1 STAT1 O VEREXPRESSION IN S YSTEMIC L UPUS E RYTHEMATOSUS C ORRELATES WITH E NHANCED B IOMARKER E XPRESSION T HERAPY R ESISTANCE, AND OCCURRE NCE OF A NEMIA By PAUL RAMON DOMINGUEZ GUTIERREZ A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 P aul R. Dominguez G utierrez
3 To my Mom Ana Maria Gutierrez Hazas; and my father, J. Ramon Dominguez Pieiro who installed in me the need for a good education who sacrificed so much and have always believed in me.
4 ACKNOWLEDGMENTS It is with my deepest gratitude that I acknowledge the following individuals for their con tributions to my development as a scientist and as a person. I am forever indebted for their assistance, guidance and support. For without these individuals, this work would have been impossible. First I would like to thank a number of faculty members for their invaluable contributions. I sincerely thank my mentor Dr. Edward K.L. Chan for the years of guidance and patience. His knowledge and insight in biomedical research and as successful academic research has taught me that a lifetime in science requires evolving with the science and learning something new each and every time. Most of all, I am grateful for his steadfast focus and patience that has guided me to this finale. I am grateful to my lab member Dr. Angela Ceribelli whose clinical perspective and guidance gave me the physician perspective that I could never obtain from the bottom of a test tube or from a pvalue. Without her guidance, this project would have never attained the level of fruition that it has today. I am especially thankful to my com mittee member Dr. Minoru Satoh. For every question I asked him, he always would have answer and riddle for me to answer. He would not only challenge my work but my very thinking and there in taught me to think in ways I never thought before and achieve that which I never knew was achievable. I am grateful to Dr. Rolf Renne for introducing me to microRNA during my first IDP rotation in his laboratory. The experience in his laboratory has been instrumental to my research endeavors and ultimately led to my int erests in RNAome and nextgen sequencing. I am thankful to Dr. Hideko Kasahara for introducing me to realtime quantitative PCR which became the keystone of my project. I would like to thank Dr. Eric Sobel for his insight
5 into lupus and his valuable experti se with the clinical data. I am grateful for James Colee and the Department of Statistics for providing JMP statistical analysis suite. I would like to think Dr. Robert Burne for the Oral Biology Department generously providing my NIH T90 training grant. I would also like to thank the Lupus Research Institute for funding my research. I would like to thank my every current and past member of the Chan lab for being my second family. Further, I wish to acknowledge Kinda Seaton, Mercedes Rivera, Shehzad Rehman, and many other friends for their continuous support and helping me overcome obstacles. I also thank my grandparents for their confidence and trust in me. Finally, I would like to thank my parents for their love and guidance. Their hard work and sacrifice have given me the opportunity to succeed in life. I would like to thank my brother, Diego, for being my best friend. I would also like to thank my wife, Dr Tania Quesada Vargas, for her tireless support of me and for our beautiful daughter and for our son due in August. I am grateful to my daughter, Victoria, who has done everything in her power to distract me from writing even when she shut off my computer making me lose two hours worth of writing for my dissertation.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 9 LIST OF FIGURES ........................................................................................................ 10 LIST OF ABBREVIATIONS ........................................................................................... 13 ABSTRACT ................................................................................................................... 16 CHAPTER 1 IMMUNITY IN SYSTEMIC LUPUS ERYTHEMATOSUS ........................................ 18 Systemic Lupus Erythematosus .............................................................................. 18 Epidemiology .................................................................................................... 19 Clinical Manifes tations ...................................................................................... 20 Immunological Aberrations ............................................................................... 21 Treatment ......................................................................................................... 23 Type I Interferon in SLE Pathogenesis ............................................................. 24 IFN I Signaling Pathway ................................................................................... 24 IFN I Induction .................................................................................................. 25 IFN I Pathogenesis in SLE ............................................................................... 26 Lupus Biomarkers ............................................................................................ 29 MicroRNA Regulate Endotoxin Tolerance ........................................................ 31 STAT1 Governs Immune Response ................................................................. 34 CCL2 Marks the Prelude to Lupus Flare .......................................................... 36 CXCL10 Indicator of Upcoming Lupus Flare .................................................... 37 2 ELEVATED STAT1 CORRELATES WITH INCREASED CCL2 AND CXCL10 LEVELS IN PERIPHERAL BLOOD OF PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS ................................................................................................ 42 Introduction ............................................................................................................. 42 Patients and Methods ............................................................................................. 46 Healthy Donors and SLE Patients Demographic Data ..................................... 46 Leukocytes and RNA Purification ..................................................................... 46 MicroRNA and Messenger RNA qRT PCR ...................................................... 46 Anti Double Stranded DNA ELISA ................................................................... 47 Complement Levels .......................................................................................... 47 IFN Score and SLE Activity .............................................................................. 47 Cell Culture and Innate Immune Ligand Stimulation ........................................ 47 Data Analysis ................................................................................................... 48
7 Results .................................................................................................................... 48 Expression of Candidate Biomarkers in the SLE Cohort .................................. 48 Biomarker Interrelationship in SLE Patients with Return Visits ......................... 51 Relationship of IFN Score to Other Biomarkers ................................................ 52 STAT1 Levels Correlate with SLE activity ........................................................ 53 STAT1 Influences the Covariation of IFN Score with ADAR, CCL2, and CXCL10 ......................................................................................................... 54 Induction of STAT1, CCL2, and CXCL10 in THP1 Cells with Type I Interferon ....................................................................................................... 55 Discussion .............................................................................................................. 56 Biomarkers Assessment ................................................................................... 57 Biomarker Connections .................................................................................... 58 3 STAT1 EXPRESSION INDICATIVE OF CCL2 AND CXCL10 RESISTANCE TO THERAPEUTIC EFFECTS OF PREDNISONE, MYCOPHENOLATE MOFETIL, AND HYDROXYCHLOROQUINE IN SYSTEMIC LUPUS ERYTHEMATOSUS PATIENTS .............................................................................................................. 74 Introduction ............................................................................................................. 74 Patients and Methods ............................................................................................. 76 Healthy Donors and SLE patients .................................................................... 76 Data Collection ................................................................................................. 77 Data Analysis ................................................................................................... 77 Results .................................................................................................................... 78 Effects of Therapy on Levels of Biomarkers ..................................................... 78 Underlying Effects of Individual Therapies ....................................................... 80 Therapeutic Influences on CCL2 and CXCL10 Association with IFN ............... 82 Individual Therapies Effects on CCL2 and CXCL10 ......................................... 83 Discussion .............................................................................................................. 8 4 4 POSITIVE CORRELATION OF STAT1 AND MIR 146A WITH ANEMIA IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS ................................. 102 Introduction ........................................................................................................... 102 Materials and Methods .......................................................................................... 104 Donor and SLE Patient Demographic Data .................................................... 104 Leukocytes and RNA Purification ................................................................... 105 MicroRNA and Messenger RNA qRT PCR .................................................... 105 IFN Score and SLE Activity ............................................................................ 105 Anemia ........................................................................................................... 106 Data Analysis ................................................................................................. 106 Results .................................................................................................................. 106 Anemia and SLE Biomarkers ......................................................................... 106 Relationship of Anemia and SLEDAI .............................................................. 108 Relationship of Anemia and Lupus Nephritis .................................................. 109 Relationship of Anemia and Race .................................................................. 110 Effects of Therapy on Anemia ........................................................................ 111
8 Discussion ............................................................................................................ 112 5 DISCUSSION AND FUTURE DIRECTIONS ........................................................ 128 High and Low STAT Populations .......................................................................... 128 High STAT1 Negates Therapy .............................................................................. 131 STAT1 and MicroRNA 146a Promote Anemia ...................................................... 132 STAT1 Enhances Pathogenesis ........................................................................... 134 Future Objectives .................................................................................................. 136 LIST OF REFERENCES ............................................................................................. 141 BIOGRAPHICAL SKETCH .......................................................................................... 161
9 LIST OF TABLES Table page 2 1 Demographic data of SLE patients and healthy donors. ..................................... 61
10 LIST OF FIGURES Figure page 1 1 ACR criteria in SLE summarized from references. ............................................ 39 1 2 Postulated T and B cell signaling and autoantigen production in SLE ............... 40 1 3 Signaling and immunological function of STAT1, CCL2, CXCL10, and miR 146a constructed by the Ingenuity Pathway Analysis software. ......................... 41 2 1 Correlation of IFN Score, STAT1, ADAR, CCL2, CXCL10, and miR 146a levels to SLE disease activity. ............................................................................ 62 2 2 Correlation of IFN Score, STAT1, ADAR, CCL2, CXCL10, and miR 146a levels to anti dsDNA autoantibodies. .................................................................. 63 2 3 Anti dsDNA level, IFN score, STAT1, CCL2, and CXCL10 in individuals with different ethnic background. ............................................................................... 64 2 4 Comp arison of SLEDAI, anti dsDNA titer, IFN score, STAT1, CCL2, and CXCL10 in patients with different ethnic background. ........................................ 65 2 5 SLE patients with two visits ranked by increasing or decreasing IFN score and STAT1 ......................................................................................................... 66 2 6 ADAR, CCL2, and CXCL10 levels correlat e with IFN score in both SLE patients and healthy donors ................................................................................ 67 2 7 Bimodal distribution of STAT1 into high and low groups .................................... 68 2 8 High levels of CCL2, CXCL10, and miR 146a compared to low STAT1 in high STAT1 SLE patients ................................................................................... 69 2 9 IFN score, CCL2, and CXCL10 in individuals with different ethnic background and STAT1 levels ............................................................................................... 70 2 10 Effect of high vs low STAT1 expression in ADAR, CCL2, and CXCL10 correlation with IFN score. .................................................................................. 71 2 11 Ethnicity effect in high and low STAT1 groups. ................................................. 72 2 12 THP ........................ 73 3 1 The effects of therapies on the levels of various clinical p arameters and biomarkers in the SLE cohort. ............................................................................ 87 3 2 High and low populations of STAT1 in both SLE and healthy donors ................. 88
11 3 3 Comparison of high and low STAT1 subsets of all treated to untreated SLE patient visits.. ...................................................................................................... 89 3 4 The effects of prednisone therapy on levels of the various biomarkers in the SLE cohort. ......................................................................................................... 90 3 5 The effects of hydroxychloroquine therapy on levels of the various biomarkers in the SLE cohort ............................................................................. 91 3 6 The effects of mycophenolate mofetil therapy on levels of the various biomarkers in the SLE cohort. ............................................................................ 92 3 7 Comparison of high and low STAT1 subsets of PDN treated patient visits to untreated patient visits. ....................................................................................... 93 3 8 Comparison of high and low STAT1 subsets of HCQ treated patient visits to untreated patient visits.. ...................................................................................... 94 3 9 Comparison of high and low STAT1 subsets of MMF treated patient visits to untreated patient visits. ....................................................................................... 95 3 10 The effects of high and low STAT1 and dosage subsets on expression levels of the various biomarkers in the SLE cohort. ..................................................... 96 3 11 Comparison of high and low STAT1 subsets segregating into low versus high dosing on levels of the various biomarkers in the SLE cohort ............................ 97 3 12 Association between CCL2, IFN score, and therapy. ......................................... 98 3 13 Association between CXCL10, IFN score, and therapy. ..................................... 99 3 14 The effect of combined therapy on the expression of CCL2 and CXCL10 in high versus low STAT1 subsets ....................................................................... 100 3 15 Separate analyses of high and low STAT1 effects on CCL2 expression in various combined therapies. ............................................................................. 100 3 16 Separate analyses of high and low STAT1 effects on CXCL10 expression in various combined therapies. ............................................................................. 101 4 1 Ge nder and Therapy by visits ........................................................................... 115 4 2 Demographic data of SLE patients and healthy donors. ................................... 116 4 3 Relationship between anemia and biomarker expression in SLE ..................... 117 4 4 The relationships of miR 146a, STAT1, and SLEDAI in non anemic versus anemic patients were essentially unchanged considering different methods of selecting only a single visit for patients with multiple visits. .............................. 118
12 4 5 Changes of anemia status were correlated with miR 146a and STAT1 mRNA levels ................................................................................................................ 119 4 6 Differential expression of IFN score, STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a in active versus inactive SLE patients with anemia ..................... 120 4 7 The relationships of miR 146a and STAT1 versus SLEDAI and anemia. miR 146a and STAT1 were analyzed with data from one visit for each patient and of those with multiple visits only anemiapositive visits were included .............. 121 4 8 The interactions of anemia and lupus nephritis on SLE biomarkers. ................ 122 4 9 African Americans SLE patients were more likely anemic and SLE patients with anemia were more likely under prednisone therapy .................................. 123 4 10 The relationship of race and anemia on STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a .............................................................................................. 124 4 11 IFN score, STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a expression in African American vs European American SLE patients with and without anemia. ............................................................................................................. 125 4 12 STAT1, CCL2, CXCL 10, miR 146a, and pri miR 146a in anemic vs nonanemic SLE patients on different treatment ...................................................... 126 4 13 The relationships of miR 146a and STAT1 versus anemia and therapy were analyzed including only data from one visit for each patient and of those with multiple visits only anemia positive visits were included ................................... 127 5 1 Relationship of STAT1, CCL2, CXCL10, miR 146a, and IFN 1 with SLE. ........ 139 5 2 STAT1 role in SLE.. .......................................................................................... 140
13 LIST OF ABBREVIATION S AA African Americans ACR A merican College of Rheumatology ADAR adenosine deaminase acting on RNA Ago Argonaute ANA antinuclear antibody APC Antigen Presenting Cells AsA Asian Americans BAFF B cell Activation Factor CCL2 C C motif chemokine ligand 2 CLRs C type Lectin Receptors CXCL10 C X C motif chemokine 10 DAI DNA dependent activator of IFN regulatory factor dsRNA double stranded RNA EA European Americans ELISA enzyme linked immunosorbent assay HD Healthy Donors HCQ H ydroxychloroquine HNPs Human N eutrophil P eptides IFN Interferon IFN I T ype I I nterferon IFNAR Interferon R eceptor IgG I mmunoglobulin G IPS1 IFN 1 IRAK Interleukin 1 R eceptor A ssociated K inase 1
14 ISGs IFN I S timulated G enes IL I nterleukin IrA Interracial Americans LA Latin Americans Lck L ymphoma K inase LDGs Low Density G ranulocytes LN Lupus N ephritis LPS L ipopolysaccharide LY6E L ymphocyte Antigen 6 Complex Locus E MDA 5 M elanomaD ifferentiation A ssociated gene 5 MHC Major Histocompatibility Complex miRNA, microRNA; MMF Mycophenolate M ofetil MX1 M yxovirus R esistance 1 MyD88 M yeloi d D ifferentiation factor 88 NET N eutrophil E xtracellular T rap Nuclear F actor kappalight chain enhancer of activated B cells NOD Nucleotidebinding Oligomerization Domain NLRs NODLike Receptors NSAIDs Non steroidal anti inflammatory drugs OA O steoarthri tis OAS1 2,5 oligoadenylate synthetase OD O ptical D ensity; PAMPs PathogenAssociated Molecular Patterns PBMCs P eripheral B lood M ononuclear C ells
15 pDC plasma Dendritic Cells PDN P rednisone Pri miRNA Primary microRNA RA R heumatoid A rthritis RIG I R etinoic acidinducible gene I RNP R ibonucleoprotein siRNA small interference RNA SLE S ystemic L upus E rythematosus SLEDAI SLE disease A ctivity I ndex STAT Signal T ransducers and A ctivators of T ranscription STING S timulator of I nterferon G enes TANK TRAF fa mily member associated NF kappaB activator TBK1 TANKbinding kinase 1 TCR T Cell Receptors Th0 T Helper Cells Th2 T Helper 2 Cells TLR T oll L ike R eceptor TNF Tumor Necrosis Factor TRAF TNF R eceptor A ssociated F actor TRIF TIR domaincontaining adapter UTR U ntranslated R egion.
16 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STAT1 OVEREXPRESSI ON IN SYSTEMIC LUPUS ERYTHEMATOSUS CORRELATES WITH ENHA NCED BIOMARKER EXPRESSION, THERAPY RESISTANCE, AND OCCUR RENCE OF ANEMIA By Paul R Dominguez Gutierrez May 2013 Chair: Edward K.L. Chan Major: Medical Science Molecular Cell Biology Systemic lupus erythematosus (SLE) is a chronic autoimmune disorder affecting multiple organ systems. Although genetic and environmental factors have been implicated, the etiology of SLE remains unclear. Recent evidence suggests that typeI interferon is integral to the pathogenesis of SLE. We investigated several of the most recently published biomarkers in SLE: CCL2, CXCL10, STAT1, and miR 146a. Peripheral blood leukocytes were collected from 65 healthy donors (HD) and 103 patients with over 180 visits. We validated that CCL2, CXCL10, and STAT1 were overexpressed in SLE. We showed for the first time that two populations of STAT1 level in both SLE patients and HD. The high STAT1 population displayed elevated CCL2 and CXCL10 levels compared to low STAT1 population of both SLE and HD. Furthermore the high STAT1 patients displayed higher correlation between CCL2 and IFN score as well as CXCL10 and IFN score compared to low STAT1 patients. We also showed for the first time that CCL2 and CXCL10 were decreased in patients rece iving therapy. We identified that CCL2 and CXCL10 decrease was
17 occurring in low STAT1 patients; however, high STAT1 patients appeared to be resistant to therapy and displayed no significant decrease in CCL2 and CXCL10. This was further validated by compari ng the association of CCL2 and CXCL10 with IFN score. High STAT1 treated and untreated patients had stronger correlations of CCL2 and CXCL10 with IFN score than low STAT1 patients. Finally, we showed for the first time that elevated STAT1 and miR 146a were associated with anemia. Anemic patients displayed significantly higher STAT1, miR 146a, CCL2, CXCL10, and IFN score. However regardless of disease activity, lupus nephritis, or race, anemic patients displayed significantly elevated levels of STAT1 and miR 146a compare to nonanemic patients indicating they may play a role in anemia in SLE patients. Taken together, these findings illustrate the importance of STAT1 in the pathogenesis of SLE and provide novel insight on the efficacy of therapy.
18 CHAPTER 1 IMMUNITY IN SYSTEMIC LUPUS ERYTHEMATOSUS This chapter introduces the current understanding of systemic lupus erythematosus (SLE) including epidemiology, clinical manifestations, and aberrations of the immune system. Furthermore, this chapter will discuss the basic functions of type I interferon (I IFN) and interferon inducible genes with strong focus on interferoninducible transcription facto rs and chemokines. This introduction will also provide the rationale for studying the relationship of I IFN and chemokine pathway in SLE. Systemic Lupus Erythematosus SLE is a chronic autoimmune disease that can affect multiple parts of the body either concurrently or independently pending the nature of the active state [ 1 ] SLE is characterized into two states: one being the a ctive state which is referred to as flare ups and the second being the inactive state which is referred to as remission. In healthy individuals, the immune system protects the host from foreign pathogens. In order to defend the host from foreign pathogens, it must first be able to discriminate invader from self. This is a much more complex task than it seems, especially if the invader is an intracellular pathogen such as a virus. The immune system has evolved with multiple screening defenses. The innate i mmune response is the first layer of response; it is a nonspecific inflammatory response that depends upon pathogen associated molecular patterns (PAMPs). PAMPs such as Toll like receptors (TLRs), C type lectin receptors (CLRs), nucleotidebinding oligomer ization domain (NOD) like receptors (NLRs), and retinoic acid inducible gene I (RIG I) receptors are ligand specific sensors that are germ line encoded. When a PAMP encounters a ligand molecule, it will trigger a nonspecific inflammation response composed of cytokines and
19 chemokines that recruit antigen presenting cells such as monocytes and dendritic cells, which will direct the adaptive immune response to produce antigen specific T and B cells against the foreign antigen from the invading pathogen. T her e are multiple safe guards to prevent the adaptive immune system from generating autoreactive B and T cells to self antigens; however, there is a breach in the innate and adaptive immune system in SLE. For reasons not completely understood, the loss of imm une tolerance in SLE accompanied by autoantibodies, I IFN, antiviral chemokines, cytokines, and involvement of multiple cell types leading to a vicious cycle of chronic inflammation occludes the actual mechanism that leads to this breach. This is further complicated by the varying clinical manifestations and by periods of remission and flares further complicated by absence of knowing the trigger. Both genetics and environment have been implicated in the etiology of SLE, but the actual causes still remain unclear. The absence of this knowledge has severe consequences for patients. Patients treatments are focused on the symptoms of SLE rather than the cure, and until the etiology of SLE is fully understood, it is unlikely a cure within reach. Epidemiology The number of Americans with SLE vary from 300,000 to 4 million individuals with the most recent statistic from the CDC estimating an occurrence of 241 per 100,000 Americans [ 2 ] Estimates worldwide are about 5 million individuals with SLE but likely on the low end due to difficulties in diagnosis [ 3 ] SLE is primarily a disease of women occurring 610:1 in women to men [ 4 ] SLE can occur in infancy and even in the elderly age with a peak age between 15 to 40 years of age commonly referred to as childbearing age [ 4 ] It remains unclear the selective occurrence of SLE in women and its peak occurrence during childbearing age.
20 E qually not clear is the increased occurrence and severity of disease in African Americans as well as in Latin Americans and AsianAmerican population than in European Americans. The CDC reports that there is increased occurrence of SLE in Native Americans and has instituted a study to validate this observation. Sex and ethnicity indicates a genetic component; however, familiar occurrences are sporadic with only 1012% occurring within the first degree relatives [ 5 ] A stronger association is seen with twins. A concordance of 26% 40% was found in monozygotic twins and only 5% in dizygotic twins [ 4 6 ] Clinical Manifestations Due to the broad spectrum of immunological and clinical manifestations expressed by SLE patients, the American College of Rheumatology (ACR) developed 11 criteria for the classification of SLE as summarized graphically in Figure 1 [ 7 8 ] The immunological criteria include complement levels, antinuclear antibodies, and other autoantibodies such as anti double stranded (anti ds) DNA, anti Sm, and anti phospholipid. Clinical criteria consist of photosensitivity, discoid rash, malar rash, oral ulcers, serositis, arthritis, renal complications, neurological complications, and hematologic aberrations. Patients must meet a minimum of 4 of t he 11 criteria to be diagnosed with SLE; however, patients classified with less than 4 criteria can be categorized as subclinical SLE displaying unique disease manifestation [ 9 ] Diagnosis using ACR criteria has up to a 90% certainty; however in clinical settings, the certainty decreased to 49 60% [ 9 ] The complicated diagnosis of SLE is in part due to the widely varying symptoms, fluctuation of disease activity, and variations in onset of immunologic manifestations such as the onset of autoantibodies preceding clinical manifestations [ 10]
21 Clinical features vary widely depending upon the manifestation and intensity of the flare. The most common clinical manifestations include anemia, photosensitivity, leucopenia, and arthritis (Figure 1). Malar rash may potentially be one of the oldest known clinical features of SLE. Lupus obtains its name from the ancient Romans who attributed the resemblance of the butterfly distribution of malar rash to a wolf (lupine) bite to the face. Anemia occurs in about 50% of SLE patients, with chroni c disease anemia being the most common form due to potentially impaired erythropoietin as well as aberrant cytokines and T lymphocytes affecting bone marrow erythropoiesis [ 11] Lupus nephritis (LN), a major cause of morbidity, occurs in about 55% of Asians, 51% of Africans, and 43% of Hispanics; however, LN occurs in only 14% of Caucasians [ 12] Other less common clinical manifestations are of cardiovascular, pulmonary, and neurological origin. Immunological Aberrations Even though the reasons for the loss of immunological tolerance are unknown, genetics and environmental factors have implicated several im munological pathways. The most prominent of these are lymphocyte hyperactivation, improper clearance of apoptotic cells, and aberrant cytokine production. It remains unclear whether the loss of tolerance is initiated by autoantibodies, self reactive lymphocytes, or I IFN. In normal healthy individuals, foreign antigens presented are processed by antigenpresenting cells (APCs) such a s macrophages and dendritic cells. APCs present the processed antigen via class II major histocompatibility complex are postulated to derive from defective clearance of apoptotic cells are being displayed by APCs to helper T cells ( Th0 cells, Figure 2). The helper T cells present the processed
22 antigens to B lymphocytes. The direct contact and cytokine secretions of the helper T cells stimulate antibody production in the B ce lls. If the T cell presents an autoantigen, specific B cells may produce autoantibodies (Figure 2). Upon activation, T cells in SLE patients exhibit aberrant signaling produced by increased lipid rafts aggregation due to decreased lymphoma kinase (Lck), abnormalities of the MAPK signaling pathway, and enhanced adhesion from alternative CD44 isoforms [ 13, 14] B cells from SLE patients exhibit abnormal signaling, enhanced proliferation, and elevated antibody production as well as expansion of memory and plasma cells [ 13, 15] The hyperactivation of lymphocytes may in part be due to aberrant production of cytokines. In lupus patients, T helper 2 (Th2) cytokines, such as IL (interleukin) 6, IL 10, and B cell activating factor (BAFF), are essential for antibody production [ 16, 17 ] IL 6 and IL10 promote B cell differentiation into plasma cells and the production of IgG respectively (Figure 2) [ 1618] Simila rly, BAFF also known as B lymphocyte stimulator overexpression promotes the proliferation and survival of B cells as well as increased production of IgG [ 16, 17] Th1 cy 12 are linked to tissue 12, and IL18 which promote autoantibody production and chemokines are associated with instigating kidney damage in LN [ 19] Another immunological aberration seen in SLE is impaired clearance via phagocytosis of cellular debris. Cellular death whether by apoptosis or injury is a continuous process requiring phagocytic cells to remove cellular debris to maintain tissue homeostasis; however, this process does not elicit an immune response. This process appears to be impaired in SLE and the cellular remnants accumulate exposing
23 antigens such as histones and GW182 that are not normally exposed to the immune system. Accumulated DNA and RNA can f unction as TLR ligands activating I IFN production [ 20, 21 ] This may be exasperated due to additional deficiencies of endogenous nucleases such as DNase I and II in the serum of SLE patients [ 2224 ] Due to the impaired clearance, injuries such as exposure to sunlight can t rigger a flare up. This may account for photosensitivity being a common clinical manifestation of SLE (Figure 1). Treatment Therapies currently focus upon treating the symptoms of patients rather than curing SLE in part due to the lack of understanding of the causes of SLE. Nonsteroidal anti inflammatory drugs (NSAIDs) and antimalarial such as hydroxychloroquine (HCQ) are effective treatments in mild cases of SLE as in fatigue, arthritis, and malar rash. HCQ is an antimalarial drug that increases the pH of the endosomal vesicles, interrupting antigen processing and the function of TLR3, 7, 8, and 9 as well as inhibiting production of interleukin1 and 6 by macrophages [ 2527 ] Glucocorticoids are often administrated during flares at high doses and reduced aft erwards to a maintenance dose. Glucocorticoids such as prednisone (PDN) inhibit monocyte and neutrophil inflammatory functions as well as B and T cell responses [ 28 ] PDN cyclophosphamide and mycophenolate mofetil (MMF) are administered to SLE patients. Mycophenolate mofetil affects T and B cell growth by inhibiting inosine monophosphate dehydrogenase, which is a critical enzyme for de novo synthesis of guanosine nucleotides [ 25] Many of these therapies possess a number of side effects from the nonspecific suppression of the immune system as well as other pathways. The next
24 generation of pharmaceuticals such as BAFF inhibitor Belimumab are coming to market but are not as commonly used as HCQ, PDN, and MMF. In part this has been facilitated by the new guidelines provided by the US Food and Drug Administration (FDA) for therapy development for SLE [ 29, 30 ] Even so, Belimumab is the first therapy to be approved by the FDA in the last 50 years [ 31] Ty pe I Interferon in SLE Pathogenesis IFN I is an anti viral cytokine with a multitude of immunological functions that has been implicated in the pathogenesis of SLE. IFN 1 was first observed to be elevated in SLE patients over 30 years ago; however, it was not until recently with the advent of microarrays that IFN 1 was discovered to play a key role in SLE [ 3235 ] Microarray studies revealed that nearly twothirds of SLE patients displayed elevated IFN I as well as a plethor a of unregulated IFN I stimulated genes (ISGs) [ 3235] Several murine models have demonstrated the pathogenic effects of IFN I and its capacity to induce lupus like disease in mice. This section summarize IFN I biology and affirms their involvement in SLE. IFN I Signaling Pathway IFN I genes cluster on chromosome 9p22 is comprised of thirteen subtypes of [ 36 39] However, only 1 species bind to the ubiquitously expressed IFNAR1 and IFNAR2 heterodimer receptor c omplex. IFN I triggers dimerization of IFNAR1 and IFNAR2 which phosphorylates Janus kinase (JAK) 1 and tyrosine kinase (Tyk) 2 that are docked on the cytoplasmic domains of IFNAR2 and IFNAR1 respectively [ 3840] This recruits and phosphorylates signal transducers and activators of transcription (STAT) complex which is translocated into the nucleus.
