA and protein, we also studied the expression of various cholesterol expression of ABCA1. (E) Silencing of IL-32γ downregulated ABCA1 significantly in THP1 cells. (F) Human carotid artery plaque tissue showed induced expression of IL-32γ mRNA compared to healthy transporters/mediators involved in cholesterol metabolism such as, ABCA1, ABCG1, tissue (p=0.0017*). (G) Human carotid artery plaque tissue showed induced expression of ABCA1 ApoA1 and LXRα. Tight regulation of low-density lipoprotein (LDL) and high-density mRNA compared to healthy tissue (p=0.0124*). lipoprotein (HDL) cholesterol is important to prevent cardiovascular disease, such as atherosclerosis. Circulating HDLc concentrations are regulated by biosynthesis and degradation processes in which the liver plays an important role22. HDL biosynthesis is regulated by the production and secretion of apoA-I by the liver and lipidation of these apolipoproteins by ABCA1 transporters expressed on liver cells23. Expression of this cholesterol transporter ABCA1 in hepatocytes plays an important role in the biosynthesis and regulation of circulating HDL cholesterol26,39,40. Furthermore, enterocytes are also involved in the biosynthesis and maintenance of HDL. Via a

96 97 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

complex network of cholesterol transporters the small intestine establishes a stimulation of hepatocytes with a viral component such as poly I:C could trigger a balance between the amount of excreted and absorbed cholesterol and in that slightly different intracellular pathway or the expression of IL-32 in these cells is of way influences RCT41,42. Moreover, downregulating absorption of cholesterol in more importance and therefore results in different effects in various cell types. the intestine has been shown to improve RCT41. Besides, degradation of HDLc is regulated by different pathways. One way, the body can get rid of circulating Besides, the changes we found on cholesterol transporters expression seem HDLc is by uptake of HDLc via the SR-B1 transporter on liver cells. An alternative functional, since intracellular lipid concentrations were lower in cells overexpressing pathway is the degradation of HDLc by CETP. A third way HDLc can be affected is IL-32 isoforms. These data show a completely new function of endogenous IL- by catabolism in the kidney via cubilin endocytosis and SRB1 in the proximal tubulus 32 in liver cells and THP-1 derived macrophages even suggesting a possible anti- of the kidney. HDLc will be reabsorbed and degraded mostly via SRB1 whereas atherosclerotic function for the pro-inflammatory cytokine. However, when cubulin and megalin endocytose essentially lipid-free apoA-I43,44. These changes in studying THP-1 macrophage foam cells or foam cells generated from human percoll HDL composition occur during the important RCT pathway. Excessive cholesterol is monocytes differentiated to macrophages, isoforms of IL-32 were reduced or transported from the peripheral tissues, such as the vessel wall back to the liver for unchanged and only ABCA1, ABCG1 and LXRα were induced. This might be explained excretion, preventing atherosclerosis. During RCT, accumulated cholesterol from by the fact that oxidized LDL is taken up by different receptors such as CD36 and macrophages in the vessel wall is removed to HDL or lipid-poor apolipoprotein scavenger receptor A which could result in a different intracellular signal cascade (apo)A1 by different mechanisms, including one which is dependent on the cell compared to poly I:C stimulation or overexpressing experiments49. membrane expression of ABCA1 on macrophages45,46. Up regulation of ABCA1 would therefore favour an anti-atherogenic environment whereas down regulation Another explanation for the new link between IL-32 and ABCA1 could be explained of ABCA1 could create an atherogenic state by reducing cholesterol efflux and HDLc by the fact that ABCA1 also has other important functions besides regulating HDL concentrations. Our aim of the study was to understand the mechanism behind cholesterol. Previous studies showed that intracellular cholesterol homeostasis was 5 the regulation of HDLc concentrations by IL-32. Our results show that induction of required for a housekeeping function of cells. Additionally, recent studies indicated IL-32γ, is correlated with induced ABCA1 mRNA expression in unstimulated human that cholesterol regulation, more specifically sterols, are dynamically regulated, primary liver cells and HepG2 cells. IL-32β expression did not show any correlation bioactive and are intrinsic players in the immune response that couples metabolism and IL-32α even showed a negative correlation with ABCA1, ABCG1 and LXRα mRNA to host defence [50]. The question whether IL-32 is anti- or pro-atherogenic is still expression. Additionally, overexpression experiments of IL-32 isoforms in HepG2 difficult to answer. For now we can conclude that IL-32 has many pro- inflammatory/ cells showed that IL-32β and IL-32γ are associated with upregulation of ABCA1 and pro-atherogenic capacities but that this study shows that IL-32γ closely associates ABCG1. Due to the fact that HepG2 cells were no longer viable after knocking down with ABCA1 and other cholesterol mediators and in that way serves as an anti- IL32, overexpressing and silencing experiments of IL-32γ were performed in THP1 atherogenic mediator. macrophages. In the presence or absence of IL-32γ, ABCA1 expression was strongly Some limitations could be envisaged in our study. One of these limitations could induced or reduced respectively. This definitely demonstrated that IL-32γ is a key be the fact we did not measure HDL functionality but only HDLc concentrations player in driving ABCA1 expression. We suggest that this effect is most likely caused even though previous studies have showed functionality being more important by the direct regulation of LXRα expression by IL-32. LXRα is known to mediate, at than HDL cholesterol concentration in determining cardiovascular disease risk19,51. least partially, the expression of ABCA147. Furthermore, we show that the most potent Furthermore, additional experiments could not be performed due to lack of sample isoform of IL-32, IL-32γ together with ABCA1, ABCG1 and LXRα are also strongly since all samples were used for the initial experiments. upregulated by Poly I:C stimulation of primary liver cells and HepG2 cells. These data are conflicting with previous data by Castrillo A et al., which showed that LXRα and To conclude, the present study shows for the first time the existence of a direct therefore ABCA1 are inhibited after activation of Toll-like receptors (TLRs) 3 and 4 association between IL-32 and cholesterol homeostasis in humans. We provide clear by microbial ligands like viruses or bacteria48. One explanation for this could be the and novel evidence that IL-32 is an important regulator of cholesterol transporters fact that two completely different cell types are used in our experiments compared ABCA1 and ABCG1, possibly explaining the variation of HDLc concentrations to the macrophages used in Castrillo’s experiments. Moreover, it is possible that previously observed in individuals bearing a SNP in IL-32 gene. Whether these

98 99 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

interactions would further translate into a higher or lower CV risk remains to be REFERENCES elucidated in future studies, as well as the gain of more knowledge and understanding of the complex interactions between inflammatory effectors, lipids homeostasis 1. Murray, CJ, AD Lopez. Global mortality, disability, and the contribution of risk factors: Global Burden of and atherosclerosis. Disease Study. Lancet. 1997;349(9063):1436-42. 2. Benjamin, EJ, MJ Blaha, SE Chiuve, M Cushman, SR Das, R Deo, et al. Heart Disease and Stroke Statistics-2017 Update: A Report From the American Heart Association. Circulation. Disclosures 2017;135(10):e146-e603. We wish to confirm that there are no known conflicts of interest associated with this 3. Libby, P. Inflammation in atherosclerosis. Nature. 2002;420(6917):868-74. publication and there has been no significant financial support for this work that 4. Cybulsky, MI, MA Gimbrone, Jr. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science. 1991;251(4995):788-91. could have influenced its outcome. 5. Ross, R. Atherosclerosis--an inflammatory disease. The New England journal of medicine. 1999;340(2):115-26. 6. Dinarello, CA. Biologic basis for interleukin-1 in disease. Blood. 1996;87(6):2095-147. Sources of funding 7. Ait-Oufella, H, S Taleb, Z Mallat, A Tedgui. Recent advances on the role of cytokines in atherosclerosis. This research was supported by grants from the Dutch Foundation for Rheumatism Arteriosclerosis, thrombosis, and vascular biology. 2011;31(5):969-79. 8. Witztum, JL, AH Lichtman. The influence of innate and adaptive immune responses on atherosclerosis. (Nr.13-03-302) and the Nijmegen Institute for Infection, Inflammation and Immunity Annual review of pathology. 2014;9:73-102. (N4i), the Netherlands. 9. Heinhuis, B, CD Popa, BL van Tits, SH Kim, PL Zeeuwen, WB van den Berg, et al. Towards a role of interleukin-32 in atherosclerosis. Cytokine. 2013;64(1):433-40. 10. Mun, SH, JW Kim, SS Nah, NY Ko, JH Lee, JD Kim, et al. Tumor necrosis factor alpha-induced interleukin-32 Author contribution is positively regulated via the Syk/protein kinase Cdelta/JNK pathway in rheumatoid synovial fibroblasts. Authors MSMAD, BH, LABJ and JCS RH, JAV, MGN, NPR and CAD contributed to Arthritis and rheumatism. 2009;60(3):678-85. the design of the study, acquisition of data, analysis and interpretation of data. All 11. Joosten, LA, MG Netea, SH Kim, DY Yoon, B Oppers-Walgreen, TR Radstake, et al. IL-32, a authors furthermore contributed to drafting and critically revising the manuscript to proinflammatory cytokine in rheumatoid arthritis. Proc Natl Acad Sci U S A. 2006;103(9):3298-303. 5 12. Solomon, DH, EW Karlson, EB Rimm, CC Cannuscio, LA Mandl, JE Manson, et al. Cardiovascular morbidity create an approved version for submission to the journal. and mortality in women diagnosed with rheumatoid arthritis. Circulation. 2003;107(9):1303-7. 13. Currier, JS, A Taylor, F Boyd, CM Dezii, H Kawabata, B Burtcel, et al. Coronary heart disease in HIV- infected individuals. Journal of acquired immune deficiency syndromes. 2003;33(4):506-12. ACKNOWLEDGEMENTS 14. Choi, JD, SY Bae, JW Hong, T Azam, CA Dinarello, E Her, et al. Identification of the most active interleukin-32 isoform. Immunology. 2009;126(4):535-42. 15. Dinarello, CA, SH Kim. IL-32, a novel cytokine with a possible role in disease. Annals of the rheumatic None diseases. 2006;65 Suppl 3:iii61-4. 16. Netea, MG, T Azam, G Ferwerda, SE Girardin, M Walsh, JS Park, et al. IL-32 synergizes with nucleotide oligomerization domain (NOD) 1 and NOD2 ligands for IL-1beta and IL-6 production through a caspase 1-dependent mechanism. Proceedings of the National Academy of Sciences of the United States of America. 2005;102(45):16309-14. 17. Nold-Petry, CA, MF Nold, JA Zepp, SH Kim, NF Voelkel, CA Dinarello. IL-32-dependent effects of IL- 1beta on endothelial cell functions. Proceedings of the National Academy of Sciences of the United States of America. 2009;106(10):3883-8. 18. Damen, MS, R Agca, S Holewijn, J de Graaf, JC Dos Santos, PL van Riel, et al. IL-32 promoter SNP rs4786370 predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis. Sci Rep. 2017;7:41629. 19. Haase, CL, A Tybjaerg-Hansen, AA Qayyum, J Schou, BG Nordestgaard, R Frikke-Schmidt. LCAT, HDL cholesterol and ischemic cardiovascular disease: a Mendelian randomization study of HDL cholesterol in 54,500 individuals. J Clin Endocrinol Metab. 2012;97(2):E248-56. 20. Rosenson, RS, HB Brewer, Jr., PJ Barter, JLM Bjorkegren, MJ Chapman, D Gaudet, et al. HDL and atherosclerotic cardiovascular disease: genetic insights into complex biology. Nat Rev Cardiol. 2018;15(1):9-19. 21. Hussain, MM. Intestinal lipid absorption and lipoprotein formation. Current opinion in lipidology. 2014;25(3):200-6.

100 101 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

22. Rader, DJ. Molecular regulation of HDL metabolism and function: implications for novel therapies. The 43. Kozyraki, R, J Fyfe, M Kristiansen, C Gerdes, C Jacobsen, S Cui, et al. The intrinsic factor-vitamin B12 Journal of clinical investigation. 2006;116(12):3090-100. receptor, cubilin, is a high-affinity apolipoprotein A-I receptor facilitating endocytosis of high-density 23. Zannis, VI, P Fotakis, G Koukos, D Kardassis, C Ehnholm, M Jauhiainen, et al. HDL biogenesis, remodeling, lipoprotein. Nat Med. 1999;5(6):656-61. and catabolism. Handb Exp Pharmacol. 2015;224:53-111. 44. Glass, C, RC Pittman, M Civen, D Steinberg. Uptake of high-density lipoprotein-associated apoprotein A-I 24. Manthei, KA, J Ahn, A Glukhova, WM Yuan, C Larkin, TD Manett, et al. A retractable lid in and cholesterol esters by 16 tissues of the rat in vivo and by adrenal cells and hepatocytes in vitro. The lecithin:cholesterol acyltransferase provides a structural mechanism for activation by apolipoprotein A-I. Journal of biological chemistry. 1985;260(2):744-50. Journal of Biological Chemistry. 2017;292(49):20313-27. 45. Yin, K, DF Liao, CK Tang. ATP-binding membrane cassette transporter A1 (ABCA1): a possible link 25. Chapman, MJ. Therapeutic elevation of HDL-cholesterol to prevent atherosclerosis and coronary heart between inflammation and reverse cholesterol transport. Molecular medicine. 2010;16(9-10):438-49. disease. Pharmacology & therapeutics. 2006;111(3):893-908. 46. Tang, CK, GH Tang, GH Yi, Z Wang, LS Liu, S Wan, et al. Effect of apolipoprotein A-I on ATP binding 26. Timmins, JM, JY Lee, E Boudyguina, KD Kluckman, LR Brunham, A Mulya, et al. Targeted inactivation of cassette transporter A1 degradation and cholesterol efflux in THP-1 macrophage-derived foam cells. hepatic Abca1 causes profound hypoalphalipoproteinemia and kidney hypercatabolism of apoA-I. The Acta biochimica et biophysica Sinica. 2004;36(3):218-26. Journal of clinical investigation. 2005;115(5):1333-42. 47. Santamarina-Fojo, S, AT Remaley, EB Neufeld, HB Brewer, Jr. Regulation and intracellular trafficking of 27. Brewer, HB, Jr., AT Remaley, EB Neufeld, F Basso, C Joyce. Regulation of plasma high-density lipoprotein the ABCA1 transporter. Journal of lipid research. 2001;42(9):1339-45. levels by the ABCA1 transporter and the emerging role of high-density lipoprotein in the treatment of 48. Castrillo, A, SB Joseph, SA Vaidya, M Haberland, AM Fogelman, G Cheng, et al. Crosstalk between LXR cardiovascular disease. Arteriosclerosis, thrombosis, and vascular biology. 2004;24(10):1755-60. and toll-like receptor signaling mediates bacterial and viral antagonism of cholesterol metabolism. Mol 28. Laffitte, BA, SB Joseph, R Walczak, L Pei, DC Wilpitz, JL Collins, et al. Autoregulation of the human liver X Cell. 2003;12(4):805-16. receptor alpha promoter. Mol Cell Biol. 2001;21(22):7558-68. 49. Kunjathoor, VV, M Febbraio, EA Podrez, KJ Moore, L Andersson, S Koehn, et al. Scavenger receptors class 29. Costet, P, Y Luo, N Wang, AR Tall. Sterol-dependent transactivation of the ABC1 promoter by the liver X A-I/II and CD36 are the principal receptors responsible for the uptake of modified low density lipoprotein receptor/retinoid X receptor. The Journal of biological chemistry. 2000;275(36):28240-5. leading to lipid loading in macrophages. The Journal of biological chemistry. 2002;277(51):49982-8. 30. Williams, DL, MA Connelly, RE Temel, S Swarnakar, MC Phillips, M de la Llera-Moya, et al. Scavenger 50. Fessler, MB. The Intracellular Cholesterol Landscape: Dynamic Integrator of the Immune Response. receptor BI and cholesterol trafficking. Current opinion in lipidology. 1999;10(4):329-39. Trends in immunology. 2016;37(12):819-30. 31. Wang, N, D Lan, W Chen, F Matsuura, AR Tall. ATP-binding cassette transporters G1 and G4 mediate 51. Santos-Gallego, CG. HDL: Quality or quantity? Atherosclerosis. 2015;243(1):121-3. cellular cholesterol efflux to high-density lipoproteins. Proceedings of the National Academy of Sciences of the United States of America. 2004;101(26):9774-9. 5 32. McCall, MBB, LJ Wammes, MCC Langenberg, GJ van Gemert, J Walk, CC Hermsen, et al. Infectivity of Plasmodium falciparum sporozoites determines emerging parasitemia in infected volunteers. Sci Transl Med. 2017;9(395). 33. Heinhuis, B, TS Plantinga, G Semango, B Kusters, MG Netea, CA Dinarello, et al. Alternatively spliced isoforms of IL-32 differentially influence cell death pathways in cancer cell lines. Carcinogenesis. 2016;37(2):197-205. 34. Heinhuis, B, MI Koenders, FA van de Loo, MG Netea, WB van den Berg, LA Joosten. Inflammation- dependent secretion and splicing of IL-32{gamma} in rheumatoid arthritis. Proceedings of the National Academy of Sciences of the United States of America. 2011;108(12):4962-7. 35. Maess, MB, B Wittig, S Lorkowski. Highly efficient transfection of human THP-1 macrophages by nucleofection. J Vis Exp. 2014(91):e51960. 36. Dos Santos, JC, B Heinhuis, RS Gomes, MS Damen, F Real, RA Mortara, et al. Cytokines and microbicidal molecules regulated by IL-32 in THP-1-derived human macrophages infected with New World Leishmania species. PLoS neglected tropical diseases. 2017;11(2):e0005413. 37. Repnik, U, M Knezevic, M Jeras. Simple and cost-effective isolation of monocytes from buffy coats. J Immunol Methods. 2003;278(1-2):283-92. 38. Westra, HJ, MJ Peters, T Esko, H Yaghootkar, C Schurmann, J Kettunen, et al. Systematic identification of trans eQTLs as putative drivers of known disease associations. Nat Genet. 2013;45(10):1238- 43. 39. Aiello, RJ, D Brees, OL Francone. ABCA1-deficient mice: insights into the role of lipid efflux in HDL formation and inflammation. Arteriosclerosis, thrombosis, and vascular biology. 2003;23(6):972-80. 40. Lee, JY, JS Parks. ATP-binding cassette transporter AI and its role in HDL formation. Current opinion in lipidology. 2005;16(1):19-25. 41. Lee-Rueckert, M, F Blanco-Vaca, PT Kovanen, JC Escola-Gil. The role of the gut in reverse cholesterol transport--focus on the enterocyte. Prog Lipid Res. 2013;52(3):317-28. 42. Abumrad, NA, NO Davidson. Role of the gut in lipid homeostasis. Physiol Rev. 2012;92(3):1061-85.

