Dynamic role of plasma ferritin during pseudomonas infection insights from the limulus

DYNAMIC ROLE OF PLASMA FERRITIN DURING PSEUDOMONAS INFECTION: INSIGHTS FROM THE LIMULUS ONG SEK TONG DERRICK DEPARTMENT OF BIOLOGICAL SCIENCES NATIONAL UNIVERSITY OF SINGAPORE 2003/2004 DYNAMIC ROLE OF PLASMA FERRITIN DURING PSEUDOMONAS INFECTION: INSIGHTS FROM THE LIMULUS ONG SEK TONG DERRICK (Bachelor of Science (Hon)) A THESIS SUBMITTED TO THE FOR THE DEGREE OF MASTER OF SCIENCES DEPARTMENT OF BIOLOGICAL SCIENCES THE NATIONAL UNIVERSITY OF SINGAPORE 2004 ACKNOWLEDGEMENTS I would like to express my heartfelt gratitude to: Professors Ding and Ho for their constant support, patience and guidance in this project. I really thank them for their understanding and assistance for my years in their labs; Dr Zhu, Lihui, Patricia and Sean, for imparting me their wealth of knowledge, giving me uncountable precious advices and directing me when I am lost. In particular, Lihui for helping me with the preparation of the huge amount of RNA, teaching me of the electrophoretic mobility shift assay and in the preparation of my manuscript; Nicole, who has assisted me and will be furthering some other aspects in this project; My family members, for their unfailing support; Han Chong, Sook Yin and Bee Ling, for helping me with the arrangement of the ultracentrifuge facility; Last but not least, all my lab-mates for their kind help extended from time to time. i TABLE OF CONTENTS Acknowledgements Table of Contents List of Abbreviations List of Figures List of Tables Page i ii vi vii x Summary xi 1. INTRODUCTION 1 1.1 Iron in biological systems and its toxic effects 1 1.2 Iron in host-pathogen interaction during infections 2 1.2.1 Iron, microbial pathogens and sepsis 2 1.2.2 Adaptive immunity and its iron-dependent nature 3 1.2.3 Innate immunity 5 1.2.3.1 The iron-withholding strategy as a component of innate immunity 6 1.3 The iron-binding proteins involved during infection and inflammation 9 1.3.1 The transferrin family 9 1.3.2 The vertebrate ferritins 10 1.3.2.1 Cytosolic ferritins 11 1.3.2.2 Secreted ferritins 13 Invertebrate ferritins 14 1.3.3.1 Cytosolic ferritins 14 1.3.3.2 Secreted ferritins 15 1.3.3 1.4 Model for host-pathogen competition for iron 16 1.4.1 The horseshoe crab 16 1.4.2 Pseudomonas aeruginosa- a model pathogen for iron piracy ii study 20 1.5 The Rationale and Aims of this Thesis 23 2. MATERIALS AND METHODS 25 2.1 Materials 25 2.2 Infection of horseshoe crab and preparation of cell-free haemolymph/ plasma 25 Oxidative activity of plasma during P. aeruginosa infection 25 2.3.1 25 2.3 2.4 2.5 Purification and identification of plasma ferritin 26 2.4.1 Purification and resolution of plasma ferritin 26 2.4.2 Two dimensional gel electrophoresis 27 2.4.3 Edman degradation for N-terminal sequencing of ferritin 29 2.4.4 Mass Spectrometry analysis of proteins 29 Cloning of horseshoe crab ferritin cDNA 29 2.5.1 Degenerate RT-PCR of CrFer-H1 30 2.5.2 5’ and 3’ Rapid Amplification of cDNA ends (RACE) of CrFer-H1 30 PCR amplification and cloning of CrFer-H2 cDNA 31 2.5.3 2.6 2.7 2.8 Supercoil relaxation assay Regulation of ferritin during P. aeruginosa infection 32 2.6.1 Northern analysis of CrFer-H2 32 2.6.2 Western analysis 32 Strategy employed by P. aeruginosa to ‘steal’ host iron 33 2.7.1 Reaction between plasma ferritin and P. aeruginosa 33 2.7.2 Quantification of total and labile iron pool in the plasma 33 2.7.3 Measurement of redox potential and pH 34 Plasma ferritin-DNA interaction 34 iii 2.8.1 Electrophoretic mobility shift of DNA by ferritin 34 2.8.2 Fluorescence measurement of ferritin complex-DNA interaction 35 2.9 Statistical analysis 35 2.10 Homology modeling methods 36 3. RESULTS 38 3.1 Iron contributes to the high oxidative activity of plasma. 38 3.2 Horseshoe crab plasma ferritin is made up of subunits of 21 kDa. 42 3.3 CrFer-H1 has 2 possible transcripts that are translated to plasma ferritin. 45 3.4 Another ferritin gene, CrFer-H2, codes for a secretory protein that is apparently absent in the plasma 52 3.5 Common features of the 3 ferritin cDNAs 55 3.6 CrFer-H2 is ubiquitously expressed and its transcription is responsive to LPS and bacterial infection. 