25 These STAT complexes include STAT1, STAT3, STAT4, STAT5, and STAT6 homodimers as well as STAT1/STAT2, STAT1/STAT3, STAT1/STAT4, STAT1/STAT5, STAT2/STAT3, and STAT5/STAT6 heterodimers [ 38, 39 ] However, the STAT1/STAT2 heterodimer is the classical signaling molecule that is believed t o be responsible for IFN I production. The STAT1/STAT2 dimer associates with IFN regulatory factor (IRF) 9 to form a heterotrimeric complex known as IFN stimulated gene factor (ISGF) 3 which can transcriptionally activate hundreds of ISGs by binding to the IFN stimulated response element (ISRE) [ 38] P38 mitogen activated protein kinase (MAPK) is phosporilated in an IFN I dependent manner but independent of Jak STAT signaling; it is also required for the formation of ISGF3 and tra nscriptional activation of ISREs [ 38, 40] IFN I I nduction IFN I production is an essential response component of the innate immune system. TLRs are the early response mechanism of the immune system monitoring for PAMPs of foreign pathogens. TLR 3, 4, 7, 8, and 9 induce production of IFN I when triggered by the appropriate ligand. Endosome localized TLR 3 recognizes viral doublestranded RNA; however, TLR 4 is localized on the ce ll surface and recognizes bacterial lipopolysaccharide (LPS) and certain viral proteins. Both TLRs induce IFN myeloid differentiation factor 88 (MyD88) independent TIR domaincontaining adapter [ 41] TLR 7/ 8 recognizes viral single stranded RNA, and TLR 9 recognizes unmethylated CpG DNA found in bacteria. All three TLR 7, 8, and via MyD88 [ 42 ] 3, 5, and [ 42]
26 IFNAR to induce IFN 1 pro and IRF 7 establishes a positivefeedback loop for signal amplification. IFN 1 has been shown to be produced in a TLR independent manner. Cytoplasmic helicase proteins retinoic acid inducible prote in I (RIG I) and melanomadifferentiation associated gene 5 (MDA 5) are PAMP sensors that detect viral RNA [ 43] RIGI and MDA 5 signal through the adaptor IPS 1 (IFN 1; also known as MAVS, VISA, or Cardif), initiating signaling cascades that lead to IRF3, IRF7, B and AP 1 activation and IFN I expression [ 43] Finally, DNA dependent activator of IFN regulatory factor (DAI, a dsDNA helicase) and stimulator of interferon genes (STING, an endoplasmic reticulum resident transmembrane protein) are cytoplasmic sensors for bacterial and viral DNA [ 4447 ] DAI and STING signal via TANK binding kinase1 (TBK 1) to activate IRF 3 to produce IFN 1 [ 48 ] Additionally, NOD2 (sensor of muramyl dipeptide, peptidoglycan), a sensor fo r ssRNA, also signals via TANK binding kinase1 (TBK 1) to activate IRF 3 to produce IFN 1 [ 48 ] IFN I P athogenesis in SLE The effects of IFN I on DC maturation, T cell survival, and antibody production are well known; however, the role of IFN I in disease pathogenesis is not well understood. The onset of SLE like disease, Hashimoto thyroiditis, hemolytic anemia, and psoriasis [ 4951] Approximately 22% of patients antinuclear autoantibodies [ 52] SLE patient sera were documented over 30 years ago [ 53] yet it was not considered significant until microarrays analysis of SLE peripheral blood mononuclear cells (PBMCs) revealed overexpression of IFN inducible genes (IIG) [ 32, 54 56]
27 IIGs overexpressed in SLE include anti r eceptors genes as well as overexpression of interferon inducible chemokines such as CCL2 and CXCL10. From half to twothirds of SLE patients display aberrant elevations for IIG. This interferon signature is associated with disease activity and severity as well as LN, dsDNA autoantibodies, and RNA associated nuclear antigens such as Sm/nRNP, SSA/Ro, and SSB/La [ 57, 58] IFN I is produced in response to an invading pathogen where it and IIGs peak and return back to normal levels in healthy individuals; however in SLE, it is unclear what is the trigger for the upregulation and overexpression of IFN I and IIGs, but the mechanism that regulates this interferon signature and prevents from entering a pathogenic cycle is not well understood and impaired in SLE patients. Potentially the perpetual auto reactive response to self antigens promotes the pathogenic cycle of IFN I aberrant production. Bacterial DNA and dsDNA antibody immune complexes stimulate IFN I in normal PBMCs [ 59] ; however, mammalian DNA does not [ 60] TLR 9 is activated by unmethylated CpG DNA such as in bacteria, but mammalian DNA is methylated and is not recognized [ 61] A similar process occurs with RNA recognition by TLR 7; however, TLR 7 is capable of recognizing some mammalian RNAs to induce IFN I production [ 62 6 4 ] These self activating TLR 7 ligands include hum an U1 RNA (associated with snRNP particles) and Y RNAs (associated with Ro60 protein) which induce IFN I production through an endosomal MyD88dependent pathway [ 63, 65] A critical question of why DNA and RNA are antigenic in SLE patients but not in healthy indivi duals may be due to
28 the impaired clearance mechanism in SLE patients which leadslead to an accumulation of apoptotic/necrotic cellular debris that contain DNA and RNA [ 66] Recently, a newly identified mechanism revealed how intracellular antigens are exposed. NETosis is the process by which neutrophils expel their genomic or mitochondrial chromatin to create a NET (neutrophil extracellular trap) to entrap invading pathogens [ 6769] NETosis can be triggered by live and even dead neutrophils, but neutrophils that undergo NETosis do not undergo necrosis indicating that NETosis induced cell death differs from apopt osis and necroptosis [ 67 ] Furthermore, neutrophils that undergo NETosis continue to be viable and capable of chemotaxis and phagocytosis even without a nucleus [ 70 72] NETs are not created equally. When neutrophils are treated with LPS, they will predominantly create mitochondri al NETs; however, mitochondrial NETs lack antimicrobial histones [ 67] The mitochondrial DN A does not produce efficient NETs but is capable of activating TLR 9 to induce an immune response to invading pathogens [ 73 ] Anti neutrophil antibodies are also capable of inducing NETosis [ 67] In SLE patients, neutrophil expansion is a common occurrence; furthermore, a subpopulation of highly inflammatory neutrophils (low density granulocytes, LDGs) capable of producing IFN I were discovered in SLE patients [ 74 ] In SLE patients, neutrophil derived antimicrobial peptide LL37 and human neutrophil peptides (HNPs) were shown to be in complex with DNA and anti DNA autoantibodies that were capable of activating TLR 9 in plasma dendritic cells (pDCs) and induce NETosis [ 75] Furthe rmore, IFN I would bring LL37 and HNPs to the surface of neutrophils which would be recognized by anti LL37 and anti HNPs autoantibodies
29 inducing NETosis [ 75 ] Anti LL37 and anti HNPs antibodies could also stimulate pDCs to produce IFN I creating a positivefeedback loop of NETosis and IFN I production [ 75] In pediatric SLE, neutrophils were shown to undergo NETosis in the presences of anti 7 dependent manner and brought both LL37 and alarmin high mobility group box 1 (HMGB1) to the surface [ 76] The release of NETs stimulated pDC to produce IFN 1; however, inhibition of TLR [ 76] again indicating a positive feedback loop between NETosis and IFN I. LDGs derived from SLE patients appear to spontaneously undergo NETosis 50% of the time when placed in culture and LDG derived NETs stimulate IFN I and IL 17 production in pDC [ 77 ] LDGs displayed upregulation of NE T proteins including LL37 [ 77] Patients with NET like materials detected in renal and skin lesions were more likely to have elevated anti dsDNA autoantibody titers indicating that NETs act as antigenic materials [ 77] In SLE, NET clearance appears to be impaired. NETs show enhanced stability in SLE sera [ 78 ] This prolonged stability of NET derived, antigenic material mi ght exacerbate SLE and explain the numerous autoantibodies to neutrophils [ 74, 79] During the process of NETosis, H3 Histones are citrullinated and may provide a mechanism for the generation of autoantibodies to citrullinated proteins. This may explain the occurrence of anti citrullinat ed protein autoantibodies in 1015% of SLE patients [ 80 81 ]. Lupus Biomarkers SLE is possibly the most clinically diverse disease [ 82] with some manifestations that can resemble AIDS, systemic tuberculosis, and even cancer. This complexity has been further undermined by a lack of definite set of biomarkers for SLE which has
30 impaired the development of therapies [ 82] A biomarker is defined as measurement of a marker whose alterations correlate with disease. This marker may be but not limited to genetic, molecular, biochemical, biological, or imaging event that is associated with the occurrence of a pathological event. The selected biomarkers discussed in this section may not be specific for SLE but are associated with it and will provide the reasoning for the subsequent studies in the following chapters. ADAR editing RNA From literature search, 2,5 oligoadenylate synthetase (OAS1), myxovirus resistance 1 (MX1), and lymphocyte antigen 6 complex locus E (LY6E) are the most co mmonly used ISGs for observing changes in 1IFN. Like many of the ISGs observed to be elevated in SLE, they are antiviral response genes. One such ISG that is not only important antiviral gene but is also immune modulator is adenosine deaminase acting on R NA (ADAR). ADAR is an enzyme that catalyzes adenosine (A) to inosine (I) in double stranded RNA (dsRNA) substrate. [ 83 84] The conversion of A to I by ADAR can impact premRNA splicing, RNA structure, RNA degradation, RNA replication of viruses, and mRNA translation. For in depth review of ADAR mechanism and function, Nishikura 2010 and George et al. 2011 are recommended [ 85, 86] There are three ADAR genes: ADAR1, ADAR2, and ADAR3, but only ADAR1 and ADAR2 are known to have A to I editing function. Furthermore only the long of isoform of ADAR1 (p150) is interferoninducible while the shor t form (p110) is constituently expressed [ 87 ] Homozygous deletion of ADAR1 in mice has been demonstrated to be embryonic lethal as well as impair hematopoiesis and cause liver disintegration but deletion of ADAR2 demonstrated shortened lifespan and neurologic abnormalities [ 88 92] Subsequently,
31 the loss of hematopoiesis in ADAR1/ mice is a result of the lack of suppr ession of interferon signaling [ 93] The increase of interferon in ADAR1/ mice may partially be due to accumulation of dsRNA which may be triggering cytosolic sensors such as MDA 5 and RIG 1 [ 94] Furthermore, ADAR1 has been observed to suppress IRF 3 and PKR activation [ 95 97] The ability of ADAR1 to respond and regulate interferon production makes it an intriguing ISG to look at in SLE. Up to now, ADAR1 has only been observed in T cells of SLE in a limited number of studies. Laxminarayana et al. 2002 observed in 8 SLE patients and 8 healthy donors that ADAR1 was approximately upregulated by 3 [ 98] In the subsequent study, Laxminarayana et al. 2007 observed increase editing of ADAR2 by ADAR1 in T cells of SLE patients [ 99] Additionally, Orlowski et al. observed a 2.6 fold increase of Phospodiesterase 8A1 (PDE8A1) gene due to increased ADAR1 in 10 SLE pat ients relative to 10 healthy donors [ 100] ADAR also functions to regulate microRNA maturation and function. MicroRNA R egulate E ndotoxin T olerance MicroRNAs (miRNAs) are the newest identified biomarkers in SLE. MiRNAs are small nonencoding RNAs that regulate translation of mRNA by binding to the 3 untranslated region (UTR) of mRNA causing translational repression and/or degradati on of the transcript. Mature miRNA, which are 20 to 23nt long, can originate via Drosha dependent pathway [ 101 ] or via a Drosha independent pathway [ 102104] Primary microRNA (pri miRNA) is transcribed by RNA polymerase II. The pri miRNA may originate from the transcript of intron coding protein or from a noncoding transcript. The pri miRNA is processed by DiGeorge syndrome critical region gene 8 (DCGR8) and Drosha to generate the precursor miRNA (pre miRNA), an approximately 70nt long
32 hairpin. The Drosha independent generated miRNA are referred to as mirtrons, but the majority of miRNA are generated via the Drosha dependent pathway. Exportin 5 transports the pre miRNA from the nucleus to the cytoplasm. In the cytoplasm, Dicer in complex with TAR RNA binding protein (TRBP) cleaves the stem loop of the pre miRNA. After Dicer/TRBP processing, the mature miRNA is incorporated into Argonaute (Ago) where one s trand is selected to be the miRNA and the other is degraded. In humans, there are four Argonautes, but only Argonaute 2 (Ago2) has endonucleolytic activity [ 105107] A single miRNA has the potential to regulate thousands of mRNA targets to varying degrees. By bioinformatics, it is estimated that 30% to 60% of the proteinome is regulated by miRNAs [ 108, 109] One reason for the ability of miRNA to regulate numerous messages stems from their imperfect binding to the 3 UTR of target mRNAs. Unlike small interference RNA ( siRNA ) miRNA are not required to bind perfectly to suppress tr anslation; the targeting of miRNA is dependent upon the seed sequence match [ 109113] This property of being able to re gulate multiple mRNA has lended it important roles from development, hematopoiesis, and immune system regulation. The role of miR 146a in inflammation may potentially make it one of the most significant miRNAs in regulating immune response and tolerance. m iR 146a was first shown to be involved in toll like receptor (TLR) 4 where Taganov et al. determined that it is transcriptionally regulated by NF kB and that it can downregulate both TNF receptor associated factor ( TRAF ) 6 and i nterleuki n 1 receptor associa ted kinase ( IRAK) 1 [ 114] This yielded insight into TLR signaling. TLR4 stimulation by LPS leads to NF kB activation through TRAF6 and IRAK1; th e activation of NF kB transcriptionally activates immune response factors,
33 which include miR 146a (miR 155 and miR 132 to a lesser extent) which regulates the same signaling components that lead to its activation [ 114 115] Furthermore, miR 146a has been demonstrated as a critical regulator of endotoxinendotoxin induced tolerance and cross tolerance [ 116 118 ] MiR 146a would function to attenuate the immune response preventing detrimental immune over activation and have potential role to regulate inflammation in nor mal immune response and autoimmune disorders. To date, miR 146a has been observed to have a strong association with autoimmune disease. In psoriasis, miR 146a is overexpressed in lymphocytes while the skin specific miR 203 was overexpressed which regulates suppressor of cytokine signaling 3 (SOCS 3) [ 119, 120 ] This discovery was followed by the identification of miR 146a being overexpressed in rheumatoid arthritis (RA) [ 121 123] Stanczyk et al. observed miR 146a and miR 155 to be overexpressed in RA synovial fibroblasts compared to osteoarthritis (OA) synovial fibroblasts; furthermore they observed similar results for miR 155 for RA s ynovial tissue compared to OA synovial tissue [ 123 ] When comparing RA synovial fibroblasts to RA peripheral blood monocytes, Stanczyk et al. observed that miR 155 was elevated in the synovial fibroblasts [ 123] The sustained expression of miR 155 in synovial fibroblasts may regulate matalloproteinase 3 which potentially allows miR 155 to downregulate tissue damage [ 123] Similar results were reported by Nakasa et al. where miR 146a was elevated in RA synovial tissue compared to normal and OA synovial fluid [ 121] The source of the high miR 146a was identified predominately to be CD68+ macrophages and in some CD+ T cells and CD79a+ B cells [ 121]
34 From our group, Paul ey et al. examined miRNA expression from PBMCs from sixteen RA patients, nine healthy donors, and four disease controls [ 122 ] The PBMCs of RA patients displayed between 1.8 to 2.6fold increase in miR 146a, miR 155, miR 132, and miR 16 compared to healthy controls. When miRNA expression was compared to disease activity, age, race, and medication, the increase in miR 146a and miR 16 expression correlated with disease activity [ 122 ] Interestingly, TRAF6 and IRAK1 both are known targets of miR 146a but show no significant difference in mRNA or protein levels between RA patients and healthy controls. This loss of regulation led us to speculate that a 3UTR modification, such as the role of single nucleotide polymorphisms (SNPs) [ 124127 ] or shortening to escape miRNA regulation [ 128 ] may be occurring in RA patients, but further work is needed. Tang et al. reported that miR 146a was underexpressed in PBMCs of Chi nese SLE patients compared to healthy controls; in addition, Tang et al. observed that there is an inverse correlation between the miR 146a expression and interferon score [ 129 ] Additionally, Tang et al. demonstrated that miR 146a downregulates STAT1 and IRF5 and that the reduction of miR 146a may lead to enhanced signaling leading to elevated levels of interferon [ 129] The reduced levels of miR 146a observed in Chi nese SLE patients could potentially explain elevation of interferon by loss of regulation of STAT1 expression. STAT1 G overns I mmune R esponse STAT1 is involved in type I, II, and III interferon signaling as well as in cytokine and chemokine signaling [ 130 ] In human monocytes, IFN II priming or overexpression of STAT1 sensitizes cells and leads to increased production of chemokines [ 131 ] Similarly in mice, IFN I priming sensitized mice to subsequent LPS challenges leading
35 / ) mice, this sensitization was nullified indicating that STAT1 was required for immune sensitization [ 132] Subsequently, STAT1 / mice are unable to respond or produce I IFN and display a protection phenotype from pristine induced lupus [ 133] ; furthermore, it is required for autoantibody production as well as TLR 7/9 activation in B cells of lupus pristine mouse model [ 134] STAT1 itself is an interferon inducible gene which both mRNA and protein levels are overexpressed in SLE [ 130] This pivotal role in immune response and association with SLE makes it a potentially interesting biom arker. STAT1 may also play a role in therapy response. In cancer, STAT1 elevated expression has been associated with therapy resistance. STAT1 overexpression protects cancers from DNA damaging agents including radiation therapies and chemotherapies in dif ferent cancer types [ 135 ] Radioresistant nu61 derived from radiosensitve SCC61 tumors displayed 49 overexpressed genes; of the 49, 31 were ISGs including STAT1 [ 136] Furthermore, when STAT1 was overexpressed in SCC61 cells, they displayed radioresistance [ 137 ] Similarly, human fibroblasts repeatedly exposed to IFN I displayed radio resistance [ 138 ] In 10 cancer cell lines, STAT1 expression correlated with resistance to doxorubicin and topoisomeraseII inhibitors [ 139] In addition, 14 ovarian cancer lines were observed for resistance to platinum compounds where STAT1 was associated with resistance to cisplatin and AMD473 [ 140] These associations between therapy resistance and STAT1 in cancer may also have an impact of SLE therapy. Interferon regulated chemokines (IRCs) have been shown to be biomarkers of SLE activity were C C motif chemokine ligand 2 (CCL2) and
36 C X C motif chemokine 10 (CXCL10) [ 141 142 ] and require STAT1 for transcription [ 131] CCL2 M arks the P relude to L upus F lare CCL2 formerly referred as monocyte chemotactic protein 1 (MCP 1) is a potent recruiter of monocytes, T cells, basophils and dendritic cells to site of infection or tissue damage, but not neutrophils or eosinophils unless the N terminus is cleaved [ 143147] Monocyte, macrophages, and dendritic cells primarily secrete CCL2. CCL2 binds cell surface receptors CCR2 and CCR4 [ 148, 149] [ 150, 151 ] Pathway s such as AP 1, ERK1/2, Ras, and p38 MAPK activate CCL2, but more importantly, CCL2 is also activated by the JAK/STAT signaling pathway [ 152, 153] Even though the role of CCL2 is beneficial in clearing pathogens, it can also exacerbate tissue damage. CCL2 was demonstrated to cause crescent formation and interstitial fibrosis in crescentic nephritis (glomerulonephritis mouse model) and by blocking CCL2 with anti CCL2 antibodies reduced crescent formation, renal impairment, and scarring as well as T cell and macrophage infiltration [ 154 ] Furthermore, in male type 2 diabetic db/db mice, CCR2 antagonist, which inhibits CCL2 from binding to CCR2, was able to prevent glomerulosclerosis and renal failure [ 155 ] The implication of CCL2 in autoimmune d iseases such as psoriasis, rheumatoid arthritis, and multiple sclerosis as well as other disorders has incited interest in its role in SLE [ 156] In a study of 81 SLE patients and 42 healthy controls, CCL2 was identified as one of 12 upregulated proteins (CCL3, CCL7, CCL 8, CCL19, CXCL9, CXCL10, CXCL11, CXCL13, IL 2SRA, IL 6, and IL 15) from a survey of the serologic proteome by antibody microarray [ 142 ] As seen above, CXCL10 was also elevated in SLE.
37 CXCL10 I ndicator of Up coming Lupus F lare CXCL10, formerly known as Interferon gammainduced protein 10 (IP 10), is a chemokine of the C X C motif family closely related to the C C motif family which CCL2 is from. Similarly, CXCL10 is potent attractor of monocytes, macrophages T cells, NK cells, and dendritic cells to site of tissue damage or infection [ 157 159] CXCL10 is the ligand for CCR3 receptor [ 160 162] Even though CXCL10 is an interferon response cytokine which implies activation via Jak/STAT pathway, CXCL10 can also be transcriptionally activated by p38, JNK, ERK, Akt, and B [ 163168 ] Though CXCL10 is a potent immune responder for bacterial and viral infections and even a potential biomarker for transplantations, it has not been demonstrated as in the case of CCL2 to be involved in the pathogenesis of autoimmune disease [ 166, 169 ] In the same study that identified CCL2 to be upregulated in serum proteome, CXCL10 was one of the 12 upregulated proteins [ 142] This same group followed up 3 years later with a longitude study where 267 patients with 1167 visits were followed for ~1 year [ 141] In this study, Bauer et al. observed that CXCL10 was elev ated in active SLE patients compared to inactive patients. Furthermore, Bauer et al. demonstrated by combining CXCL10, CCL2, and CCL19 into a chemokine score that this score was elevated in active SLE compared to inactive SLE and that it could be used to predict upcoming flares [ 141 ] STAT1, CCL2, CXCL10, and miR 146a are only a few of the many biomarkers identified to potentially be involved in SLE H owever by using the Ingenuity Pathway Analysis software to data mine the literature for immune function, these biomarkers demonstrate an important r ole in regulating how the immune system may respond and develop (Figure 3). Furthermore, STAT1, CCL2, CXCL10, and miR 146a are involved
38 in numerous immunological pathways with overlapping regulation (Figure 3). These multiple roles in the immune system make these biomarkers likely candidates for further study to understand their role in SLE.