102 103 Supplementary Figure 2 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

SUPPLEMENTARY FIGURES A(A) B (B) Supplementary Figure 1 Correlation Correlation A (A) 100 100 IL-8 mRNA 48 h 10 48 h 48 h 48 h 48 h 3000 10 24 h 24 h 24 h 1 48 h 1 48 h 24 h x1000 ) H 1 48 h 24 h 4 h Ct AB CA AB CA PD 24 h 4 h -d 2000 1 24 h 24 h (2 GA 0.1 4 h r 4 h on fo

ed essi 1000 0.1 0.01 Supplementary Figure 1 110100 110100 1000 rrect ex pr

co IL-32 IL-32 (A) ve ti

IL-8 mRNA la 0 Spearman r: -0.04762 Spearman r: -0.6923

Re d C p = 0.9349 (ns) p = 0.0155 (*) 3000 Me F TN Poly I: x1000 ) H B (B)(CC) (D)(E) Supplemental Fig. 2 Correlation between basal expression of IL-32α or IL-32β and cholesterol Ct

PD transporter ABCA1 in human primary liver cells. (A) Correlation between IL-32α and ABCA1 at different

-d 2000 time points showed no correlation. (2 GA r

on (B) Correlation between IL-32β and ABCA1 at different time points showed a negative correlation fo 5 (p=0.0155, r=- 0.6923). ed 1 1 essi 1000 ApoA 1 LXRα ABCA ABCG rrect ex pr co ve ti

la 0

Re d C IL-32γ IL-32γ IL-32γ IL-32γ Me F TN Poly I: (B)(C) D (D)(EE) 1 1 ApoA 1 LXRα ABCA ABCG

IL-32γ IL-32γ IL-32γ IL-32γ

Supplemental Fig. 1 mRNA expression in human primary liver cells and correlations of IL-32γ mRNA with ABCA1,ABCG1, LXRα, ApoA1 mRNA after TNFα stimulation. (A) IL-8 mRNA expression in human primary liver cells afterrhTNFα and poly I:C stimulation. (B) Positive correlation between IL-32γ and ABCA1 (Spearman r=0.7964, p=0.0004). (H) Positive correlation between IL-32γ and ABCG1(Spearman r=0.9179, p<0.0001***). (I) Positive correlation between IL-32γ and LXRα(Spearman r=0.9134, p<0.0001***). (J) Positive correlation between IL-32γ and ApoA1 (Spearmanr=0.9571, p<0.0001***).

104 105 Supplementary Figure 3 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1

(A) (B)

A(A) (B) B (C)C A 30 B 100

SR-B1 mRNA (24h) ApoA1 mRNA (24h) ApoA1 mRNA (48h) y 800 80 4 800 20 x1000) H x1000) H x1000) H 600 600 60 Ct

3 Ct Ct PD PD PD -d -d -d (2 GA (2 GA (2 GA r r r 400 0 (10^3) on 2 40 40 on on fo fo

d fo 10 essi ed ed te essi essi 1 200 200 20 rrec ex pr % BODIPY Positive cells rrect rrect ex pr ex pr co Mean Fluorescence Intensit ve co co ve ve ti ti ti la 0 0 0 0 0 la la

Re β γ β γ β γ

Re Re IL32 IL32 WT WT α β γ N α 2β γ eGFP IL32α eGFP IL32α IL32 IL32 eGFP IL32α IL32 IL32 3 32 CON WT CO WT 32 eGFP IL32 IL32 IL32 eGFP IL IL IL pCDNA3 overexpression pCDNA3 overexpression pCDNA3 overexpression pCDNA3 overexpression pCDNA3 overexpression (D) (E) D E ApoA1 (24h) ApoA1 (48h) 1500 2000 Supplemental Fig. 4. Mean Fluorescence Intensity and % BODIPY positive THP-1 cells overexpressing IL-32 isoforms IL-32α, IL-32β and IL-32γ. (A) Mean FLuorescence Intensity (10^3) of THP-1 cells 1500 overexpressing IL-32 isoforms and fed ox-LDL for 24h prior to BODIPY staining. (B) % BODIPY positive 1000 L L TH-1 cells overexpressing various isoforms of IL- 32 and fed ox-LDL for 24h prior to BODIPY staining. 1000

500 apoA1 ng/m apoA1 ng/m 500

0 0 γ β γ P β WT WT GF 32α IL32 eGFP IL32α IL32 IL32 e IL IL32 5 pCDNA3 overexpression pCDNA3 overexpression

Supplemental Fig. 3 mRNA and protein expression of various mediators involved in cholesterol metabolism in HepG2 cells overexpressing IL-32 isoforms IL-32α, IL-32β and IL-32γ. (A) SR-B1 mRNA expression in HepG2 cells overexpressing various isoforms of IL-32. (B) ApoA1 mRNA expression after 24h in HepG2 cells overexpressing various isoforms of IL-32. (C) ApoA1 mRNA expression after 48hin HepG2 cells overexpressing various isoforms of IL-32. (D) Protein expression of ApoA1 after 24h in supernatants of HepG2 cells overexpressing various isoforms of IL-32. (E) Protein expression of ApoA1 after 48h measured in supernatants of HepG2 cells overexpressing various isoforms of IL-32.

106 107 Chapter 5 – Interleukin-32 upregulates the expression of ABCA1 and ABCG1 Supplementary Figure 5 A B C D (A) (B) (C) (D) SUPPLEMENTARY TABLE S1

ABCA1 ABCG1 ApoA1 LXRa IL-32α (poly I:C) r=-0.3000 r=-0.08333 r=-0.2667 r=-0.03833 p=ns p=ns p=ns p=ns IL-32α (TNFα) r=-0.3319 r=-0.3187 r=-0.02418 r=-0.3187 p=ns p=ns p=ns p=ns E (E) F G (F) (G) IL-32β (poly I:C) r=-0.4000 r=0.3500 r=0.1167 r=-0.08333 p=ns p=ns p=ns p=ns IL-32β (TNFα) r=-0.01429 r=0.2357 r=0.6143 r=0.2357 p=ns p=ns p=0.0148* p=ns

Table S1. Correlations between IL-32 isoform expression and cholesterol mediator expression in human primary liver cells stimulated with poly I:C and rhTNFα.

H I J K (H) (I) (J) (K)

J K L 5

L (L)

N O

Supplemental Fig. 5. mRNA expression of THP-1 cells fed ox-LDL for 24h and human isolated percoll monocytes fed ox-LDL for 24h. (A) mRNA expression of IL-32α in THP-1 cells generated foam cells. (B) mRNA expression of IL-32β in THP-1 cells generated foam cells. (C) mRNA expression of IL-32γ in THP-1 cells generated foam cells. (D) mRNA expression of ABCA1 in THP-1 cells generated foam cells. (E) mRNA expression of ABCG1 in THP-1 cells generated foam cells. (F) mRNA expression of LXRα in THP-1 cells generated foam cells. (G) mRNA expression of ApoA1 in THP- 1 cells generated foam cells. (H) mRNA expression of IL-32β in human percoll derived monocytes generated foam cells. (I) mRNA expression of IL-32γ in human percoll derived monocytes generated foam cells. (J) mRNA expression of ABCA1 in human percoll derived monocytes generated foam cells. (K) mRNA expression of ABCG1 in human percoll derived monocytes generated foam cells. (L) mRNA expression of LXRα in human percoll derived monocytes generated foam cells.

108 109 CHAPTER 6 HIV-infected individuals on successful antiretroviral treatment show differences in cardiovascular risk profile when stratified for IL-32 promoter SNP rs4786370 6 Submitted

Michelle S.M.A. Damen1 Wouter van der Heijden1 Quirijn de Mast1 Mihai G. Netea1 Charles A. Dinarello1,2 Calin D. Popa3,4 Andre van de Ven1 Leo A.B. Joosten1*

110 111 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

ABSTRACT INTRODUCTION

Background and aims: Subjects successfully treated for the human immunodeficiency During chronic HIV-1 infection, natural killer (NK) cells, monocytes, macrophages virus (HIV-1) are facing persistent immune activation and dyslipidemia which may and especially T-cells are affected. These immune cells produce pro-inflammatory contribute to the increased risk for cardiovascular diseases. IL-32 is known to play cytokines and chemokines, such as interleukin (IL)-6, tumour necrosis factor a role in HIV replication, inflammation and lipid metabolism. The aim of our study (TNF)-α, IL-1β, IL-8, interferon (IFN)-γ, resulting in hyper-immune activation and was to explore the expression pattern of IL-32 (in)dependent of the IL-32 promoter intense depletion of CD4+ T lymphocytes, a decrease in naive CD8+ T lymphocytes SNP and its effects on inflammatory response of PBMCs and cholesterol metabolism and increased viral load [1- 4]. Despite clinical benefits of the potent combined dysfunction within HIV-infected individuals on stable antiretroviral therapy antiretroviral therapy (cART), HIV infected individuals still suffer from persistent compared to healthy individuals. immune activation and dyslipidemia contributing to the increased metabolic and cardiovascular morbidity and mortality seen in HIV infection [5, 6]. This increased Methods: Peripheral mononuclear cells (PBMCs) from ART-treated HIV-infected cardiovascular disease risk can be linked to various processes such as lipodystrophy, individuals (n=40) and healthy individuals (n=18) were stimulated with various endothelial dysfunction, accelerated atherosclerosis and dyslipidemia triggered by ligands to study the mRNA and protein expression of IL-32 and other cytokines. either the inflammation itself or treatment [7-10]. Conversely, changes in (intra) mRNA expression of cholesterol transporters (ABCA1 and ABCG1) and chaperone cellular cholesterol load of HIV-1 susceptible cells may play an essential role during protein calnexin were determined. Additionally, data was stratified for the IL-32 HIV infection, since replication of enveloped viruses at the cell membrane, like HIV, SNP (C/T) and possible correlations between IL32 SNP, IL-32β/γ isoforms expression, critically depend on cholesterol[11, 12]. ABCA1, ABCG1 and calnexin expression were explored. Recently, various reports have highlighted a role for IL-32 in HIV [13-16]. IL-32 serum Results: Stimulation of PBMCs from HIV-infected and healthy individuals resulted levels and IL-32 expression in gut and lymphatic tissue were found to be elevated in in similar mRNA expression of IL-32β, IL-32γ, ABCA1, ABCG1 and calnexin. HIV infected patients compared to healthy individuals. Furthermore, it was shown Furthermore, inflammatory cytokine concentrations after stimulation were also that IL-32 in vitro could suppress viral replication [17]. Additionally, IL- 32 expression similar. Stratification for the IL-32 promoter SNP showed a tendency toward higher is linked to TNFα and vice versa, which suggests IL-32 as an important cytokine 6 IL- 32γ mRNA expression in HIV-infected bearing the TT genotype which was linked involved in the immune response against HIV. Recently, a promoter single nucleotide to the CC genotype in healthy controls. mRNA expression of ABCA1 and ABCG1 was polymorphism (SNP) in IL-32 was shown to be associated with higher concentrations significantly different depending the promoter SNP. mRNA expression of chaperone of IL-32 protein and increased high-density lipoprotein concentration (HDLc) in RA protein calnexin showed a similar trend although not significant. Finally, positive patients [14]. Possibly, by modulating transmembrane cholesterol transporters. correlations between IL-32β/γ and ABCA1/ABCG1, IL-32β/γ and calnexin were found. The present study aims to investigate the role of IL-32 and this SNP in HIV-infected Conclusion: IL-32 mRNA of isoforms IL-32β and IL-32γ were differentially affected individuals, focusing on the inflammatory response of peripheral blood mononuclear by the promoter SNP in HIV-infected individuals compared to healthy controls. cells (PBMCs) and cholesterol metabolism dysfunction. Additionally, variation was observed for ABCA1, ABCG1 and calnexin mRNA expression linked to the SNP and IL-32 isoform expression. These SNP dependent differences could provide new insight in the role of IL-32 in the cardiovascular METHODS risk in HIV-infected individuals and suggest new targets for treatment to improve cardiovascular complications. Patients and controls The HIV-infected individuals enrolled in our study have been participating in a single-center, open-label, randomized, controlled trial (RAPID trial) in adult HIV- infected study participants with undetectable (<40 copies/mL) viral load who have

112 113 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

been receiving a standard backbone of two NRTI’s (either TDF/FTC or ABC/3TC) with Statistical analysis either a NNRTI (EFV or RPV) or a boosted PI (DRV/r, ATZ/r or LPV/r) as initial regimen. Normality was tested using the D’Agostino normality test. Continuous variables Written informed consent was obtained from all participants prior to inclusion. Blood are presented as mean and standard deviation (SD). The differences IL-32 mRNA was taken at baseline to study laboratory measurements. In addition, N= 18 controls expression, IL-32 protein concentrations and cytokine concentrations were analyzed have been enrolled. The was registered at clinicaltrails.gov (NCT02383355). This using the Mann-Whitney test. A p-value less than 0.05 was considered statistically trial was conducted according the principles of the declaration of Helsinki and was significant (*p<0.05 and **p<0.01). Data was analyzed using GraphPad Prism v5.3. approved by the local ethics committee (CMO Arnhem-Nijmegen).

DNA isolation and Taqman genotyping RESULTS Blood was obtained from all HIV-infected individuals and healthy controls included in the study. Genomic DNA was isolated from whole blood using the Qiagen (Valencia, Patient characteristics CA, USA) isolation kit and following the standard protocol[18]. The genotype for the Forty HIV-infected individuals were included in the study. In addition, 18 healthy IL32 promoter (rs4786370) polymorphism was screened by the TaqMan SNP assay individuals were included in this study also no differences we observed with respect C_27972515_20, (Applied Biosystems, Foster City, CA, USA). The TaqMan qPCR to demographical aspects (Table 1). Differences were observed between groups for assays were performed on the AB StepOnePlus polymerase chain reaction system age, BMI and gender distribution (Table 1). (Applied Biosystems). HIV-infected characteristics total Healthy controls characteristics Difference between groups Quantitative PCR Age (years), median 48 (43-58) 27 (23-30) ** (IQR) Peripheral blood mononuclear cells (PBMCs) from HIV-infected and healthy Female sex, n (%) 1 (2.5%) 11 (55%) ** individuals were stimulated with Poly I:C (50 µg/ml) (Invivogen), recombinant human Body mass index, mean 25.5 (±3.6) 22.9 (±2.8) ** (rh)TNFα (10 ng/ml) (R&D Systems) or E.coli lipopolysaccharide (LPS) (1ng/mL) for 24 (SD) (kg / m2) hours. Subsequently, supernatants were collected for cytokine measurements and Current smoker, n (%) 10 (25%) 0 (0%) cells were stored in TriZol reagent at -200C until further use. RNA was isolated using Diabetes mellitus, n (%) 2 (5%) 0 (0%) 6 Trizol reagent (Invitrogen) according to a protocol supplied by the manufacturer. cART regimens N.A Backbone (2 NRTIs), n After RNA isolation, mRNA was transformed into cDNA by applying an iScript kit (%) TDF/FTC ABC/3TC TDF/3TC 30 (75%) (Bio-Rad) to transform mRNA into cDNA. IL-32 primers were previously developed 3TC/FTC 8 (30%) and other primers sequences were extracted from the Harvard Primerbank database. 1(2.5%) 1 (2.5%) Primers were produced by Biolegio (Nijmegen, The Netherlands) and a StepOnePlus Boosted PI, n (%) 6 (15%) qPCR system (Applied Biosystems) was used to analyze relative mRNA expression. NNRTI, n (%) 33 (82.5%) Relative expression was calculated by normalizing for GAPDH and 2^-dCt method. Zidovudine, n (%) 1 (2.5%) Viral load(<50 40 (100%) Determination of cytokine levels copies/mL), n (%) Nadir CD4 (106 cells/L), 280 (230-340) Human TNF-α, IL-6, IL-8, IL-1β, IL-1Ra, IL-10, IFN-γ, IL-17 and IL-22 were determined median (IQR) in culture supernatants using commercial Enzyme-Linked Immunosorbent Assay CD4 count (106 cells/L), 665 (542-792) (ELISA) kits (Sanquin, and R&D Systems), according to the manufacturer’s protocol. median (IQR) Human IL-32 was determined in cell lysates in Triton X 100 0.5%, using a commercial HIV1 diagnosis (years), 9 () median (IQR) ELISA kit (R&D Systems). cART exposure (years), 6 () median (IQR)

Table 1. Characteristics of HIV-infected vs healthy individuals.

114 115 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

IL-32β and IL-32γ mRNA expression differs between PBMCs from HIV-   # infected vs healthy individuals dependent on IL-32 promoter SNP A   B   First, peripheral blood mononuclear cells (PBMCs) of HIV-infected and healthy   individuals were isolated and mRNA expression of IL-32β and IL-32γ was determined.      