61 LPS and iron can regulate ferritin protein synthesis during Pseudomonas infection. 64 3.8 P. aeruginosa ‘steals’ host iron by degrading plasma ferritin. 64 3.9 Ferritin switches from a DNA-binding to non-DNA-binding conformer during infection. 69 Uninfected and infected plasma ferritin complexes contain different ferritin isoforms. 69 4. DISCUSSIONS 73 4.1 Plasma ferritin is directly involved in innate immune response. 73 3.7 3.10 4.1.1 4.1.2 4.2 The horseshoe crab plasma ferritin evades degradation by P. aeruginosa to prevent iron loss. 73 Regulation of plasma ferritin may contribute to iron homeostasis and constant free radical level. 75 A dynamic role of ferritin during Pseudomonas infection. 76 iv 4.3 5. Insights into the role of plasma ferritin: from horseshoe crab to mammalian plasma ferritin 80 CONCLUSION AND FUTURE PERSPECTIVES 82 v LIST OF ABBREVIATIONS 2DE Ame CFH CFU CrFer-H DTT ES-Q-TOF EST Fur Hep HT Hpi IAA INT IRE LIP LPS MALDI-TOF MS-BLAST MUS ORF PCR PMF PVDF RACE RT-PCR SDS-PAGE STC TPTZ TSB UTR Two dimensional gel electrophoresis Amebocytes Cell-free hemolymph Colony forming unit Carcinoscorpius rotundicauda ferritin heavy chain Dithiothreitol Electrospray-Quadrupole-Time-of-flight Expressed sequence tags Ferric uptake regulator Hepatopancreas Heart Hour post-infection Iodoacetamide Intestine Iron responsive element Labile iron pool Lipopolysaccharides Matrix-Assisted Laser Desorption Ionization Time-of-flight Mass spectrometry-Basic Local Alignment Search Tools Muscles Open-reading frame Polymerase chain reaction Peptide mass fingerprint Polyvinylidene fluoride Rapid Amplification cDNA end Reverse Transcription-Polymerase Chain Reaction Sodium Dodecylsulphate-Polyacrylamide Gel Electrophoresis Stomach 2,4,6-tripyridyl-s-triazine Tryptone soy broth Untranslated region vi LIST OF FIGURES Fig. No. 1.1 1.2 1.3 1.4 2.1 2.2 3.1 3.2 3.3 Title Confocal microscopic images of GFP-labeled P. aeruginosa in biofilm flow cells perfused with lactoferrin free (a-d) and lactoferrin-containing (20 µg/ml) (e-h) media. Human H-chain and human L-chain both have 5 α-helices and the heavy chain subunits then assemble into a apoferritin complex of 24 subunits in 432 symmetry viewed down the a four-fold axis. (A) A hypothetical scenario to the coagulation-based clotting mechanism and containment of foreign invaders. (B) The LPS- and glucan-mediated pathways require the serine protease Factor C and Factor G respectively. Gram staining of Pseudomonas aeruginosa and colony morphology on agar plates. Schematic view of the partial purification and enrichment of horseshoe crab plasma ferritin. Diagrammatic representation of the iron assay to measure total plasma iron and LIP. Plasma iron regulates free radical-induced DNA nicking. (A) The role of iron as a catalyst in the Haber-Weiss reaction. (B) Concentration dependent DNA nicking activity by naïve plasma and the effect of nicking buffer. (C) Effect of glycerol as a free radical scavenger in the highly oxidative plasma. (D) Effect of metal chelators on oxidative activity of plasma using EDTA, potassium ferrocyanide and ferricyanide. (E) Oxidative activity of plasma during P. aeruginosa infection of the horseshoe crab. Identification of limulus plasma ferritin complex and its subunits in the plasma. (A) The native state of limulus plasma ferritin was detected by Prussian blue staining. (B) The ferritin complex is made up of 21 kDa subunits. (C) Protein sequencing of 21 kDa band. (A) PCR products of the same size (~ 240 and 330 bp) were obtained from degenerate RT-PCR using naïve heart, intestine and stomach cDNA as template. (B) Alignment of the deduced amino acid sequence of the 240 and 330 bp PCR products show that they may be encoded by the same gene. (C) Design of 5’ and 3’ RACE primers using the partial ferritin DNA sequence. (D) 5’ and 3’ RACE products of ferritin gene. (E) Screening of positive clones by EcoRI digestion of Page 8 12 17 17 21 28 37 39 41 43 46 47 48 49 50 vii 3.4 3.5 3.6 3.7 3.8 3.9 3.10 4.1 pGEM-T Easy vector. (F) Nucleotide sequence and deduced amino acid sequence of CrFer-H1a and -H1b. Cloning of CrFer-H2. (A) Screening of clones that harbor the 3’ fragment of CrFer-H2 after EcoRI digestion of pGEM-T Easy vector. (B) 