39 Figure 1 1 ACR criteria in SLE summarized from references. The most common criteria used by ACR to diagnose SLE patients are graphically depicted [ 7 8 ]
40 Figure 1 2 Postulated T and B cell signaling and autoantigen production in SLE. Antigen presenting cells (APCs) boun d to self antigens from apoptotic cells interact with T cells which in turns help B cells to produce autoantibodies. Modified maps were generated by Kegg pathways from the Kanehisa laboratories: http://www.polygenicpathways.co.uk/KEGGhivinteractions.htm
41 Figure 1 3 Signaling and immunological function of STAT1, CCL2, CXCL10, and miR 146a constructed by the Ingenuity Pathway Analysis software. The literature data mining and plotting of interactions and functions by the Ingenuity Pathway Analysis software shows the important role that these biomarkers play in the immune system and potentially in the pathogenesis of SLE. Blue lines represent direct relationships while blue dotted lines represent indirect relationships. Grey lines show connection fro m bubbles to specific cellular functions represented by Fx
42 CHAPTER 2 ELEVATED STAT1 CORRELATES WITH INCREASED CCL2 AND CXCL10 LEVE LS IN PERIPHERAL BLOOD OF PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS Introduction Systemic lupus erythematosus is a chronic systemic autoimmune disease characterized by periods of increased disease activity, referred to as flare ups, and periods of remission. Several genetic and environmental factors have been implicated in SLE etiopathogenesis, but in recent years increased type I interferon (IFN despite being known for over 30 years that it is elevated in SLE patients [ 32 35] Bec one common way to evaluate IFN I expression is to examine the levels of common interferon inducible genes, such as 2,5 oligoadenylate synthetase (OAS1), myxovirus resistance 1 ( MX1), and lymphocyte antigen 6 complex locus E (LY6E); the mRNA levels of these IFN I inducible genes are then used to calculate the interferon score [ 32, 5456] Another interferon inducible gene that plays an important antiviral and immunomodulatory function is the adenosine deaminase acting on RNA (AD AR). ADAR is an enzyme that catalyzes the conversion from adenosine (A) to inosine (I) in double stranded RNA (dsRNA) substrate [ 83 84] with an impact on RNA at different levels, such as mRNA splicing and degradation [ 85, 86] Furthermore, ADAR1 has been observed to suppress interferon regulatory factor (IRF) 3 and protein kinase RNA activated therefore blocking IFN induction [ 9597] The ability of ADAR1 to respond and regulate IFN I production makes it an intriguing IFN I inducible gene to examine in SLE. Up to now, ADAR1 expression has only been observed in T cells of SLE patients, as
43 shown in a limited number of studies [ 98100] In fact, Laxminarayana et al. showed that ADAR1 is upregulated by approximately 3 fold in SLE patients [ 98 ] The same group later observed the increased editing of ADAR2 by ADAR1 in T cells of SLE patients [ 99] Additionally, due to increased ADAR1 in SLE patients, Orlowski et al. observed an incre ase of p hosphodiesterase 8A1, which participates in the termination of cyclic nucleotide signaling by hydrolyzing cAMP and cGMP and is activated by interferon and enhances T cell adhesion [ 100 ] Other IFN I inducible genes include STAT (signal transducers and activators of transcription) 1 and 2. STAT1 is involved in type I, II, and III IFN signaling and has been observed to be elevated in SLE [ 170] In response to type I I FN, STAT1 causes interferon receptor (IFNAR) 1 and 2 dimerization, activation and phosphorylation of IFNAR by Tyk2 and Jak1, and thus docking and phosphorylation of STAT1 and STAT2 [ 171] The heterodimer STAT1S TAT2 is then translocated into the nucleus where it can bind specific promoters playing a key role in IFN signaling and production [ 130] Besides STAT1 and ADAR, interferon regulated chemokines have become another important topic of research in recent years [ 142 ] Two of these chemokines have been shown to be SLE biomarkers, and they are called C C motif chemokine ligand 2 (CCL2) and C X C motif chemokine 10 (CXCL10) [ 141 ] CCL2, formerly referred to as monocyte chemotactic protein1 (MCP 1), is a potent recruiter of monocytes, T cells, basophils, and dendritic cells to site of infection or tissue damage, but it has no effect on neutrophils or eosinophils unless the N terminus of CCL2 is cleaved [ 170 172] Some cell types such as monocytes, macrophages, and dendritic cells can primarily secrete CCL2 that signals via the cell surface receptors CCR2 and
44 CCR4 and is upregulated by [ 150, 151] The role of CCL2 is beneficial in clearing pathogens, but it has also been involved in some pathological processes. In a glomerulonephritis mouse model, CCL2 plays a role in cresc ent formation and interstitial fibrosis supported by the observation that anti CCL2 antibodies can reduce crescent formation, renal impairment, and scarring, as well as T cell and macrophage infiltration [ 154] CCL2 has been observed in the recruitment of T cells and monocytes/macrophages in lupus nephritis and bl ockade of CCL2 ameliorates lupus nephritis in MRL(Fas)lpr mice [ 172, 173] In a serologic proteome study by antibody microarray in SLE, CCL2 was identified as one of the 12 upregulated proteins; furthermore CCL2 was one of three chemokines that would precede lupus flare indicat ing that they are good predictors of increased SLE activity [ 142 ] CXCL10, also known as Interferon gammainduced protein 10 (IP 10), is a chemokine of the C X C motif family. Similarly to CCL2, CXCL10 is a potent attractor of monocytes, macrophages, T cells, NK cells, and dendritic cells to sites of tissue damage and inf ection [ 157 158 ] CXCL10 is an interferon response cytokine that binds its CCL3 receptor and acts via Jak/STAT pathway activation [ 161 163 ] Even though CXCL10 is a potent immune responder for bacterial and viral infections and a critical biomarker for organ transplant rejection, its role in the pathogenesis of autoimmune diseases is not clear [ 166 169 ] Furthermore, the combination of CXCL10 and CCL2 protein levels could be useful as prediction factor for upcoming flares [ 141 ] The reason behind upregulation and control of IFN in SLE is not known, but some studies have recently focused on the possible role played by selected microRNAs (miRNAs ). MiRNAs are small non encoding 2023 nucleotide long RNAs, that regulate
45 their target mRNA by binding to the 3 untranslated region (UTR), causing translational repression and/or degradation of targets. miR 146a is one of the most significant miRNAs in r egulating innate immune response and tolerance [ 174] and it was first shown to be involved in toll like receptor (TLR) regulation through the NF B pathway [ 114] miR 146a would function to attenuate the immune response and regulate inflammation in normal immune response and autoimmune disorders, and it is also a critical regulator of endotoxin induced tolerance and cross tolerance [ 116 175 176] To date, miR 146a has been found in association with autoimmune diseases such as Sjgrens syndrome [ 177] psoriasis [ 119, 120 ] and rheumatoid arthritis [ 121 123 178] Tang et al. reported that miR 146a was underexpressed in peripheral blood mononuclear cells (PBMCs) of Chinese SLE patients [ 129 ] MiR 146a was significantly lower in active SLE patients with proteinuria compared to inactive SLE patients [ 129] Additionally, SLE patients displayed an inverse correlation between miR 146a expression and IFN score [ 129] Tang et al. also demonstrated that reduction of miR 146a may enhance the signaling due to elevated levels of STAT1 and IRF5 which leads to increased production of IFN [ 129] The reduced levels of miR 146a observed in Chinese SLE patients could potentially explain elevation of IFN by loss of regulation of STAT1 expression. Our present study evaluates the interaction among STAT1, ADAR, CCL2, CXCL10, and miR 146a in SLE patients and healthy controls, demonstrating that all except for miR 146a correlate to IFN Score in both SLE patients and healthy donors.
46 Patients a nd Methods Healthy D onor s and SLE P atient s D emographic D ata Whole blood was collected from a total of 103 SLE patients and 65 healthy controls enrolled in the University of Florida Center for Autoimmune Diseases registry from 20082011. Healthy donors were selected based on no history of autoimmune disease, while all SLE patients satisfied the ACR criteria [ 7 ] Healthy donors only visited the clinic once; theref ore, they represent a single visit. There were a to tal of 18 0 SLE visits with sequential samples collected in 60 SLE patients (Table 1). SLE patients and healthy controls were segregated by ethnic profile (Table 1). All human blood samples were obtained from enrolled individuals with the approval of instit utional review board at the University of Florida. Leukocytes and RNA Purification Peripheral blood leukocytes were collected from whole blood using Ambion LeukoLOCK kit (Ambion, Austin, TX). LeukoLOCK filters were washed twice with 3 ml of PBS and stabili zed with 3 ml of RNAlater solution. Stabilized filters were stored for a minimum of 24 h at 80C before collecting total RNA. Total RNA, including small RNAs, was collected using the Alternative Protocol (version 0602, Ambion) for the extraction of RNA from cells captured on LeukoLOCK filters using TRI reagent. MicroRNA and Messenger RNA qRT PCR OAS1, MX1, LY6E, CCL2, CXCL10, and miR 146a levels were analyzed by qRT PCR. MiRNA qRT PCR was performed using the TaqMan MicroRNA Reverse Transcription Kit, TaqMan Fast Advance PCR Master Mix, and TaqMan MicroRNA primers (Applied Biosystems, Foster City, CA). mRNA qRT PCR was performed using the TaqMan HighCapacity cDNA Reverse Transcription Kit, TaqMan Fast Advance
47 PCR Master Mix, and TaqMan mRNA assay primers (Applied Biosystems). All reactions were analyzed using StepOne Real Time PCR System (Applied Biosystems). Anti Double Stranded DNA ELISA After the collection of leukocytes with the LeukoLOCK filters, the leukocyte free blood was transferred to 10 ml Vacutainer SST plus blood collection tubes (BD, Franklin Lakes, NJ). Blood was centrifuged at 1000 g for 20 minutes. The plasma was transferred to a 15 ml conical tube and stored at 20oC. Anti dsDNA ELISA was performed as previously described [ 179 ] Complement L evels C3 and C4 complement levels were obtained from clinical data. C3 levels less than 90 mg/dl and C4 levels less than 15 mg/dl were considered as low complement levels in the analysis. IFN Score and SLE Activity The expression of three known typeI IFN signa ture genes, MX1, OAS1, and LY6E, was z transformed into IFN score as previously shown [ 32, 180] SLEDAI was used to classify the patients into active (SLEDAI > 4) or inactive (SLEDAI of 4 or less) at the time of the visit (Table 1) [ 181, 182 ] Cell C ulture and I nnate Immune L igand S timulation Human THP 1 cells we re obtained from the American Type Culture Collection (ATCC, Manassas, VA). THP 1 cells were maintained in RPMI containing 10% (v/v) FBS (Mediatech, Manassas, VA) and 100 U/ml penicillinstreptomycin (Mediatech ). For analysis of THP 1 monocyte response to ligand in vitro log phase cells were seeded at 5x105 cells/ml in a 24 well plate. Cells were stimulated with the following agonists: 1000 ng/ml of LPS from S. enterica serotype Minnesota Re595 (LPS Se, TLR4 li gand,
48 4 ligands were reconstituted in endotoxin free water and used at concentrations as reported before [ 175 ] free PBS with 1 mg/ml BSA to make 5 g/ml stocks stored in 80oC. Data A nalysis The copy number of miR 146a was normalized to total loaded RNA whereas mRNA levels were normalized to 18S RNA. Copy number of miR 146a was determined using a standard curve with synthetic miR 146a (Integrated DNA Technologies Inc., Coralvill e, IA) [ 183 ] Relative expression of mRNA compared to controls was T method [ 184] Analyses were performed using SAS version 9.2 and JMP Genomics version 5 (SAS, Cary, NC). The Wilcoxon/Kruskal Wallis test was used to evaluate significance between groups, whereas Wilcoxon Signed Rank test for match ed pairs was used to evaluate SLE patients with two visits. P values < 0.05 were considered significant. The coefficient of determination (r2) was used to determine linear correlations. Significance between slopes was evaluated by analysis of covariance (ANCOVA). Results Expression of C andidate B iomarkers in the SLE C ohort To determine whether previously reported biomarkers were elevated in our SLE patient cohort, we measured the biomarker expression levels in healthy donors (HD), active SLE, and inactive SLE patient visits (Figure 2 1). The SLE cohort was segregated by disease activity index (SLEDAI) into a ctive SLE (45 visits, SLEDAI > 4) and i nactive I was estimated by quantifying the
49 expression of interferon inducible genes. The IFN score, STAT1, ADAR, CCL2, and CXCL10 levels were significantly elevated in both active and inactive SLE patient visits compared to healthy donors (Figure 2 1A E) establishing and confirming that these biomarkers were aberrantly overexpressed in our SLE patients. To explore if these biomarkers were capable of distinguishing disease activity status, active and inactive patient visits were compared to one another. No significant difference was observed between active and inactive SLE patient visits for IFN score (Figure 2 1A, 62.76.1 units versus 57.84.9 units), ADAR (Figure 2 1C, 5.270.31 fold versus 5.270.23 fold), and CXCL10 (Figure 2 1E, 158.126.6 fold versus 120.010.5 fold), but STAT1 (Figure 2 1B, 44.810.7 fold versus 34.46.6 fold, p=0.033) and CCL2 (Figure 2 1D, 18.23.1 fold versus 9.96 1.42 fold, p=0.0061) were significantly elevated in active SLE compared to among the three groups (Figure 2 1F). Similarly miR 146a did not display any significant difference among active SLE, inactive SLE, and HD (Figure 2 1G). To validate this, we determined the levels of the primary transcript of miR 146a (pri mir 146a) which also did not demonstrate any significant difference among active SLE, inactive SLE, and HD. With the exception of miR 146a, these results are consistent with reports on SLE patients with elevated IFN score compared to HD [ 32, 180] as well as upregulated levels of IFN signature genes (STAT1 and ADAR) [ 98 100] and chemokines ( CCL2 and CXCL10) [ 142] The clinical and expression data was correlated with anti double strand DNA (dsDNA) autoantibody level, which is an indicator for patients disease activity in certain
50 patients [ 185 188 ] Decreases in C3 and C4 levels correlated with SLE activity and renal damage as well as increased levels of anti dsDNA antibodies [ 189] Anti dsDNA autoantibody level s have also been used for subclassification of SLE patients [ 190, 191 ] SLE patient visits and HD were segregated into anti dsDNA(+) and anti dsDNA( ). Patient visits that were anti dsDNA(+) displayed higher SLEDAI and decreased C3 and C4 levels (Figure 2 2A C). The results for the remaining biomarkers (Figure 2 2D K) closely resembled those from active versus inactive SLEDAI results (Figure 2 1). The influence of race in anti dsDNA, IFN score, STAT1, CCL2, and CXCL10 were also examined. African Americans (AA) and European Americans (EA) contributed to 83.3% of the visits, followed by Latin Americans (LA) and Asian Americ ans (AsA) for 15%, and Interracial Americans (IrA) for less than 2% of patient visits (Table 1). Due to the small sample size, IrA were excluded in all subsequent analyses. In general, results show that higher levels of anti dsDNA, IFN score, STAT1, CCL2, and CXCL10 were observed in all race groups analyzed (Figure 2 3 ). The lack of statistically significant differences between SLE and HD in certain groups, such as LA, might be due to the smaller sample sizes. By comparing patients of different race to each other (Figure 2 4 ), the levels of the parameters examined were all higher in AA than EA, LA, and AsA. In particular, AA had significantly higher SLEDAI (p=0.024), anti dsDNA level (p=0.044), IFN score (p=0.0005), STAT1 (p=0.0011), CCL2 (p=0.0004), and CXCL10 (p=0.0004) than EA. CXCL10 than LA (Figure 2 4B F). Also in this case, the lack of additional statistically
51 significant results for LA and AsA might be due to the small sample sizes. However, AA clearly displayed increase in biomarker levels more than any other race. Biomarker I nterrelationship in SLE P atients with R eturn V isits To expand upon the interrelationship of these biomarkers, data from SLE patients with two consecutive visits were segregated for analyses by increasing or decreasing IFN score of at least 50% from the first to the second visit. Patients with increasing IFN score from one visit to the next (N =13; p= 0.0001, Figure 2 5 ) displayed significant increase in STAT1 (p= 0.0017), CCL2 (p= 0.0086), CXCL10 (p=0.038), and miR 146a (p= 0.0034). Similarly, for SLE patients with increasing STAT1 by at least 50% between first and second visit (N=25; p<0.0001, Figure 2 5 B ), significant increases were observed for IFN score (p=0.027), CCL2 (p<0.0001), CXCL10 (p= 0.0003), and miR 146a (p=0.0078). The strong correlation of STAT1, CCL2, and CXCL10 were expected; however, correlation of IFN sco re to increasing STAT1 was weaker than expected. This may be indicating that high STAT1 levels do not necessarily translate into high levels of IFN I. The highly significant correlation between miR 146a levels and IFN score in the return visits was unexpec ted since the level of miR 146a in SLE was not significantly different from HD (Figure 2 1H, 2H) and also previously it was reported to be decreased in SLE and inversely correlated to IFN score in a Chinese SLE cohort [ 129 ] SLE patients who had decreasing IFN score by at least 50% between first and se cond visit (N=32; p<0.0001, Figure 2 5C ) displayed a significant decrease in STAT1 (p=0.0002) and CXCL10 (p= 0.0002), but not in CCL2 and miR 146a. Similarly, decreasing STAT1 SLE patients (N=13; p= 0.0001, Figure 2 5 D ) had significant decrease in IFN scor e (p= 0.0001) and CXCL10 (p= 0.0004), while no significant changes in CCL2 and miR 146a were observed. By ranking patients according to
52 decreasing IFN score or STAT1, the reversal of the results from ranking by increasing IFN score or STAT1 should have been observed ideally. Interestingly, the exception was observed only for CCL2 and miR 146a (Figure 2 3C,D). Relationship of IFN S core to O ther B iomarkers To better understand whether the association of IFN score with the other biomarkers in paired patient visits could be expanded, levels of ADAR, CCL2, and CXCL10 from the entire cohort of SLE patient visits and HD were correlated to IFN score (Figure 2 6 ). ADAR, CCL2, and CXCL10 displayed coefficients in both SLE and HD (Figure 2 6 ). The consistent significant correlations of these genes to IFN from the low levels observed in HD (Figure 2 6 right panels) to aberrantly high pathogenic levels of IFN in SLE patient visits (Figure 2 6 left panels) was indicative of a normal intrins ic response of ADAR, CCL2, and CXCL10 to IFN production. Contrary to an earlier report showing that the level of miR 146a was negatively correlated with IFN score [ 129] miR 146a as well as pri miR 146a did not display any significant correlation with IFN score in either HD or SLE patients (data not shown). Surprisingly, in the same type of analysis, STAT1 did not display a significant correlation to IFN score either (data not shown). Further analysis of STAT1 expression revealed two populations after applying a log10 transformation (Log[STAT1]) in both HD and SLE patients (Figure 2 7 A ). Using an arbitrary cut off of 1.50 Log[STAT1] to segregate STAT1 results, values below 1.50 were referred as the low STAT1 group and above 1.50 were the high STAT1 group (Figure 2 7B C). In the low STAT1 group, SLE patient visits displayed significantly higher expression of STA T1 compared to HD (2.44 fold, p<0.0001, Figure 2 7 B ), but in the high STAT1 group, no significant difference was observed (Figure 2 7C ). Furthermore, the low STAT1 group displayed
53 significant positive association of STAT1 with IFN score for both HD (Figure 2 7D ) and SLE patients (Figure 2 7 E ). In contrast, the high STAT1 group showed no correlation of STAT1 to IFN score (data not shown). STAT1 L evels C orrelate with SLE activity The effects of high and low STAT1 on IFN score and ADAR appeared to be related t o the active versus inactive status of SLE ( SLEDAI Figure 2 1A,C) and anti dsDNA(+) vs ( ) patients (Figure 2 2A,C) where IFN score and ADAR were significantly higher than HD, but not significantly different between high and low STAT1 SLE patient visits ( Figure 2 8A ,B). CCL2 was significantly different between active and inactive SLE, and between HD and active and inactive SLE as well (Figure 2 1D), what resembles the results of anti dsDNA(+ versus ) (Figure 2 2D) and high versus low STAT1 comparisons (F igure 2 8D ). Similar observations are valid for CCL2, with the addition that there is a difference in CCL2 expression between high and low STAT1 SLE (Figure 2 8C ). Since both SLEDAI active and anti dsDNA (+) are indicators of increased disease activity th ese results indicate that high STAT1 patients are also in a more active disease state than low STAT1 patient visits. To determine whether ethnicity could be a confounding factor for the effects of high and low STAT1, IFN score, CCL2, and CXCL10 levels wer e segregated based on ethnicity and high and low STAT1 (Figure 2 9A C). Overall, high STAT1 patient visits did not show significant difference among AA, EA, and LA. However, low STAT1 AA patients showed significantly higher IFN score, CCL2 and CXCL10 compared to other groups These results indicated that high and low STAT1 groups were identified essentially in all ethnicity and differences in IFN score, CCL2, CXCL2 levels were observed among low STAT1 groups but not among the high STAT1 groups
54 STAT1 I nfluences the Covariation of IFN S core with ADAR, CCL2, and CXCL10 To determine whether high versus low STAT1 levels affected the correlation between IFN score and the other biomarkers, we analyzed these parameters in patients with high versus low STAT1 ex pression (Figure 2 10 ). Even though ADAR expression was reported to be STAT1 independent [ 192, 193] low STAT1 SLE patient visits (red, r2=0.29, p<0.0001), high STAT1 SLE (blue, r2=0.35, p=0.0002) patients, and low STAT1 HD (black, r2=0.24, p<0.0001) displayed significant association between ADAR and IFN score (Figure 2 10 A ). Similarly, CCL2 was significantly associated with IFN score in low STAT1 SLE patient visits (r2=0.07, p<0.0010), high STAT1 S LE patient visits (r2=0.76, p<0.0001), and low STAT1 HD (r2=0.08, p=0.0002); also CXCL10 displayed significant association with IFN score for low STAT1 SLE patients (r2=0.09, p=0.0003), high STAT1 SLE patient visits (r2=0.30, p=0.0008), and low STAT1 HD (r2=0.08, p=0.027, Figure 2 10B ,C). The slope of the linear regression represents the rate of change of ADAR, CCL2, and CXCL10 per unit of change of IFN score. This led to the intriguing possibility that high STAT1 patient visits have a higher slope than low STAT1 patient visits. An analysis of covariance was used to test if the slopes were significantly different (Figure 2 10). ADAR/IFN scores were not significantly different between high and low STAT1 patients (Figure 2 10A p value not shown), but CCL2/IFN score and CXCL10/IFN score slopes were significantly higher in the high STAT1 patients compared to the low STAT1 (LS) patients (Figure 2 10B ,C). This suggests that high STAT1 levels may enhance CCL2 and CXCL10 expression potentially induced by IFN. Next, we studied whether ethnic background could influence the association of IFN score with CCL2 and CXCL10 and altered the effects of high and low STAT1 ( Figure 2 -
55 11). Influence on ethnic background appeared to be minimal on CCL2 in high STAT1 patient visits. CCL2 in high STAT1 AA, EA, and LA displayed very good linear correlation (r2 2 11A ,C,E). Low STAT1 EA and LA also showed good linear correlation (r2 2 11C E); however, low STAT1 AA did not display a linear correlation between CCL2 and IFN score (Figure 2 11A ). CXCL10 had a significant correlation (r2 score for high STAT1 AA and EA (Figure 2 11B ,D); however, CXCL10 had significant corr elation (r2 2 11A ,D,F). AsA could not ascertain significant correlations for CCL2/IFN score and CXCL10, probably due to the small sample size (data not shown). Induction of STAT1, CCL2, and CX CL10 in THP1 C ells with Type I Interferon TLRs have been implicated to play a role in SLE pathogenesis. To model the response of STAT1, CCL2, and CXCL10 as well as IFN I, TLR 4 was stimulated in human monocytic THP 1 for 24 h with 1000 ng/ml of LPS. IFN sc ore increased around 4 h and peak around 8 h (Figure 2 12A treated THP 1 cells, IFN score displayed a similar trend as in LPS treatment (Figure 2 12F eased till 12 h (Figure 2 8K) while 0.1 ng/ml treated cells displayed little change (Figure 2 12F ). This results demonstrated THP 1 responsiveness to IFN I as well as that they were capable of IFN I production. Interestingly whereas LPS displayed a gradual long term increase of CCL2 and of CCL2 and CXCL10. After LPS stimulation, STAT1 did not increase till 4 h and reached its peak expression at 8 h (Figure 2 12B ); however in THP 1 cells stimulated
56 2 12G ,L). CCL2 increases at 2 h in LPS treated THP1 cells and continues increase during the 24 h period (Figure 2 12C ); however not till after STAT1 reached its maxim um expression (Figure 2 12B ), CCL2 began to rapidly increase (Figure 2 12C ). CCL2 increased at 2 h 1 cells, but it peaked at 4 h and began to decrease rapidly (Figure 2 12H ). For 1.0 ng/ml IF THP 1 cells shifted the peak by 2 h so that CCL2 peaked at 2 h and began to rapidly decrease (Figure 2 12M ). CXCL10 displayed a trend similar to CCL2 for 1.0 ng/ml 2 12I ,N). In 1.0 ng/ml I treatment of THP 1 cells, CXCL10 continued till 8 h (Figure 2 12N ). These results indicated that CCL2 and CXCL10 rapidly responded rapidly while TLR 4 stimulation appeared to induce slow gradual increase, but then rapidly increased after STAT1 reached its maximum expression. miR 146a appeared to differ in its response from the other biomarkers. LPS upregulated miR 146 by 3 fold and it rapidly reached a peak of 11 fold increase at 12 h (Figure 2 12E ). miR treated cells showed a modest peak of 3 to 4 fold at 8 h potentially indicating that IFN I did not induce significant production of miR 146a (Figure 2 12J ,O). Discussion In this study, expression of previously identified SLE biomarkers was examined and correlated with demographic and clinical parameters, focusing on the analysis of a possible correlation among them.