A statistical significant difference for IL-32β after stimulation with synthetic double- ""  ""    ! !   & &   stranded RNA ligand poly I:C was observed (p<0.0001) with HIV-infected individuals  

%  %

      # # expression being higher compared to healthy individuals (Fig. 1A-B). Subsequently,     we determined the mRNA expression of IL-32β and IL-32γ when stratified for the                 IL-32 promoter SNP genotype (TT-, CT- and CC-genotype) in both groups, in (un)   $ $   $ $           stimulated PBMCs with poly I:C (TLR3 agonist), rhTNFα, LPS 1ng/mL (TLR4 agonist)            '  '  '  ' (Fig. 1C-H). We found that HIV-infected individuals with the TT-genotype showed   # significantly higher IL-32β mRNA expression compared to healthy individuals C   #'  D   #'  (p=0.0041) (Fig. 1C). Moreover, HIV-infected individuals showed a significant increase   in IL-32γ mRNA expression after rhTNFα (p=0.0153) stimulation, with a similar    "" ""   

  ! !

trend in the other conditions (RPMI, poly I:C and LPS)(Fig.1D). Additionally, HIV-    & &     

infected subjects bearing the CT genotype showed a significantly increased IL-32β % %       # #  

mRNA expression after poly I:C stimulation (p=0.0082) (Fig. 1E). In contrast, healthy   individuals bearing the CC genotype showed significantly different levels of IL-32γ                 $ $     $ $   mRNA compared to HIV-infected individuals after poly I:C and rhTNFα stimulation of                   PBMCs (p=0.0095 and p=0.0190 respectively)(Fig. 1H). The expression pattern of IL-  '  '  '  '

32γ mRNA was therefore completely opposite in HIV-infected individuals compared E    #'  F    #'  to healthy individuals depending on the SNP genotype (TT vs CC respectively).  

  6 "" ""     ! !     & &     % %

       # #    

> Figure 1. mRNA expression of IL32 isoforms IL-32β and IL-32γ in PBMCs from HIV-infected individuals   versus healthy individuals (in)dependent of the IL-32 promoter SNP. A) IL-32β mRNA expression in               $ $     $ $           HIV-infected versus healthy individuals independent of the IL-32 promoter SNP (n=37 HIV patients,           n=18 HC) (* poly I:C stimulation p<0.0001). B) IL-32γ mRNA expression in HIV-infected versus healthy  '  '  '  ' individuals independent of the IL-32 promoter SNP (n=36 HIV, n=18 HC). C) IL-32β mRNA expression G    #'  H    #'  in HIV-infected vs healthy individuals stratified for the IL-32 promoter SNP TT-genotype (n=14 HIV,    n=9 HC) (*poly I:C stimulation p=0.0041). D) IL-32γ mRNA expression in HIV-infected versus healthy     "" ""  individuals stratified for the IL-32 promoter SNP TT genotype, *rhTNFα stimulation (p=0.0153) (n=14   ! !     & HIV, n=9 HC). E) IL-32β mRNA expression in HIV-infected vs healthy individuals stratified for the IL-  &      % 32 promoter SNP CT-genotype (n=17 HIV, n=4 HC *poly I:C stimulation p=0.0082). F) IL-32γ mRNA   %       # #     expression in HIV-infected vs healthy individuals stratified for the IL-32 promoter SNP CT-genotype   (n=17 HIV , n=4 HC). G) IL-32β mRNA expression in HIV-infected vs healthy individuals stratified for   the IL-32 promoter SNP CC-genotype (n=6 HIV, n=4 HC). H) IL-32γ mRNA expression in HIV-infected               $ $     $ $           vs healthy individuals stratified for the IL-32 promoter SNP CC- genotype (*poly I:C and *rhTNFα           stimulation (p=0.0095, p=0.0190) (n=6 HIV, n=4 HC)). Data was shown as mean + SEM and analysis  '  '  '  ' was performed by Mann-Whitney U-test.

116 117 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

A B  IL32 promoter SNP does not affect cytokine production in HIV-infected     individuals and healthy individuals       After observing differences in IL-32 isoforms expression according to genotype, we      explored whether these differences also translate into modifications of inflammation        by changes in cytokine production. PBMCs from HIV-infected individuals as well as   

    healthy individuals were stimulated with various ligands to study the capacity of ex-    vivo cytokine production by PBMCs. Figures 2A-2E show that no clear differences   were observed between PBMCs isolated from HIV -infected or healthy individuals                    independent of the promoter SNP. Additionally, T-cell derived cytokines IFNγ, IL-17              and IL-22 did not differ between the groups (Suppl. Fig. 2A-I). Yet, after observing     changes in IL-32 isoform mRNA expression dependent on the promoter SNP, we C  D  were interested whether the different genotypes would also be associated with     changes in the production of pro-inflammatory cytokines by PBMCs (Suppl. Fig.   1A-O). Significant differences were not observed when cytokine production was        stratified for the promoter SNP.     

       IL-32 promoter SNP genotype differentially affects mRNA expression     of cholesterol transporters ABCA1 and ABCG1 in HIV-infected versus                 healthy individuals                  It is known that dyslipidemia occurs during HIV-1 infection. Moreover, previous       observations by our group showed that IL-32 and the IL-32 SNP are associated with high-density lipoprotein cholesterol concentrations [14]. In addition, cholesterol E  in general and cholesterol transporter ABCA1 have been described to play a role in  6 HIV infection [19, 20].Therefore, we were curious for the effects of the promoter 

SNP on cholesterol mediators in HIV infection. Figure 3A and 3B show that there are   

no differences in mRNA expression between HIV-infected and healthy individuals  

in general. Nevertheless, when mRNA expression data was stratified for the IL-32   promoter SNP, clear statistical significant changes were observed for ABCA1 mRNA expression after poly I:C stimulation in the TT genotype (Fig. 3C) (p=0.0326 poly I:C HIV        vs HC). A similar trend was observed for the other stimuli and was also visible in ABCG1        mRNA expression in the groups bearing the TT genotype (Fig. 3D). Furthermore, ABCG1     mRNA expression was elevated in unstimulated PBMCs from healthy individuals bearing the CT genotype (p=0.0353)(Fig. 3F). In addition, ABCG1 mRNA expression Figure 2. Cytokine expression of PBMCS from HIV-infected individuals versus healthy individuals after 24hour was significantly higher expressed in PBMCs from healthy individuals stimulated with stimulations, independent on IL-32 SNP genotype. A) IL-6 cytokine expression in HIV-infected individuals versus healthy individuals (n=40 HIV, n=18 HC). B) IL-8 cytokine expression in HIV-infected individuals versus healthy either poly I:C or rhTNFα compared to HIV-infected individuals (p=0.0381, p=0.0381 individuals (n=40 HIV, n=18 HC). C) IL-1β cytokine expression in HIV-infected individuals versus healthy individuals respectively)(Fig. 3h). Finally, we studied whether HDLc, LDLc and total cholesterol (n=40 HIV, n=18 HC). D) IL-10 cytokine expression in HIV-infected individuals versus healthy individuals (n=40 HIV, (TC) concentrations in HIV-infected were affected by the promoter SNP, however no n=18 HC). E) TNFα cytokine expression in HIV-infected individuals versus healthy individuals (n=40 HIV, n=18 HC). Data was shown as mean + SEM and analysis was performed by Mann-Whitney U-test. significant differences were observed (Data not shown).

118 119 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

IL-32 promoter SNP does not affect Calnexin mRNA expression in HIV-  "$""' infected or healthy individuals A  B  HIV can affect its cellular uptake by targeting chaperone protein Calnexin, which    under normal circumstances, regulates cholesterol transporter folding of ABCA1 in  #"   #" && &&    % the nucleus affecting the signalling of ABCA1 to the cell membrane [21]. Since HIV %     *$ *$

     ) )  may affect IL-32 isoform mRNA expression when stratified for the promoter SNP  ' '    ! !     and this SNP was shown to influence the expression of ABCA1 and ABCG1, we were  interested to investigate whether the promoter SNP also affects the expression           ( (   ( (   of calnexin in HIV-infected versus healthy individuals. Figure 4A again shows         "" "" that independent of the IL-32 promoter SNP, there are no differences in calnexin #!+  #!+  #!+  #!+  $""' mRNA expression levels between groups. Contrary to these findings, when calnexin mRNA expression was stratified for the promoter SNP, expression levels of calnexin C  "#'+$ D  "#'+$ showed the same trend linked to the genotype as seen for the mRNA expression of     

ABCA1 and ABCG1 in both groups (Fig. 4B-C). In HIV-infected individuals bearing #" #"

&&  &&    % %  

the TT-genotype both calnexin and ABCA1/ABCG1 mRNA expression was highest.   *$ *$     ) )  

Conversely, healthy individuals showed the highest calnexin and ABCA1/ABCG1     !' !'   mRNA expression to the CC-genotype.                    (  (   " "    (  (   " "     > Figure 3. mRNA expression of cholesterol transporters ABCA1 and ABCG1 in PBMCs from HIV !+  !+  !+  !+  infected versus healthy individuals after 24hours stimulation and stratified for the IL-32 promoter # # # # SNP. A) ABCA1 mRNA expression independent on IL-32 promoter SNP in HIV-infected versus E  "#'+$ F  "#'+$ healthy individuals (n=37 HIV, n=17 HC). B)ABCG1 mRNA expression independent of IL-32 promoter   SNP in HIV-infected versus healthy individuals (n=37 HIV, n=17 HC). C) ABCA1 mRNA expression in #"  #"

&& &&    HIV-infected individuals vs healthy individuals stratified for the IL-32 promoter SNP TT genotype. % % 6   *$ *$

   Statistical significant differences were observed between HIV-infected and healthy individuals after   ) )   ' '

   ! ! poly I:C stimulation p=0.0326 (n=14 HIV, n=9 HC). D) ABCG1 mRNA expression in HIV-infected vs     healthy individuals stratified for the IL-32 promoter SNP TT genotype (n=14 HIV, n=9 HC). E) ABCA1   mRNA expression in HIV-infected vs healthy individuals stratified for the IL-32 promoter SNP CT               genotype (n=17 HIV, n=4 HC). F) ABCG1 mRNA expression in HIV-infected vs healthy individuals    (  (   " "    (  (   " "     !+  !+  !+  !+  stratified for the IL-32 promoter SNP CT genotype. Statistical significant differences were observed # # # # after RPMI stimulation p=0.0353 (n=17 HIV, n=4 HC). G) ABCA1 mRNA expression in HIV-infected vs  "#'+$ G H  "#'+$ healthy individuals stratified for the IL-32 promoter SNP CC genotype (n=6 HIV, n=4 HC). H) ABCG1     mRNA expression in HIV-infected vs healthy individuals stratified for the IL-32 promoter SNP CC   #" genotype. Statistical significant differences were observed after poly I:C stimulation p=0.0381 and #" &&  &&   %  %   rhTNFα stimulation p=0.0381 (n=6 HIV, n=4 HC). Data was shown as mean + SEM and analysis was *$  *$     )  )  performed by Mann- Whitney U-test. ' 

 '  !   !   

                 (  ( ( (   " "       !+  !+  "" # # #!+  #!+ 

120 121 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

A B ABCA1/ABCG1 and calnexin are correlated with different inflammatory  ! !    isoforms of IL-32         

Observing a similar trend in mRNA expression of the cholesterol transporters         and calnexin linked to the IL-32 promoter SNP genotypes, we explored whether a   !    certain isoform of IL-32 is associated with their expression levels. Both IL-32β and     IL-32γ are known to be potent (pro-inflammatory) isoforms that can induce varies   other immune mediators [22-24]. IL-32γ mRNA expression in PBMCs was strongly               ! ! !    "  "    "  "  correlated with ABCA1 and ABCG1 mRNA expression (Table 1.) (r=0.828 p=4.52E-10     "  ! !     "   " !   for ABCA1 and r=0.843 with p=1.14E-10 for ABCG1) but showed a negative correlation with calnexin (Table 1) (r=-0.460 p=6.19E-03). In contrast, IL-32β showed a positive C  !" D !" correlation with calnexin mRNA expression (Table 2.) (r=0.732 p=1.96E-06) and a ! !  ! !      negative correlation with ABCA1 and ABCG1 (Table 2)( r=-0.726 with p=1.17E-06 for                 ABCA1 and r=-0.587 p=2.59E-4 for ABCG1).            

     

 

                    ! !   ! !            

   " "      " "      " " ! !    " " ! !

Figure 4. Calnexin expression in PBMCs from HIV-infected individuals stratified for the IL-32 promoter SNP versus expression in PBMCs from healthy individuals. A). Calnexin mRNA expression independent of IL-32 promoter SNP in HIV-infected versus healthy individuals (n=34 HIV, n=17 HC). B) Calnexin mRNA expression in HIV-infected versus healthy individuals bearing the TT-genotype for the IL- 32 promoter SNP (n=14 HIV, n=9 HC). C) Calnexin mRNA expression in HIV-infected versus healthy 6 individuals bearing the CT-genotype (n=14 HIV, n=4 HC). D) Calnexin mRNA expression in HIV-infected versus healthy individuals bearing the CC-genotype (n=6 HIV, n=4 HC). Data was shown as mean + SEM and analysis was performed by Mann-Whitney U-test.

122 123 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

IL32y mRNA Calnexin mRNA ABCA1 mRNA ABCG1 mRNA IL32b mRNA Calnexin mRNA ABCA1 mRNA ABCG1 mRNA expression expression expression expression expression expression expression expression baseline baseline baseline baseline baseline baseline baseline baseline IL32y mRNA Correlation Coefficient 1,000 -,429* ,808** ,842** IL32b Correlation 1,000 ,768** -,687** -,535** expression baseline mRNA Coefficient Sig. (2-tailed) 1,1E-02 1,52E-09 6,63E-11 expression baseline Sig. (2-tailed) 2,95E-07 5,0E-06 1,0E-03 N 37 34 37 37 N 35 32 35 35 Calnexin mRNA Correlation Coefficient -,429* 1,000 -,519** -,465** ** ** ** expression baseline Calnexin Correlation ,768 1,000 -,519 -,465 Sig. (2-tailed) 6,19E-03 2,00E-03 6,00E-03 mRNA Coefficient N 34 34 34 34 expression baseline Sig. (2-tailed) 2,95E-07 2,00E-03 6,00E-03 ABCA1 mRNA Correlation Coefficient ,808** -,519** 1,000 ,943** N 32 34 34 34 expression baseline Sig. (2-tailed) 1,52E-09 2,00E-03 2,41E-18 ABCA1 Correlation -,687** -,519** 1,000 ,943** N 37 34 37 37 mRNA Coefficient expression baseline Sig. (2-tailed) 5,0E-06 2,00E-03 2,41E-18 ABCG1 mRNA Correlation Coefficient ,842** -,465** ,943** 1,000 expression baseline N 35 34 37 37 ABCG1 Correlation -,587** -,465** ,943** 1,000 Sig. (2-tailed) 6,63E-11 6,00E-03 2,41E-18 mRNA Coefficient N 37 34 37 37 expression baseline Sig. (2-tailed) 1,0E-03 6,00E-03 2,41E-18 N 35 34 37 37 Table 1. Spearman correlation of IL-32 isoform IL-32γ with Calnexin, ABCA1, ABCG1 mRNA expression in unstimulated PBMCs from HIV-infected individuals. Table 3. Spearman correlation of IL-32 isoform IL-32β with Calnexin, ABCA1, ABCG1 mRNA expression in unstimulated PBMCs from HIV-infected individuals.

IL32y mRNA Calnexin mRNA ABCA1 mRNA ABCG1 mRNA expression expression expression expression baseline baseline baseline baseline IL32b mRNA Calnexin mRNA ABCA1 mRNA ABCG1 mRNA IL32y mRNA Correlation Coefficient 1,000 -,488* ,803** ,862** expression expression expression expression 6 expression baseline baseline baseline baseline baseline Sig. (2-tailed) 4,70E-02 1,81E-04 1,8E-05 IL32b mRNA Correlation Coefficient 1,000 ,765** -,362 -,444 N 19 17 16 16 expression baseline Sig. (2-tailed) 3,50E-04 1,69E-01 8,48E-02 Calnexin mRNA Correlation Coefficient -,488* 1,000 -,391 -,556* N 19 17 16 16 expression baseline Sig. (2-tailed) 4,70E-02 1,34E-01 2,50E-02 Calnexin mRNA Correlation Coefficient ,765** 1,000 -,391 -,556* N 17 17 16 16 expression baseline Sig. (2-tailed) 3,50E-04 1,34E-01 2,54E-02 ABCA1 mRNA Correlation Coefficient ,803** -,391 1,000 ,924** N 17 17 16 16 expression baseline Sig. (2-tailed) 1,81E-04 1,34E-01 3,34E-07 ABCA1 mRNA Correlation Coefficient -,362 -,391** 1,000 ,924** N 16 16 16 16 expression baseline Sig. (2-tailed) 1,69E-01 1,34E-01 3,34E-07 ABCG1 mRNA Correlation Coefficient ,862** -,556* ,924** 1,000 N 16 16 16 16 expression baseline Sig. (2-tailed) 1,8E-05 2,50E-02 3,34E-07 ABCG1 mRNA Correlation Coefficient -,444 -,556* ,924** 1,000 N 16 16 16 16 expression baseline Sig. (2-tailed) 8,48E-02 2,54E-02 3,34E-07 N 16 16 16 16 Table 2. Spearman correlation of IL-32 isoform IL-32γ with Calnexin, ABCA1, ABCG1 mRNA expression in unstimulated PBMCs from healthy individuals. Table 4. Spearman correlation of IL-32 isoform IL-32β with Calnexin, ABCA1, ABCG1 mRNA expression in unstimulated PBMCs from healthy individuals.