5’ RACE product of CrFer-H2 at various annealing temperatures using naïve cardiac cDNA as template. (C) Nucleotide sequence and deduced amino acid sequence of CrFer-H2. (A) Predicted secondary structures of CrFer IRE at the 5’ UTR and the alignment of CrFer IRE sequence with that from other organisms. (B) Both CrFer-H1 and –H2 are predicted to possess the typical 5 α-helices of ferritins. (C) Phylogenetic analysis of CrFer-H1 and –H2. (D) Central region of subunits of human H-ferritin. (E) CrFer-H1 and –H2 share ~ 72 % identity (top) and there are likely to be other isoforms of plasma ferritin (bottom). (A) Northern analysis to study differential expression of CrFer-H2 in various naïve, 3h LPS-induced and 3h FeSO4 induced tissues of the limulus. (B) Quantitative analyses of CrFer-H expression using ImageMaster software. (C) Northern analysis to study kinetics of CrFer-H2 expression in various tissues after infection with P. aeruginosa and (D) the change in fold of CrFer-H2 normalized against actin 3. Regulation of limulus plasma ferritin at the protein level during P. aeruginosa infection. Strategy employed by P. aeruginosa to obtain host iron. (A) P. aeruginosa can degrade ferritin in vitro. (B) Pseudomonas infection does not result in hypoferraemia of the limulus. (C) Pseudomonas does not lower plasma redox potential or pH to ‘steal’ host iron. Uninfected plasma ferritin but not infected plasma ferritin can bind to DNA in a sequence-independent manner. (A) Using LDorThr and LkBCom as probes, plasma ferritin from naïve, 3, 6 and 72 hpi individuals, as well as 3 h iron-loaded individuals were incubated and run on 4 % PAGE gel. (B) Fluorescence measurement of ferritin complex-DNA interaction. Uninfected and infected plasma ferritins consist of different 21 kDa isoforms. Proposed model for the dynamic role of plasma ferritin 51 53 54 56 57 58 59 60 62 63 65 67 70 72 78 viii during infection. ix LIST OF TABLES Table No. 1 2 3 Title Effect of iron deficiency and iron overload on various immunological functions. Defense molecules found in the hemoctyes and haemolymph plasma of the horseshoe crab. Summary of primers used in the cloning of ferritin genes. Page 4 19 31 x Summary Ferritin, normally found intracellularly in vertebrates, is responsible for iron storage and detoxification, although it has been isolated from plasma in trace amount. Plasma ferritins serve as extracellular iron storage molecules and loss of plasma iron to pathogen is detrimental to the host during infection. Interestingly, the horseshoe crab plasma iron level is 8-10-fold higher than human plasma. In this study, horseshoe crab plasma ferritin complex was purified, characterized and its dynamic role in innate immune defense was investigated using Pseudomonas aeruginosa as a model pathogen. We demonstrate the interesting phenomenon that on one hand, Pseudomonas attempts to degrade the host ferritin in order to usurp the host iron for its survival. On the other hand, the host maintains iron homeostasis by tightly regulating its level of plasma ferritin, plasma redox potential and pH that keeps the plasma free radicals in check. Between 6-48 h of infection, the host plasma ferritin evades Pseudomonas-mediated degradation by transiting from extracellular to intracellular space, during which different ferritin isoforms constitute the ferritin complex. Our data show that the host recovers its level of plasma ferritin by 72 h. Furthermore, we demonstrate that contrary to the naïve ferritin, which binds the host DNA sequence-independently and probably protects the host genome, infection somehow disables the ferritin complex from binding host DNA. We propose that the plasma ferritin plays dual roles: (i) pathogen evasion and (ii) DNA protection or chromatin remodeling after nuclear translocation. xi 1. Introduction 1.1 Iron in biological systems and its toxic effects Iron is an abundant metal, being the fourth most plentiful element in the earth’s crust. It can be found in the first row of transition metals in the periodic table. It exists mainly in one of the two readily reversible redox states: the reduced Fe2+ ferrous form and the oxidized Fe3+ ferric form. Depending on its ligand environment, both ferrous