57 Biomarkers Assessment Our results show that ADAR, STAT1, CCL2, and CXCL10 levels are significantly elevated in the SLE cohort as expected. This is in part validated by previously published results showing increased levels of these biomarkers and their correlation to IFN I production in SLE patients [ 32, 33 55, 141 142 194] Furthermore, our study shows that increase of CCL2, and 700 fold increase of CXCL10, confirming that these genes respond to IFN I stimulation. Tang et al reported miR 146a underexpression in SLE PBMCs [ 129] while we did not observe a decrease or a difference between active and inactive SLE patients for miR 146a expression in peripheral blood leukocytes of SLE patients in our cohort. Luo et al. [ 195 ] hypothesize that a functional variant in the miR 146a promoter may be responsible for decreased levels of miR 146a in SLE, so the pri miR 146a levels should be decreased in our populati on; however, no significant differences in pri miR 146a expression were observed in our population. Furthermore, Tang et al. reported an inverse correlation between miR 146a and IFN score in their SLE cohort, while we did not observe a significant correlat ion in our cohort. A significant increase in miR 146a was observed only in SLE patients with increasing IFN score between initial and second visit [ 114 ] Other possible explanations for the discrepancy between the two data sets could be the difference in cell populations and racial composition in our cohort versus the one examined by Tang et al and Luo et al [ 129, 195 ] As for the THP 1 monocyte cell model, IFN I weakly stimulated miR 146a expression compared to LPS. All these results suggest that the role of miR 146a in regulating IFN I in our cohort of SLE patients may be limited
58 Biomarker Connections Previous reports have demonstrated the involvement of ADAR, CCL2, and CXCL10 in SLE [ 98 100 141] In published literature and in our cohort, a positive association with IFN score was observed for ADAR, CCL2, and CXCL10 [ 86, 141, 142, 196 ] This was observed not only in SLE but in HD as well, potentially indicating that these genes are responding normally to IFN even when at levels aberrantly elevated. Unlike what reported in previous studies, STAT1 did not correlate well with the IFN score in the SLE patient population [ 131 197] Instead, low STAT1 SLE patients and low STAT 1 HD expression was associated with IFN score. Patients paired by two visits that were ranked by increasing IFN score demonstrate a strong covariation with STAT1, but the co variation of IFN score to increasing STAT1 appeared to be weaker. In paired SLE patient visits, decreasing IFN score or STAT1 are accompanied by a decrease of the other biomarkers suggesting that STAT1 and IFN I may be driving factors. When SLE patient visits are grouped into high and low STAT1, high STAT1 SLE patient visits showed si gnificantly higher levels of CCL2 and CXCL10. After grouping by high and low STAT1, the high STAT1 patient visits showed a significantly increased slope for CCL2/IFN and CXCL10/IFN scores compared to low STAT1 SLE patient visits. This enhanced response by CCL2 and CXCL10 to IFN I in high STAT1 patients may be due in part to the role of STAT1 in activation of CCL2 and CXCL10 [ 198 200] Hence, STAT1 levels appear to be enhancing chemokine response to IFN I. IFN score, STAT1, CCL2, CXCL10 and miR 146a were upregulated in a time
59 decreased levels of CCL2 and CXCL10 shortly after reaching their peak expression, while LPS treatment displayed a steady increase of CCL2 and CXCL10 with a less rapid induction compared to their exp peak expression in LPS treated THP 1 cells, CCL2 and CXCL10 expression rapidly and CXCL10 both started decreasing after reaching their peak expression, while STAT1 and CXCL10 respond differently to TLR4 stimulation compared to IFN signaling. It also indicates that CCL2 and CXCL10 response to IFN I is very rapid but very short compared to TLR signaling. Since in the high STAT1 patients, IFN score correlates with greater increase of CCL2 and CXCL10 than in low STAT1 patients. The results of TLR 4 stimulation suggest that at least in the hi gh STAT1 patient population CCL2 and CXCL10 are being driven by TLR signaling rather than IFN I directly since IFN I stimulation caused a rapid increase followed by an equally rapid decrease of CCL2 and CXCL10 independent of STAT1 expression. It is unclear why STAT1 was elevated to such high levels in some of the SLE patients and HD. One possibility may be TLR activation as seen in the LPS stimulations but those levels simulated the expression levels seen in the low STAT1 patients. Another possibility is the impairment of miR 146a, which is known to target STAT1 [ 129] In the paired SLE patients visits, miR 146a might be increased as a response to STAT1 increases, but it is unable to downregulate STAT1. One potential reason that miR 146a is unable to downregulate STAT1 is due to alternative splicing. STAT1 exists as a long form (STAT1a) and short form (STAT1b). According to the TargetScan.com
60 (bioinformatics, miRNA prediction site), STAT1b has a shorter 3UTR compared to STAT1a 3UTR. The smaller 3UTR of STAT1b lacks miR 146a binding sites which would prevent miR 146a downregulation of STAT1b. Several HDs also displayed very high STAT1 levels, however CCL2 and CXCL10, even though elevated compared to low STAT1 HD, were significantly lower than SLE patients. A potential reason may be that IFN I drives CCL2 and CXCL10 expression, and high STAT1 primes the immune system to amplify CCL2 and CXCL10 expression when IFN I is present. Without IFN I, the high STAT1 levels may still prime the immune system but they lack t he ignition to drive the process forward. The results of this study indicate that STAT1 may be an important driver of pathogenesis in SLE, and this needs further analysis in future studies.
61 Table 2 1. Demographic data of SLE patients and healthy donors. Features SLE HD 1 Number of cases 103 65 Number of visits (2 or more) 180 (60) 65 (0) Mean age, years (range) 44 (25 68) 33 (19 59) Sex (Females/Males/unknown) 90/13/0 31/9/24 Race (AA/EA/LA / AsA / IrA / u nknown) 1 35/50/12/3/3/0 10/18/6/3/2/25 Race by visit (AA/EA/LA / AsA / IrA / u nknown) 1 64/86/20/7/3/0 10/18/6/3/2/25 Active/inactive by SLEDAI 49/131 N/A Malar Rash 9/94 N/A Discoid Lesions 2/101 N/A Photosensitivity 8/95 N/A Oral Ulcers 7/96 N/A Arthritis 17/86 N/A Serositis 4/99 N/A Pleuritis 5/98 N/A Nephritis 49/54 N/A Seizures 2/101 N/A Psychosis 1/102 N/A 1 A A, African Americans ; AsA, Asian Americans ; E A, European Americans ; HD, healthy donors with no history of autoimmune disease; IrA inter racial Americans ; LA Latino Americans ; unknown, undisclosed race.
62 Figure 21 Correlation of IFN Score, STAT1, ADAR, CCL2, CXCL10, and miR 146a le vels to SLE disease activity. A D) IFN Score, STAT1, ADAR, CCL2, and E. CXCL10 were significantly elevated in SLE patient visits (active and inactive disease activity are indicated as + and respectively) compared to healthy donors (HD). No statistical difference was detected between active and inactive SLE for I FN score, ADAR, and CXCL10. F H) miR 146a, and miR 146a did not show any significant difference among the groups.
63 Figure 22 Correlation of IFN Score, STAT1, ADAR, CCL2, CXCL10, and miR 146a levels to anti dsDNA autoantibodies. A) SLEDAI scores were significantly higher in anti dsDNA(+) than anti dsDNA( ) pat ient visits B,C) C3 and C4 were significantly lower in anti dsDNA(+) patients than anti dsDNA( ) patients. K. SLEDAI scores were significantly higher in anti dsDNA(+) than anti dsDNA( ) patient visits D H) IFN Score, STAT1, ADAR, CCL2, and CXCL10 were s ignificantly elevated in SLE patient visits compared to HD, but no statistical differences were detected between anti dsDNA(+) and anti dsDNA( ) patients for IFN score, ADAR, and CXCL10. I K) miR 146a, and miR 146a did not show any significant di fference among the groups.
64 Figure 23 Anti dsDNA level, IFN score, STAT1, CCL2, and CXCL10 in individuals with different ethnic background. A D ) Anti dsDNA titer was significantly higher in SLE patient visits of AA, EA, and LA race compared to HD of the respective race E L,Q T ) EA, and AsA patients than HD, while no significant difference was observed for LA M P) CCL2 was only significantly higher in AA but not in EA, LA, and AsA
65 Figur e 24 Comparison of SLEDAI, anti dsDNA titer, IFN score, STAT1, CCL2, and CXCL10 in patients with different ethnic background. To examine whether difference of race could affect clinical and biomarkers, AA, EA, LA, and AsA patient visits were compared to each other. A F) Overall, AA patient visits displayed generally higher SLEDAI, anti dsDNA titer, IFN score STAT1 CCL2, and CXCL10 than SLE patient visits of any other races.
66 Figure 25 SLE patients with two visits ranked by increasing or decreasing IFN score and STAT1. Data from first and second visits for each patient is denoted by an individual color line. A ) SLE patients ranked by increasing IFN score from the first to the second visit showed significant increase in STAT1, CCL2, CXCL10, and miR 146a. B ) Patients ranked by increasing STAT1 also showed significant increase in IFN score, CCL2, CXCL10, and miR 146a. C ) SLE patients ranked by decreasing IFN score from the first to the second visit showed significant decrease only in STAT1 and CXCL10. D ) Patients ranked by decreasing STAT1 showed significant decrease in IFN score and CXCL10.
67 Figure 26 ADAR, CCL2, and CXCL10 levels correlate with IFN score in both SLE patients and healthy donors. A ) ADAR, B) CCL2, and C ) CXCL10 display direct and sign for both SLE visits and healthy donors.
68 Figure 27 Bimodal distribution of STAT1 into high and low groups. A ) The log10 transformation of STAT1 shows a bimodal distribution of STAT1 with two populations (high and low STAT1 groups) with cut off at 1.5 Log[relative fold change] of STAT1 (Log[STAT1]) for both HD and SLE patient visits. B ) The low STAT1 groups displayed significant difference of STAT1 in SLE patient visits compared to HD. C ) On the other hand, the high STAT1 groups showed no significant diff erence between SLE and HD. D, E) In the low STAT1 group, STAT1 levels display a direct correlation to the IFN Score in SLE patient visits and healthy donors.
69 Figure 28 High levels of CCL2, CXCL10, and miR 146a compared to low STAT1 in high STAT1 SLE patients. A F) IFN score, ADAR, CCL2, CXCL10, pri miR 146a, and miR 146a were compared in SLE patient visits and HD which were segregated by high and low STAT1 levels demonstrating that high STAT1 SLE patients expressed higher levels of CCL2, CXCL10, and miR 146a than low STAT1 patients.
70 Figure 29 IFN score, CCL2, and CXCL10 in individuals with different ethnic background and STAT1 levels A C) IFN score, CCL2, and CXCL10 levels were compared segregated based on race and high and low STAT1. A. IFN scores were not significantly different between high and low STAT1 groups of the same race. In high STAT1 patient visits did not display significant di fferences among the different ethnic groups. There were too few patients in AsA high STAT1 group to be included. A) However, in the low STAT1 patient visits, IFN scores were higher in AA than EA and LA, EA were higher than LA, and LA had the lowest levels. B C ) High STAT1 patient visits displayed higher CCL2 and CXCL10 compared to their corresponding low STAT1 groups of the same race. Ethnic differences were not observed for AA, EA, and LA in the high STAT1 patients; however, in low STAT1 patient visits, si milar difference appeared to have the same effect on CCL2 and CXCL10 expression as described above for IFN scores.
71 Figure 210. Effect of high vs low STAT1 expression in ADAR, CCL2, and CXCL10 correlation with IFN score. A C) ADAR, CCL2, and CXCL10 displayed a significant linear coefficient of determination (r2) with IFN score for high STAT1 (HS, blue, left panels). The low STAT1 (LS, red, left panels) SLE patient visits r2 resembled that of the LS HD (black, right panels). B,C ) The slopes of high S TAT1 for CCL2 and CXCL10 were significantly higher than those for low STAT1 patient visits (left panels).
72 Figure 211. Ethnicity effect in high and low STAT1 groups. An analysis similar to that in Figure 7 was performed, but in addition, patient visits were segregated by race and by high and low STAT1. The effects of high and low STAT1 on the association of CCL2 and CXCL10 with IFN score did not appear to be influenced by A B) AA, C,D) EA, and E,F) LA ethnicity.
73 Figure 212 THP 1 cells harvested at various times from 2 to 24 h for RN A isolation and analyses. A,F,K) IFN score and the expression of B,G,L) STAT1, C,H,M) CCL2, D,I,N) CXCL10, and E, J,O) miR
74 CHAPTER 3 STAT1 EXPRESSION IND ICATIVE OF CCL2 AND CXCL10 RESI STANCE TO THERAPEUTIC EFFECTS OF PREDNISONE, MYCOPHENOLATE MOFETI L, AND HYDROXYCHLOROQUINE I N SYSTEMIC LUPUS ERYTHEMATOSUS PATIENTS Introduction Systemic Lupus Erythematosus (SLE) is a systemic autoimmune rheumatic disease affecting multiple systems and organs in the body. Several genetic and environmental factors have been implicated in SLE etiopathogenesis. Even though type I interferon (IFN 30 years ago to be elevated in SLE patient serum, it is only in recent years their increased expression has been rediscovered and postulated to play a key role in disease pathogenesis in the majority of patients [ 32 35] In addition to IFN I, STAT1 ( signal transducers and activators of transcription 1), an interferon inducible gene, is involved in type I, II, and III IFN signaling and is reported to be elevated in SLE [ 170 ] Besides STAT1, interferon regulated chemokines also p lay a role in SLE pathogenesis [ 142] C C motif chemokine ligand 2 (CCL2) and C X C motif chemokine 10 (CXCL10) have been implicated in SLE as good indicators of potential flares [ 141 ] The role of CCL2 in diseases such as psoriasis, rheumatoid arthri tis, and multiple sclerosis has incited additional interest o n its role in SLE [ 156 ] Both CCL2 and CXCL10 depend upon the Jak/STAT pathway activation for induction by interferon [ 161 163 ] and these two chemokines were identified as one of the 12 upregulated proteins in SLE [ 142 ] The role of microRNAs (miRNAs) has also been implicated in autoimmunity [ 201, 202 ] miR 146a was reported to be underexpressed in peripheral blood mononuclear cells of Chinese SLE patients [ 129 ] The function of miR 146a is now
75 known to regulate innate immune response and endotoxin tolerance [ 114, 116 118, 203] miR 146a has also been reported to be overexpressed in Sjgrens syndrome [ 177] psoriasis [ 119 120] and rheumatoid arthritis [ 121123 ] Recently, we described high and low STAT1 visits in SLE patients [ 204 ] In the low STAT1 population, levels of STAT1 correlated well with IFN score; however in the high STAT1 population, they did not. More impor tantly, high STAT1 patients displayed elevated expression of CCL2 and CXCL10, but no significant differences were observed linear regression representing the rate of change of CCL2 or CXCL10 per unit of change of IFN score was analyzed, the slopes of CCL2/IFN score and CXCL10/IFN score were significantly greater in the high STAT1 patients compared to the low STAT1 patients indicating that STAT1 potentially enhanced CCL2 and CXCL10 response to IFN I [ 204 ] The current therapies for SLE primarily aim to relieve symptoms and suppress the autoimmune response. Commonly used therapies include prednisone (PDN), mycophenolate mofetil (MMF), and hydroxychloroquine (HCQ ) PDN is a synthetic glucocorticoid that suppresses inflammation by inhibiting NF It inhibits monocyte and neutrophil inflammatory functions as well as B and T cell responses [ 28 ] Synthetic glucocorticoid, such as dexamethasone and PDN can inhibit phosphorylation of STAT1 and potentially blocks IFN induction by suppressing IFNAR signaling [ 205] ; however it has been shown that dexamethasone also upregulates STAT1 transcription. This inhibition of STAT1 function while increasing its transcription appears to be counterintuitive but may represent a case of cell adapting to compensate for the loss of
76 functional STAT1. Increases in STAT1 levels may lead to undesired consequences [ 206] MMF is a cytotoxic drug commonly used to prevent organ rejection after transplantation and also to treat autoimmune diseases such as SLE. MMF is a reversible inhibitor of inosine monophosphate dehydrogenase that blocks the de novo synthesis of guanosine nucleotides [ 25] The latter is requ ired for growth and proliferation of T and B cells, as they lack the scavenger pathway and are unable to compensate for the inhibition of de novo synthesis of guanosine. Inhibition of T and B cell growth blocks autoimmune response and leads to decrease in autoantibody production and T cell mediated tissue damage. The antimalarial drug HCQ functions by increasing the pH of endosomal vesicles. This disrupts antigen processing and inhibiting toll like receptor (TLR) 3, 7, 8, and 9 activity [ 25 27] ; furthermore, HCQ can inhibit macrophage production of interleukin 1 and interleukin6 [ 25] Since TLR7/9 have been implicated in inciting IFN I production due to recognition of self RNA/DNA, the blockade of these TLR could be attenuating IFN I production and antigen processing for presentation of T cells by antigen prese nting cells such as dendritic cells. In this study, we address the effects of commonly used drugs PDN, MMF, and HCQ on the expression of various biomarkers, including STAT1, ADAR, CCL2, CXCL10, and miR 146a, in our SLE cohort. P atients and M ethods Health y D onor s and SLE patient s Patient information are as described in accompanied manuscript [ 204] In brief, w hole blood was collected from a total of 103 SLE patients and 65 healthy donors enrolled in the University of Florida Center for Autoimmune Diseases registry from 20082011. Heal thy donors were selected based on no history of autoimmune disease, while all
77 SLE patients satisfied the American College of Rheumatology criteria [ 7 ] There were a total of 18 0 SLE visits with sequential samples collected in 60 SLE patients [ 204] Healthy donors only visited the clinic once; therefore, they represent a single visit. Among the total of 180 visits, SLE patients were active in 49 visits according to the SLEDAI score >4 All human blood samples were obtained from enrolled individuals with the approval of institutional review board at the Universi ty of Florida. Data C ollection RNA samples were isolated from peripheral blood leukocytes using Ambion LeukoLOCK kit (Ambion, Austin, TX) for each patient visit LeukoLOCK filters were washed twice with 3 ml of PBS and stabilized with 3 ml of RNAlater solution. Stabilized filters were stored at 80C for a minimum of 24 h before collecting total RNA. Total RNA including small RNAs w as collected using the Alternative Protocol (version 0602, Ambion) for the extraction of RNA from cells captured on LeukoLOCK filters using TRI reagent. Anti dsDNA levels, C3 and C4 complement levels, IFN Score, and SLE Disease Activity Index (SLEDAI) were obtained as described [ 204 ] Data Analysis TaqMan realtime PCR assays was used to measure gene expression. The copy number of miR 146a was normalized to total loaded RNA whereas mRNA levels were normalized to 18S RNA. Copy number of miR 146a was determined using a standard curve with synthetic miR 146a (Integrated DNA Technologies Inc., Coralville, IA) [ 207 ] T method [ 184] SLE patients were primarily treated with PDN, MMF, and/or HCQ. Correlations of all therapies during each patient visit were analyzed with levels of different SLE biomarkers. No patient in
78 our SLE cohort was treated with Belimumab, a BAFF inhibitor approved by the FDA for SLE [ 31] Analyses were performed using SAS version 9.2 and JMP Genomics version 5 (SAS, Cary, NC). Wilcoxon/Kruskal Wallis test was used to evaluate statistical significance between groups. P values less than 0.05 were consider ed significant. Biplot of normal mixtures was used to demonstrate the bimodal nature STAT1 in SLE patient and HD visits [ 208] associations in the study. Coefficient of determination (r2) was used to determine linear correlations. Significance between slopes was evaluated by analysis of covariance (ANCOVA). Results Eff ects of T herapy on L evels of B iomarkers Changes in C3, C4, and anti dsDNA antibody levels in SLE patient visits, and various mRNA biomarkers in peripheral blood leucocytes were examined for possible effects of therapy (Figure 3 1). As expected, SLEDAI (Figure 1A) and anti dsDNA autoantibody (Figure 3 11 D ) levels were significantly lower in treated (Tx) than untreated (UTX) patients while C3 (Figure 3 11B ) and C4 (Figure 3 1C) were significantly higher in Tx than UTX patients Overall, anti dsDNA autoantibody, IFN scores, ADAR, STAT1, CCL2, and CXCL10, were significantly lower in healthy donors (HD) than either UTX or Tx SLE patient visits (Figure 3 1D I). However, there were no significant differences among the groups for miR 146a (Figure 3 (Figure 3 1L) expression. pri miR 146a showed significantly higher level only in UTX compared to HD.
79 A biplot of STAT1 and the log[STAT1] using the normal mixtures method revealed that two groups, the high and low STAT1 populations in both SLE patients a nd HD (Figure 3 2A ,B). Interestingly, while there was no significant difference in STAT1 levels between high STAT1 SLE and HD low STAT1 SLE was significantly higher in STAT1 than low STAT1 HD ( Figure 3 2C ). To further elucidate the influence of high and l ow STAT1 populations, UTX and HD from Figure 3 1 were further examined by comparing the high (blue) and low (red) STAT1 groups (Figure 3 2D N). As expected, regardless of STAT1 levels, UTX was significantly higher in anti dsDNA, IFN score, ADAR, CCL2, and CXCL10 than HD (Figure 3 2D F, H,I) while there was no difference in STAT1, miR 146a, pri miR 3 2G ,J N ). High STAT1 HD displayed higher levels of STAT1, CCL2, and CXCL10 (Figure 3 2G I ) than low STAT1 HD; however for the remaining bio markers, there were no significant differences (Figure 3 2D F,L N). UTX patient visits were not significantly different by STAT1 levels (Figure 3 2G D F,H N) with the exception of STAT1 (Figure 3 2G ). Due to the lack of significant difference in levels of biomarkers between high and low STAT1 UTX patients, UTX were not separated in any subsequent analysis. Next various biomarker levels in treated patients with high versus low STAT1 visits were compared with UTX and HD. Overall two very important outcomes became apparent. First the lack of significant difference between UTX and high STAT1 for SLEDAI, IFN score, ADAR, CCL2, and CXCL10 (Figure 3 3A F, H,I) potentially indicating that the pathology of high STAT1 Tx patients resembled that of UTX patients. Second, high STAT1 Tx patient visits displayed significantly higher CCL2 and CXCL10 (Figure 3 3H ,I) than low STAT group, which might be indicators of increased
80 pathological activity. miR 146a also showed the same trend however, high STAT1 Tx patients have higher levels of miR 146a than UTX (Figure 3 3J ). Interestingly, pri miR 146a appeared to have an opposite trend ( Figure 3 3K ). Underlying E ffects of I ndividual T herapies Since many patients were on more than one medication, we wanted to observe the effects of individual drug by simple exclusion of patients not the particular therapy As for PDN (Figure by excluding patients not receiving PDN from the Tx group, there were no statistical significant difference between PDN Tx and UTX with SLEDAI, C3, and C4 (Figure 2A C). However SLE patients receiving PDN were more likely to be inactive (p=0.0071; likelihood ratio: 7.44) than active by SLEDAI score. The remaining biomarkers (Figure 3 4D L) showed similar significant trends as seen in the Tx population (Figure 3 1D L ). To appreciate these results, HCQ and MMF were also analyzed in the same manner (Figure 3 5,3 6 ). SLEDAI, C3, and C4 were significantly different between HCQ patients and UTX (Figure 3 5A C); however only SLEDAI and C4 we re significantly different between MMF and UTX patient visits (Figure 3 6A C). The results for SLEDAI were consistent with SLE patient visits treated with HCQ (p=0.0002; likelihood ratio: 13.9) or with MMF (p<0.0001; likelihood ratio: 16.1) were more likel y to be in inactive states The remaining biomarkers for HCQ (Figure 3 5D L) and MMF (Figure 3 6D L) resembled those on the entire population (Figure 3 1D L) reinforcing the hypothesis that the observed differences were results of therapy. After establishing the basic role of high and low STAT1, their correlation was further was significantly decreased in the low STAT1 PDN patient visits relative to UTX and HD; however high STAT1 PDN patient vis its were not significantly different (Figure 3 7L). This trend was not observed
81 for either HCQ or MMF patients (Figure 3 8L,3 9L). High and low STAT1 patients under PDN therapy ( Figure 3 7A C) did not display any significant differences for SLEDAI, C3, and C4 which resembled the earlier results (Figure 3 2A C). This differed for HCQ and MMF where low STAT1 patient visits were significantly lower than UTX patient visits for SLEDAI ; however higher in C3, and C4 ( Figure 3 8A C, 3 9A C). In PDN, HCQ, and MMF pat ient visits, CCL2 and CXCL10 was significantly elevated in the high STAT1 population compared to the low STAT1, but significantly different from UTX (Figure 3 7H ,I; 3 8H ,I; 3 9H ,I). This resembled what was observed earlier in high/low STAT1 Tx patients ( Fi gure 3 3H ,I) that high STAT1 patients were shielded from the effects of therapy regardless of the therapy used. The relationship between miR 146a and pri miR 146 was particularly revealing when the analyses took into account of the difference in high ST AT1 versus low STAT1. Where miR 146a did not show any significant difference in PDN, HCQ, and MMF patient visits (Figure 3 4J ,3 5J,3 6J ), high versus low STAT1 Tx patient visits as well as patients treated with PDN, HCQ, and MMF revealed that high STAT1 patient visits were significantly higher in miR 146a than low STAT1 patient visits, UTX, and HD (Figure 37J,3 8J,3 9J). However, the reverse is seen for pri miR 146a levels were significantly lower in high STAT1 patient visits than in low STAT1 patient visi ts, UTX, and HD for high/low STAT1 Tx patient visits (Figure 3 3 K) as well as patients treated with PDN, HCQ, and MMF (Figure 34K,3 5 K, 3 6 K). The reverse trend seen between pri miR 146a and miR 146a was probably due to differences in conversion from primary to mature miRNA.