124 125 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

DISCUSSION controls [15, 31]. The fact that we only found significant differences in IL-32β could be explained by the study design in which we studied the expression of IL-32 mRNA In the present study, we investigated the role of IL-32 and a promoter SNP in IL-32 in PBMCs from virologically suppressed HIV-infected individuals and not in patients in HIV-infected individuals versus healthy individuals, focusing on the inflammatory during acute HIV infection or treatment naive [15]. Alternatively, discrepancies at response of PBMCs and the cardiovascular risk profile associated with cholesterol IL-32 protein level might be related to IL-32alfa, another IL-32 isoform, even if no metabolism dysfunction. More specifically, we explored whether the promoter SNP differences are found at mRNA level. would be involved in regulation of important mediators responsible for cholesterol transport and could be involved in the persistent immune activation due to chronic In the present report we extended our previous observations on the influence of IL-32 low-grade inflammation. We have shown that differences were observed in mRNA SNP (C/T) on the expression IL-32 isoforms and the production of pro-inflammatory expression in the important pro-inflammatory isoform IL-32β, but not in IL-32γ, cytokines by PBMCs. When mRNA expression levels were stratified for the IL-32 between HIV-infected and healthy individuals. Furthermore, when mRNA expression promoter SNP a significant difference was observed for IL-32γ mRNA expression was stratified for the promoter SNP genotypes, there was a statistical significant between HIV-infected individuals and healthy individuals bearing the TT- or CC- difference in IL-32γ mRNA expression between the two groups depending on the genotype respectively. This is the first time the IL-32 promoter SNP is shown to be TT- or CC-genotype. These observed differences however did not translate into important for the specific isoform production of IL-32 in HIV-infected versus healthy differences in cytokine production by PBMCs of IL-6, IL-8, IL-10, IL-1β, TNFα, IFNγ, individuals and could suggest that this SNP could be important to predict disease IL-17 and IL- progression due to the highly potent IL-32γ isoform in decreasing HIV-1 viral load or 22. Interestingly, we found a clear effect of the promoter SNP on cholesterol could possibly explain the persistent immune activation in these patients [32]. Yet, transporters ABCA1 and ABCG1 following the same trend as IL-32γ mRNA expression these differences at mRNA expression level did not result in differences in cytokine between the TT- and CC-genotypes in HIV-infected versus healthy individuals. production by PBMCs between HIV-infected individuals and healthy individuals. Moreover, a tendency towards IL-32 SNP dependent mRNA expression of chaperone protein calnexin was depicted. Finally, strong expression correlations were detected Previous literature on the interaction of HIV, inflammation, lipids and cardiovascular between IL- 32γ and cholesterol transporters ABCA1/ABCG1 and between IL-32β risk on the one hand, and our recent observations on IL-32 SNP and changes in and calnexin in both HIV-infected and healthy individuals. high-density lipoprotein cholesterol (HDLc) concentrations in rheumatoid arthritis 6 patients on the other, we explored the latter in the HIV-infected individuals [14]. Various isoforms of IL-32 have been shown to be involved in cancer, rheumatoid Interestingly, lipid profile and metabolism of chronically HIV-infected individuals is arthritis but also HIV [25-28]. The two most potent and pro-inflammatory isoforms altered. are IL-32β and IL-32γ due to alternative splicing, IL-32γ can be spliced into IL-32β [29]. Nevertheless, no differences in HDLc were observed in HIV-infected individuals (in) Unfortunately, the exact mechanism, expression pattern and role of these main two dependent of the promoter SNP. Low grade persistent inflammation is associated isoforms in HIV has not been fully explored and studies trying to determine effects with early onset of atherosclerosis resulting in an increased risk for cardiovascular of IL-32 in general showed conflicting data regarding the beneficial effect of IL-32 in disease [33]. Important changes associated with dyslipidemia may include HIV infection [16, 17, 30]. Furthermore, it is known that both IL-32β and IL-32γ are hypertriglyceridemia, increased levels of low-density lipoprotein cholesterol (LDLc) involved in the induction of other pro-inflammatory mediators such as, IL-6, IL-1β, and decreased levels of high-density lipoprotein cholesterol (HDLc). Unfortunately, IL-8, TNFα and IFN-inducible . IL-32 isoforms could therefore play an important the exact mechanism behind the occurrence of these changes remains unclear [34]. role in the defence against or maintenance of persistent immune activation in HIV. In ABCA1 and ABCG1 are important transmembrane cholesterol transporters playing our study, we found differences between mRNA expression of IL-32 isoforms IL-32β/ a role in determining HDL circulating levels and have also been described to be IL-32γ (in)dependent of promoter SNP genotype in PBMCs from HIV-infected versus important for controlling HIV infection [35-37]. ABCA1 and ABCG1 are responsible healthy individuals. Expression of IL-32β mRNA was significantly increased after poly for cholesterol efflux from human macrophages to prevent foamcell formation I:C stimulation in PBMCs of HIV-infected. This finding is in line with previous reports and atherogenesis. However, high intracellular concentrations of cholesterol are indicating higher IL-32 mRNA and protein levels in HIV-infected individuals versus important for HIV replication, therefore maintaining a high intracellular cholesterol

126 127 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

concentration is beneficial for the virus and its replication [38]. Moreover, lipid rafts Some limitations could be envisaged in our study. One of these might be related are used by the virus to enter targeted cells and therefore high levels of lipid rafts to the low number of HIV- infected per genotype (n= 16 TT vs n= 7 CC) for the IL-32 on the cell membrane would favour HIV replication. Interestingly, the virus is able to promoter SNP. In addition, there were statistical significant differences between the affect intracellular mechanisms increasing cholesterol load on the lipid rafts, using HIV-infected and healthy individuals regarding to age, BMI and gender distribution. a mechanism involving HIV-1 Nef protein and calnexin at the endoplasmic reticulum We know from a previous study that gender does not affect the distribution of the (ER) level, eventually leading to lower ABCA1 function [11, 21]. By suppressing IL-32 promoter SNP. Yet, age and BMI could be confounding factors when exploring ABCA1, cholesterol efflux from cells is decreased resulting in high cellular lipid cholesterol metabolism [14]. Moreover, the absence of protein measurements contents and lipid rafts on the plasma membrane [39]. Our results show that the related to IL-32, ABCA1 and calnexin leave us with mRNA expression patterns (in) promoter SNP in IL-32 strongly affects the mRNA expression of ABCA1 and ABCG1. dependent of the IL-32 promoter SNP. This however does not necessarily translate Furthermore, we were able to detect a strong positive correlation between ABCA1/ to similar protein expression patterns. Finally, it would be interesting to study ABCG1 expression and the expression of the IL-32γ isoform. This could suggest Nef expression in relation to the promoter SNP and whether the effect of the that HIV-infected individuals bearing the TT-genotype for the IL-32 promoter SNP promoter SNP on calnexin expression actually affects ABCA1 functionality therefore produce more IL-32γ and therefore show an increased ABCA1 expression possibly decreasing cholesterol efflux. resulting in decreased intracellular cholesterol concentrations. This could lead to an environment in which it is more difficult for the virus to infect cells. In conclusion, our study shows that IL-32β but not IL-32γ mRNA expression differs between HIV-infected and healthy individuals, and shows important differences Cholesterol is essential for viral replication that it has developed a way to interfere with when stratified for the IL-32 SNP, especially for the IL- 32γ isoform. Additionally, the cholesterol metabolism. HIV infection, by the activity of viral protein Nef, impairs the promoter SNP (rs4786370) is associated with the expression of cholesterol cholesterol efflux mediated by ABCA1 [37]. Nef can interact with endoplasmatic reticulum transporters ABCA1 and ABCG1. Our data suggests an important role for IL-32 chaperone calnexin, which regulates folding and maturation of glycosylated protein isoforms in HIV infection. The exploration of the role of IL-32 and the regulation including, ABCA1 [21]. The interaction between calnexin and ABCA1 is interrupted by of specific isoforms of IL-32 in HIV should be continued in future research in order Nef which resulted in dysfunctional ABCA1 and reduced cholesterol efflux [21, 40]. In our to provide novel insights in unravelling the underlying mechanism for CVD and study we were able to show a significant association of IL-32β expression and calnexin persistent inflammation. 6 expression, which however was not dependent on the IL-32 SNP could. Most likely, the two isoforms of IL-32 serve alternate cellular functions which is in line with previous Competing interests studies [22, 24]. In addition, IL-32γ is known to be spliced to IL-32β and later to IL-32α The authors declare that they have no competing interests. WvdH received a travel as a less potent inflammatory isoform to reduce inflammation [29]. Therefore, IL-32γ grant from Merck Sharpe and Dohme inc. and IL- 32β might be related to other immune regulatory functions against HIV infection than IL-32α. The question whether IL-32 expression and in more detail the two isoforms Funding IL-32β/IL-32γ are important for host defence or cause a detrimental immune response This research was supported by grants from the Dutch Foundation for Rheumatism favouring the virus and cardiovascular problems, still remains to be answered. At the (Nr.13-03-302) and the Nijmegen Institute for Infection, Inflammation and Immunity moment, we can conclude that IL-32 and its’ isoforms have many capacities and that (N4i), the Netherlands. The trial was supported by a research grant from the this study shows two distinct functions involved in HIV infection dependent on which Investigator Initiated Studies Program of Merck Sharp & Dohme Corp. MGN was isoform is being expressed. In the case of expression of IL-32γ cholesterol transporters funded by a Spinoza Grant of the Netherlands Organization for Scientific Research, are upregulated. This could be associated with an increased cholesterol efflux and and a Competitiveness Operational Program Grant of the Romanian Ministry of protection against cardiovascular problems. In contrast, when IL-32β is expressed, European Funds (FUSE). cholesterol transporter is reduced but higher levels of calnexin are being expressed which could add to a beneficial environment for the virus and increase the risk forCVD.

128 129 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

ACKNOWLEDGEMENTS REFERENCES

We would like to thank HIV nurses Marjolein Bosch, Bert Zomer en Karin Grintjes for 1. Catalfamo, M., C. Le Saout, and H.C. Lane, The role of cytokines in the pathogenesis and treatment of HIV infection. Cytokine Rev, 2012. 23(4-5): p. 207-14. assisting in patient inclusion and recruitment. 2. Breen, E.C., Pro- and anti-inflammatory cytokines in human immunodeficiency virus infection and acquired immunodeficiency syndrome. Pharmacol Ther, 2002. 95(3): p. 295-304. 3. Decrion, A.Z., et al., HIV and inflammation. Curr HIV Res, 2005. 3(3): p. 243-59. 4. Fauci, A.S., D. Mavilio, and S. Kottilil, NK cells in HIV infection: paradigm for protection or targets for ambush. Nat Rev Immunol, 2005. 5(11): p. 835-43. 5. French, M.A., et al., Serum immune activation markers are persistently increased in patients with HIV infection after 6 years of antiretroviral therapy despite suppression of viral replication and reconstitution of CD4+ T cells. J Infect Dis, 2009. 200(8): p. 1212-5. 6. Lederman, M.M., et al., Immunologic failure despite suppressive antiretroviral therapy is related to activation and turnover of memory CD4 cells. J Infect Dis, 2011. 204(8): p. 1217-26. 7. Fourie, C.M., et al., Lipid abnormalities in a never-treated HIV-1 subtype C-infected African population. 8. Lipids, 2010. 45(1): p. 73-80. 9. Riddler, S.A., et al., Impact of HIV infection and HAART on serum lipids in men. JAMA, 2003. 289(22): p. 2978-82. 10. Carr, A. and D.A. Cooper, Images in clinical medicine. Lipodystrophy associated with an HIV-protease inhibitor. N Engl J Med, 1998. 339(18): p. 1296. 11. Lake, J.E. and J.S. Currier, Metabolic disease in HIV infection. Lancet Infect Dis, 2013. 13(11): p. 964-75. 12. Zheng, Y.H., et al., Nef increases infectivity of HIV via lipid rafts. Curr Biol, 2001. 11(11): p. 875-9. 13. Liao, Z., et al., Lipid rafts and HIV pathogenesis: host membrane cholesterol is required for infection by HIV type 1. AIDS Res Hum Retroviruses, 2001. 17(11): p. 1009-19. 14. Damen, M., et al., Interleukin-32 in chronic inflammatory conditions is associated with a higher risk of cardiovascular diseases. Atherosclerosis, 2017. 264: p. 83-91. 15. Damen, M.S., et al., IL-32 promoter SNP rs4786370 predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis. Sci Rep, 2017. 7: p. 41629. 6 16. Rasool, S.T., et al., Increased level of IL-32 during human immunodeficiency virus infection suppresses HIV replication. Immunol Lett, 2008. 117(2): p. 161-7. 17. Smith, A.J., et al., The immunosuppressive role of IL-32 in lymphatic tissue during HIV-1 infection. J Immunol, 2011. 186(11): p. 6576-84. 18. Nold, M.F., et al., Endogenous IL-32 controls cytokine and HIV-1 production. J Immunol, 2008. 181(1): p. 557-65. 19. Lionakis, M.S., et al., CX3CR1-dependent renal macrophage survival promotes Candida control and host survival. J Clin Invest, 2013. 123(12): p. 5035-51. 20. Rappocciolo, G., et al., Alterations in cholesterol metabolism restrict HIV-1 trans infection in nonprogressors. MBio, 2014. 5(3): p. e01031-13. 21. Cui, H.L., et al., HIV-1 Nef mobilizes lipid rafts in macrophages through a pathway that competes with ABCA1-dependent cholesterol efflux. J Lipid Res, 2012. 53(4): p. 696-708. 22. Jennelle, L., et al., HIV-1 protein Nef inhibits activity of ATP-binding cassette transporter A1 by targeting endoplasmic reticulum chaperone calnexin. J Biol Chem, 2014. 289(42): p. 28870-84. 23. Kim, Y.G., et al., Interleukin-32gamma enhances the production of IL-6 and IL-8 in fibroblast-like synoviocytes via Erk1/2 activation. J Clin Immunol, 2010. 30(2): p. 260-7. 24. Shoda, H., et al., Interactions between IL-32 and tumor necrosis factor alpha contribute to the exacerbation of immune-inflammatory diseases. Arthritis Res Ther, 2006. 8(6): p. R166. 25. Kang, J.W., et al., A proinflammatory cytokine interleukin-32beta promotes the production of an anti- inflammatory cytokine interleukin-10. Immunology, 2009. 128(1 Suppl): p. e532-40.

130 131 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

26. Khawar, M.B., M.H. Abbasi, and N. Sheikh, IL-32: A Novel Pluripotent Inflammatory Interleukin, towards SUPPLEMENTARY FIGURES Gastric Inflammation, Gastric Cancer, and Chronic Rhino Sinusitis. Mediators Inflamm, 2016. 2016: p. 8413768. 27. Tsai, C.Y., et al., Interleukin-32 increases human gastric cancer cell invasion associated with tumor A B C progression and metastasis. Clin Cancer Res, 2014. 20(9): p. 2276-88. 28. Joosten, L.A., et al., IL-32, a proinflammatory cytokine in rheumatoid arthritis. Proc Natl Acad Sci U S A, 2006. 103(9): p. 3298-303. 29. El-Far, M., et al., Proinflammatory isoforms of IL-32 as novel and robust biomarkers for control failure in HIV-infected slow progressors. Sci Rep, 2016. 6: p. 22902. 30. Heinhuis, B., et al., Inflammation-dependent secretion and splicing of IL-32{gamma} in rheumatoid arthritis. Proc Natl Acad Sci U S A, 2011. 108(12): p. 4962-7. 31. Kang, J.W., et al., Intracellular interaction of interleukin (IL)-32alpha with protein kinase Cepsilon D E F (PKCepsilon ) and STAT3 protein augments IL-6 production in THP-1 promonocytic cells. J Biol Chem, 2012. 287(42): p. 35556-64. 32. Li, Q., et al., Microarray analysis of lymphatic tissue reveals stage-specific, gene expression signatures in HIV-1 infection. J Immunol, 2009. 183(3): p. 1975-82. 33. Zepp, J.A., et al., Protection from RNA and DNA viruses by IL-32. J Immunol, 2011. 186(7): p. 4110-8. 34. Myerson, M., C. Malvestutto, and J.A. Aberg, Management of lipid disorders in patients living with HIV. 35. J Clin Pharmacol, 2015. 55(9): p. 957-74. 36. Simon, V., D.D. Ho, and Q. Abdool Karim, HIV/AIDS epidemiology, pathogenesis, prevention, and treatment. Lancet, 2006. 368(9534): p. 489-504. 37. Pushkarsky, T., et al., Short Communication: Accumulation of Neutral Lipids in Liver and Aorta of Nef- Transgenic Mice. AIDS Res Hum Retroviruses, 2017. 33(1): p. 57-60. 38. Mukhamedova, N., et al., Analysis of ABCA1 and Cholesterol Efflux in HIV-Infected Cells. Methods Mol Biol, 2016. 1354: p. 281-92. G H I 39. Mujawar, Z., et al., Human immunodeficiency virus impairs reverse cholesterol transport from macrophages. PLoS Biol, 2006. 4(11): p. e365. 40. Campbell, S.M., S.M. Crowe, and J. Mak, Virion-associated cholesterol is critical for the maintenance of HIV- 1 structure and infectivity. AIDS, 2002. 16(17): p. 2253-61. 6 41. Landry, Y.D., et al., ATP-binding cassette transporter A1 expression disrupts raft membrane microdomains through its ATPase-related functions. J Biol Chem, 2006. 281(47): p. 36091-101. J K L 42. Hunegnaw, R., et al., Interaction Between HIV-1 Nef and Calnexin: From Modeling to Small Molecule Inhibitors Reversing HIV-Induced Lipid Accumulation. Arterioscler Thromb Vasc Biol, 2016. 36(9): p. 1758- 71.