and ferric forms can adopt different spin states. As a result of these properties, iron is an extremely attractive prosthetic component for incorporation into proteins as a biocatalyst or electron carrier during evolution of early life (Andrews et al., 2003). Iron plays an indispensable role in various physiological processes, such as photosynthesis, nitrogen fixation, methanogenesis, hydrogen production and consumption, respiration, the trichloroacetic acid cycle, oxygen transport, gene regulation and DNA biosynthesis. The incorporation of iron into proteins allows its local environment to be regulated such that iron can adopt the necessary redox potential (-300 to +700 mV), geometry and spin state for realization of its prescribed function (Andrews et al., 2003). Unfortunately with the appearance of oxygen on earth ~ 2.2 to 2.7 billion years ago, two major problems arose. One was the production of toxic oxygen species and the other, a drastic decrease in iron availability (Touati, D., 2000). In its reduced ferrous form, iron potentiates oxygen toxicity by converting the less reactive hydrogen peroxide to the more reactive oxygen species, hydroxyl radical and ferryl iron, via the Fenton reaction (O2- + H2O2 → HO + OH- + O2; iron as a catalyst). Conversely, superoxide favours the Fenton reaction by releasing iron from ironcontaining molecules. It is widely accepted that tight regulation of iron assimilation prevents an excess of free intracellular iron that could lead to oxidative stress. 1 1.2 Iron in host-pathogen interaction during infection 1.2.1 Iron, microbial pathogens and sepsis Sepsis has been a challenge to humans and it has steadily worsened in recent years. In the United States alone, there are ~ 500,000 incidents each year with a death rate of 35-65 % (Dellinger et al., 1997; Bone et al., 1997). Amongst the numerous complex interactions between host and pathogen, one common and essential factor is the ability to invade and multiply successfully within host tissues. Proliferation of pathogen is critical to establishing an infection and this mediates the pathogen to produce the full arsenal of virulence determinants required for pathogenicity (Bullen and Griffiths, 1999). The availability of iron in the host environment and its effects on bacterial growth is one of the best studied aspects in pathogenicity. Humans are equipped with a well-developed natural resistance against bacterial infection. Currently, some of the understood mechanisms involved are the antibacterial properties of tissue fluids and the phagocytic abilities of cells (Bullen et al., 2000). However, research has revealed that these mechanisms require a virtual iron-free environment for proper function (Ward et al., 1999). In normal human plasma, the high affinity constant for Fe3+ (10-36 M) and the unsaturated state of the iron-binding protein transferrin ensure that the amount of free ferric iron is ~ 10-18 M (Bullen, et al., 1978). In vivo, bacterial growth is inhibited by strong bactericidal and bacteriostatic mechanisms in the plasma. These include unsaturated transferrin, antibody, and complement, which function in the virtual absence of freely available iron. Even though freely available iron in normal body fluids is virtually absent, pathogenic bacteria are able to multiply successfully in vivo to establish an infection. The observation that all known bacterial pathogens require iron to multiply suggests that they must adapt to the iron-free extracellular environment in vivo and develop 2 mechanisms to acquire protein-bound iron. Thus, pathogens have evolved various ways to compete for the host iron. The production of low molecular mass ironchelating compounds (siderophores), expression of transferrin and lactoferrin receptors, proteolytic cleavage of iron-binding glycoproteins, disruption of ironbinding site, reduction of ferric to ferrous complex to effect ferrous iron release and utilization of iron in haem compounds are some of the strategies developed during coevolution of host and pathogen (Bullen and Griffiths, 1999). The invading pathogens could also migrate into local environments where iron is more readily available, such as inside some cells. Low environmental iron levels can signal pathogens to induce their virulence genes (Litwin et al., 1993) and this has been extensively demonstrated in the opportunistic human pathogen, Pseudomonas aeruginosa, which employs a Fur protein as an iron sensor to induce cytotoxic exotoxin A and extracellular proteases under iron-depleted conditions (Bullen et al., 1978). 