82 Therapy dosage could vary based on disease manifestation and severity To examine the effects of therapy dosage, the PDN, MMF, and HCQ treated patients were separated by dosage (Figure 3 10). Unexpectedly, as dosage increased so did the levels of the biomarkers. This might be attributed to the way therapy was administered. As patients became more ill or higher active disease activity, prescription of higher doses of therapy might be expected. Essentially, therapy dosage might act as a marker of disease activity. Interestingly, the high STAT1 patient visits (blue) appeared to show higher levels of STAT1, CCL2 and CXCL10 than in low STAT1 patient visits as therapy dose increased (Figure 3 10, 3 11). Therapeutic I nfluences on CCL2 and CXCL10 A ssociation with IFN The accumulated evidence so far indicated that high levels of STAT1 were shielding CCL2 and CXCL10 from the effects of therapy; we tested how STAT1 levels affected the association of CCL2 and CXCL10 with I FN score. Since CCL2 and CXCL10 are known to be induced by interferon, this would suggest a positive covariation where CCL2 and CXCL10 increase as IFN score increases. The slope of CCL2/IFN score and CXCL10/IFN score thus represents the association between CCL2 and CXCL10 with IFN score. By comparing the slope between groups, the effects of therapy on the association of CCL2 and CXCL10 with IFN score could be examined. For example, when the slope of CCL2/IFN score was greater for UTX than that of a particul ar therapy, it suggested that the decreased association in CCL2/IFN score for the treated patients was a result of that particular therapy. When the association of CCL2 with IFN score was plotted as shown in Figure 3 12A three items were noted. First, both UTX and Tx were monotonic and increased as displayed a
83 linear component as described by the coefficient of determination (r2) and UTX has a greater linearity than Tx. Third, UTX had a greater slope for CCL2/IFN score than Tx potentially indicating that therapy decreased CCL2 responsiveness to IFN I. In Figure 3 12B Tx was segregated into high and low STAT1. Similarly, high STAT1 Tx and low STAT1 Tx were monotonic, increasing and linear. High STAT1 Tx displayed a significantly higher slope than low STAT1 Tx (Figure 3 12B ) and significantly higher slope than Tx (Figure 3 12A ) indicating that CCL2 responsiveness to IFN I in high STAT1 patients was more similar to that of the UTX patients Overall similar results were observed for PDN, MMF, and HCQ (Figure 3 12C H). The same analysis was performed for CXCL10 (Figure 3 13). The results were similar to those of CCL2 (Figure 3 12) with the exception for PDN and MMF in the high versus low STAT1 patient visits (Figure 3 13D ,F). PDN in high STAT1 patient visits was not significantly different than low STAT1; in addition, high STAT1 PDN was significantly lower than UTX, PDN which might indicate that PDN affected CXCL10 response to IFN 1 (F igure 3 13D ). For MMF, high STAT1 patient visits had significantly higher slope than low STAT1 patient visits (Figure 3 13F ); however, high STAT1 MMF was not significantly different in CXCL10 from MMF patient visits (Figure 3 13E). Individual T herapies E ffects on CCL2 and CXCL10 Finally, the effects of all possible therapy combinations (MMF, PDN, HCQ, HCQ+MMF, PDN+MMF+HCQ, PDN+MMF, PDN+HCQ, and UTX) on the expression of all biomarkers were compared. No significant differences were observed between UTX and the various treatments upon IFN score, STAT1, ADAR, pri miR 146a, and mature miR 146a (data not shown). However, CCL2 and CXCL10 displayed significant trends (Figure 3 14). For nearly every treatment, CCL2 was decreased compared to UTX
84 (Figure 3 12A ). In contrast, HD had significantly lower CCL2 levels than HCQ, HCQ+MMF, and PDN+MMF while the other therapy groups were not significantly different (Figure 3 12A ). Lack of significant difference from HD with the added significant decrease from UTX patient vi sits indicated that therapy may be decreasing CCL2 to normal levels ; however this was not true for high STAT1 patients. High STAT1 patient visits (blue) were significantly higher in CCL2 than the low STAT1 patients for nearly every treatment (Figure 3 12A ) The low STAT1 patients appear to be responsive to therapy; they were significantly lower than UTX and the majority was not significantly different from HD (Figure 3 15A ). This was reversed with the high STAT1 patients where HD were significantly lower than treated patients and the majority were not significantly different from UTX patients (Figure 3 15b). The results for CXCL10 were not as consistent as CCL2. UTX patients were significantly higher in CXCL10 than any treated patient (Figure 3 14B ). Both th e treated patient visits, high STAT1 patient visits, and the majority low STAT1 patient visits were significantly lower than UTX (Figure 3 14A,316). While the low STAT1 patient visits were significantly lower in CXCL10 than UTX, the high STAT1 were not sig nificantly different from UTX (Figure 3 16 ) potentially again indicating that high STAT1 levels shielded CXCL10 from the effects of therapy. Discussion Our study focused on the analysis of possible effects of therapy on SLE biomarkers and their relationshi p with interferon, CCL2 and CXCL10. Type I IFN and interferon signature genes were reported to be elevated both at the mRNA level based on data from microarray analyses and even at the protein level in the serum of SLE patients [ 35 5456] Not surprisingly our results reaffirm the elevated expression of ADAR, STAT1, CCL2, and CXCL10 in SLE patients [ 32, 33 55, 141 142 194] The
85 majority of SLE patient visits were receiving therapy at the time of sample collection. SLE patient visits using PDN, MMF, and HCQ as well as therapy combinations displayed no significant decrease of IFN score, STAT1, ADAR, pri miR 146a, and mature m iR 146a. PDN is a glucocorticoid that suppresses NF [ 209] It is unclear how or even if PDN suppresses IFN production. Glucocorticoids have been reported to suppress STAT1 phosphorylation (pSTAT1) [ 205] but depending upon cell type and profile, they can also lead to changes in the transcription of STAT1 [ 206, 210 ] STAT1 is important for CCL2 and CXCL10 induction by interferon [ 198 200 ] Furthermore, the decrease in pSTAT1 could explain why CCL2 and CXCL10 decreased in the low STAT1 patients. The increase in STAT1 expression may be an attempt to compensate for decreased pSTAT1 levels and may possibly explain the occurrence of the high ST AT1 patients. This may also be the reason for CCL2 and CXCL10 increase in high STAT1 patients and why CCL2 and CXCL10 are not as responsive to therapy in the high STAT1 patients compared to the low STAT1 patients. On the other hand, CCL2 and CXCL10 expres sion in SLE patients displayed significant decrease with therapy. PDN has been previously reported to decrease CCL2 and CXCL10 expression [ 211213 ] If PDN reduces pSTAT1 levels, this may explain in part the decrease of CCL2 and CXCL10 expression due to the role of STAT1 in chemokine signaling [ 198 200 ] In high STAT1 SLE patients, CCL2 and CXCL10 do not significantly change from UTX SLE patients, possibly indicating that the elevated levels of STAT1 a re facilitating pathogenic pattern occurring in the UTX patients. In part, STAT1 may be increasing to compensate for inhibition of STAT1 phosphorylation and
86 maintain CCL2 and CXCL10 levels as in the UTX patients. STAT1 has been associated with therapy resi stance in cancer. STAT1 overexpression protects cancers from DNA damaging agents including radiation therapies and chemotherapies in different cancer types [ 135] Radioresistant nu61 derived from radiosensitve SCC61 tumors displayed 4 9 overexpressed genes; of the se 49 genes 31 were ISGs also including STAT1 [ 136 ] Furthermore when STAT1 was overexpressed in SCC61 cells, it displayed radioresistance [ 137 ] Similarly, human fibroblas ts repeatedly exposed to IFN I displayed radioresistance [ 138] In 10 cancer cell lines, STAT1 expression correlated with resistance to doxorubicin and topoisomeraseII inhibitors [ 139 ] In addition, 14 ovarian cancer lines were observed for resistance to platinum compounds where STAT1 was associated with resistance to cisplatin and AMD473 [ 140 ] These associations between therapy resistance and STAT1 in cancer may explain the association of STAT1 levels with higher CCL2 and CXCL10 and the apparent lack of therapy sensitivity in high STAT1 patients. Increases in CCL2 and CXCL10 have been associated with SLE patients entering a state of flare activity [ 141, 142] We consider reduction of CCL2 and CXCL10 as good indicators of successful therapy, while elevation in STAT1 levels may indicate therapy resistance. Further work is needed to determine the role that STAT1 plays in therapy, but this study gives insight to a potentially new role for STAT1 in SLE.
87 F igure 31 The effects of therapies on the levels of various clinical parameters and biomarkers in the SLE cohort. A L) Disease activity, complement levels, anti dsDNA antibody levels IFN Score, ADAR, STAT1, CCL2, CXCL10, miR 146a, pri miR in treated (Tx) and untreated (UTX) SLE patient visits as well as healthy donors (HD ). Average trend lines for high STAT1 (blue) and low STAT1 (red) patient visit subsets are shown for comparison. All groups were compared among them and only significant p values are shown indicating each specific comparison.
88 Figure 32 High and low populations of STAT1 in both SLE and healthy donors. A,B) In the biplot of STAT1 and the log[STAT1] using the normal mixtures method, two populations of STAT1 were observed in HD and SLE patient visits. C ) Visits with levels above 1.5 log[STAT1] were considered as high STAT1, which displayed no significant difference between HD and SLE. Visits with levels below 1.5 log[STAT1] were classified as low STA T1 which showed SLE patient visits had significantly higher level than HD. D N) When analyzing different candidate biomarkers there was no significant difference between high and low STAT1 UTX patients, with the exception of STAT1. G I) Comparing high and low STAT1 HD showed significant difference in STAT1, CCL2 and CXCL10. D F,L N) No significant difference for anti dsDNA, IFN score, ADAR, miR 146a, pri miR 146a D I) Regardless of STAT1 levels, UTX patients were significantly higher in anti d sDNA, IFN score, ADAR, STAT1, CCL2, CXCL10 than HD
89 Figure 33 Comparison of high and low STAT1 subsets of all treated to untreated SLE patient visits. Data from Figure 1 are replotted to examine significant differences between the high and low STAT1 subsets.
90 Figure 34 The effects of prednisone therapy on levels of the various biomarkers in the SLE cohort. Data were analyzed as in Figure 1 except patients not receiving prednisone wer e excluded from the treated patient population.
91 Figure 35 The effects of hydroxychloroquine therapy on levels of the various biomarkers in the SLE cohort. Data were analyzed as in Figure 1 except patients not receiving mycophenolate mofetil were excluded from the treated patient population.
92 Figure 36 The effects of mycophenolate mofetil therapy on levels of the various biomarkers in the SLE cohort. Data were analyzed as in Figure 1 except patients not receiving mycophenolate mofetil were excluded from the treated patient population.
93 Figure 37 Comparison of high and low STAT1 subsets of PDN treated patient visits to untreated patient visits. Data from Figure 2 are replotted to examine significant differences between the high and low STAT1 subsets
94 Figure 38 Comparison of high and low STAT1 subsets of HCQ treated patient visits to untreated patient visits. Data from Figure 3 are replotted to examine significant differences between the high and low STAT1 subsets.
95 Figure 39 Comparison of high and low STAT1 subsets of MMF treated patient visits to untreated patient visits. Data from Figure 4 are replotted to examine significant differences between the high and low STAT1 subsets.
96 Figure 310. The effects of high and low STAT1 and dosage subsets on expression levels of the various biomarkers in the SLE cohort. Effects between doses were significant in SLEDAI (A) where SLEDAI scores were higher for PDN dose of 2060 mg/day compared to the 218 mg/day dose.
97 Figure 311 Comparison of high and low STAT1 subsets segregating into low versus high dosing on levels of the various biomarkers in the SLE cohort. Data from Figure 5 are replotted to examine significant differences between the high and low STAT1 subsets.
98 Figure 312. Association between CCL2, IFN score, and therapy. A ) The relationship of CCL2 versus IFN score presented as a slope was analyzed in untreated (UTX, black) and treated SLE patient visits (Tx, green). C,E,G) Similar analyses were carried out for PDN treated, MMF treated and HCQ t reated patient visits as well as for high STAT1 (blue) and low STAT1 (red) for B) Tx, D) PDN treated, F) MMF treated, and H) HCQ treated patient visits
99 Figure 313. Association between CXCL10, IFN score, and therapy. Data were analyzed as in Figure 3 12except CCL2 was substituted by CXCL10.
100 Figure 314. The effect of combined therapy on the expression of CCL2 and CXCL10 in high versus low STAT1 subsets. A ) CCL2 levels in HD, UTX, and other patient visits under treatment with different combinations of PDN, MMF, and HCQ were plotted (black bars). Only significant differences comparing each treatment group to either HD or UTX are indicated as black lines with p value shown. Data segregating into high STAT1 (blue line) and low STAT1 (red line) subsets are also shown and significant B ) CXCL10 data were analyzed similarly. Figure 315 Separate analyses of high and low STAT1 effects on CCL2 expression in various combined therapies. Data from Figure 3 14A are replotted to examine significant differences between the high and low STAT1 subsets.
101 Figure 316 Separate analyses of high and low STAT1 effects on CXCL10 expression in various combined therapies. Data from Figure 3 14B are replotted to examine significant differ ences between the high and low STAT1 subsets.
102 CHAPTER 4 POSITIVE CORRELATION OF STAT1 AND MIR 146A WITH ANEMIA IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS Introduction Systemic Lupus Erythematosus (SLE) is a chronic systemic autoimmune disease with complex clinical manifestations that can affect multiple organs or different organs at different periods of activity. Around 50% of SLE patients suffer from anemia [ 11] Commo nly anemia arises from iron deficiency, but several forms of anemia can be found in SLE patients, which include anemia of chronic disease, uremic anemia, megaloblastic anemia and hemolytic anemia [ 11] The most common cause of anemia in SLE [ 11] is anemia of chronic disease, which is result of prolonged inflammation leading to the production of hepcidin by the liver [ 214] Hepcidin upregulates ferroportin causing the sequestration of iron in the bone marrow which leads to impaired erythropoiesis. Similarly, uremic anemia is a result of decreased erythr opoietin caused by accumulation of nitrogenous waste such as urea due to renal failure. Megaloblastic anemia is resulted from impaired DNA synthesis due to a deficiency of vitamin B12. Hemolytic anemia is caused by the autoantibodies to erythrocytes via complement activation and hemolysis. Environmental and genetic factors have been implicated in SLE etiopathogenesis. Type I interferon (I IFN) expression has been discovered to play a key role in the pathogenesis of SLE [ 215 216] Due to the presence of common way to evaluate this component is to measure IFN response genes [ 32, 54 55] IFN response genes such as OAS1, MX1, and LY6E can be specific for type I IFN signaling, or can respond to multiple types of IFN signaling as STAT1 (signal transducers and activators of transcription 1). STAT1 is involved in type I, II, and III IFN
103 signaling and has been observed to be elevated in SLE [ 170] I n response to type I IFN, STAT1 causes interferon receptor (IFNAR) 1 and 2 dimerization, activation and phosphorylation of IFNAR by Tyk2 and Jak1, and thus docking and phosphorylation of STAT1 and STAT2 (signal transducers and activators of transcription 2 ) [ 171 ] The heterodimer STAT1STAT2 is then translocated into the nucleus where it can bind specific promoters playing a key role in IFN signaling and production [ 130] STAT1 is also involved in multiple chemokine and cytokine signaling pathways [ 217, 218 ] Chemokines such as C C motif chemokine ligand 2 (CCL2) and C X C motif chemokine 10 (CXCL10) have become another important topic of research in SLE in recent years [ 141 142 ] CCL2 is a potent recruiter of monocytes, T cells, basophils, and dendritic cells to site of infection or tissue damage [ 170 172] Roles of CCL2 have been implicated in psoriasis, rheumatoid arthritis, and multiple sclerosis [ 156 ] In SLE, CCL2 was reported to be upregulated in the serum [ 142] CXCL10, a potent attractor of monocytes, macrophages, T cells, NK cells, and dendritic cells to site of tissue damage and infection [ 157 158] was identified to be upregulated in serum of SLE patients [ 141, 142 ] Both CCL2 and CXCL10 with CCL19 were shown to be good indicators of SLE activity [ 141 142 ] miR 146a is a small noncoding RNA involved in the regulation of endotoxin induced tolerance and cross tolerance [ 116118 ] miR 146a was reported to be underexpressed in PBMCs of Chinese SLE patients resulting in deregulation of STAT1 and IRF5 that could potentially result to overexpression of IFN [ 129 ] Our present study observes the interaction among STAT1, CCL2, CXCL10, and miR 146a in SLE patients with and without anemia, demonstrating that these
1 04 biomarkers are overexpressed in anemic SLE patients and that anemia appears to be partially dependent on cl inical features and lupus nephritis but not with therapy and race. This study explores the potential upregulation of STAT1 and miR 146a in SLE and the association of anemia with STAT1 and miR 146a. Materials and Methods Donor and SLE Patient Demographic D ata All studies have been performed in accordance with the 1964 Declaration of Helsinki and its later amendments. All human blood samples were obtained from enrolled individuals in the University of Florida Center for Autoimmune Diseases registry from 20082011 with the approval of institutional review board at the University of Florida. Whole blood was collected from a total of 101 SLE patients and 39 healthy controls. All persons gave their informed consent prior to their inclusion in the study. There wer e a total of 180 SLE visits including sequential samples collected in 58 SLE patients (43 patients with 2 visits, 10 patients with 3 visits, and 5 patients with 4 visits) and 41 patients with a single visit. Females to males ratio in SLE were approximately 6:1 by both individuals ( Figure 4 1A) and visits ( Figure 4 1B). The ethnic profile of SLE patients and healthy controls were segregated into African Americans (AA), Asian Americans (AsA), European Americans (EA), Latino Americans (LA), and Multiethnic Ame ricans (MA, Figure 4 1C, 2A,B). Average age of SLE patient was 43 year old and the average age of HD was 33 years old ( Figure 4 1D). SLE patients were active in 43/180 visits, according to the SLEDAI score greater than 4. SLE patients were receiving 0 60 m g/day of prednisone (PDN), 0 400 mg/day of hydroxychloroquine (HCQ), and 0 3500 mg/day of mycophenolate mofetil (MMF, Figure 4 2C E) and many of them were on multiple medications ( Figure 4 1D).
105 Leukocytes and RNA Purification Peripheral blood leukocytes were collected from whole blood using Ambion LeukoLOCK kit (Ambion, Austin, TX). LeukoLOCK filters were washed twice with 3 ml of PBS and stabilized with 3 ml of RNAlater solution. Stabilized filters were stored at 80C for a minimum of 24 h before collec ting total RNA. Total RNA including small RNAs was collected using the Alternative Protocol (version 0602, Ambion) for the extraction of RNA from cells captured on LeukoLOCK filters using TRI reagent. Micro RNA and Messenger RNA qRT PCR OAS1, MX1, LY6E, C CL2, CXCL10, and miR 146a levels were analyzed by qRT PCR. MiRNA qRT PCR was performed using the TaqMan MicroRNA Reverse Transcription Kit, TaqMan Fast Advance PCR Master Mix, and TaqMan MicroRNA primers (Applied Biosystems, Foster City, CA). mRNA qRT PCR was performed using the TaqMan HighCapacity cDNA Reverse Transcription Kit, TaqMan Fast Advance PCR Master Mix, and TaqMan mRNA assay primers (Applied Biosystems). All reactions were analyzed using StepOne Real Time PCR System (Applied Biosystems). IFN Sc ore and SLE Activity The expression of three known type1 IFN signature genes, MX1, OAS1, and LY6E, were z transformed into Interferon Scores (IFN score) as previously shown [ 32, 180] The SLE Disease Activity Index (SLEDAI) was used to classify patients into active or inactive at the time of the visit. SLE patients with a SLEDAI greater than 4 were considered active and patients with SLEDAI of 4 or less were considered to be inactive.
106 Anemia Anemia was defined by hemoglobulin level <12 gm/dl for women and <13 gm/dl for men. Evaluation of anemia for a few pregnant pat ients was excluded. Twelve patients displayed a change in anemia status between visits. Only one patient of the 51 anemic patients was diagnosed to have hemolytic anemia. Data Analysis The copy number of miR 146a was normalized to total RNA loaded whereas mRNA levels were normalized to 18S RNA. Copy number of miR 146a was determined using a standard curve with synthetic miR 146a (Integrated DNA Technologies Inc., T method [ 184] where the cycle threshold (CT) values, corresponding to the P CR cycle number at which fluorescence emission reaches a threshold above baseline emission, were determined for the miRNA or mRNA expression relative to untreated controls. Analyses were performed using SAS version 9.2 and JMP Genomics version 5 (SAS, Cary NC). Wilcoxon/Kruskal Wallis test was used to evaluate significance between groups. Wilcoxon Signed Rank test for matched pairs was used to evaluate SLE patients with two visits. Fishers Exact Test was used to determine significance between groups in contingency tables. Data from patients with two visits showing changes in their anemia status from one visit to next were compared using paired t test. P values less than 0.05 were considered significant. Results Anemia and SLE B iomarkers The SLE cohort w as segregated into patient visits that were clinically diagnosed with anemia (n=52, 28%) and without anemia (n=128, 72%) according to medical
107 records. The IFN scores in anemic SLE patients were elevated compared to nonanemic SLE patients (p<0.0001) and healthy donors (HD, p<0.0001, Figure 4 3 A). Also the IFN score of nonanemic SLE patients was higher than HD (p=0.017). Similarly, STAT1 (Fig ure 4.3 B), CCL2 (Fig ure 4 3 C), CXCL10 (Figure 4 3 D), and miR 146a (Fig ure 4 3 E) were significantly elevated in anemic compared to non anemic SLE patients and HD. In addition, STAT1 (p<0.0001), CCL2 (P<0.0001), and CXCL10 (p<0.0001, Figure 4 3 B D) were elevated in nonanemic patients compared to HD. In contrast, primary microRNA 146a (pri miR 146a) was higher in nonanemi c than anemic SLE patients and HD (Figure 4 3 F). Primary miRNA is converted into the precursor miRNA and finally to the mature miRNA. The reverse in expression of pri miR 146a vs miR 146a may be indicative of problems in pri miR 146a being converted to mat ure miR SLE, was not significantly different among all the groups (data not shown). To rule out any potential bias in the contribution of patients with multiple visits, miR 146a and STAT1 data using only a single visit per patient were also analyzed (Fig ure 4 4 ). When data from the first visit of patients with multiple visits were selected along with patients with only a single visit, miR 146a, STAT, and SLEDAI were signi ficantly elevated in anemic patients compared to nonanemic patients ( Figure 4 4 A C). Similarly, when data from only the second visits were selected instead of the first visit, miR 146a, STAT, and SLEDAI were also significantly elevated in anemic patients compared to nonanemic patients ( Figure 4 4 D F). Finally, when data from visits that were anemic were selected over nonanemic in the multiplevisit patients to enrich the population for anemia, as seen before, miR 146a, STAT, and SLEDAI were significantly
108 elevated in anemic patients compared to nonanemic patients ( Figure 4 4 G I). These results demonstrated that considering visits rather than individual patients did not affect the observed outcome for anemia. To further understand the effects of anemia, there were twelve patients with two visits showing changes in anemia status from one visit to the next were further analyzed ( Figure 4 5 ). Anemic patients displayed significant increase in miR 146a and STAT1 compared to the visits that they were not anemic ( Figure 4 5 A,B); however, IFN score, CCL2, CXCL10, and SLEDAI did not display significant change (data not shown). To our knowledge, the observation that these biomarkers are significantly elevated in anemic compared to nonanemic SLE patients is a novel finding. Relationship of Anemia and SLEDAI To examine whether anemia has any relationship with SLE disease activity, patients visits with and without anemia were compared to visits with 43 active and 135 inactive SLEDAI statuses of the patients. Active S LE patients were more likely to be anemic than non anemic (anemia ( ), likelihood ratio: 5.0, Fisher exact test: p=0.020, Fig ure 4 6 A).Additionally anemic SLE patients displayed higher SLEDAI scores than nonanemic SLE patients (p=0.0028, Figure 4 6 B). T o further investigate the interplay between anemia and SLEDAI, the expression of IFN score, STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a was analyzed in patients classified into 4 groups by a combination of anemia and SLE activity (SLEDAI >4 for SLEDAI ure 4 6 C H). In inactive SLE visits, IFN score was higher in anemic compared to nonanemic patients (p=0.021) while no other groups displayed significant differences (Figure 4 6 C). CCL2 was higher in anemic ac tive and inactive SLE (p<0.031) compared to nonanemic inactive SLE
109 visits (Fig ure 4 6 E). Also, CXCL10 was elevated in anemic inactive SLE compared to nonanemic active and inactive SLE visits (p<0.026, Figure 4 6 F). In contrast, STAT1 was significantly higher in anemic compared to nonanemic patient visits (p<0.025) regardless of active or inactive status (Figure 4 6 D). Similarly, miR 146a was significantly higher in the anemic SLE compared to nonanemic SLE patient visits (p<0.045), regardless of activ e or inactive status (Figure 4 6 G). Pri miR 146a was significantly lower in anemic active SLE compared to nonanemic active and inactive SLE patient visits; in addition, anemic inactive SLE displayed significantly lower pri miR 146a than non anemic active SLE patient visits (Figure 4 6 H). Even when excluding multiple visits, similar results were observed for miR 146a and STAT1 (Figure 4 7 5). In summary, SLE disease activity and anemia are related; STAT1 and miR 146a were exclusively elevated in anemic patients regardless of SLE disease activity. Relationship of Anemia and Lupus Nephritis Lupus nephritis (LN) is one of the most serious pathological manifestations in SLE. Additionally, anemia has been shown to be associated with LN as to be a good predict ing factor for a nephritis flare [ 219, 220] To examine the link between anemia and LN similar analysis was performed as above. It was not surprising to observe that LN patients were more likely to be anemic than nonanemic also in our cohort (likelihood ratio: 5.0, Fisher exact test: p=0.024, Figure 4 8 A). To further analyze the etiological association between anemi a and LN, SLEDAI, IFN score, STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a expression were compared (Figure 4 8 B H). SLEDAI scores were higher in anemic patients with LN than nonanemic SLE patients with (p=0.011) or without LN (p=0.028, Figure 4 8 B). IF N score was higher in anemic SLE patients without LN compared to nonanemic patients with (p=0.0021) or without LN (p=0.049)
110 and also in anemic LN patients compared to nonanemic LN patients (p=0.040, Figure 4 8 C). STAT1 was elevated in anemic patients com pared to nonanemic patients regardless of LN (p<0.025, Fig ure 4 8 D). CCL2 was elevated in anemic LN patients compared to nonanemic SLE patients with (p=0.013) or without LN (p=0.037, Figure 4 8 E). CXCL10 was significantly elevated in anemic LN patients c ompared to non anemic LN patients (p=0.024), but no other group was significantly different (Figure 4 8 F). miR 146a was higher in anemic patients compared to nonanemic patients regardless of LN (Fig ure 4 8 G). Finally, pri miR 146a was lower in anemic LN patients compared to LN patients with (p=0.012) or without anemia (p=0.0002, Figure 4 8 H). Additionally, pri miR 146a was elevated in non anemic non LN patients compared to nonanemic LN patients (p=0.014, Figure 4 8 H). Similar to what was observed for the analysis of anemia and SLEDAI (Figure 4 6 ), the results shown in Fig ure 4 8 indicate a link between anemia and LN; STAT1 and miR 146a specifically elevated in anemic patients regardless of LN. Relationship of Anemia and Race African Americans (AA) and Eur opean Americans (EA) were selected since they were the majority of SLE patients (Fig ure 4 2A). Additionally, AA were significantly more likely to be anemic than EA (likelihood ratio: 8.5, p=0.0032, Figure 4 9 A). AA and EA were separated to compare anemic and non anemic patients. In AA patients, STAT1 (p=0.020) and miR 146a (p=0.019) were elevated in anemic SLE patients compared to nonanemic patients, while pri miR 146a was decreased (p=0.018) in anemic patients (Fig. 4A, D, E). Similar results were observ ed in EA patients for STAT1 (p=0.0065), miR 146a (p=0.0099), and pri miR 146a (p=0.0077, Fig.