132 133 Chapter 6 –HIV-infected individuals, IL-32 and cardiovascular risk profile

 !  " Supplementary Figures Suppl. S1 A-O Supplementary Figure S1. Cytokine expression of PBMCS from HIV-infected individuals versus healthy individuals dependent of IL-32 SNP genotype. A) IL-6 production by PBMCs from HIV-infected and A  B  C  healthy individuals after stimulation with various ligands for 24h stratified for the TT genotype     (n=16 HIV, n=10 HC). B) IL-6 production by PBMCs from HIV-infected and healthy individuals after      stimulation with various ligands for 24h stratified for the CT genotype (n=17 HIV, n=4 HC). C) IL-6   

! production by PBMCs from HIV-infected and healthy individuals after stimulation with various ligands ! !              for 24h stratified for the CC genotype (n=7 HIV, n=4 HC). D) IFNγ production by PBMCs from HIV-    infected and healthy individuals after stimulation with various ligands for 7days stratified for the TT-    genotype (n=13 TT HIV, n=9 TT HC). E) ) IL-17 production by PBMCs from HIV-infected and healthy                                                               # #   # #   # #       individuals after stimulation with various ligands for 7days stratified for the TT-genotype (n=13 TT HIV,   $ $   $ $   $ $             n=9 TT HC). F) ) IL-22 production by PBMCs from HIV-infected and healthy individuals after stimulation             with various ligands for 7days stratified for the TT-genotype (n=13 TT HIV, n=9 TT HC). G) )IFNγ production by PBMCs from HIV-infected and healthy individuals after stimulation with various ligands !  " for 7days stratified for the CT-genotype (n=15 CT HIV, n=4 CT HC). H) IL-17 production byPBMCs from HIV-infected and healthy individuals after stimulation with various ligands for 7days stratified for the CT-genotype (n=15 CT HIV, n=4 CT HC). I) IL-22 production by PBMCs from HIV- infected and D  E F     healthy individuals after stimulation with various ligands for 7days stratified for the CT-genotype (n=15 

 CT HIV, n=4 CT HC). J) IFNγ production by PBMCs from HIV-infected and healthy individuals after       stimulation with various ligands for 7days stratified for the CC-genotype (n=6 CC HIV, n=4 CC HC). K)  ! ! 

  !     IL-17 production by PBMCs from HIV-infected and healthy individuals after stimulation with various       ligands for 7days stratified for the CC-genotype (n=6 CC HIV, n=4 CC HC). L) IL-22 production by PBMCs

   from HIV-infected and healthy individuals after stimulation with various ligands for 7days stratified for                     the CC-genotype (n=6 CC HIV, n=4 CC HC).                             # #      # #         $ $   $ $      $# $#                 Figure S2A-L Supplementary Figure S2. T-cell derived cytokine expression of PBMCS from HIV infected individuals   G H  I   versus healthy individuals (in)dependent of IL-32 SNP genotype. A) IFNγ production by PBMCs from  6  HIV-infected and healthy individuals after stimulation with various ligands for 7days (n=34 HIV, n=17       HC). B) IL-17 production by PBMCs from HIV-infected and healthy individuals after stimulation with  ! ! 

  ! 

 various ligands for 7days (n=34 HIV, n=17 HC). C) IL-22 production by PBMCs from HIV-infected and         healthy individuals after stimulation with various ligands for 7days (n=34 HIV, n=17 HC). D) IFNγ production by PBMCs from HIV-infected and healthy individuals after stimulation with various                                    ligands for 7days stratified for the TT-genotype (n=13 TT HIV, n=9 TT HC). E) ) IL-17 production by                    # #     # #        $ $   $ $   PBMCs from HIV-infected and healthy individuals after stimulation with various ligands for 7days    $# $#    

            stratified for the TT- genotype (n=13 TT HIV, n=9 TT HC). F) ) IL-22 production by PBMCs from HIV- infected and healthy individuals after stimulation with various ligands for 7days stratified for the TT- J K L    genotype (n=13 TT HIV, n=9 TT HC). G) ) IFNγ production by PBMCs from HIV-infected and healthy    individuals after stimulation with various ligands for 7days stratified for the CT-genotype (n=15 CT     HIV, n=4 CT HC). H) IL-17 production by PBMCs from HIV-infected and healthy individuals after     ! !  !  stimulation with various ligands for 7days stratified for the CT-genotype (n=15 CT HIV, n=4 CT HC). I)            IL-22 production by PBMCs from HIV-infected and healthy individuals after stimulation with various   ligands for 7days stratified for the CT-genotype (n=15 CT HIV, n=4 CT HC). J) IFNγ production by   

  PBMCs from HIV-infected and healthy individuals after stimulation with various ligands for 7days                                                 # #   # #   # #       stratified for the CC-genotype (n=6 CC HIV, n=4 CC HC). K) IL-17 production by PBMCs from HIV-   $ $   $ $   $ $             infected and healthy individuals after stimulation with various ligands for 7days stratified for the             CC-genotype (n=6 CC HIV, n=4 CC HC). L) IL-22 production by PBMCs from HIV- infected and healthy individuals after stimulation with various ligands for 7days stratified for the CC-genotype (n=6 CC HIV, n=4 CC HC).

134 135 CHAPTER 7 Summary, General discussion & Future perspectives

7

136 137 Chapter 7 – Summary, General discussion & Future perspectives

SUMMARY AND GENERAL DISCUSSION to the high CVD burden in these patients. Therefore, in chapter 2, we examined for the first time whether IL-32 and more specifically a promoter SNP in the IL32 gene Even though various new discoveries led to effective treatments such as the use of could affect important components such as high-density lipoprotein cholesterol statins, other lipid-lowering drugs and anti-inflammatory strategies like anti-IL-1b (HDLc) associated with (prevention of) CVD in RA patients compared to non-RA antibodies (e.g. canakinumab), cardiovascular diseases (CVD) are still a major cause patients. First of all, we observed no significant differences in the allelic distribution of death globally (1). The single most important pathological process accountable for for the IL-32 promoter SNP between RA patients and non-RA patients. Interestingly, CVD is atherosclerosis. At the start of exploring the pathogenesis of atherosclerosis, we did find a significant increase of HDLc in RA patients when stratified for the it was thought to be a relatively simple vascular lipid storage disease (2, 3). However, genotypes of the IL-32 promoter SNP. This effect did not seem to be influenced later it was shown that many inflammatory processes equally contributed to its by disease activity in RA patients. In contrast to HDLc, other lipid concentrations development and was therefore characterized as a chronic inflammatory disorder including LDLc and total cholesterol (TC) were not affected by the promoter SNP. of the arterial vessel wall (4). Ongoing atherosclerosis can eventually lead to either Interestingly, atherosclerotic plaques were less frequently encountered in individuals stroke, myocardial infarction or peripheral artery disease, with different immune bearing the CC genotype, although not statistically significant, independent of their mechanisms playing a role in all stages of atherosclerosis (5). HDLc level. Finally, RA patients who previously experienced a CV event had lower HDLc levels as expected, with the patients bearing TT- genotype having the lowest Interestingly, recent epidemiological studies have identified common inflammatory concentrations of them all. LDLc and TC concentrations were reduced in all patients mechanisms shared by many chronic inflammatory (autoimmune) diseases such with a prevalence of CVD, independent of genotype. as rheumatoid arthritis (RA), systemic lupus erythematousus (SLE) and human immunodeficiency virus (HIV-1) infection on the one hand and atherosclerosis on Following this study, we questioned the exact mechanisms behind the observed the other hand (6-8). Individuals suffering from these diseases are known to have an association of IL-32 and HDLc levels in RA patients. Two major hypotheses evolved, increased risk to develop CVD compared to the general population that may not fully one implying the modulatory effect of general inflammation and disease activity be captioned by traditional CVD risk factors (9, 10). Because there is still a strong on lipoprotein levels (15), in which IL-32 would indirectly affect lipids through other need to improve the understanding of the underlying inflammatory mechanisms inflammatory markers already indicated to do that e.g. TNF, IL-1. The other hypothesis in atherosclerosis, exploring the inflammatory mechanisms contributing to would postulate a direct effect of IL-32 on lipoprotein metabolism, by modulating diseases like RA and HIV could help to understand and improve the treatment of their production or clearance in circulation. To explore the first hypothesis we turned atherosclerosis and its serious complications. our attention to two chronic inflammatory diseases: rheumatoid arthritis and HIV- infection. In chapter 3 the IL-32 promoter SNP and the different IL-32 isoforms 7 Therefore, in this thesis, we aimed to explore the hypothesis that interleukin (IL)-32 (IL-32α, IL-32β and IL-32γ) and their relation to TNFα/TNFα-inhibitor treatment in is one of the important inflammatory mediators contributing to the increased risk RA patients have been studied. TNFα is one of the major cytokines involved in RA, for CVD in chronic inflammatory diseases. The last few years IL-32 has gained a lot linked to disease activity and is strongly correlated to IL-32 in a positive feedback of interest, reports indicating that the cytokine may be involved in the pathogenesis loop being able to enhance each other’s expression. Moreover, previous studies of various inflammatory diseases of which most are associated with an increased have looked at SNPs in cytokines such as IL-6, IL-1β or TNFα and their association cardiovascular risk. The results of this thesis propose that IL-32 could play a role in to the response to treatment in RA and found positive correlations (16-20). We the development of CVD in patients suffering from chronic low-grade inflammatory therefore explored the role of IL-32 and the IL-32 promoter SNP in RA patients diseases. with respect to their clinical response to treatment with TNF- blocking agents. By performing ex vivo stimulation with Candida albicans or Pam3Cys we found that IL- Rheumatoid arthritis is a chronic inflammatory autoimmune disease associated with 32β and IL-32γ mRNA expression tended to be higher in PBMCs from RA patients an increased risk for CVD (11, 12). Recent studies have shown a role for IL-32 in disease compared to healthy individuals. Further analysis also showed that intracellular severity of RA and a strong correlation with TNFα expression, a major cytokine in RA IL-32 protein production was increased in RA patients, especially in those bearing (13, 14). However, never before has IL-32 been explored as a potential contributor the CC genotype for the promoter SNP. Even though a tendency towards higher

138 139 Chapter 7 – Summary, General discussion & Future perspectives

concentrations of pro-inflammatory cytokines was observed, this did not translate overexpressing the various isoforms of IL-32 in HepG2 cells and THP-1 cells resulted to higher disease activity scores (DAS28-CRP). Most importantly, ex vivo stimulation in lower intracellular lipid content suggesting increased lipid efflux from the cells. of PBMCs in the presence of anti-TNF agents etanercept or adalimumab was able to Indeed, THP-1 cells overexpressing IL-32γ showed higher ABCA1 expression and potentially identify responders to treatment according to clinical criteria, dependent when IL-32γ was silenced expression of ABCA1 was strongly decreased. Besides in on the genotype for the IL-32 promoter SNP. As IL-32 genotypes are not related with carotid plaque tissue from humans both IL-32γ and ABCA1 mRNA expression was disease activity and response to TNF blockade, the results of this study indirectly induced. Eventually, this study added more data to support the hypothesis of the suggest that other mechanisms might be involved for the initial association between direct interactions between IL-32 and intermediary metabolism pathways. IL-32 and HDLc concentration. Finally, we were interested to explore whether the results suggested by the animal We further turned our attention to our second hypothesis, which propose a and in vitro models would be confirmed by in vivo human studies. In order to do direct effect of IL-32 on metabolic pathways. In order to investigate that, several that, we turned our attention to another chronic inflammatory condition, known to approaches have been explored. We firstly studied in a mice model the role of IL-32 be associated with higher cardiovascular risk: HIV infection. The risk for developing on lipid metabolism and obesity. In general, mice lack the IL32 gene and thereby CVD in HIV-infected individuals is at least two times higher than that of non-HIV- it is an interesting animal species to study the black and white differences of the infected (22, 23). From previous studies, it has been described that IL-32 is involved absence/presence of IL-32 in the system. Recently, IL-32 has been suggested to in disease progression since high levels of endogenous IL-32 prevented further viral be directly implicated in insulin-resistance and diabetes, underlining that this replication (24, 25). Subsequently, knocking down IL-32 expression resulted in a cytokine may somehow be connected with intermediary metabolism pathways highly increased viral load (24). Our next approach was to study the role of IL-32 (21). In chapter 4 we therefore studied the effects of IL-32 on lipid metabolism and and its promoter SNP on cardiovascular risk mediators in chronically HIV-infected obesity in an in vivo model of human IL-32 transgenic mice on regular chow diet. under stable treatment with cART compared to healthy individuals. In chapter 6 Moreover, the effect of IL-32 presence in mice on their adipose tissue phenotype we selected a group of HIV-infected individuals together with a group of healthy and possible inflammatory state was studied. Interestingly, adipose tissue cell size individuals. We found that IL-32β and IL-32γ mRNA expression in PBMCs from HIV- was significantly increased in IL-32 transgenic mice and circulating concentrations infected was different than the expression in PBMCs of healthy individuals dependent of leptin were also increased. Furthermore, a trend towards higher circulating on the IL-32 promoter SNP genotype. Moreover, we observed IL-32 SNP dependent insulin and cholesterol concentrations in IL-32tg mice was observed. These findings variation in cholesterol transporter mRNA expression of ABCA1 and ABCG1. In HIV- suggest that IL-32 plays an important role in lipid metabolism and that studying infected, ABCA1 and ABCG1 were induced in the TT-genotype whereas healthy these animals under a high fat diet regimen would result in even more interesting individuals showed this induction when bearing the CC-genotype. Furthermore, 7 data. Moreover, this data prompted us to continue to explore this hypothesis using we studied whether a chaperone protein calnexin could be the explanation for the different approaches. differences in ABCA1/ABCG1 expression dependent on genotype. However, this was not the case since ABCA1/ABCG1 were correlated with IL-32γ expression whereas In chapter 5 we performed a more basic science study, trying to explore the role calnexin expression was correlated with the IL-32β isoform. This final study adds of IL-32 and its main isoforms IL-32α,β,γ on mediators involved in lipoprotein to the previously described body of evidence suggesting a direct modulatory effect metabolism. Important mediators such as, ABCA1 and ABCG1, apolipoprotein of IL-32 on intermediary metabolic pathways, specifically HDL production and A1 (main component of HDLc) and LXRα which controls transcriptional programs lipoprotein metabolism. involved in lipid homeostasis/inflammation were studied in human primary hepatocytes, HepG2 cells, THP-1 derived macrophages and carotid plaque tissue. We were able to show that in human primary hepatocytes and HepG2 cells IL-32γ DISCUSSION mRNA expression is positively correlated with the expression of ABCA1, ABCG1, LXRα and ApoA1 in these cells. In contrast, IL-32β and IL-32α did not show any Taken together, the results presented in this thesis support a role for IL-32 in correlation or showed a slightly negative correlation. Furthermore, we showed that affecting lipoprotein metabolism pathways and HDL cholesterol. Nevertheless, a

140 141 Chapter 7 – Summary, General discussion & Future perspectives

question which arose from the observed data and still remains to be investigated is; To conclude how can a pro-inflammatory cytokine be associated with increased levels of HDLc In this thesis, we have shown that IL-32 and a promoter SNP in the IL32 gene are and possible protection against CVD? important in regulation of lipoprotein metabolism. We have showed that HDLc concentrations differ between genotypes and that by stratifying clinical responding From many previous studies it is known that lipid profiles can become pro- RA patients to the promoter SNP genotype, the CC-genotype could serve as a atherogenic during inflammation thereby augmenting the risk of developing CVD possible new biomarker to predict the response to anti-TNF treatment in RA. (26-30). Both in patients suffering from RA or HIV, lipid profiles are changed due to Furthermore, we were able to show that IL-32 transgenic mice on regular chow diet the inflammatory status resulting in decreased HDLc and/or ApoA1 and overall higher already had enlarged adipocytes and increased circulating leptin concentrations ratios between LDLc and HDLc (31, 32). As it has been described earlier, IL-32 can compared to WT mice, suggesting a disturbed metabolic state as compared with the induce pro-inflammatory cytokines and chemokines involved in these diseases and wild-type mice. Next, we showed that the IL-32 promoter SNP affects cholesterol is therefore thought of as a novel player in promoting CVD. Interestingly, dependent mediators differentially in HIV-infected individuals compared to healthy individuals on the IL-32 promoter SNP genotype, HDL cholesterol levels were modulated in RA and showed variation in IL-32 isoform expression dependent on the genotype. Lastly, patients treated with biological disease-modifying anti-rheumatic drugs (bDMARDs) we were able to unravel part of the underlying mechanism by showing that IL-32γ and non-RA patients but not in HIV-infected patients on stable cART. Literature has is the main isoform responsible for regulation of cholesterol transporter expression shown that long-term treatment with TNF inhibitors only results in a minor increase and expression of important mediators such as LXRα and ApoA1. We have therefore in HDLc concentrations and no effect on LDLc concentrations or atherogenic index provided evidence supporting the concept that IL-32 is a candidate to be studied in (AI) (33). Moreover, treatment of RA patients with bDMARDs could improve the lipid the development of CVD in chronic inflammatory diseases. profile due to improved HDLc functions/ cholesterol efflux and overall decrease of inflammation but data is still conflicting (34-37). In contrast, antiretroviral therapy in Future perspectives HIV has clearly been linked to a worsened CVD outcome (38-40). This could explain We now know that a promoter SNP in the IL32 gene is associated with increased why the promoter SNP does not show any effects on HDLc in the HIV-infected concentrations of HDLc and that this SNP might be able to distinguish RA patients individuals. Besides, LDLc and TC concentrations were not affected by the IL-32 clinically responding to etanercept or adalimumab after ex vivo stimulation of their promoter SNP in RA patients, which could suggest that the SNP effect is specific PBMCs. Furthermore, human IL-32 transgenic mice, which normally lack IL-32 for HDLc metabolism/regulation. Moreover, overexpressing IL-32 in HepG2 cells expression, already show to have increased adipocytes and circulating leptin levels showed to be correlated with increased cholesterol transporter expression, which on a regular chow diet. Besides, the IL-32 promoter SNP seemed to affect ABCA1 could induce cholesterol efflux and lower intracellular lipid content. This might and ABCG1 mRNA expression differentially in HIV-infected individuals compared to 7 explain the selected effect on HDLc regulation. What we, however, did not show healthy individuals. Lastly, mechanistically this effect seemed to be associated with in chapter 5 was whether the reduced cellular lipid content is actually caused by specific isoform IL-32γ, which was strongly correlated to mRNA expression of ABCA1, increased cholesterol efflux due to increased function of cholesterol transporters. LXRα and ApoA1 in various cell types. However, the exact and causal role of the Besides, ABCA1 transporters not only play a role in reverse cholesterol transport promoter SNP and IL-32 isoforms in the risk to develop CVD still needs to be proven. (RCT), they also play a role in protection against viral infection (41, 42). Enveloped Therefore, the precise function of the promoter SNP on IL-32 isoform expression and viruses depend on cholesterol for their uptake and replication (43). However, when the effect of treatment on IL-32 isoform expression in different inflammatory diseases ABCA1 cholesterol transporters are highly expressed and functional, intracellular compared to healthy individuals should be studied in more detail. lipid content decreases preventing viral infection. Il-32 expression and especially the specific isoforms could therefore serve multiple roles within the regulation of When trying to explore the role of IL-32 isoforms IL-32α, IL-32β and IL-32γ in lipoprotein metabolism. This should be the focus of future research. inflammation and lipoprotein metabolism, the first question that needs to be approached is what is the exact function of each isoform intracellularly. To study this, it would be important to establish an inducible knock- out cell line (for example a macrophage cell-line) in which each individual isoform can be induced/silenced