111 4G, J, K); in addition, CCL2 (p=0.021) and SLEDAI (p=0.021) were elevated in anemic patients (Figure 4 10H, L). By segregating our cohort into anemic and nonanem ic SLE patients, AA and EA were compared. In nonanemic SLE patients, AA expressed significantly elevated IFN score (p=0.0012), STAT1 (p=0.0038), CCL2 (p=0.0001), CXCL10 (p=0.0022), and SLEDAI (p=0.032) but decreased miR 146a (p=0.029) compared to EA patients (Fig ure 4 11A E, G), but in anemic SLE patients there were no significant differences between AA and EA patients (Fig ure 411H N). Thus, while race plays a role in the expression of SLE biomarkers in non anemic patients, it does not appear to play a significant role in SLE patients with anemia. Effects of Therapy on Anemia The majority of SLE patients were receiving therapy in the form of PDN, HCQ, and MMF (Fig ure 4 2C E). PDN is a glucocorticoid that suppresses inflammation by blocking NF ients receiving PDN (PDN+) were more likely to be anemic than non anemic (likelihood ratio: 9.8, p=0.0015, Figure 4 9 B). HCQ is an antimalarial drug that increases the pH of the endosomal vesicles interrupting antigen processing and the function of TLR3, 7, 8, and 9 as well as inhibiting production of interleukin1 and 6 by macrophages [ 2527 ] MMF is an immunosuppressant that affects T and B cell growth by inhibiting inosine monophosphate dehydrogenase, which is critical enzyme for de novo synthesis of guanosine nucleotides [ 25] PDN treated anemic patients displayed higher STAT1 (p=0.025), miR 146a (p=0.023), and SLEDAI (p=0.020) but lower pri miR 146 (p=0.013) than nonanemic SLE patients (Fi gure 412A, D F). In MMF treated SLE patients, anemic patients displayed elevated STAT1 (p=0.0002), CCL2 (p=0.0089), CXCL10 (p=0.0029) and
112 miR 146a (p=0.048) but decreased pri miR 146a (p=0.033) compared to non anemic patients (Fi gure 412G K). HCQ treated patients displayed similar results as those seen in MMF, with the exception of CXCL10 (Fi gure 412M R). When patients instead of vis its were analyzed, PDN, MMF, and HCQ were pooled into treated (Tx) grouping. Anemic Tx patients displayed significantly higher miR 146a and STAT1 than nonanemic Tx patients; however anemic untreated (UTX) patients were not significantly different from non anemic UTX patients (Fi gure 413 ). The lack of statistical significance for UTX may be due to the small sample size. However, these results indicated that in anemic SLE patients the expression of STAT1 and miR 146a was consistently higher than in nonane mic patients and this was not significantly affected by the most common therapies for SLE. Discussion In this study, expression of previously identified SLE biomarkers was examined and correlated with demographic and clinical parameters. Anemic SLE patients displayed significantly higher levels of these biomarkers vs. nonanemic patients, particularly STAT1 and miR 146a, independent of disease activity and association with LN. In general, anemia is also considered as a good predictor of SLE flares [ 221 ] Anemia can lead to reduced levels of oxygen even resulting in hypoxia in organs [ 222] Low oxygen conditions as sociated with anemia can promote neutrophils, monocytes, and macrophage survival [ 223, 224] macrophages and dendritic cell pro inflammatory response, and direct activation of monocytes [ 225227 ] Pro inflammatory transcription factors such as NF 1 are activated in hypoxic conditions [ 225, 226 ] Since a link exists between hypoxia, anemia, and inflammation via hepcidin [ 228] it is possible that hypoxia and anemia act as factors that modulate the immune
113 response to enhance inflammation to expedite the clearance of foreig n pathogens. Our study shows that anemic SLE patients display significantly higher IFN score, STAT1, CCL2, CXCL10 and miR 146a levels in anemic patients than nonanemic patients and HD. Active SLE patients with LN displayed significantly elevated STAT1 and miR 146a exclusively in those with anemia. Even the role of race did not influence the effects of anemia on STAT1 and miR 146a observed in this study. In a study using zebra fish models for hematopoiesis, STAT1 was required for hematopoiesis and its incr ease was associated with increase in myeloid lineage but decrease in erythroid lineage [ 229 ] In both humans and mice, hepcidin is primarily produced by the liver in response to IL 6 [ 214 ] but has also been shown to be produced [ 230] STAT1 [ 230] which may indicate that macrophages contribute to anemia via production of hepcidin by IFN stimulation. The present study shows that miR 146a can be considered as another biomarker consistently elevated in anemic SLE patients. In addition, miR 146a is transcriptionally regulated by NF 1, and PU.1 [ 114 195 231] In zebra fish, transcriptional regulation of miR 146a by PU.1 has been shown to be critical for macrophage development and hematopoiesis and loss of miR 146 led to a loss of macrophages [ 232] In mice where miR 146a was overexpressed, erythropoiesis and bone marrow reconstitution capacity decreased [ 233 ] Activation of NF 1 under hypoxic conditions [ 225 226 234] may in part explain why miR 146a levels are elevated in anemic SLE patients, but it is unclear why elevated levels of miR 146a are not
114 decreasing levels of IFN in SLE. TRAF6 and IRAK1/2 are known targets of miR 146a, SLE as described in rheumat oid arthritis patients [ 118, 122 ] miR 146a has been reported to target STAT1, but miR 146a was also shown to be reduced in SLE patients in the same study which is not the case in our study and may be due to differences between ethnic groups and differences between PBMCs vs mononuclear cells [ 129] In SLE patients, there may be a loss of function of miR 146a that becomes unable to regulate STAT1. Furthermore, STAT1 has two isoforms (STAT1a and STAT1b). STAT1a is the long isoform which accord ing to Targetscan is predicted to be targeted by miR 146a while the short form STAT1b is not predicted to be regulated by miR 146a. STAT1b is the same isoform that was shown to decrease erythropoiesis in zebra fish [ 229] Th e elevated expression of STAT1 and miR 146a in anemia represents a novel observation in SLE that requires further investigation.
115 Figure 41 Gender and Therapy by visits. A,B) The composition of gender by individual and by visit were plotted for both HD and SLE patients. C) Similarly, the ethnic composition of both HD and SLE patient were plotted by visit. D) The therapy regimen per visit were plotted. E) SLE patients were si gni ficantly older than the HD
116 Figure 42 Demographic data of SLE patients and healthy donors. A,B) Patients and healthy donors were composed of African Americans (AA), Asian Americans (AsA), European Americans (EA), Latin Americans (LA), a nd Multiethnic Americans (MA). C) Of the 78 patients receiving prednisone (PDN), the median dose was 12.5 mg/day and 102 patients did not receive any PDN. D) For (E), For hydroxychloroquine (HCQ), 124 patients were receiving a median dose of 400 mg.day and 56 patients were not receiving any HCQ. E) For mycopheloate moefetil (MMF), 71 patients were receiving a median dose of 3000 mg/day and 109 patients were not receiving MMF.
117 Figure 43. Relationship between anemia and biomarker expression in SLE. A D) IFN score, STAT1, CCL2 and CXCL10 in both SLE patients with and without anemia displayed significantly higher expression than in healthy donors (HD), and they were also significantly higher in SLE patients with than without anemia. E) miR 146a in anemic SLE patients is significantly higher than in nonanemic SLE patients and HD. F) pri miR 146a is significantly higher in nonanemic SLE patients than in anemic SLE patients and HD.
118 Figure 44. The relationships of miR 146a, STAT1, and SLEDAI in non anemic versus anemic patients were essentially unchanged considering different methods of selecting only a single visit for patients with multiple visits. A C) Data from the first visit of the 58 patients with multiple visits were analyzed together with the 4 3 patients with single visits. D F) Similar to the above except the data from first visit was substituted by data from the second visit of the 59 patients with multiple visits. G I) For the 58 patients with multiple visits, data from visits positive for anemia were selected for analysis. Nonanemic patients that remained nonanemic from one visit to the next did not show significant change in miR 146a and STAT1.
119 Figure 45. Changes of anemia status were correlated with miR 146a and STAT1 mRNA levels. Twelve patients with two visits where their anemia status changed from nonanemic to anemic were compared with paired t test. A,B) Both miR 146a and STAT1 displayed significant increase when the patients changed from nonanemic to anemic status.
120 Figure 46. Differential expression of IFN score, STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a in active versus inac tive patients with anemia. A) A contingency plot to compare active and inactive patients against anemic and nonanemic patients revealed that anemic patients were more likely (likelihood rati o: 5.0) to be active patients. B) SLEDAI scores were significantly elevated in anemic patients. C) Anemic inactive patients displayed significantly higher IFN score compared to inactive nonanemic patients. D) Ane mic patients display higher STAT1 scores compared to nonanemic patients regardless of SLEDAI. E) For CCL2 expression, anemic active and inactive patients were higher than nonanemic inactive patients. F) CXCL10 was higher in anemic inactive patients compared to nonanemic act ive and inactive patients. G) miR 146a was higher in anemic patients compared to nonanemic SLE patients regardless of SLEDAI status. H) Pri miR 146a was higher in nonanemic active SLE patients compared to anemic active and inactive patients. Non anemic inactiv e SLE patients also displayed higher pri miR 146a than anemic active SLE patients.
121 Figure 47. The relationships of miR 146a and STAT1 versus SLEDAI and anemia. miR 146a and STAT1 were analyzed with data from one visit for eac h patient and of those with multiple visits only anemiapositive visits were included. A) Only anemic inactive patients displayed significantly higher miR 146a than nonanemic inactive patients. B) STAT1 was significantly lower in healthy controls than any of the four groups of patients. In addition, non anemic inactive patients were lower than anemic active and anemic inactive patients.
122 Figure 48. The interactions of anemia and lupus nephritis on SLE biomarkers. A) A contingency plot to compare patients with and without lupus nephritis (LN) against anemic and non anemic SLE patients revealed that anemic patients were more likely to have LN. B) SLEDAI scores were elevated in anemic LN patients compared to nonanemic SLE patients with and without LN. C) IFN score was higher in anemic patients without LN compared to nonanemic patients with or without LN. IFN scores in anemic LN patients were significantly highe r than non anemic LN patients. D) Anemic patients display higher STAT1 levels compared to nonan emi c patients with or without LN. E) Anemic LN patients expressed higher CCL2 than non anemi c patients with or without LN. F) CXCL10 was higher in anemic LN patients compared to nonanemic patients without LN. G) miR 146a was higher in anemic patients comp ared to nonanemic patients regardless of LN. H) Pri miR 146a was higher in nonanemic patients without LN compared to non anemic LN and anemic patients without LN. Anemic patients without LN also displayed a higher pri miR 146a level than anemic patients with LN.
123 Figure 49. African Americans SLE patients were more likely anemic and SLE patients with anemia were more lik ely under prednisone therapy. A) A contingency plot of African Americans (AA) and European Americans (EA) to compare anemic and non anemic SLE patients showed that anemic patients are more likely to be AA than EA. B) A contingency plot of patients receiving and not receiving PDN showed that anemic patients were more likely to be receiving PDN than nonanemic patients.
124 Figure 410. The relationship of race and anemia on STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a. A,D,E) Anemic African American (AA) SLE patients displayed significantly higher STAT1 and miR 146a levels compared to nonanemic SLE patients, but pri miR 146a was el evated in nonanemic co mpared to the anemic patients. B,C,F) CCL2, CXCL10, and SLEDAI were not significantly different between anem ic and nonanemic AA patients. G,H,J L) European American SLE patients (EA) display significantly higher STAT1, CCL2, miR 146a, and SLEDAI in anemic cases, but the reverse was observed for pri miR 146a. C,I) CXCL10 was not significantly different between anemic and non anemic for either AA or EA.
125 Figure 411. IFN score, STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a expression in African American vs European American SLE pat ients with and without anemia. A E,G) In non anemic SLE patients, African Americans (AA) display significantly higher IFN score, STAT1, CC L2, CXCL10, miR 146a, and SLEDAI score than European Americans (EA). F) No significant difference was observed for pri miR 146a between non anemic AA and EA. H N) In anemic SLE patients no significant differences were observed between AA and EA for all biomarkers examined.
126 Figure 412. STAT1, CCL2, CXCL10, miR 146a, and pri miR 146a in anemic vs non anemic SLE patients on different treatment. In PDN, MMF, and HCQ treatments, levels of STAT1 (A,G,H) and miR 146a (D,J,P) were significantly elevated in anemic compared to nonanemic SLE patients, but pri miR 146a (E,K,Q) was significantly decreased in anemic versus nonanemic SLE patients. CCL2 and CXCL10 were not significantly different in patients treated with PDN (B,C), but were significantly elevated in anemic patient receiving MMF (H,I), and only CCL2 was elevated in HCQ (N). SLEDAI scores were significantly higher in PDN (F) patients with anemia but no significant difference was observed for MMF (L) and HCQ (R) treated patients.
127 Figure 413. The r elationships of miR 146a and STAT1 versus anemia and therapy were analyzed including only data from one visit for each patient and of those with multiple visits only anemia positive visits were included. A,B) SLE patients were separated into treated (Tx) and untreated (UTX) patients with and without anemia. Only anemic Tx patients displayed significantly higher miR 146a and STAT1 than nonanemic Tx patients.
128 CHAPTER 5 DISCUSSION AND FUTURE DIRECTIONS Although environmental and genetic factors have been implicated in SLE, the precise etiology remains an unknown. Currently, both the adaptive and innate immunity are believed to play a role in the autoimmune condition. This now spans from abnormalities of B and T lymphocytes to the involvement of TLRs and I FN I in SLE. The work presented in the previous chapters exploit the interconnectivity among IFN I, STAT1, CCL2, and CXCL10 as well as miR 146a. Though the adaptive and innate immunity are looked at as separate paradigms, the reality is that they are heavi ly interconnected. STAT1 plays a key role in how IFN I upregulates CCL2 and CXCL10. STAT1 also plays a role on how therapy affects CCL2 and CXCL10 expression. Finally, STAT1 and miR 146a seem to work together in anemia in SLE. In spite of these nover High and L ow STAT P opulations For the validation of relatively newer biomarkers in our patient cohort, the focus was on IFN score, STAT1, ADAR, CCL2, CXCL10, and miR 146a compared to established clinical markers, including SLEDAI, anti dsDNA, C3, and C4. Our results show that IFN score, ADAR, STAT1, CCL2, and CXCL10 levels were significantly elevated in the SLE patients as compared to healthy donors. This is consistent with the current literature that many of these biomarkers are upregulated in SLE [ 32, 33, 55 141, 142, 194 ] However, STAT1 displayed unique property not previously described in the literature where two populations of STAT1 exist. The high STAT1 population of patients displayed h igher levels of CCL2 and CXCL10 as well as
129 reaffirm the finding in patients that interferon induces upregulation of STAT1, CCL2 and CXCL10 comparable to those found in low ST AT1 patients However, STAT1 appeared to be more intriguing than the other markers due to the capacity to stratify patients and healthy donors into the two high and low STAT1 populations based upon their expression of STAT1. The population of STAT1 that proximate what would be considered as near normal levels which we refer to as low STAT1, and population that is 10 to 100 fold higher which we refer to as high STAT1 population. The discovery of these two populations of STAT1 evolved from attempting to explain why STAT1 was not correlating with IFN score in a linear manner. One possible explanation was that STAT1 did not exhibit a Gaussian distribution and was skewed causing a nonlinear or even nonmonotonic relationship between STAT1 and IFN score. Indeed, the distribution appeared to be heavily skewed, but the distribution of the log[STAT1] revealed possibly two populations of STAT1 and further confirmed by k clustering. The low population of STAT1 maintained a linear relationship with IFN score; however, the high STAT1 population demonstrated a nonmonotonic relationship which displayed no covariation with IFN score. The high STAT1 population displayed several interesting features. High STAT1 in both HD and SLE expressed higher levels of ADAR, CCL2 and CXCL10 compared to their respective low STAT1 HD and SLE. In the SLE patients, the high STAT1 patients appear to express more CCL2 and CXCL10 per change of IFN score compared to the low STAT1 patients. This phenomenon was observed by comparing the means of the slopes between high and low STAT1 CCL2 as well as CXCL10. Whether a direct or indirect effect of STAT1, it is not possible to say unless protein levels of STAT1 and
130 pSTAT1 were measured and correlated to expression levels of CCL2 as well as CXCL10; even then, it may not be totally conclusive. Even the association between IFN score with CCL2 and CXCL10 may not be a di rect effect since both of these chemokines can be activated via NF kB as well as others. However, there is some sort of relationship with increased STAT1 levels which do not correlates with IFN score, but somehow enhance the relationship between IFN score and CCL2 as well as CXCL10. More puzzling than the relationship among STAT1, CCL2, CXCL10, and IFN score, is why STAT1 was elevated in some of the patients. From the data of the paired, STAT1 can be observed going from high to low from one visit to the next. Most likely, this high STAT1 is not from some allele defect. One short coming that may be considered is that the total blood lymphocytes were collected. For instance, it is common for neutrophils population to expand in SLE patients. The high levels of STAT1 could be an anomaly due to a change in cell population. This may be true, but since the high STAT1levels were also observed in healthy donors who are known not to have any autoimmune disorder due to screening, the change in cell population would seem very unlikely to also occur in healthy individuals. High STAT1 levels did not seem to be race dependent. What drives STAT1 levels to such extremely high levels is unclear but appear to influence CCL2 and CXCL10 expression. CCL2 and CXCL10 with elevated in high STAT1 patients; furthermore, both were also elevated in the high STAT1 healthy donors. It is unclear why STAT1 was elevated in the healthy donors even though their IFN score were in the normal range, but it appears that this elevated STAT1 may on i ts own not be pathogenic. However in the presence of IFN I, CCL2 and CXCL10 enhanced covariation with IFN score in high STAT1 patients than in low STAT1
131 patients. We hypothesize that high STAT1 may enhance CCL2 and CXCL10 response to IFN 1 and may act as indicator of increased activity. In addition STAT1, may play a role in how patients respond to therapy. High STAT1 Negates Therapy STAT1 levels appear to play a role in how therapy affects CCL2 and CXCL10 levels. CCL2 and CXCL10 were decreased in treated patients compared to untreated patients. When the treated patients were separated into high and low STAT1 patients, CCL2 and CXCL10 were significantly higher in the high STAT1 patients compared to the low STAT1 patients. When untreated patients were separated in to high and low STAT1, there was no significant difference in CCL2 and CXCL10 levels. CCL2 and CXCL10 were higher in the high STAT1 healthy donors than the low STAT1 healthy donors; however, the high STAT1 healthy donors are still lower than the low STAT1 SLE patients. This most likely indicates that there is some sort of mechanism in healthy donors that inhibits CCL2 and CXCL10 to increase out of control. The missing factor may be interferon. Since the healthy donors display normal levels of interfe ron, CCL2 and CXCL10 are not upregulated, but in patients, CCL2 and CXCL10 continue to climb in response to interferon and more so in the patients that have high levels of STAT1. This was observed when the slopes of CCL2/IFN score and CXCL10/IFN score wer e compared. In CCL2 and CXCL10, the slope of high STAT1 treated patients resembled that of the untreated patients while the low STAT1 slopes of the treated where much lower. This leads to the interesting possibility that high STAT1 patients are not affected as much by therapy due to how they resemble untreated patients. Could high STAT1 levels indicate how well a patient will respond to therapy? When considering the results of the therapy dose, something somewhat unexpected occurs. If
132 dose increases, we would expect the expression of CCL2 and CXCL10 to decrease or at least stay the same; however, this is not what happens. As dose increases, CCL2 and CXCL10 appear to increase as well. The reason is that higher doses of therapy are prescribed to patients who are severely ill. The more ill the patient, the higher the dose and possibly more medications are given. Indicating that dose was an indicator of disease activity. It is interesting to see the trends of the low and high STAT1 patients by dose. It appears t hat the high STAT1 patients express more CCL2 and CXCL10 as the dose of the therapies increase; however, the low STAT1 patients do not change as much. This may indicate that high STAT1 may be indicator of increased activity or prelude to increase activity in patients. In either case, high levels of STAT1 appear to facilitate CCL2 and CXCL10 upregulation; therefore, circumventing the effects of therapy. How STAT1 circumvents CCL2 and CXCL10 being downregulated by therapy is unclear. It is possible that the h igh levels of STAT1 enhance interferon's upregulation of CCL2 and CXCL10 that the therapy is unable to compensate and which leads to increase of dosage in response of the patient becoming more ill. STAT1 and M i croRNA 146a Promote Anemia One symptom of increased or prolonged activity of SLE is anemia. Anemia comes in many flavors ranging for hemolytic anemia to therapy induced anemia. Due to lack of clear classifications, anemia could not be stratified into different categories. For practi cal purposes, anemia would was kept as a single category. Even if each type of anemia was known, sample size would have become a critical issue. Sample size limitations are a common problem with SLE due to the heterogeneity of the disease. The anemic pat ients appear to have higher IFN score, STAT1, CCL2, CXCL10, and miR 146a than patients without anemia. Prevailing paradigm is that anemia in SLE
133 primary arises from prolong disease activity. When observing the role of SLEDAI and lupus nephritis, there was an association between how many patients were anemic compared to how many were active or had lupus nephritis; however, this did not seem to reflect upon IFN score, CCL2, and CXCL10 no clear indication that activity or lupus nephrites influenced anemia, but both STAT1 and miR 146a were elevated in anemic patients regardless of SLEDAI and lupus nephritis. This seems to be true when carried with race and therapy. Again STAT1 seems to appear to have some role in anemia. STAT1 an important signaling molecule and transcription factor for interferon that is also inducible by interferon has also been implicated in anemia. In a zebra fish model for hematopoiesis, STAT1 expression, in particular isoform STAT1b, which both isoforms (long form STAT1a and short form S TAT1b) are conserve with between humans and zebra fish, was required for hematopoiesis and that increase in STAT1b was associated with increase in myeloid lineage but a decrease in erythroid lineage [ 229 ] Furthermore in mouse RAW264.7 macrophage cells, STAT1 and IFN B induced hepcidin expression. Potentially in SLE patients, STAT1 is reducing hematopoiesis and in addition induce hepcidin in monocytes/macrophages. In part, this is what confirmed in Fanconi anemia. Fanconi gene FANCC has been shown to bind STAT1 and causing it to be phosphorylated in the bone marrow leading to a depletion of haemopoietic progenitors. Anemia is a serious complication of SLE, but what if anemia is can exacerbate SLE? All the biomarkers were overexpressed in the anemic SLE patients which may be an indicator that anemia is making the patient more ill. STAT1 appears to be associated with disease activity in this study. Anemia has previously been demonstrated
134 to be a good predictor of activity for patients [ 221] which may indicate that anemia is occurring before increase in SLE activity. Additionally, low oxygen conditions such as hypoxia promote neutrophils, monocytes, and macrophage survival [ 223 224] Furthermore, hypoxia enhances macrophages and dendritic cell proinflammatory response and directly activates monocytes [ 225, 226, 235 ] Pro inflammatory transcription factors such as NF 1 are activated hypoxic conditions [ 225 226 ] Hypoxia activation of monocytes, NF 1 may explain the elevated levels of miR 146a. The increase of miR 146a may promote anemia by decreasing erythropoiesis. Since a link existed between hypoxia, anemia, and inflammation via hepcidin [ 228 ] it is possibl e that, hypoxia and anemia are part of the immune response to enhance inflammation to resolve systemic infection. Whether anemia exacerbates SLE pathogenesis is unclear. It is interesting to note that both anemia and SLE have higher occurrences in women t han males. Could there be a common link such as STAT1 and miR 146a? Could anemia lead to increased activity in SLE by enhancing inflammation? Whether anemia exacerbates SLE pathogenesis and STAT as well as miR 146a will require further studies to resolve t heir role. STAT1 Enhances Pathogenesis STAT1 may be the key factor connecting the elevated biomarker expression, lack of downregulation of CCL2 and CXCL10 by therapy, and the association of anemia. The relationships among IFN I, STAT1, CCL2, CXCL10, and miR 146a that were mapped with Ingenuity Pathway Analysis based upon the known published literature (Figure 51, blue lines). STAT1 directly involved with CCL2 and CXCL10 (Figure 51, blue solid lines) as well as indirectly with IFN I (Figure 5 1, blue dotted lines). As seen by our data and the literature, IFN 1, CCL2, and CXCL10 are involved in the pathogenesis of SLE.
135 STAT1 is at the apex of the pathway and as our data indicates directly related to SLE (Figure 51, orange). Based upon our findings, we purpose the following model (Figure 5 2). Several genetic and environmental factors have been implicated (Figure 52). There is strong evidence that toll like receptors (TLRs) are involved in the activation of IFN I which in turn upregulates STAT1 express ion (Figure 52). Our own data, demonstrated that TLR 4 activation by LPS could lead to upregulation of IFN I and of STAT1. Therefore, TLRs could activate STAT1 directly or indirectly (Figure 52). STAT1 has two potential pathogenic avenues by which it can affect a patient. One the increase of STAT1 can induce anemia (Figure 52). The second STAT1 enables and enhances CCL2 and CXCL10 production which in turn ignite inflammation at the source of the autoimmune response (Figure 5 2). Inflammation itself can induce anemia by upregulation of such factors as hepcidin or as we purpose via miR 146a (Figure 52). The combination of STAT1 and miR 146a causes the decrease erythropoiesis which promotes anemia and an increase of hematopoiesis. Anemia leads to decreas e of oxygen and possibly hypoxia at the site of inflammation which potentially prime if not cause self activation of lymphocytes; therefore, we hypothesize that anemia promotes and enhances inflammation in SLE (Figure 5 2). Therapy led to a decrease of C CL2 and CXCL10 (Figure 5 2). Similarly, we had observed that PDN led to a decrease in TNF. However, therapy did not appear to significantly impact IFN score or STAT1. This leads us to conclude that therapy could be impacting inflammation rather than IFN I production (Figure 52). This is exemplified in high STAT1 patients. High STAT1 patients appeared to be resistant to therapy
136 (Figure 52). In high STAT1 patients, therapy is unable to affect CCL2 and CXCL10 because STAT1 levels are somehow compensating or overcoming the effects of therapy allows CCL2 and CXCL10 expression to increase as interferon increases. Taken together, we present a model that may explain in part how therapy functions in SLE patients. Furthermore, this model yields incite that interferon role in SLE may be assisted and enhanced by STAT1. This enhancement comes in the form of increased production of CCL2 and CXCL10, decreased effectiveness of therapy to regulate CCL2 and CXCL10, and finally by leading to the development of anemia which in itself may play a role in enhancing inflammation. Future Objectives Complexity of SLE with the added nuance of a patients manifesting different manifestation of SLE from one flare to the next requires a study to have sufficient number of patients to acc urately observe outcomes. In the current study we observed the expression of STAT1 in SLE patients. To better understand the role of STAT1, the subsequent study would observe not only the expression but as well protein levels and phosphorylated levels of S TAT1. Since two isoforms of STAT1 exist and the short form (STAT1b) possibly plays a role in anemia, both isoforms of STAT1 should also be observed. However since STAT1b is identical to the portion of STAT1a (long form), it is not possible to directly observe STAT1b expression of protein level. An assay either for expression or protein would have implore probes specific for a homologous region for both STAT1a and STAT1b as well as region specifically for STAT1a that is not found in STAT1b. By taking a ratio n of [STAT1a+STAT1b]:[STAT1a], the level of STAT1b could be inferred. Furthermore, CCL2 and CXCL10 will also be observed to determine their
137 relationship with STAT1. In addition to observing expression of CCL2 and CXCL10, protein levels in the serum will al so observed. Another important objective will be to understand how STAT1 levels influence the patients response to therapy. If as in some cancers STAT1 confers drug resistance, this would be a very important aspect for how therapy is used in patients. I n addition to collecting the medication patients are receiving, other aspects relating to disease such autoantibody titiers, SLEDAI, other clinical information will be collected. CCL2 and CXCL10 expression and protein levels will be correlated to STAT1 lev els and use of therapy to understand their relationships. In addition to observing STAT1, other issues need to be addressed. The current study observed total peripheral blood leukocytes. Even though this is a somewhat common practice, it overlooks the potential contribution of individual cell types as well the possible skewing that may result from cell population differences due to activity. For instance, it is common for neutrophils to expand in SLE patients especially those that are active. Due to limit ations of the amount of blood that can be collected, the study would focus on cell types that are relatively abundant such monocytes and neutrophils. A pilot study will be needed to determine if sufficient number of B and T cells can be collected for analy sis. The future study will also need to address sample size to be collected. The current study collected 102 SLE patients with 181 visits. The analysis focused on visits rather than individual patients due to statistic limitations. The contribution of p atients with multiple visits is a disconcerting due to potential bias contribution of a patient. In the future study at least 180 patients will be required based on the results of the current
138 study. In addition, a longitudinal study to follow patients ove r multiple visits would be another objective. Most patients visits are every 3 to 6 months assuming that they donate at every other visit, it potentially will 3 years to collect 3 consecutive visits. Multiple visits will be important to understand how STA T1, CCL2, CXCL10, and IFN I change over time as well as to be able to take into account how therapy over time may change the expression of these markers. An emphasis on cataloging each patients anemia status as well as anemia type and duration would be another of the main objectives. This will be critical to validate the current studys observations for STAT1 and miR 146a elevations in anemic patients. A particular interest will to identify patients that receive treatments to improve their anemia particul arly if resolving anemia would The objectives above should address the weaknesses in the current study and expand the knowledge base by observing the role of STAT1 in multiple aspects such isoform levels and phos phorylation status in SLE in multiple cell types. In addition to determining if STAT1 levels confer drug resistance or reduced response to therapy would be an important contribution to the treatment of SLE. It is important to note that there is not cure for SLE and that all current therapies focus on the treatment of the symptoms which essentially allowing a patient to better cope with SLE and experience life as close to normal as possible.
139 Figure 51 Relationship of STAT1, CCL2, CXCL10, miR 146a, a nd IFN 1 with SLE. The relationships among STAT1, CCL2, CXCL10, miR 146a, and IFN 1 were mapped with Ingenuity Pathway Analysis based upon known literature. Solid blue lines represent direct relationships while dotted blue lines represent indirect relationships. Orange line represents the contribution of the current study.
140 Figure 52 STAT1 role in SLE. The summary of the current work demonstrates the relationships that were identified in our study and how they are influencing pathogenesis of SLE.
141 LIST OF REFERENCES 1. Reeves WH, Narain S, Satoh M (2004) Autoantibodies in systemic lupus erythematosus. In: Koopman WJ, Moreland LW, editors. Arthritis and Allied Conditions: A Textbook of Rheumatology. Philadelphia, PA: Lippincott Williams & Wilkins. 2. Lim SS, Drenkard C, McCune WJ, Helmick CG, Gordon C, et al. (2009) Populationbased lupus registries: advancing our epidemiologic understanding. Arthritis Rheum 61: 14621466. 3. Hanh BH (2005) Systemic Lupus Erythematosus. ; Kasper DL, editor. New York City: McGraw Hill. 2750 p. 4. Pisetsky DS (2008) Systemic Lupus Erythematosus B. Epidemiology, Pathology, and Pathogenesis; Klippel JH, Crofford LJ, Stone JH, editors. Atlanta, GA: Spri nger. 721 p. 5. Hahn BH (2012) The Pathogenesis of SLE. In: Wallace D, Hahn BH, editors. Dubois' Lupus Erythematosus and Related Syndromes: Expert Consult Online. 8 ed: Elsevier Health Sciences. 6. Deapen D, Escalante A, Weinrib L, Horwitz D, Bachman B et al. (1992) A revised estimate of twin concordance in systemic lupus erythematosus. Arthritis Rheum 35: 311 318. 7. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, et al. (1982) The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 25: 12711277. 8. Hochberg MC (1997) Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 40: 1725. 9. Jolly M, Francis S, Sequeira W (2012) Differ ential Diagnosis and Disease Associations. In: Wallace D, Hahn BH, editors. Dubois' Lupus Erythematosus and Related Syndromes: Expert Consult Online. 8 ed: Elsevier Health Sciences. 10. Arbuckle MR, McClain MT, Rubertone MV, Scofield RH, Dennis GJ, et a l. (2003) Development of autoantibodies before the clinical onset of systemic lupus erythematosus. N Engl J Med 349: 15261533. 11. Giannouli S, Voulgarelis M, Ziakas PD, Tzioufas AG (2006) Anaemia in systemic lupus erythematosus: from pathophysiology to clinical assessment. Ann Rheum Dis 65: 144148.
142 12. de Zubiria Salgado A, Herrera Diaz C (2012) Lupus nephritis: an overview of recent findings. Autoimmune Dis 2012: 849684. 13. Nagy G, Koncz A, Perl A (2005) T and B cell abnormalities in systemic lupus erythematosus. Crit Rev Immunol 25: 123140. 14. Moulton VR, Tsokos GC (2011) Abnormalities of T cell signaling in systemic lupus erythematosus. Arthritis Res Ther 13: 207. 15. Dorner T, Giesecke C, Lipsky PE (2011) Mechanisms of B cell autoimmunity in S LE. Arthritis Res Ther 13: 243. 16. Dolff S, Abdulahad WH, Bijl M, Kallenberg CG (2009) Regulators of B cell activity in SLE: a better target for treatment than B cell depletion? Lupus 18: 575 580. 17. Jacob N, Stohl W (2011) Cytokine disturbances in systemic lupus erythematosus. Arthritis Res Ther 13: 228. 18. Cross JT, Benton HP (1999) The roles of interleukin 6 and interleukin10 in B cell hyperactivity in systemic lupus erythematosus. Inflamm Res 48: 255261. 19. Calvani N, Tucci M, Richards HB, Tartaglia P, Silvestris F (2005) Th1 cytokines in the pathogenesis of lupus nephritis: the role of IL18. Autoimmun Rev 4: 542548. 20. Santiago Raber ML, Baudino L, Izui S (2009) Emerging roles of TLR7 and TLR9 in murine SLE. J Autoimmun 33: 231238. 21. Rahman AH, Eisenberg RA (2006) The role of toll like receptors in systemic lupus erythematosus. Springer Semin Immunopathol 28: 131 143. 22. Skiljevic D, Jeremic I, Nikolic M, Andrejevic S, SefikBukilica M, et al. (2012) Serum DNase I activity in syste mic lupus erythematosus: correlation with immunoserological markers, the disease activity and organ involvement. Clin Chem Lab Med: 19. 23. Ahn J, Gutman D, Saijo S, Barber GN (2012) STING manifests self DNA dependent inflammatory disease. Proc Natl Acad Sci U S A 109: 1938619391. 24. Shin HD, Park BL, Cheong HS, Lee HS, Jun JB, et al. (2005) DNase II polymorphisms associated with risk of renal disorder among systemic lupus erythematosus patients. J Hum Genet 50: 107 111. 25. Dall'era M, Chakravarty EF (2011) Treatment of mild, moderate, and severe lupus erythematosus: focus on new therapies. Curr Rheumatol Rep 13: 308316.
143 26. Chaturvedi A, Pierce SK (2009) How location governs toll like receptor signaling. Traffic 10: 621 628. 27. Takeda K, Kaisho T Akira S (2003) Toll like receptors. Annu Rev Immunol 21: 335 376. 28. Guiducci C, Gong M, Xu Z, Gill M, Chaussabel D, et al. (2010) TLR recognition of self nucleic acids hampers glucocorticoid activity in lupus. Nature 465: 937 941. 29. Administration FaD (2010) Guidance for industry: systemic lupus erythematosus developing medical products for treatment. 30. Fulmer T (2010) Sparing steroids in lupus. Science Business eXchange 3: 1. 31. Herbst R, Liu Z, Jallal B, Yao Y (2012) Biomarkers for systemic lupus erythematosus. Int J Rheum Dis 15: 433444. 32. Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, et al. (2003) Interferon inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc Natl Acad Sci U S A 100: 2610 2615. 33. Bennett L, Palucka AK, Arce E, Cantrell V, Borvak J, et al. (2003) Interferon and granulopoiesis signatures in systemic lupus erythematosus blood. J Exp Med 197: 711 723. 34. Crow MK (2005) Interferon pathway activation in systemic lupus erythematosus. Curr Rheumatol Rep 7: 463468. 35. Preble OT, Black RJ, Friedman RM, Klippel JH, Vilcek J (1982) Systemic lupus erythematosus: presence in human serum of an unusual acidlabile leukocyte interferon. Science 216: 429431. 36. Hardy MP, Owczarek CM, Jermiin LS, Ejdeback M, Hertzog PJ (2004) Characterization of the type I interferon locus and identification of novel genes. Genomics 84: 331 345. 37. Fountain JW, Karayiorgou M, Taruscio D, Graw SL, Buckler AJ, et al. (1992) Genetic and physical map of the interferon region on chromosome 9p. Genomics 14: 105 112. 38. Platanias LC (2005) Mechanisms of typeI and typeII interferon mediated signalling. Nat Rev Immunol 5: 375386. 39. Takaoka A, Yanai H (2006) Interferon signalling netw ork in innate defence. Cell Microbiol 8: 907922.
144 40. Stark GR, Darnell JE, Jr. (2012) The JAK STAT pathway at twenty. Immunity 36: 503514. 41. PalssonMcDermott EM, O'Neill LA (2004) Signal transduction by the lipopolysaccharide receptor, Toll like rece ptor 4. Immunology 113: 153162. 42. Uematsu S, Akira S (2007) Toll like receptors and Type I interferons. J Biol Chem 282: 1531915323. 43. Baccala R, Hoebe K, Kono DH, Beutler B, Theofilopoulos AN (2007) TLR dependent and TLR independent pathways of ty pe I interferon induction in systemic autoimmunity. Nat Med 13: 543551. 44. Barber GN (2011) Cytoplasmic DNA innate immune pathways. Immunol Rev 243: 99108. 45. Furr SR, Chauhan VS, Moerdyk Schauwecker MJ, Marriott I (2011) A role for DNA dependent activator of interferon regulatory factor in the recognition of herpes simplex virus type 1 by glial cells. J Neuroinflammation 8: 99. 46. Ishikawa H, Ma Z, Barber GN (2009) STING regulates intracellular DNA mediated, type I interferon dependent innate im munity. Nature 461: 788 792. 47. Parker D, Martin FJ, Soong G, Harfenist BS, Aguilar JL, et al. (2011) Streptococcus pneumoniae DNA initiates type I interferon signaling in the respiratory tract. MBio 2: e0001600011. 48. Parker D, Prince A (2011) Type I interferon response to extracellular bacteria in the airway epithelium. Trends Immunol 32: 582588. 49. Ioannou Y, Isenberg DA (2000) Current evidence for the induction of autoimmune rheumatic manifestations by cytokine therapy. Arthritis Rheum 43: 14311442. 50. Ronnblom LE, Alm GV, Oberg KE (1991) Autoimmunity after alphainterferon therapy for malignant carcinoid tumors. Ann Intern Med 115: 178 183. 51. Wilson LE, Widman D, Dikman SH, Gorevic PD (2002) Autoimmune disease complicating antiviral t herapy for hepatitis C virus infection. Semin Arthritis Rheum 32: 163173. 52. Kalkner KM, Ronnblom L, Karlsson Parra AK, Bengtsson M, Olsson Y, et al. (1998) Antibodies against double stranded DNA and development of polymyositis during treatment with int erferon. QJM 91: 393399.
145 53. Hooks JJ, Moutsopoulos HM, Geis SA, Stahl NI, Decker JL, et al. (1979) Immune interferon in the circulation of patients with autoimmune disease. N Engl J Med 301: 58. 54. Nikpour M, Dempsey AA, Urowitz MB, Gladman DD, Barnes DA (2008) Association of a gene expression profile from whole blood with disease activity in systemic lupus erythaematosus. Ann Rheum Dis 67: 10691075. 55. Crow MK, Kirou KA, Wohlgemuth J (2003) Microarray analysis of interferonregulated genes in SLE. Autoimmunity 36: 481490. 56. Kirou KA, Lee C, George S, Louca K, Papagiannis IG, et al. (2004) Coordinate overexpression of interferon alpha induced genes in systemic lupus erythematosus. Arthritis Rheum 50: 39583967. 57. Zhuang H, Narain S, Sobel E, Lee PY, Nacionales DC, et al. (2005) Association of anti nucleoprotein autoantibodies with upregulation of Type I interferoninducible gene transcripts and dendritic cell maturation in systemic lupus erythematosus. Clin Immunol 117: 238250. 58. Kirou KA, Lee C, George S, Louca K, Peterson MG, et al. (2005) Activation of the interferonalpha pathway identifies a subgroup of systemic lupus erythematosus patients with distinct serologic features and active disease. Arthritis Rheum 52: 14911503. 59. Vallin H, Perers A, Alm GV, Ronnblom L (1999) Anti double stranded DNA antibodies and immunostimulatory plasmid DNA in combination mimic the endogenous IFN alpha inducer in systemic lupus erythematosus. J Immunol 163: 63066313. 60. Means TK, Latz E, Hayashi F, Murali MR, Golenbock DT, et al. (2005) Human lupus autoantibody DNA complexes activate DCs through cooperation of CD32 and TLR9. J Clin Invest 115: 407 417. 61. Barrat FJ, Meeker T, Gregorio J, Chan JH, Uematsu S, et al. (2005) Nucleic acids of mammalian origin can act as endogenous ligands for Toll like receptors and may promote systemic lupus erythematosus. J Exp Med 202: 11311139. 62. Nacionales DC, Kelly KM, Lee PY, Zhuang H, Li Y, et al. (2006) Type I interferon production by tertiary lymphoid tissue developing in response to 2,6,10,14tetramethyl pentadecane (pristane). Am J Pathol 168: 12271240. 63. Savarese E, Chae OW, Trowitzsch S, Weber G, Kastner B, et al. (2006) U1 small nuclear ribonucleoprotein immune complexes induce type I interferon in plasmacytoid dendritic cells through TLR7. Blood 107: 3229 3234.
146 64. Bave U, Magnusson M, Eloranta ML, Perers A, Alm GV, et al. (2003) Fc gamma RIIa is expressed on natural IFN alpha producing cells (plasmacytoid dendritic cells) and is required for the IF N alpha production induced by apoptotic cells combined with lupus IgG. J Immunol 171: 32963302. 65. Marshak Rothstein A (2006) Toll like receptors in systemic autoimmune disease. Nat Rev Immunol 6: 823835. 66. Ronnblom L, Eloranta ML, Alm GV (2006) The type I interferon system in systemic lupus erythematosus. Arthritis Rheum 54: 408 420. 67. Remijsen Q, Kuijpers TW, Wirawan E, Lippens S, Vandenabeele P, et al. (2011) Dying for a cause: NETosis, mechanisms behind an antimicrobial cell death modality. Ce ll Death Differ 18: 581588. 68. Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, et al. (2004) Neutrophil extracellular traps kill bacteria. Science 303: 1532 1535. 69. Lee WL, Grinstein S (2004) Immunology. The tangled webs that neutrophils w eave. Science 303: 14771478. 70. Leavy O (2012) Innate immunity: Multitasking NET makers. Nat Rev Immunol 12: 684685. 71. Peschel A, Hartl D (2012) Anuclear neutrophils keep hunting. Nat Med 18: 13361338. 72. Yipp BG, Petri B, Salina D, Jenne CN, Scott BN, et al. (2012) Infection induced NETosis is a dynamic process involving neutrophil multitasking in vivo. Nat Med 18: 1386 1393. 73. Zhang Q, Itagaki K, Hauser CJ (2010) Mitochondrial DNA is released by shock and activates neutrophils via p38 map kinase. Shock 34: 5559. 74. Knight JS, Kaplan MJ (2012) Lupus neutrophils: 'NET' gain in understanding lupus pathogenesis. Curr Opin Rheumatol 24: 441450. 75. Lande R, Ganguly D, Facchinetti V, Frasca L, Conrad C, et al. (2011) Neutrophils activate plasmacytoid dendritic cells by releasing self DNA peptide complexes in systemic lupus erythematosus. Sci Transl Med 3: 73ra19. 76. Garcia Romo GS, Caielli S, Vega B, Connolly J, Allantaz F, et al. (2011) Netting neutrophils are major inducers of type I I FN production in pediatric systemic lupus erythematosus. Sci Transl Med 3: 73ra20.
147 77. Villanueva E, Yalavarthi S, Berthier CC, Hodgin JB, Khandpur R, et al. (2011) Netting neutrophils induce endothelial damage, infiltrate tissues, and expose immunostimul atory molecules in systemic lupus erythematosus. J Immunol 187: 538552. 78. Hakkim A, Furnrohr BG, Amann K, Laube B, Abed UA, et al. (2010) Impairment of neutrophil extracellular trap degradation is associated with lupus nephritis. Proc Natl Acad Sci U S A 107: 9813 9818. 79. Bosch X (2011) TRAPping the cellular mechanisms of lupus. EMBO Mol Med 3: 578580. 80. Kakumanu P, Sobel ES, Narain S, Li Y, Akaogi J, et al. (2009) Citrulline dependence of anti cyclic citrullinated peptide antibodies in systemic lupus erythematosus as a marker of deforming/erosive arthritis. J Rheumatol 36: 26822690. 81. Qing YF, Zhang QB, Zhou JG, Yuan GH, Wei J, et al. (2009) The detecting and clinical value of anti cyclic citrullinated peptide antibodies in patients with syst emic lupus erythematosus. Lupus 18: 713 717. 82. Ahearn JM, Liu CC, Kao AH, Manzi S (2012) Biomarkers for systemic lupus erythematosus. Transl Res 159: 326 342. 83. Bass BL, Weintraub H (1988) An unwinding activity that covalently modifies its doublestr anded RNA substrate. Cell 55: 1089 1098. 84. Wagner RW, Smith JE, Cooperman BS, Nishikura K (1989) A doublestranded RNA unwinding activity introduces structural alterations by means of adenosine to inosine conversions in mammalian cells and Xenopus eggs. Proc Natl Acad Sci U S A 86: 26472651. 85. Nishikura K (2010) Functions and regulation of RNA editing by ADAR deaminases. Annu Rev Biochem 79: 321349. 86. George CX, Gan Z, Liu Y, Samuel CE (2011) Adenosine deaminases acting on RNA, RNA editing, and i nterferon action. J Interferon Cytokine Res 31: 99117. 87. Patterson JB, Samuel CE (1995) Expression and regulation by interferon of a doublestrandedRNA specific adenosine deaminase from human cells: evidence for two forms of the deaminase. Mol Cell Bi ol 15: 5376 5388. 88. Hartner JC, Schmittwolf C, Kispert A, Muller AM, Higuchi M, et al. (2004) Liver disintegration in the mouse embryo caused by deficiency in the RNA editing enzyme ADAR1. J Biol Chem 279: 48944902.
148 89. Wang Q, Miyakoda M, Yang W, Khillan J, Stachura DL, et al. (2004) Stress induced apoptosis associated with null mutation of ADAR1 RNA editing deaminase gene. J Biol Chem 279: 4952 4961. 90. Wang Q, Khillan J, Gadue P, Nishikura K (2000) Requirement of the RNA editing deaminase ADAR1 gene for embryonic erythropoiesis. Science 290: 17651768. 91. XuFeng R, Boyer MJ, Shen H, Li Y, Yu H, et al. (2009) ADAR1 is required for hematopoietic progenitor cell survival via RNA editing. Proc Natl Acad Sci U S A 106: 1776317768. 92. Higuchi M, Maas S, Single FN, Hartner J, Rozov A, et al. (2000) Point mutation in an AMPA receptor gene rescues lethality in mice deficient in the RNA editing enzyme ADAR2. Nature 406: 7881. 93. Hartner JC, Walkley CR, Lu J, Orkin SH (2009) ADAR1 is essential for t he maintenance of hematopoiesis and suppression of interferon signaling. Nat Immunol 10: 109115. 94. Vitali P, Scadden AD (2010) Doublestranded RNAs containing multiple IU pairs are sufficient to suppress interferon induction and apoptosis. Nat Struct M ol Biol 17: 10431050. 95. Nie Y, Hammond GL, Yang JH (2007) Double stranded RNA deaminase ADAR1 increases host susceptibility to virus infection. J Virol 81: 917 923. 96. Toth AM, Li Z, Cattaneo R, Samuel CE (2009) RNA specific adenosine deaminase ADAR1 suppresses measles virus induced apoptosis and activation of protein kinase PKR. J Biol Chem 284: 29350 29356. 97. Li Z, Wolff KC, Samuel CE (2010) RNA adenosine deaminase ADAR1 deficiency leads to increased activation of protein kinase PKR and reduced v esicular stomatitis virus growth following interferon treatment. Virology 396: 316 322. 98. Laxminarayana D, Khan IU, Kammer G (2002) Transcript mutations of the alpha regulatory subunit of protein kinase A and upregulation of the RNA editing gene transc ript in lupus T lymphocytes. Lancet 360: 842849. 99. Laxminarayana D, O'Rourke KS, Maas S, Olorenshaw I (2007) Altered editing in RNA editing adenosine deaminase ADAR2 gene transcripts of systemic lupus erythematosus T lymphocytes. Immunology 121: 359 369. 100. Orlowski RJ, O'Rourke KS, Olorenshaw I, Hawkins GA, Maas S, et al. (2008) Altered editing in cyclic nucleotide phosphodiesterase 8A1 gene transcripts of systemic lupus erythematosus T lymphocytes. Immunology 125: 408 419.
149 101. Lee Y, Ahn C, Han J Choi H, Kim J, et al. (2003) The nuclear RNase III Drosha initiates microRNA processing. Nature 425: 415419. 102. Berezikov E, Chung WJ, Willis J, Cuppen E, Lai EC (2007) Mammalian mirtron genes. Mol Cell 28: 328336. 103. Okamura K, Hagen JW, Duan H, Tyler DM, Lai EC (2007) The mirtron pathway generates microRNA class regulatory RNAs in Drosophila. Cell 130: 89100. 104. Ruby JG, Jan CH, Bartel DP (2007) Intronic microRNA precursors that bypass Drosha processing. Nature 448: 8386. 105. Miyoshi K, T sukumo H, Nagami T, Siomi H, Siomi MC (2005) Slicer function of Drosophila Argonautes and its involvement in RISC formation. Genes Dev 19: 28372848. 106. Tolia NH, Joshua Tor L (2007) Slicer and the argonautes. Nat Chem Biol 3: 36 43. 107. Peters L, Meis ter G (2007) Argonaute proteins: mediators of RNA silencing. Mol Cell 26: 611 623. 108. Filipowicz W, Bhattacharyya SN, Sonenberg N (2008) Mechanisms of post transcriptional regulation by microRNAs: are the answers in sight? Nat Rev Genet 9: 102114. 109. Friedman RC, Farh KK, Burge CB, Bartel DP (2009) Most mammalian mRNAs are conserved targets of microRNAs. Genome Res 19: 92105. 110. Doench JG, Sharp PA (2004) Specificity of microRNA target selection in translational repression. Genes Dev 18: 5045 11. 111. Grimson A, Farh KK, Johnston WK, Garrett Engele P, Lim LP, et al. (2007) MicroRNA targeting specificity in mammals: determinants beyond seed pairing. Mol Cell 27: 91105. 112. Lewis BP, Burge CB, Bartel DP (2005) Conserved seed pairing, often fl anked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120: 15 20. 113. Nielsen CB, Shomron N, Sandberg R, Hornstein E, Kitzman J, et al. (2007) Determinants of targeting by endogenous and exogenous microRNAs and siRNAs. RNA 13: 18941910. 114. Taganov KD, Boldin MP, Chang KJ, Baltimore D (2006) NF kappaB dependent induction of microRNA miR 146, an inhibitor targeted to signaling proteins of innate immune responses. Proc Natl Acad Sci U S A 103: 12481 12486.
150 115. Perry MM, Moschos SA, Williams AE, Shepherd NJ, Larner Svensson HM, et al. (2008) Rapid changes in microRNA 146a expression negatively regulate the IL 1betainduced inflammatory response in human lung alveolar epithelial cells. J Immunol 180: 56895698. 116. Nahid MA, Rivera M, Lucas A, Chan EK, Kesavalu L (2011) Polymicrobial infection with periodontal pathogens specifically enhances microRNA miR 146a in ApoE / mice during experimental periodontal disease. Infect Immun 79: 15971605. 117. Nahid MA, Satoh M, Chan EK (2011) Mechanistic role of microRNA 146a in endotoxin induced differential cross regulation of TLR signaling. J Immunol 186: 17231734. 118. Nahid MA, Pauley KM, Satoh M, Chan EK (2009) miR 146a is critical for endotoxin induced tolerance: IMPLICATION IN INNATE IMMUNITY. J Biol Chem 284: 3459034599. 119. Sonkoly E, Stahle M, Pivarcsi A (2008) MicroRNAs: novel regulators in skin inflammation. Clin Exp Dermatol 33: 312 315. 120. Sonkoly E, Wei T, Janson PC, Saaf A, Lundeberg L, et al. (2007) MicroRNAs : novel regulators involved in the pathogenesis of psoriasis? PLoS One 2: e610. 121. Nakasa T, Miyaki S, Okubo A, Hashimoto M, Nishida K, et al. (2008) Expression of microRNA 146 in rheumatoid arthritis synovial tissue. Arthritis Rheum 58: 12841292. 122. Pauley KM, Satoh M, Chan AL, Bubb MR, Reeves WH, et al. (2008) Upregulated miR 146a expression in peripheral blood mononuclear cells from rheumatoid arthritis patients. Arthritis Res Ther 10: R101. 123. Stanczyk J, Pedrioli DM, Brentano F, Sanchez Perna ute O, Kolling C, et al. (2008) Altered expression of MicroRNA in synovial fibroblasts and synovial tissue in rheumatoid arthritis. Arthritis Rheum 58: 10011009. 124. Jazdzewski K, de la Chapelle A (2009) Genomic sequence matters: a SNP in microRNA 146a can turn anti apoptotic. Cell Cycle 8: 1642 1643. 125. Jazdzewski K, Murray EL, Franssila K, Jarzab B, Schoenberg DR, et al. (2008) Common SNP in premiR 146a decreases mature miR expression and predisposes to papillary thyroid carcinoma. Proc Natl Acad S ci U S A 105: 7269 7274. 126. Xu T, Zhu Y, Wei QK, Yuan Y, Zhou F, et al. (2008) A functional polymorphism in the miR 146a gene is associated with the risk for hepatocellular carcinoma. Carcinogenesis 29: 21262131.