142 143 Chapter 7 – Summary, General discussion & Future perspectives

separately. A complete knock-out cell line of the whole IL32 gene by CRISPR/ Modulation of IL-32 isoforms to improve cardiovascular disease CAS9 in HepG2 cells was not viable probably due to the importance of IL-32 in cell outcome in chronic inflammatory disorders homeostasis (44, 45). A relevant clinical question that remains is the following: could strategies regulating IL-32 isoform expression contribute to novel therapies to lower atherosclerosis risk By generating an inducible knock-out we could study the specific effects of each and additional complications? It is know that SNPs influence the expression/function isoform in cells important for inflammatory responses and lipoprotein metabolism. of proteins. One way this can occur is by interfering with alternative splicing (47). Macrophages expressing the various isoforms could be loaded with ox-LDL to study Various cytokines and other proteins are known to have multiple isoforms including foam cell formation and expression/function of cholesterol transporter activity IL-1, IL-6, IL-12, Toll-Like receptors (TLRs) and many transcription factors (48-50). dependent on IL-32 isoform. Furthermore, expressing only one isoform could help us These isoforms are mostly generated through alternative splicing and about 94% of to explore the exact mechanism behind the regulation of ABCA1, ABCG1, LXRα and human genes is subjected to alternative splicing (50). Alternative splicing increases ApoA1 in these cells. Does IL-32γ directly affect LXRα therefore increasing ABCA1/ the diversity of mRNAs, which leads to profound functional effects important for ABCG1 or is this an indirect affect and what happens in the presence of IL-32β/IL- innate immune responses (50, 51). It can regulate localization of proteins, the 32α? Besides studying the isoforms in an in vitro model, it would be interesting to binding to other proteins or membranes, the enzymatic activity and interaction with study the precise expression pattern of IL-32 isoforms depending on the promoter ligands such as receptors. Regarding IL-32, alternative splicing results in variation of SNP genotype in PBMCs from healthy individuals (ex vivo), The SNP seems to isoform expression. The SNP could affect binding of the spliceosome to the promoter activate the promoter region resulting in more IL-32 protein expression (46). region, creating differences in splicing of IL-32 mRNA and post-transcriptional changes in isoform expression. Moreover, an online prediction model showed that However, how the change of one nucleotide could lead to a change in the type of IL- the change from a T-allele to a C- allele could interfere with histone deacetylase 32 isoform being expressed still remains an interesting topic of research. 2 (HDAC2) binding (52). This could suggest that epigenetic changes occur in IL-32 once individuals bear the CC genotype but how this affects the expression of splice Another question which remains to be addressed is: how does pharmacological variants remains unknown. Subsequently, this could provide a new way to be able treatment in the various chronic inflammatory diseases affect IL-32 isoform to try and regulate specific isoform expression of IL-32. To elucidate this, in vitro and expression, and does that have an impact on HDLc levels or CV risk? In addition, possible in vivo studies should be performed with the addition of HDAC2 inhibitor it would be interesting to study the effect of statins or the new anti-IL-1β drug Romidepsin to study the effect of HDAC2 on IL-32 isoform expression and possible canakinumab on IL-32 isoform expression. These drugs are known to lower the risk effects on lipoprotein metabolism. for CVD and might influence IL-32 isoform expression in a way that could explain a 7 new working mechanism of these drugs. Furthermore, it would be highly interesting Another topic of interest is why would HIV-infected individuals show a different to study PBMCs from RA patients and HIV patients from the beginning of the disease expression of IL-32 isoforms dependent on the IL-32 SNP compared to healthy when they are still naïve for treatment and then follow them over the course of individuals? From recent unpublished data we know that poly I:C (dsRNA TLR3 time. This way we could study the role of the disease on IL- 32 expression and study agonist) somehow affects splicing resulting in mostly IL- 32γ expression but only the effect of treatment on expression of IL-32 by PBMCs. Regarding the last two little expression of IL-32β/IL-32α. HIV-infected individuals are infected with a virus questions, we should try to include a large number of patients to be able to stratify which can trigger a TLR3 response possibly changing the splicing of IL-32 isoforms results for each genotype of the IL-32 promoter SNP. Finally, creating a new ELISA and changing the isoform expression dependent on genotype. However, these able to measure protein expression of the various isoforms would provide a better effects should be studied in untreated/acute HIV-infected patients compared to understanding of which isoforms represent the total IL-32 protein concentrations healthy individuals since treatment with cART could have strong effects on IL-32 (in)dependent of the SNP genotype. expression.

Besides observed changes in IL-32 splicing due to poly I:C stimulation or cART treatment, there is no knowledge on the effect of commonly used RA therapies

144 145 Chapter 7 – Summary, General discussion & Future perspectives

including methotrexate (MTX) or bDMARDs. MTX was initially used to treat cancer Finally, despite many years of attempts to prevent atherosclerosis and cardiovascular until in 1962 the first study of MTX treatment in RA was performed with positive effects disease it still remains a global burden. Using novel strategies, like affecting (53). Low doses of MTX inhibit 5-aminoimidazole-4- carboxamide ribonucleotide cholesterol transporters via new cytokines such as IL-32 and screening patients for (AICAR) formylotransferase, which affects intracellular adenosine concentrations SNPs that could predict response to treatment, we as researchers can continue to and result in increased anti-inflammatory responses such as decreased TNFα and work towards the goal of full prevention of CVD in the future. IL- 6 secretion (54, 55). In that way, MTX can decrease disease activity and lower cardiovascular disease risk. Unpublished data from our group showed that MTX also influences splicing of IL-32 just like poly I:C. Even more interesting, when we treated PBMCs with poly I:C + MTX (10oo-10ng/mL), splicing of IL-32 seemed to be blocked completely and only IL-32γ mRNA was produced. It would therefore be highly interesting to study the effect of MTX on IL-32 in RA patients negative for MTX (or any other treatment) and follow these patients over time to observe possible changes in expression profiles. MTX has been shown to lower CVD risk, which could be due to its anti- inflammatory effects but additionally could affect cholesterol transporter expression due to strong regulation of IL-32γ expression.

IL-32 promoter SNP as biomarker for personalized medicine Bearing the CC-genotype for the IL-32 promoter SNP could predict the eventual ex vivo response of PBMCs to Candida albicans or Pam3Cys in addition to either etanercept or adalimumab. This is a completely new finding within the field. We therefore hypothesize that the IL-32 promoter SNP could serve as a novel biomarker in predicting response to treatment in RA patients on either etanercept or adalimumab. Since etanercept is the first choice of bDMARDS RA patients are given, this suggests that the IL-32 SNP could make a distinction between patients that will respond or will switch to another type of bDMARD and prevents the trouble suffering from ongoing disease activity due to lack of response. To obtain an unbiased group 7 of RA patients to study response to treatment, preferably untreated RA patients should be included and followed over a time course of at least 6 months according to EULAR guidelines (56). After 3-months the first follow-up should be performed measuring DAS28-CRP to check effect of therapy, however response to treatment can still change after three months making a 6 months follow-up necessary.

In line with this patients should be separated on MTX positive or negative since we know from unpublished data that MTX could interfere with IL-32 isoform expression. Furthermore, during follow-up lipid profiles should be determined to study possible changes over the course of time.

146 147 Chapter 7 – Summary, General discussion & Future perspectives

REFERENCES

1. centre WM. Cardiovascular diseases (CVDs) Fact sheet. UPDATED 2017 ed. WHO webpage: WHO; 2017. 21. Catalan V, Gomez-Ambrosi J, Rodriguez A, Ramirez B, Valenti V, Moncada R, et al. Increased Interleukin-32 2. Munro JM, Cotran RS. The pathogenesis of atherosclerosis: atherogenesis and inflammation. Lab Invest. Levels in Obesity Promote Adipose Tissue Inflammation and Extracellular Matrix Remodeling: Effect of 1988;58(3):249-61. Weight Loss. Diabetes. 2016;65(12):3636-48. 3. Holman RL, Mc GH, Jr., Strong JP, Geer JC. The natural history of atherosclerosis: the early aortic lesions 22. Triant VA. Epidemiology of coronary heart disease in patients with human immunodeficiency virus. Rev as seen in New Orleans in the middle of the of the 20th century. Am J Pathol. 1958;34(2):209-35. Cardiovasc Med. 2014;15 Suppl 1:S1-8. 4. Ross R. Atherosclerosis is an inflammatory disease. Am Heart J. 1999;138(5 Pt 2):S419-20. 23. Zanni MV, Schouten J, Grinspoon SK, Reiss P. Risk of coronary heart disease in patients with HIV infection. 5. Frostegard J. Immunity, atherosclerosis and cardiovascular disease. BMC Med. 2013;11:117. Nat Rev Cardiol. 2014;11(12):728-41. 6. Hollan I, Scott H, Saatvedt K, Prayson R, Mikkelsen K, Nossent HC, et al. Inflammatory rheumatic disease 24. Nold MF, Nold-Petry CA, Pott GB, Zepp JA, Saavedra MT, Kim SH, et al. Endogenous IL-32 controls and smoking are predictors of aortic inflammation: a controlled study of biopsy specimens obtained at cytokine and HIV-1 production. J Immunol. 2008;181(1):557-65. coronary artery surgery. Arthritis Rheum. 2007;56(6):2072-9. 25. Rasool ST, Tang H, Wu J, Li W, Mukhtar MM, Zhang J, et al. Increased level of IL-32 during human 7. Turesson C, Jacobsson LT, Matteson EL. Cardiovascular co-morbidity in rheumatic diseases. Vasc Health immunodeficiency virus infection suppresses HIV replication. Immunol Lett. 008;117(2):161-7.2 Risk Manag. 2008;4(3):605-14. 26. Park YB, Lee SK, Lee WK, Suh CH, Lee CW, Lee CH, et al. Lipid profiles in untreated patients with 8. Kabanny MTJ, A.A.;Mehta, N.N.; FACC. Cardiovascular diseases in chronic inflammatory disorders. rheumatoid arthritis. J Rheumatol. 1999;26(8):1701-4. American College of Cardiology. 2016. 27. Park YB, Choi HK, Kim MY, Lee WK, Song J, Kim DK, et al. Effects of antirheumatic therapy on serum lipid 9. Mirjafari H, Al-Husain A, Bruce IN. Cardiovascular risk factors in inflammatory arthritis. Curr Opin Lipidol. levels in patients with rheumatoid arthritis: a prospective study. Am J Med. 2002;113(3):188-93. 2011;22(4):296-301. 28. Kim JY, Lee EY, Park JK, Song YW, Kim JR, Cho KH. Patients with Rheumatoid Arthritis Show Altered 10. Amaya-Amaya J, Sarmiento-Monroy JC, Mantilla RD, Pineda-Tamayo R, Rojas-Villarraga A, Anaya JM. Lipoprotein Profiles with Dysfunctional High-Density Lipoproteins that Can Exacerbate Inflammatory and Novel risk factors for cardiovascular disease in rheumatoid arthritis. Immunol Res. 2013;56(2-3):267-86. Atherogenic Process. Plos One. 2016;11(10):e0164564. 11. Solomon DH, Goodson NJ, Katz JN, Weinblatt ME, Avorn J, Setoguchi S, et al. Patterns of cardiovascular 29. Cui HL, Ditiatkovski M, Kesani R, Bobryshev YV, Liu Y, Geyer M, et al. HIV protein Nef causes dyslipidemia risk in rheumatoid arthritis. Ann Rheum Dis. 2006;65(12):1608-12. and formation of foam cells in mouse models of atherosclerosis. FASEB J. 2014;28(7):2828-39. 12. Avina-Zubieta JA, Thomas J, Sadatsafavi M, Lehman AJ, Lacaille D. Risk of incident cardiovascular 30. Crowe SM, Westhorpe CL, Mukhamedova N, Jaworowski A, Sviridov D, Bukrinsky M. The macrophage: the events in patients with rheumatoid arthritis: a meta-analysis of observational studies. Ann Rheum Dis. intersection between HIV infection and atherosclerosis. J Leukoc Biol. 2010;87(4):589-98. 2012;71(9):1524-9. 31. Estrada V, Portilla J. Dyslipidemia Related to Antiretroviral Therapy. Aids Rev. 2011;13(1):49-56. 13. Joosten LA, Netea MG, Kim SH, Yoon DY, Oppers-Walgreen B, Radstake TR, et al. IL-32, a proinflammatory 32. Arts E, Fransen J, Lemmers H, Stalenhoef A, Joosten L, van Riel P, et al. High-density lipoprotein cholesterol cytokine in rheumatoid arthritis. Proc Natl Acad Sci U S A. 2006;103(9):3298-303. subfractions HDL2 and HDL3 are reduced in women with rheumatoid arthritis and may augment the 14. Heinhuis B, Koenders MI, van Riel PL, van de Loo FA, Dinarello CA, Netea MG, et al. Tumour necrosis cardiovascular risk of women with RA: a cross-sectional study. Arthritis Res Ther. 2012;14(3):R116. factor alpha-driven IL-32 expression in rheumatoid arthritis synovial tissue amplifies an inflammatory 33. Daien CI, Duny Y, Barnetche T, Daures JP, Combe B, Morel J. Effect of TNF inhibitors on lipid profile in cascade. Ann Rheum Dis. 2011;70(4):660-7. rheumatoid arthritis: a systematic review with meta-analysis. Ann Rheum Dis. 2012;71(6):862-8. 15. Popa C, Netea MG, van Riel PL, van der Meer JW, Stalenhoef AF. The role of TNF-alpha in chronic 34. Low AS, Symmons DP, Lunt M, Mercer LK, Gale CP, Watson KD, et al. Relationship between exposure to inflammatory conditions, intermediary metabolism, and cardiovascular risk. J Lipid Res. 2007;48(4):751- tumour necrosis factor inhibitor therapy and incidence and severity of myocardial infarction in patients 62. with rheumatoid arthritis. Ann Rheum Dis. 2017;76(4):654-60. 7 16. Miceli-Richard C, Comets E, Verstuyft C, Tamouza R, Loiseau P, Ravaud P, et al. A single tumour necrosis 35. Gabay C, McInnes IB, Kavanaugh A, Tuckwell K, Klearman M, Pulley J, et al. Comparison of lipid and factor haplotype influences the response to adalimumab in rheumatoid arthritis. Ann Rheum Dis. lipid-associated cardiovascular risk marker changes after treatment with or adalimumab in 2008;67(4):478-84. patients with rheumatoid arthritis. Ann Rheum Dis. 2016;75(10):1806-12. 17. Padyukov L, Lampa J, Heimburger M, Ernestam S, Cederholm T, Lundkvist I, et al. Genetic markers 36. Tam LS, Tomlinson B, Chu TT, Li TK, Li EK. Impact of TNF inhibition on insulin resistance and lipids levels in for the efficacy of tumour necrosis factor blocking therapy in rheumatoid arthritis. Ann RheumDis. patients with rheumatoid arthritis. Clin Rheumatol. 2007;26(9):1495-8. 2003;62(6):526-9. 37. Singh JA, Beg S, Lopez-Olivo MA. Tocilizumab for rheumatoid arthritis: a Cochrane systematic review. J 18. Davila-Fajardo CL, Marquez A, Pascual-Salcedo D, Moreno Ramos MJ, Garcia-Portales R, Magro C, et Rheumatol. 2011;38(1):10-20. al. Confirmation of -174G/C interleukin-6 gene promoter polymorphism as a genetic marker predicting 38. Brown TT, Glesby MJ. Management of the metabolic effects of HIV and HIV drugs. Nat Rev Endocrinol. antitumor necrosis factor treatment outcome. Pharmacogenet Genomics. 2014;24(1):1-5. 2011;8(1):11-21. 19. Harrison P, Pointon JJ, Chapman K, Roddam A, Wordsworth BP. Interleukin-1 promoter region 39. Mallon PW. Impact of nucleoside reverse transcriptase inhibitors on coronary heart disease. Rev polymorphism role in rheumatoid arthritis: a meta-analysis of IL-1B-511A/G variant reveals association Cardiovasc Med. 2014;15 Suppl 1:S21-9. with rheumatoid arthritis. Rheumatology (Oxford). 2008;47(12):1768-70. 40. Nguyen KA, Peer N, Mills EJ, Kengne AP. A Meta-Analysis of the Metabolic Syndrome Prevalence in the 20. O’Rielly DD, Roslin NM, Beyene J, Pope A, Rahman P. TNF-alpha-308 G/A polymorphism and Global HIV-Infected Population. Plos One. 2016;11(3):e0150970. responsiveness to TNF-alpha blockade therapy in moderate to severe rheumatoid arthritis: a systematic 41. Cui HL, Grant A, Mukhamedova N, Pushkarsky T, Jennelle L, Dubrovsky L, et al. HIV-1 Nef mobilizes lipid review and meta-analysis. Pharmacogenomics J. 2009;9(3):161-7. rafts in macrophages through a pathway that competes with ABCA1-dependent cholesterol efflux. J Lipid Res. 2012;53(4):696-708.

148 149 Chapter 7 – Summary, General discussion & Future perspectives

42. Mukhamedova N, Brichacek B, Darwish C, Popratiloff A, Sviridov D, Bukrinsky M. Analysis of ABCA1 and Cholesterol Efflux in HIV-Infected Cells. Methods Mol Biol. 2016;1354:281-92. 43. Campbell SM, Crowe SM, Mak J. Virion-associated cholesterol is critical for the maintenance of HIV-1 structure and infectivity. AIDS. 2002;16(17):2253-61. 44. Heinhuis B, Netea MG, van den Berg WB, Dinarello CA, Joosten LA. Interleukin-32: a predominantly intracellular proinflammatory mediator that controls cell activation and cell death. Cytokine. 2012;60(2):321-7. 45. Nold-Petry CA, Rudloff I, Baumer Y, Ruvo M, Marasco D, Botti P, et al. IL-32 promotes angiogenesis. J Immunol. 2014;192(2):589-602. 46. Westra H-J, Peters MJ, Esko T, Yaghootkar H, Schurmann C, Kettunen J, et al. Systematic identification of trans eQTLs as putative drivers of known disease associations. Nature Genetics. 2013;45:1238. 47. Malik M, Simpson JF, Parikh I, Wilfred BR, Fardo DW, Nelson PT, et al. CD33 Alzheimer’s risk-altering polymorphism, CD33 expression, and exon 2 splicing. J Neurosci. 2013;33(33):13320-5. 48. Sahoo A, Im SH. Interleukin and diversity: role of alternative splicing. Int Rev Immunol. 2010;29(1):77-109. 49. Wang ET, Sandberg R, Luo S, Khrebtukova I, Zhang L, Mayr C, et al. Alternative isoform regulation in human tissue transcriptomes. Nature. 2008;456(7221):470-6. 50. Carpenter S, Ricci EP, Mercier BC, Moore MJ, Fitzgerald KA. Post-transcriptional regulation of gene expression in innate immunity. Nat Rev Immunol. 2014;14(6):361-76. 51. Kelemen O, Convertini P, Zhang Z, Wen Y, Shen M, Falaleeva M, et al. Function of alternative splicing. Gene. 2013;514(1):1-30. 52. Ward LD, Kellis M. HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res. 2012;40(Database issue):D930-4. 53. Black RL, O’Brien WM, Vanscott EJ, Auerbach R, Eisen AZ, Bunim JJ. Methotrexate Therapy in Psoriatic Arthritis; Double-Blind Study on 21 Patients. JAMA. 1964;189:743-7. 54. Swierkot J, Szechinski J. Methotrexate in rheumatoid arthritis. Pharmacol Rep. 2006;58(4):473-92. 55. Cutolo M, Sulli A, Pizzorni C, Seriolo B, Straub RH. Anti-inflammatory mechanisms of methotrexate in rheumatoid arthritis. Ann Rheum Dis. 2001;60(8):729-35. 56. Smolen JS, Landewe R, Breedveld FC, Buch M, Burmester G, Dougados M, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease- modifying antirheumatic drugs: 2013 update. Ann Rheum Dis. 2014;73(3):492-509. 7

150 151 Chapter 8 Nederlandse samenvatting

8

152 153 Chapter 8 – Nederlandse samenvatting

Atherosclerose als oorzaak van hart- en vaatziekten (toen en nu) In mijn thesis heb ik onderzocht hoe een relatief nieuwe immuunstof genaamd Iedereen kent wel iemand in zijn of haar omgeving die lijdt aan hart- en vaatziekten interleukine-32 (IL-32) en een genetische verandering in IL-32, betrokken is bij het (HVZ). De meest voorkomende onderliggende oorzaak is atherosclerose, ook wel ontstaan van een verhoogd risico op hart- en vaatziekten in patiënten met een bekend als aderverkalking (Figuur 1). chronische ontstekingsreactie. Daarnaast heb ik gekeken of we kunnen voorspellen of iemand meer of minder risico loopt in de aan/afwezigheid van deze stof en de Vroeger dacht men dat atherosclerose niets meer was dan de opstapeling genetische verandering in gezonde individuen en in patiënten met chronische van cholesterol in de vaatwand. Atherosclerose ontstaat echter doordat witte ontstekingsreacties. bloedcellen van het immuunsysteem cholesterol opeten. Wanneer de hoeveelheid cholesterol in de bloedbaan te hoog is kunnen witte bloedcellen dit niet goed Mijn hypothese en projecten in het lab verwerken waardoor ze gaan plakken aan de binnenkant van de bloedvaten. Dit De grote vraag die ik tijdens mijn PhD heb proberen te beantwoorden was: ‘Op resulteert in een ontstekingsreactie. Tijdens een ontstekingsreactie produceren welke manier is interleukine-32 en zijn verschillende isovormen (IL-32a, IL-32b en witte bloedcellen bepaalde stoffen (o.a. cytokines en chemokines) die meer witte IL-32y) betrokken bij het ontstaan van het verhoogde risico op hart-en vaatziekten/ bloedcellen aantrekken en ervoor zorgen dat de veroorzaker van de ontsteking atherosclerose in patiënten die lijden aan chronische ontstekings ziekten?’ (normaal gesproken een bacterie of virus) wordt verwijderd. In het geval van Gedurende de jaren van mijn PhD onderzocht ik dit in verschillende ziektebeelden atherosclerose zorgen de geproduceerde immuunstoffen er echter voor dat meer en op verschillende niveaus. witte bloedcellen stapelen en een plaque ontstaat. Afhankelijk van de hoeveelheid van cholesterol in het bloed en de duur van de aanwezigheid kan een plaque zich Reumatoïde artritis en IL-32 verder uitbreiden en leiden tot verstopping van een bloedvat of een scheuring van Een deel van mijn onderzoek focuste zich op het effect van IL-32 in HVZ in reuma de vaatwand. patiënten. Door middel van eerdere studies is bekend dat IL-32 een rol speelt in de ziekte activiteit van RA patiënten. Ook is IL-32 sterk gecorreleerd met een andere zeer belangrijke immuunstof in RA, namelijk TNF-alfa. Wanneer de ene immuunstof toeneemt, stimuleert dit de toename van de ander. Dit wordt ook wel een positive feedback loop genoemd. In dit geval resulteert dit in een hogere ziekte activiteit en Plaque een slechtere prognose voor de patiënt. De rol van IL-32 in HVZ in deze patiënten is echter nog nooit eerder bestudeerd.

Genetische verandering in het IL32 gen Figuur 1. Overzicht van de verschillende fases van atherosclerose. Een van de meest gebruikten methoden om te zien of iemand een verhoogd risico loopt op het ontwikkelen van hart- en vaatziekten is het bepalen van het cholesterol. Tegenwoordig weten we dat de ontstekingsreactie in atherosclerose vergelijkbare In patiënten met een chronische ontsteking is de samenstelling van het cholesterol mechanismen vertoond met ontstekingsreacties bestudeerd in chronische vaak verstoord. Deze verstoring kan bijdragen aan het verhogen van het risico op 8 ontsteking (auto-immuun-) ziektes zoals reumatoïde artritis (RA), systemische lupus hart- en vaatziekten. Een groot deel van mijn onderzoek richtte zich op een genetisch erytematodus (SLE) en humaan immunodeficiëntie virus (HIV). Patiënten die lijden verandering in het IL-32 gen. Deze verandering beïnvloedt cholesterol concentraties aan een van deze ziektes hebben een verhoogd risico op het ontwikkelen van hart- en daardoor factoren die het risico op hart- en vaatziekten verhogen. In het eerste en vaatziekten/atherosclerose. Verschillende factoren kunnen een rol spelen in de deel van mijn thesis heb ik gekeken naar het effect van deze genetische verandering ontwikkeling van deze ziektes en atherosclerose zoals, leefstijl (roken, overgewicht/ op HDL cholesterol (het “goede” cholesterol) concentraties en risico op HVZ in lichamelijke inactiviteit) of onderliggende genetische veranderingen/aandoeningen. het bloed van reuma patiënten vergeleken met niet-reuma patiënten. Interessant genoeg hadden reuma patiënten, met de genetische verandering, een hogere concentratie HDL in het bloed dan niet-reuma patiënten. Dit was onafhankelijk

154 155 Chapter 8 – Nederlandse samenvatting

van de ziekte activiteit. Verder hadden individuen met de genetische verandering aanwezigheid van IL-32 en de verschillende isovormen op belangrijke componenten minder plaques. in het lipoproteïne metabolisme. Naast bepaalde immuun cellen (macrofagen) is de lever een belangrijk orgaan in dit proces. Verschillende lever cellijnen zijn daarom IL-32 en hart- en vaatziekten bestudeerd. Essentiele mediatoren in het lipoproteïne metabolisme zoals ABCA1, Om een beter beeld te krijgen van de manier waarop IL-32 cholesterol concentraties ABCG1 (cholesterol transport kanalen), LXRa en ApoA1, waren positief geassocieerd kan moduleren, hebben we twee aparte vragen proberen te beantwoorden. De met IL-32y expressie, maar negatief met IL-32a. Na het tot over expressie brengen eerste hypothese keek naar de mogelijke indirecte manier waarin IL-32, door het van deze isovormen in zowel lever cellijnen als een macrofaag cellijn constateerden reguleren van andere inflammatoire stoffen, een verandering in cholesterol tot we dezelfde resultaten. Deze resultaten bevestigden daarmee de tweede hypothese. stand kon brengen. De tweede vraag keek juist naar een potentieel direct effect van de aanwezigheid van IL-32 op cholesterol concentraties en cholesterol metabolisme. Als laatste studie hebben we daarom gekeken of de resultaten die we verkregen hadden door in vitro studies met cellen en een muizen studie, ook tot stand kwamen Om de eerste vraag te beantwoorden is er gekeken naar twee groepen patiënten in een humane studie. Hiervoor hebben we ons gericht op een andere patiënten met chronische ontsteking, namelijk reuma patiënten en HIV patiënten. Het is populatie met een chronische inflammatoire aandoening en verhoogd risico op bekend van genetische veranderingen in andere immuunstoffen dat ze een de hart- en vaatziekten, namelijk HIV patiënten. Voorgaande studies hebben laten expressie van de stof kunnen beïnvloeden en zo een rol spelen in het verhoogde zien dat IL-32 een rol speelt in de immuun reactie tegen HIV infectie. In mijn thesis risico voor HVZ. Daarnaast dragen enkele van deze veranderingen ook bij aan een hebben we gekeken naar de rol van IL-32 en de verschillende isovormen in langdurig verslechterde response op medicatie. In reuma patiënten werd daarom gekeken naar behandelde HIV patiënten. Afhankelijk van het genotype voor de genetische mutatie de response op medicatie (TNF remmers) afhankelijk van de genetische verandering in IL-32, was de expressie van de isovormen IL-32b en IL-32y anders in PBMCs van in IL-32 (genotype). Om te beginnen werden mononucleaire cellen uit het perifere HIV patiënten ten opzichte van gezonde individuen. Verder, vonden we dezelfde bloed (PBMCs) geïsoleerd om de immuun response van deze cellen te bestuderen. link met IL-32y en ABCA1/ABCG1. Echter, welk genotype voor de mutatie in IL-32 PBMCs van reuma patiënten met de genetische mutatie produceerde meer IL-32 hiermee geassocieerd was bleek tegenovergesteld te zijn in HIV patiënten versus en lichtelijk meer pro-inflammatoire immuunstoffen dan patiënten met de andere gezonde individuen. Ondanks dit verschil afhankelijk van genotype, bevestigde genotypen of gezonde individuen. De meeste belangrijke bevinding was dat PBMCs deze data dat IL-32 daadwerkelijk een belangrijke modulerende rol speelt in het gestimuleerd met Candida albicans en behandeld met TNF remmers ( etanercept of lipoproteïne metabolisme en de immuun respons die bijdragen aan het risico op adalimumab) een variatie in immuun response lieten zien afhankelijk van het IL-32 hart- en vaatziekten. genotype. Het genotype kan daarom mogelijk voorspellen hoe iemand zal reageren op de medicatie. Daarnaast gaf dit antwoord op de eerste hypothese doordat IL-32 Het idee voor de toekomst inderdaad een indirect effect kan uitoefenen op cholesterol concentraties. Uiteindelijk is het natuurlijk het idee om specifieke kenmerken te ontdekken in verschillende patiënt groepen met een verhoogd risico op hart- en vaatziekten. Om de tweede vraag te beantwoorden hebben we meerdere experimenten Op die manier kan de patiënt hopelijk persoonlijke medicatie krijgen die het risico uitgevoerd. Ten eerste hebben we in een muis model bestudeert of de aanwezigheid verkleinen of helemaal wegnemen en daarnaast de chronische ontsteking aanpakken. 8 van IL-32 veranderingen kon veroorzaken in het lipiden metabolisme en het Momenteel is het helaas nog steeds een “trial and error” methode. De truc om de ontstaan van obesitas. Andere studies laten zien dat IL-32 geassocieerd is met juiste combinatie van medicijnen te vinden waar de patiënt de minste bijwerkingen insuline-resistentie en diabetes, wat kan betekenen dat IL-32 een rol speelt in ons van ondervind en die ervoor zorgen dat de ziekte en risico’s niet verergeren kan metabolisme. Muizen met IL-32 hadden grotere vetcellen en circulerende hormonen soms heel lang duren. In veel gevallen is het zelfs nog steeds zo dat de medicatie (leptine) geassocieerd met obesitas waren verhoogd. Dit gaf ons reden om deze niet goed/volledig werkt en de patiënt nergens goed op reageert. Het ontdekken directe link van IL-32 met hart-en vaatziekten verder te onderzoeken. van persoonlijke kenmerken en patronen zal een wereld van verschil maken en voor Om meer te weten te komen over hoe IL-32 het lipoproteïne metabolisme kan heel veel mensen bijdragen aan een betere kwaliteit en duur van het leven. beïnvloeden hebben we in cellen gekeken naar het effect van meer of minder

156 157 Addendum – Acknowledgements, Curriculum Vitae

DANKWOORD/ACKNOWLEDGEMENTS Dr. Heinhuis, beste Bas. Bedankt voor alle hulp in het lab en de begeleiding tijdens mijn PhD. Je hebt me de nodige nieuwe technieken geleerd. Daarnaast was het En toen was daar ineens het einde... indrukwekkend hoe je altijd rustig te werk ging en zelden tot nooit gestrest leek/ Inmiddels woon ik alweer een paar maanden in Cincinnati waar ik bezig ben met was. Ik hoop dat je inmiddels een coole plek gevonden hebt om les te geven! Succes mijn postdoc en met een naderende promotie werd het tijd om mijn thesis volledig met je nieuwe carrière! af te ronden door het schrijven van mijn dankwoord. Dit onderdeel is waarschijnlijk het lastigste om te schrijven, maar daarentegen wel het meest gelezen en mogelijk Prof. Dr. Dinarello, dear Charles, as an expert in not only in IL-1 but also IL-32 you belangrijkste onderdeel van het proefschrift. Graag wil ik iedereen bedanken voor helped me to try and explore the function of this interesting cytokine. Thank you for de directe of indirecte bijdrage, steun en input aan dit proefschrift en steun aan mij the possibility to perform part of the IL32 transgenic study within your lab. It was a persoonlijk. honor to be able to work with you.