151 127. Abelson JF, Kwan KY, O'Roak BJ, Baek DY, Stillman AA, et al. (2005) Sequence variants in SLITRK1 are associated with Tourette's syndrome. Science 310: 317320. 128. Sandberg R, Neilson JR, Sarma A, Sharp PA, Burge CB (2008) Proliferating cells express mRNAs with shortened 3' untranslated regions and fewer microRNA target sites. Science 320: 16431647. 129. Tang Y, Luo X, Cui H, Ni X, Yuan M, et al. (2009) MicroRNA 146A contributes to abnormal activation of the type I interferon pathway in human lupus by targeting the key signaling proteins. Arthritis Rheum 60: 10651075. 130. Ivashkiv LB (2003) Type I interferon modulation of cellular responses to cytokines and infectious pathogens: potential role in SLE pathogenesis. Autoimmunity 36: 473479. 131. Hu X, Herrero C, Li WP, Antoniv TT, Fal ck Pedersen E, et al. (2002) Sensitization of IFN gamma Jak STAT signaling during macrophage activation. Nat Immunol 3: 859866. 132. Doughty L, Nguyen K, Durbin J, Biron C (2001) A role for IFN alpha beta in virus infectioninduced sensitization to endotoxin. J Immunol 166: 26582664. 133. Durbin JE, Hackenmiller R, Simon MC, Levy DE (1996) Targeted disruption of the mouse Stat1 gene results in compromised innate immunity to viral disease. Cell 84: 443 450. 134. Thibault DL, Chu AD, Graham KL, Balboni I, Lee LY, et al. (2008) IRF9 and STAT1 are required for IgG autoantibody production and B cell expression of TLR7 in mice. J Clin Invest 118: 14171426. 135. Khodarev NN, Roizman B, Weichselbaum RR (2012) Molecular pathways: interferon/stat1 pathway: role in the tumor resistance to genotoxic stress and aggressive growth. Clin Cancer Res 18: 30153021. 136. Khodarev NN, Beckett M, Labay E, Darga T, Roizman B, et al. (2004) STAT1 is overexpressed in tumors selected for radioresistance and confers protection from radiation in transduced sensitive cells. Proc Natl Acad Sci U S A 101: 1714 1719. 137. Khodarev NN, Minn AJ, Efimova EV, Darga TE, Labay E, et al. (2007) Signal transducer and activator of transcription 1 regulates both cytotoxic and prosurviva l functions in tumor cells. Cancer Res 67: 92149220. 138. Kita K, Sugaya S, Zhai L, Wu YP, Wano C, et al. (2003) Involvement of LEU13 in interferoninduced refractoriness of human RSa cells to cell killing by X rays. Radiat Res 160: 302 308.
152 139. Rickardson L, Fryknas M, Dhar S, Lovborg H, Gullbo J, et al. (2005) Identification of molecular mechanisms for cellular drug resistance by combining drug activity and gene expression profiles. Br J Cancer 93: 483492. 140. Roberts D, Schick J, Conway S, Biade S, Laub PB, et al. (2005) Identification of genes associated with platinum drug sensitivity and resistance in human ovarian cancer cells. Br J Cancer 92: 11491158. 141. Bauer JW, Petri M, Batliwalla FM, Koeuth T, Wilson J, et al. (2009) Interferonregulate d chemokines as biomarkers of systemic lupus erythematosus disease activity: a validation study. Arthritis Rheum 60: 30983107. 142. Bauer JW, Baechler EC, Petri M, Batliwalla FM, Crawford D, et al. (2006) Elevated serum levels of interferonregulated chemokines are biomarkers for active human systemic lupus erythematosus. PLoS Med 3: e491. 143. Xu LL, Warren MK, Rose WL, Gong W, Wang JM (1996) Human recombinant monocyte chemotactic protein and other C C chemokines bind and induce directional migration of dendritic cells in vitro. J Leukoc Biol 60: 365371. 144. Carr MW, Roth SJ, Luther E, Rose SS, Springer TA (1994) Monocyte chemoattractant protein 1 acts as a T lymphocyte chemoattractant. Proc Natl Acad Sci U S A 91: 36523656. 145. Bischoff SC, Kriege r M, Brunner T, Dahinden CA (1992) Monocyte chemotactic protein 1 is a potent activator of human basophils. J Exp Med 175: 12711275. 146. Conti P, Boucher W, Letourneau R, Feliciani C, Reale M, et al. (1995) Monocyte chemotactic protein1 provokes mast cell aggregation and [3H]5HT release. Immunology 86: 434440. 147. Weber M, Uguccioni M, Baggiolini M, Clark Lewis I, Dahinden CA (1996) Deletion of the NH2terminal residue converts monocyte chemotactic protein 1 from an activator of basophil mediator re lease to an eosinophil chemoattractant. J Exp Med 183: 681 685. 148. Charo IF, Myers SJ, Herman A, Franci C, Connolly AJ, et al. (1994) Molecular cloning and functional expression of two monocyte chemoattractant protein 1 receptors reveals alternative spl icing of the carboxyl terminal tails. Proc Natl Acad Sci U S A 91: 27522756. 149. Power CA, Meyer A, Nemeth K, Bacon KB, Hoogewerf AJ, et al. (1995) Molecular cloning and functional expression of a novel CC chemokine receptor cDNA from a human basophilic cell line. J Biol Chem 270: 1949519500.
153 150. La Gruta NL, Kedzierska K, Stambas J, Doherty PC (2007) A question of self preservation: immunopathology in influenza virus infection. Immunol Cell Biol 85: 8592. 151. Marsche G, Semlitsch M, Hammer A, Frank S, Weigle B, et al. (2007) Hypochloritemodified albumin colocalizes with RAGE in the artery wall and promotes MCP 1 expression via the RAGE Erk1/2 MAP kinase pathway. FASEB J 21: 11451152. 152. Chen IY, Chang SC, Wu HY, Yu TC, Wei WC, et al. (2010) Up regulation of the chemokine (C C motif) ligand 2 via a severe acute respiratory syndrome coronavirus spikeACE2 signaling pathway. J Virol 84: 77037712. 153. Burysek L, Syrovets T, Simmet T (2002) The serine protease plasmin triggers expression of MCP 1 and CD40 in human primary monocytes via activation of p38 MAPK and janus kinase (JAK)/STAT signaling pathways. J Biol Chem 277: 3350933517. 154. Lloyd CM, Minto AW, Dorf ME, Proudfoot A, Wells TN, et al. (1997) RANTES and monocyte chemoattractant protein 1 (MCP 1) play an important role in the inflammatory phase of crescentic nephritis, but only MCP 1 is involved in crescent formation and interstitial fibrosis. J Exp Med 185: 13711380. 155. Sayyed SG, Ryu M, Kulkarni OP, Schmid H, Lichtnekert J, et al. (2011) An orally active chemokine receptor CCR2 antagonist prevents glomerulosclerosis and renal failure in type 2 diabetes. Kidney Int 80: 6878. 156. Xia M, Sui Z (2009) Recent developments in CCR2 antagonists. Expert Opin Ther Pat 19: 295 303. 157. Dufour JH, Dziejman M, Liu MT, Leung JH, Lane TE, et al. (2002) IFN gammainducible protein 10 (IP 10; CXCL10) deficient mice reveal a role for IP 10 in effector T cell generation and trafficking. J Immunol 168: 31953204. 158. Luster AD, Jhanwar SC, Chaganti RS, Kersey JH, Ravetch JV (1987) Interferoninducible gene maps to a chromosomal band associated with a (4;11) translocation in acute leukemia cells. Proc Natl Acad Sci U S A 84: 28682871. 159. Luster AD, Unkeless JC, Ravetch JV (1985) Gammainterfero n transcriptionally regulates an early response gene containing homology to platelet proteins. Nature 315: 672 676. 160. Murphy PM, Baggiolini M, Charo IF, Hebert CA, Horuk R, et al. (2000) International union of pharmacology. XXII. Nomenclature for chemokine receptors. Pharmacol Rev 52: 145 176.
154 161. Loetscher M, Loetscher P, Brass N, Meese E, Moser B (1998) Lymphocytespecific chemokine receptor CXCR3: regulation, chemokine binding and gene localization. Eur J Immunol 28: 36963705. 162. Weng Y, S iciliano SJ, Waldburger KE, SirotinaMeisher A, Staruch MJ, et al. (1998) Binding and functional properties of recombinant and endogenous CXCR3 chemokine receptors. J Biol Chem 273: 18288 18291. 163. Han C, Fu J, Liu Z, Huang H, Luo L, et al. (2010) Dipyr ithione inhibits IFN gamma induced JAK/STAT1 signaling pathway activation and IP 10/CXCL10 expression in RAW264.7 cells. Inflamm Res 59: 809816. 164. Williams R, Yao H, Dhillon NK, Buch SJ (2009) HIV 1 Tat co operates with IFN gamma and TNF alpha to incr ease CXCL10 in human astrocytes. PLoS One 4: e5709. 165. Mendez Samperio P, Perez A, Rivera L (2009) Mycobacterium bovis Bacillus Calmette Guerin (BCG) induced activation of PI3K/Akt and NF kB signaling pathways regulates expression of CXCL10 in epithelia l cells. Cell Immunol 256: 1218. 166. Liu M, Guo S, Hibbert JM, Jain V, Singh N, et al. (2011) CXCL10/IP 10 in infectious diseases pathogenesis and potential therapeutic implications. Cytokine Growth Factor Rev 22: 121130. 167. Nakamichi K, Saiki M, Sawada M, TakayamaIto M, Yamamuro Y, et al. (2005) Rabies virus induced activation of mitogenactivated protein kinase and NF kappaB signaling pathways regulates expression of CXC and CC chemokine ligands in microglia. J Virol 79: 1180111812. 168. Nakamic hi K, Inoue S, Takasaki T, Morimoto K, Kurane I (2004) Rabies virus stimulates nitric oxide production and CXC chemokine ligand 10 expression in macrophages through activation of extracellular signal regulated kinases 1 and 2. J Virol 78: 9376 9388. 169. Romagnani P, Crescioli C (2012) CXCL10: a candidate biomarker in transplantation. Clin Chim Acta 413: 1364 1373. 170. Karonitsch T, Feierl E, Steiner CW, Dalwigk K, Korb A, et al. (2009) Activation of the interferon gamma signaling pathway in systemic lupus erythematosus peripheral blood mononuclear cells. Arthritis Rheum 60: 14631471. 171. Usacheva A, Smith R, Minshall R, Baida G, Seng S, et al. (2001) The WD motif containing protein receptor for activated protein kinase C (RACK1) is required for recrui tment and activation of signal transducer and activator of transcription 1 through the type I interferon receptor. J Biol Chem 276: 2294822953.
155 172. Yadav A, Saini V, Arora S (2010) MCP 1: chemoattractant with a role beyond immunity: a review. Clin Chim A cta 411: 1570 1579. 173. Kulkarni O, Pawar RD, Purschke W, Eulberg D, Selve N, et al. (2007) Spiegelmer inhibition of CCL2/MCP 1 ameliorates lupus nephritis in MRL(Fas)lpr mice. J Am Soc Nephrol 18: 2350 2358. 174. Nahid MA, Satoh M, Chan EKL (2011) MicroRNA in TLR signaling and endotoxin tolerance. Cell Mol Immunol 8: 388403. 175. Nahid MA, Satoh M, Chan EKL (2011) Mechanistic role of microRNA 146a in endotoxin induced differential cross regulation of TLR signaling. J Immunol 186: 17231734. 176. Nahid MA, Pauley KM, Satoh M, Chan EKL (2009) miR 146a is critical for endotoxin induced tolerance: Implication in innate immunity. J Biol Chem 284: 3459034599. 177. Pauley KM, Stewart CM, Gauna AE, Dupre LC, Kuklani R, et al. (2011) Altered miR 146a expression in Sjogren's syndrome and its functional role in innate immunity. Eur J Immunol 41: 2029 2039. 178. Pauley KM, Satoh M, Chan AL, Bubb MR, Reeves WH, et al. (2008) Upregulated miR 146a expression in peripheral blood mononuclear cells from rheumatoi d arthritis patients. Arthritis Res Ther 10: R101. 179. Yamasaki Y, Narain S, Hernandez L, Barker T, Ikeda K, et al. (2006) Autoantibodies against the replication protein A complex in systemic lupus erythematosus and other autoimmune diseases. Arthritis R es Ther 8: R111. 180. Feng X, Wu H, Grossman JM, Hanvivadhanakul P, FitzGerald JD, et al. (2006) Association of increased interferoninducible gene expression with disease activity and lupus nephritis in patients with systemic lupus erythematosus. Arthrit is Rheum 54: 29512962. 181. Bombardier C, Gladman DD, Urowitz MB, Caron D, Chang CH (1992) Derivation of the SLEDAI. A disease activity index for lupus patients. The Committee on Prognosis Studies in SLE. Arthritis Rheum 35: 630640. 182. Touma Z, Gladm an DD, Urowitz MB (2012) Cinical Meausres, Metrics, and indices. In: Wallace D, Hahn BH, editors. Dubois' Lupus Erythematosus and Related Syndromes. 8 ed. Philadelphia, PA: Elsevier Health Sciences. pp. 563581.
156 183. Nahid MA, Yao B, Dominguez Gutierrez P R, Kesavalu L, Satoh M, et al. (2013) Regulation of TLR2 Mediated Tolerance and Cross Tolerance through IRAK4 Modulation by miR 132 and miR 212. J Immunol 190: 12501263. 184. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real time quantitative PCR and the 2( Delta Delta C(T)) Method. Methods 25: 402408. 185. Ho A, Magder LS, Barr SG, Petri M (2001) Decreases in anti doublestranded DNA levels are associated with concurrent flares in patients with systemic lupus erythemat osus. Arthritis Rheum 44: 23422349. 186. Bootsma H, Spronk P, Derksen R, de Boer G, Wolters Dicke H, et al. (1995) Prevention of relapses in systemic lupus erythematosus. Lancet 345: 15951599. 187. ter Borg EJ, Horst G, Hummel EJ, Limburg PC, Kallenber g CG (1990) Measurement of increases in anti double stranded DNA antibody levels as a predictor of disease exacerbation in systemic lupus erythematosus. A long term, prospective study. Arthritis Rheum 33: 634643. 188. Isenberg DA, Garton M, Reichlin MW, Reichlin M (1997) Long term follow up of autoantibody profiles in black female lupus patients and clinical comparison with Caucasian and Asian patients. Br J Rheumatol 36: 229 233. 189. Bengtsson AA, Sturfelt G, Truedsson L, Blomberg J, Alm G, et al. (200 0) Activation of type I interferon system in systemic lupus erythematosus correlates with disease activity but not with antiretroviral antibodies. Lupus 9: 664671. 190. Ching KH, Burbelo PD, Tipton C, Wei C, Petri M, et al. (2012) Two major autoantibody clusters in systemic lupus erythematosus. PLoS One 7: e32001. 191. To CH, Petri M (2005) Is antibody clustering predictive of clinical subsets and damage in systemic lupus erythematosus? Arthritis Rheum 52: 40034010. 192. Umareddy I, Tang KF, Vasudevan SG, Devi S, Hibberd ML, et al. (2008) Dengue virus regulates type I interferon signalling in a straindependent manner in human cell lines. J Gen Virol 89: 30523062. 193. Perry ST, Buck MD, Lada SM, Schindler C, Shresta S (2011) STAT2 mediates innate imm unity to Dengue virus in the absence of STAT1 via the type I interferon receptor. PLoS Pathog 7: e1001297. 194. Qing X, Putterman C (2004) Gene expression profiling in the study of the pathogenesis of systemic lupus erythematosus. Autoimmun Rev 3: 505509
157 195. Luo X, Yang W, Ye DQ, Cui H, Zhang Y, et al. (2011) A functional variant in microRNA 146a promoter modulates its expression and confers disease risk for systemic lupus erythematosus. PLoS Genet 7: e1002128. 196. George CX, Das S, Samuel CE (2008) Organization of the mouse RNA specific adenosine deaminase Adar1 gene 5' region and demonstration of STAT1independent, STAT2 dependent transcriptional activation by interferon. Virology 380: 338 343. 197. Tassiulas I, Hu X, Ho H, Kashyap Y, Paik P, et al (2004) Amplification of IFN alphainduced STAT1 activation and inflammatory function by Syk and ITAM containing adaptors. Nat Immunol 5: 11811189. 198. Fulkerson PC, Zimmermann N, Hassman LM, Finkelman FD, Rothenberg ME (2004) Pulmonary chemokine expression is coordinately regulated by STAT1, STAT6, and IFN gamma. J Immunol 173: 75657574. 199. Kok SH, Hong CY, Kuo MY, Wang CC, Hou KL, et al. (2009) Oncostatin M induced CCL2 transcription in osteoblastic cells is mediated by multiple levels of STAT 1 and STAT 3 signaling: an implication for the pathogenesis of arthritis. Arthritis Rheum 60: 14511462. 200. Valente AJ, Xie JF, Abramova MA, Wenzel UO, Abboud HE, et al. (1998) A complex element regulates IFN gammastimulated monocyte chemoattractant protein1 gene transcription. J Immunol 161: 37193728. 201. Ceribelli A, Satoh M, Chan EK (2012) MicroRNAs and autoimmunity. Curr Opin Immunol 24: 686691. 202. Ceribelli A, Yao B, Dominguez Gutierrez PR, Nahid MA, Satoh M, et al. (2011) MicroRNAs in systemic rheumatic diseases. Arthritis Res Ther 13: 229. 203. Chan EKL, Ceribelli A, Satoh M (2012) MicroRNA 146a in autoimmunity and innate immune responses. Annals of the Rheumatic Diseases 72: ii90 ii95. 204. Dominguez Gutierrez PR, Ceribelli A, Satoh M, Sobel ES, Reeves WH, et al. (2013) Elevated STAT1 correlates with increased CCL2 and CXCL10 levels in peripheral blood of patients with systemic lupus erythematosus. Arthritis Res Ther: submitted. 205. Bhattacharyya S, Zhao Y, Kay TW, Muglia LJ (2011) Glucocorticoids target suppressor of cytokine signaling 1 (SOCS1) and type 1 interferons to regulate Toll like receptor induced STAT1 activation. Proc Natl Acad Sci U S A 108: 95549559.
158 206. Aittomaki S, Pesu M, Groner B, Janne OA, Palvimo JJ, et al. (2000) Cooperation among Stat1, glucocorticoid receptor, and PU.1 in transcriptional activation of the high affinity Fc gamma receptor I in monocytes. J Immunol 164: 5689 5697. 207. Nahid MA, Yao B, Dominguez Gutierrez PR, Kesavalu L, Satoh M, et al. (2013) Regulation of TLR2 mediated tolerance and cross tolerance through IRAK4 modulation by miR 132 and miR 212. J Immunol 190: 12501263. 208. Day NE (1969) Estimating Components of a Mixture of Normal Distributions. Biometrika 56: 463 474. 209. Rhen T, Cidl owski JA (2005) Antiinflammatory action of glucocorticoids --new mechanisms for old drugs. N Engl J Med 353: 17111723. 210. Hu X, Li WP, Meng C, Ivashkiv LB (2003) Inhibition of IFN gamma signaling by glucocorticoids. J Immunol 170: 48334839. 211. Ansar i AW, Schmidt RE, Heiken H (2007) Prednisolone mediated suppression of HIV 1 viral load strongly correlates with C C chemokine CCL2: In vivo and in vitro findings. Clin Immunol 125: 1 4. 212. Matsuo H, Tamura M, Kabashima N, Serino R, Tokunaga M, et al. ( 2006) Prednisolone inhibits hyperosmolarity induced expression of MCP 1 via NF kappaB in peritoneal mesothelial cells. Kidney Int 69: 736746. 213. de Kruif MD, Lemaire LC, Giebelen IA, Groot AP, Pater JM, et al. (2008) Effects of prednisolone on the syst emic release of mediators of cell mediated cytotoxicity during human endotoxemia. Shock 29: 458461. 214. Nemeth E, Rivera S, Gabayan V, Keller C, Taudorf S, et al. (2004) IL 6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest 113: 1271 1276. 215. D'Cruz DP, Khamashta MA, Hughes GR (2007) Systemic lupus erythematosus. Lancet 369: 587 596. 216. Crow MK (2008) Collaboration, genetic associations, and lupus erythematosus. N Engl J Med 358: 956 961. 217. O'Shea JJ, Gadina M, Schreiber RD (2002) Cytokine signaling in 2002: new surprises in the Jak/Stat pathway. Cell 109 Suppl: S121131. 218. Soldevila G, Licona I, Salgado A, Ramirez M, Chavez R, et al. (2004) Impaired chemokineinduced migr ation during T cell development in the absence of Jak 3. Immunology 112: 191 200.
159 219. Fiehn C, Hajjar Y, Mueller K, Waldherr R, Ho AD, et al. (2003) Improved clinical outcome of lupus nephritis during the past decade: importance of early diagnosis and tr eatment. Ann Rheum Dis 62: 435439. 220. Moroni G, Quaglini S, Gallelli B, Banfi G, Messa P, et al. (2007) The longterm outcome of 93 patients with proliferative lupus nephritis. Nephrol Dial Transplant 22: 2531 2539. 221. Mirzayan MJ, Schmidt RE, Witte T (2000) Prognostic parameters for flare in systemic lupus erythematosus. Rheumatology (Oxford) 39: 1316 1319. 222. Peyssonnaux C, Nizet V, Johnson RS (2008) Role of the hypoxia inducible factors HIF in iron metabolism. Cell Cycle 7: 2832. 223. Hamilto n JA, Lacey DC, Turner A, de Kok B, Huynh J, et al. (2012) Hypoxia enhances the proliferative response of macrophages to CSF 1 and their pro survival response to TNF. PLoS One 7: e45853. 224. Roiniotis J, Dinh H, Masendycz P, Turner A, Elsegood CL, et al. (2009) Hypoxia prolongs monocyte/macrophage survival and enhanced glycolysis is associated with their maturation under aerobic conditions. J Immunol 182: 7974 7981. 225. Bosco MC, Puppo M, Blengio F, Fraone T, Cappello P, et al. (2008) Monocytes and dend ritic cells in a hypoxic environment: Spotlights on chemotaxis and migration. Immunobiology 213: 733749. 226. Sica A, Melillo G, Varesio L (2011) Hypoxia: a double edged sword of immunity. J Mol Med (Berl) 89: 657665. 227. Bosco MC, Varesio L (2012) Dendritic cell reprogramming by the hypoxic environment. Immunobiology. 228. Deicher R, Horl WH (2004) Hepcidin: a molecular link between inflammation and anaemia. Nephrol Dial Transplant 19: 521524. 229. Song H, Yan YL, Titus T, He X, Postlethwait JH (2011) The role of stat1b in zebrafish hematopoiesis. Mech Dev 128: 442456. 230. Sow FB, Alvarez GR, Gross RP, Satoskar AR, Schlesinger LS, et al. (2009) Role of STAT1, NF kappaB, and C/EBPbeta in the macrophage transcriptional regulation of hepcidin by mycobacterial infection and IFN gamma. J Leukoc Biol 86: 1247 1258. 231. Jurkin J, Schichl YM, Koeffel R, Bauer T, Richter S, et al. (2010) miR 146a is differentially expressed by myeloid dendritic cell subsets and desensitizes cells to TLR2 dependent act ivation. J Immunol 184: 49554965.
160 232. Ghani S, Riemke P, Schonheit J, Lenze D, Stumm J, et al. (2011) Macrophage development from HSCs requires PU.1coordinated microRNA expression. Blood 118: 22752284. 233. Starczynowski DT, Kuchenbauer F, Wegrzyn J, Rouhi A, Petriv O, et al. (2011) MicroRNA 146a disrupts hematopoietic differentiation and survival. Exp Hematol 39: 167 178 e164. 234. Labbaye C, Testa U (2012) The emerging role of MIR 146A in the control of hematopoiesis, immune function and cancer. J H ematol Oncol 5: 13. 235. Bosco MC, Varesio L (2012) Dendritic cell reprogramming by the hypoxic environment. Immunobiology 217: 12411249.
161 BIOGRAPHICAL SKETCH Paul Ramon Dominguez Gutierrez was born to Ana Maira Gutierrez Hazas and Jose Ramon Domi nguez Pieiro on April 1979 in Montreal, Canada. He moved that same year to Florida, USA with his family. Paul graduated from A Beka Video Home School Program in 1997 and received Associates in Arts degree from in 2000 from Lake Sumter Comm unity College. H e began in the s ummer of 2000 at the University of Florida as a chemistry major Paul participated in undergraduate research in Dr. Kathryn Williams laboratory developing testing methodology to study caffeine chelation of antibiotics by differential scann ing calorimeter. He also was part of the Washington Center internship program where he was an intern at the Environmental Protection Agency for his last semester. Paul graduated with dual Bachelor of Science degrees in Chemistry and Microbiology & Cell Sci ence in 2003. After graduation, he was a contractor at Environmental Protection Agency conducting efficacy testing of hospital disinfectants and the development of sporicidal efficacy assay for Homeland Security. All these positions helped develop his sk ills in laboratory techniques. After the end of his contract, Paul moved back to Gainesville, FL and worked at Regeneration Technology Inc. as a tissue receiving specialists. He received and inspected donor tissue in BSL3 settings. This tissue would be later used for graft implantation. After a year, he went to work for Quick Med Technologies Inc. as a microbiologist developing and testing bonded antimicrobial wound dressing as well as nonleeching hand sanitizer and antimicrobial under garments. Finally, Paul decided that he wanted a PhD and enrolled in interdisciplinary Program in Biomedical Science at the University of Florida College of Medicine. He
162 began the program 3 months earlier when he was selected as a recipient of the Florida Board of E ducation fellowship. He began his first laboratory rotation in Dr. Greg Shultz where he learned about wound healing. In his second rotation, Paul learned about microRNA in Dr. Rolf Rennes laboratory. From his experience in the Renne lab, Paul decided to pursue research in microRNA. Paul joined Dr. Edward K.L. Chans laboratory and began working on the role of microRNAs in systemic lupus erythematosus He expanded his work to include STAT1, CCL2, CXCL10, and microRNA 146a. Upon completing the Ph.D. degree, he plans to obtain postdoctoral training and pursue a career in academic research.