Ik wil allereerst mijn Promotoren en Copromotor bedanken, Prof. Dr. Leo Joosten, Lieve analisten Heidi, Helga, Cor, Trees, Liesbeth, Anneke en (nieuwe) Liesbeth. Prof. Dr. Niels Riksen en Dr. Calin Popa. Heidi en Helga, jullie zijn echt een goed geoliede machine om naar te kijken als jullie Prof. Dr. Leo Joosten, beste Leo. Heel erg bedankt voor de kans om mijn PhD in samen weer een project aanpakken. Samen hebben we gewerkt aan de RA BIOTOP Lab AIG te volbrengen. Het is bewonderingswaardig hoe je altijd goed gehumeurd studie en ik ben jullie dankbaar voor de vele isolaties en alle vragen die ik kon stellen met bent en iedereen weet te motiveren voor verschillende projecten. Mocht het weer is betrekking tot lipiden en de vele bestellingen. Bedankt dat ik altijd zo makkelijk bij jullie even niet zo gaan als gepland, stond je deur altijd open en bedacht je altijd wel weer kon aankloppen. Cor, een dag zonder jouw aanwezigheid in het lab was maar stilletjes. Je een manier om verder te gaan en de positieve mindset terug te brengen. Je bent een bent een echte sfeermaker, en onze vele grappen naar elkaar plus je hilarische verhalen enorme inspiratie voor velen van ons en ik ben blij dat ik onder jouw begeleiding heb tijdens de pauzes zal ik missen. Bedankt voor je gezelligheid en je hulp met “ shoppen bij mogen werken binnen Lab AIG. Ik zal je geplaatste grappen en positiviteit missen. de buren” en flowcytometrie. Trees en Liesbeth, ik leerde jullie kennen als het ELISA- duo en heb de eer gehad om jullie taak over te nemen in het coating team. Bedankt Prof. Dr. Niels Riksen, beste Niels. Onze samenwerking was van korte duur, maar ik voor alle hulp tijdens mijn masterstage en PhD. Anneke, ik leerde je kennen doordat wil je bedanken voor je input en samenwerking voor het tot stand brengen van mijn ik altijd een praatje kwam maken met Kiki in de ochtend. Je hebt me enorm geholpen proefschrift. met het isoleren van RNA uit plaque weefsel en het verwerken/stainen van de slides van pathologie. Bedankt voor al je hulp en altijd positieve instelling. (Nieuwe) Liesbeth, Dr. Calin Popa, beste Calin. Jij hebt het talent om altijd de juiste vragen te stellen helaas was onze samenwerking maar van korte duur, maar ik heb in die paar weken en complexe resultaten zodanig te beschrijven dat alles opeens duidelijk wordt. een hoop geleerd over transfecties en transfectie efficiëntie. Ik hoop dat je nieuwe plek Ik wil je bedanken voor al onze woensdagochtend meetings samen met Leo en binnen lab AIG goed bevalt, jouw kennis en praktische vaardigheden zullen zeker nog later individuele maandag meetings, waarin we efficiënt en kritisch de data en vaak van pas komen binnen het lab. experimenten bespraken, of afdwaalden naar discussies over voetbal of verhalen over auto’s en Roemenië. Daarnaast waardeer ik enorm dat je eindeloos correcties Buitenhoek roomies, Johanneke en Jaap, als nieuwe onwetende PhD-student kwam wilde toepassen op mijn manuscripten en dat je me af en toe net dat extra zetje in ik bij jullie op de kamer terecht. Twee AIO PhD’s, nogal indrukwekkend. Bedankt dat de rug gaf om iets af te ronden. Bedankt voor je vertrouwen. jullie naar mijn vele gepraat hebben geluisterd en me een goede start van mijn PhD hebben bezorgd. Johanneke, ik vergeet nooit meer wat je tegen me zei vlak nadat A je klaar was met je verdediging. Ik denk dat ik die woorden nog goed moet herhalen voordat ik binnenkort zelf daar sta. Jaap, ik weet niet hoe je het doet, maar ik heb veel respect voor je werkmentaliteit, positieve instelling en humor. Voor beiden het allerbeste voor de toekomst.

158 159 Addendum – Acknowledgements, Curriculum Vitae

Q-roomies, Martin, Andreea, Charlotte de Bree, Megan and Laszlo. I really enjoyed Lieve Siroon en Maartje, wat ben ik blij samen met jullie mijn tijd op het lab te my time in the Q-room (even though nobody of us actually still worked on Q-fever hebben doorgebracht! Eigenlijk lijken jullie heel veel op elkaar qua persoonlijkheid ;)). Thanks for all the discussions on data analysis, experimental design and of course en heb ik het geluk gehad met twee schatten van mensen te hebben mogen our many non-work related talks and enormous piles of candy to be shared :D! Good samenwerken! Bedankt voor alle steun en peptalks door de jaren heen! Ik bewonder luck with all your projects! jullie beiden voor jullie eindeloze vrolijkheid, rust, organisatie skills (behalve tijdens zwangerschaps-dementie ;)), drive en zelfverzekerdheid! Siroon, ik wens je heeeel Wouter bedankt dat ik samen met je mocht werken aan het RAPID-project. Ik heb veel succes en plezier in Australie samen met Thom. Mama Maartje, heel veel succes me prima vermaakt tijdens onze lange dagen in het RNA-hok en onze enorme ELISA met je nieuwe baan en geniet lekker van Thuur en Bart ! Dat we nog maar jaren dagen! Succes met de afronding van je PhD en de verdere opleiding. skypen/appen/bellen met elkaar!

Graag wil ik ook iedereen in Lab AIG bedanken die mij de laatste jaren hebben Dan mijn Lieve paranimfen: geholpen en hebben bijgedragen aan een geweldige tijd vol serieuze maar ook Lieve Kiki, wat had ik al die tijd zonder je moeten doen?! Je hebt me oneindig feestelijke momenten! Ajeng, Alex, Anca, Andre, Anna G, Anne A, Anne J, Annelie, vaak geholpen met van alles in het lab en me ook veel laten lachen! Ik bewonder Anouk, Arjan, Bas B, Carolien, Charlotte vd H, Daniela, Daya, Dennis, Duby, je geordendheid en vond het tof om met je samen te werken aan de RA BIOTOP. Erik, Evelien, Floor, Freek, George S, Godfrey, Hanne, Hedwig, Hinta, Inge vd Daarnaast heb ik genoten van al onze Dance-workouts waarin we weer vooraan in M, Inneke, Intan, Jacqueline (bedankt voor de pasta-avonden!), James, Janna, de zaal eindigden ;)! Het was dan ook niet zo moeilijk om te beslissen dat ik jou als Jelle, Jelmer, Jessica Q, Jorge, Julia, Katrin, Katharina, Kathrin T, Kathrin E, paranimf naast me wilde hebben! Bedankt voor de leuke gezellig tijd! Succes met je Khutso, Leonie, Lian, Lily, Marije, Mariska, Mariolina, Marlies, Mark S, Michelle, opleiding en het werk in het lab! Mihai, Quirijn, Reinout, Rinke, Rob A, Rob ter Horst, Rosanne, Ruud, Sam, Bad-Ass-Bestie (BAB) Berenice, ondanks dat we elkaar pas in het lab hebben leren Sanne, Simone, Stephan, Tania, Teske, Thalijn, Theo, Valerie, Vera, Vesla, Viola kennen (en je in eerste instantie bang voor me was ;)!) heb ik het idee dat we elkaar en Yvette, bedankt voor de toffe tijd en succes met al jullie projecten! juist al jarenlang kennen. Je oneindige energie en vrolijkheid is aanstekelijk! Verder denk ik dat ik nog nooit iemand heb ontmoet die zo gedreven (en gek) is in het doen Dear Ekta, thank you for being my supervisor during my master internship and wat je gelukkig maakt als jij! Fran en Izzy betekenen alles voor je en ik bewonder je continuing to help me when I joined the lab as a new PhD student.! I really enjoyed voor je discipline en gedrevenheid om alle dingen te doen die je wekelijks doet! Ik all the time we spent together at work and all the nice things we did outside of work. mis onze lunch-breaks en lange gesprekken over onze ideeën voor de toekomst etc! Hope you are doing well and wishing you and Chris all the best! Thanks voor alles BAB!

Sportbuddies Mark en Lisa. Bedankt voor alle keren dat we samen gingen sporten Lieve Charissa, wat kennen wij elkaar alweer lang! Ondanks dat we elkaar de laatste ’s ochtends om 07.00uur of na een lange dag werken hahah!! Mark, super bedankt jaren niet zo vaak zien en spreken weet ik dat je er altijd voor me zult zijn net zoals voor alle hulp tijdens mijn PhD en alle keren dat ik je weer om advies kwam vragen! de afgelopen jaren. Ik denk nog vaak aan al onze weekendjes, danstrainingen en Hopelijk bevalt je nieuwe lab goed! avondjes uit, plus hoe mooi het is dat ik jou en Dennis vanaf het begin van jullie relatie al ken! Geniet van de komende tijd! Dikke knuffel! IL-32 partner in crime, dear Jessica. I still remember the day you joined the lab. A nice but shy girl. I admire your positivity and drive! You grew so much on a personal Lieve Roze Olifanten (#FFCOCB), ik heb nog steeds geen idee hoe we nu uiteindelijk level as well as professional level. I had a lot of fun working with you on our IL-32 allemaal bevriend zijn geraakt, maar dat is inmiddels alweer bijna 10jaar geleden! A project! Good luck finishing your PhD and hope to visit you in Brazil some day! Onze vele Laafjes, pub quiz avondjes, Vierdaagse-feesten, zomervakanties en weekendjes hebben een grote bijdrage geleverd aan mijn mooie tijd in Nijmegen tijdens mijn studie en PhD. Daarnaast is vooral de onvoorwaardelijke steun aan elkaar iets wat onze vriendschap zo speciaal voor me maakt. Ik weet dat jullie er

160 161 Addendum – Acknowledgements, Curriculum Vitae

altijd voor me zullen zijn (ook al zit ik momenteel aan de andere kant van de wereld) Lieve Pap en Mam, zonder jullie had ik dit allemaal nooit kunnen bereiken. Jullie en daar ben ik iedereen van jullie eeuwig dankbaar voor! Ik ben ontzettend trots op hebben me altijd geleerd om dingen in ieder geval te proberen, hard te werken en jullie en wens jullie allemaal het allerbeste!! zeker niet te snel op te geven! Ik ben jullie voor eeuwig dankbaar voor de -tig keren dat jullie me hebben opgehaald en weggebracht naar het station en de keren dat Lieve Pleun, bedankt dat je (op korte termijn) mijn thesis cover wilde ontwerpen! jullie me zelfs helemaal naar Nijmegen brachten zodat het voor mij mogelijk was Dit betekent veel voor me en zo zal ik altijd aan je beschrijving herinnerd worden. om bij familiefeestjes etc aanwezig te zijn in Eindhoven. Bedankt dat jullie je altijd in bochten wringen om mij te helpen en voor al het vertrouwen en steun in het doen Lieve “gestoorde” familie, (dit wordt een lange alinea). Ik ben enorm blij onderdeel wat ik graag wil, ook al betekende dat mijn verhuizing naar Amerika! Door de jaren uit te maken van deze gekke bende. Lieve Sis, vroeger konden we elkaar lekker heen heb ik meerdere plekken “thuis” genoemd, maar mijn echte thuis zal toch altijd irriteren en kon ik doordat ik de jongste was het altijd zo draaien dat jij op je kop in Eindhoven zijn! Ik hou van jullie! kreeg haha. Inmiddels pesten we elkaar alleen voor de lol. Bedankt dat je zoveel om me geeft. Ik ben trots op je hoe je Zoë en Liam opvoedt en ben blij dat ik tante ben van deze twee apen. Lieve Broer, Ik was altijd maar je kleine zusje en totdat ik naar de universiteit ging hadden we nooit veel met elkaar behalve onze Mariokart wedstrijdjes. Ik ben blij dat we in de afgelopen jaren meer samen zijn opgetrokken en moet soms nog steeds lachen om onze droge humor/grappen tijdens onze Thailand vakantie. Ik ben trots op wat je hebt bereikt en hoop dat je een mooie toekomst tegemoet gaat samen met Cecillia. Lieve Clau, mijn stap-buddy! Ik mis onze stapavonden en lange gesprekken over van alles en nog wat wel! Ik bewonder je om je doorzettingsvermogen en kracht! Ik hoop snel weer een keer wat samen te doen en bij jullie te crashen ;). Bedankt voor je steun en vertrouwen! Lieve Sjaan, als oudste van de neven en nichten ben je waarschijnlijk ook de meest rustige. Altijd even zorgzaam en lief. Ik hoop dat je volop kan genieten van de komende tijd en zoals gezegd zal het me niks verbazen als 11-december een speciale dag wordt. Lieve Tante Corry, jouw verjaardag betekent altijd een geweldig en gezellig feest dat tegenwoordig een extra speciaal randje heeft. Ik baal ervan dat ik er dit jaar niet bij kan zijn en daarnaast ook een mogelijke lasagne-avond moet missen! Bedankt voor je steun en hopelijk brengt dit jaar mooie verandering! Lieve Amy en Sean, mijn “kleine” nichtje en neefje, ik kan me nog goed herinneren hoe we naar het dolfinarium gingen samen en samen uren lang wedstrijdjes deden op de PlayStation; en Amy hoe we natuurlijk menig neven&nichten avond hebben afgesloten op stratum. Ik ben heel benieuwd naar alle veranderingen binnenkort en wens je al het beste voor de toekomst! Lieve Tante Susan & Ome Dre, bedankt voor alle steun en memorabele feestjes de afgelopen jaren haha! Fijn dat alles de laatste tijd positief is! Veel geluk en plezier gewenst, en geniet van de nieuwe ontwikkelingen! A

162 163 Addendum – Acknowledgements, Curriculum Vitae

LIST OF PUBLICATIONS CURRICULUM VITAE

1. Lachmandas E, van den Heuvel CN, Damen MS, Cleophas MC, Netea MG, van Crevel R. Diabetes Michelle Damen was born in Eindhoven on the 23rd of March in 1990. She spent her Mellitus and Increased Tuberculosis Susceptibility: The Role of Short-Chain Fatty Acids. J Diabetes Res. youth in Eindhoven and attended the Stedelijk College high school in Eindhoven were 2016;2016:6014631. she graduated in 2008 with the main profile in Physics and Biology. Subsequently,

2. Damen MSMA, Popa CD, Netea MG, Dinarello CA, Joosten LAB. Interleukin-32 in chronic inflammatory she started her Bachelor in Biomedical sciences later in 2008 at the Radboud conditions is associated with a higher risk of cardiovascular diseases. Atherosclerosis. 2017;264:83-91. University in Nijmegen. During her bachelor study, she performed an internship at the Department of Physiology in the Radboud Institute for Molecular Life Sciences 3. amen MS, Agca R, Holewijn S, de Graaf J, Dos Santos JC, van Riel PL, et al. IL-32 promoter SNP rs4786370 predisposes to modified lipoprotein profiles in patients with rheumatoid arthritis. Sci Rep. 2017;7:41629. in Nijmegen investigation the role of SPC (Signal Peptidase Complex) in the cleavage of the CNNM2 Signal Peptide (supervision Dr. J. De Baaij, Prof. J. Hoenderop, Prof. 4. Dos Santos JC, Damen MSMA, Oosting M, de Jong DJ, Heinhuis B, Gomes RS, et al. The NOD2 receptor R. Bindels). is crucial for immune responses towards New World Leishmania species. Sci Rep. 2017;7(1):15219.

5. Dos Santos JC, Heinhuis B, Gomes RS, Damen MS, Real F, Mortara RA, et al. Cytokines and microbicidal After obtaining her bachelor degree in 2011 she followed the master track Pathobiology molecules regulated by IL-32 in THP-1-derived human macrophages infected with New World Leishmania with a minor in Infectious diseases within the Biomedical Sciences Study at the species. PLoS Negl Trop Dis. 2017;11(2):e0005413. Radboud University Medical Center. During her master study she performed two internships. The first internship was related to her minor infectious diseases and was 6. van Laarhoven A, Dian S, Ruesen C, Hayati E, Damen MSMA, Annisa J, et al. Clinical Parameters, Routine Inflammatory Markers, and LTA4H Genotype as Predictors of Mortality Among 608 Patients With performed within the Department of Medical Microbial Pathogenesis at the Karolinska Tuberculous Meningitis in Indonesia. J Infect Dis. 2017;215(7):1029-39. Institute in Stockholm Sweden (Supervision dr. S. Muschiol, Prof. Dr. B. Henriques- Normark). Within this internship she studied the role of streptococcus pneumoniae 7. Damen MSMA, Dos Santos JC, Hermsen R, Adam van der Vliet J, Netea MG, Riksen NP, et al. Interleukin-32 upregulates the expression of ABCA1 and ABCG1 resulting in reduced intracellular lipid concentrations in and mycobacterium marinum in inflammasome activation in macrophages. In her primary human hepatocytes. Atherosclerosis. 2018;271:193-202. second internship at the Department of Internal Medicine under supervision of Dr. E. Lachmandas and Prof. Dr. R. Van Crevel, she investigated the role of aspirin and 8. Dos Santos JC, Damen MSMA, Joosten LAB, Ribeiro-Dias F. Interleukin-32: An endogenous danger signal short chain fatty acids in modulating the immune response against mycobacterium or master regulator of intracellular pathogen -Focus on leishmaniases. Semin Immunol. 2018. Tuberculosis. In 2014 she obtained her Master of Science degree at the Radboud 9. Damen MSMA, van der Heijden W, de Mast Q, Netea MG, Dinarello CA, Popa CD, van de Ven A, Joosten University Nijmegen. LAB. HIV-infected individuals on successful antiretroviral treatment show differences in cardiovascular risk profile when stratified for IL-32 promoter SNP rs4786370. Submitted 2018. Quickly after obtaining her Master of Science degree, she started a PhD project at

10. Damen MSMA, Ballak D, Sapinsley Z, Bai X, Chan E, Seals DR, Dinarello CA, Popa CD, Joosten LAB. the Department of Internal Medicine at the RadboudUMC. Within this project she Human IL-32 transgenic mice develop adipokine profiles resembling those of obesity- induced metabolic and her colleagues studied the role of IL-32 and its various isoforms in various chronic changes. Submitted 2018. inflammatory diseases and the risk for cardiovascular diseases, as described in this

11. Damen MSMA, Schraa K, Tweehuysen L, den Broeder AA, Netea MG, Popa CD, Joosten LAB. Genetic thesis. Her work was presented at national as well as international conferences. variant rs4786370 of IL-32 is associated with the 45 ex vivo cytokine production in PBMCs treated with αTNF therapy in rheumatoid arthritis patients. Submitted 2018. After finishing her PhD she moved to Cincinnati Ohio and started working as a Postdoctoral fellow within the Division of Immunobiology at the Cincinnati Children’s Hospital Medical Center in Cincinnati Ohio (USA) under supervision of A dr. S. Divanovic.

164 165