1.2.2 Adaptive immunity and its iron-dependent nature Adaptive immunity can be classified as humoral immunity, mediated by antibodies which are produced by activated B lymphocytes, and cell-mediated immunity, mediated by T lymphocytes. The immune system is activated when an antigen is recognized and processed by an antigen-presenting cell such as macrophage, dendritic cell, or a B lymphocyte. Subsequently, the T and/or B lymphocytes are activated and this leads to cell division, phenotypic changes and protein synthesis. The cytokines activate the phagocytic cells and lymphocytes to exert increased microbicidal and cytotoxic activities (Brock, 1999). Iron is critical for many metabolic processes and since immunological activation involves various metabolic events, iron bioavailability has been believed to influence the immune system. This 3 link has been supported by studies of various immune functions in humans and experimental animals that reveal defects associated with abnormalities of iron metabolism, as well as in vitro studies that illustrate the iron-dependent nature of the immune system. Some of these effects are summarized in Table 1. 4 1.2.3 Innate immunity The innate immune system is believed to predate the adaptive immune response. The innate immune system represents a frontline defense that targets microbial pathogens by recognizing molecular structures that are shared by large groups of pathogens, the pathogen-associated molecular patterns via pattern recognition proteins or the pattern recognition receptors. The pathogen-associated molecular patterns are conserved products of microbial metabolism and they are essential for the survival or pathogenicity of the microorganisms (Medzhitov and Janeway, 1997). Examples of pathogen-associated molecular patterns include lipopolysaccharides (LPS) of all gram-negative bacteria, lipoteichoic acids of all gram-positive bacteria and the mannans of yeasts /fungi. A key feature of these microbial patterns is their polysaccharide chains that vary in length and carbohydrate composition (Franc and White, 2000). The invertebrates have a defense system centered on both cellular and humoral immune response. The former is known to include encapsulation, phagocytosis (Foukas et al., 1998), and nodule formation, while the humoral response includes the coagulation system of arthropods (Iwanaga et al., 1998), the synthesis of a broad spectrum of potent antimicrobial proteins in many insects and crustaceans (Hoffman et al., 1996), and the prophenoloxidase activating system (proPO system) (Soderhall et al., 1998). In the vertebrates, innate immunity provides a first line of host defense against pathogens and the signals that are needed for the activation of the adaptive immunity (Fearson and Locksley, 1996). The vertebrate innate immunity was suggested to resemble a mosaic of invertebrate immune responses. For example, the effectors, receptors and regulation of gene expression of insects in acute immune response are 5 similar to those of humans. Some antibacterial peptides and immune stimulators have originated from the processing of neuropeptide precursors (Salzet, 2001). The vertebrate pathogen recognition receptors are displayed by particular cell types, such as macrophages, natural killer cells, and probably also epithelial and endothelial cells in the lung, kidney, skin and gastrointestinal tract (Wright, 1991). Similar to the invertebrate innate immune molecules, expression of the vertebrate innate immune molecules works on a broad-based specificity targeted at broad classes of pathogens and their corresponding pathogen-associated molecular patterns. A number of mammalian pathogen recognition receptors have been characterized and these include the macrophage mannose receptor, scavenger receptors, integrins, collectins, and some clusters of differentiation antigens (Epstein et al., 1996; Wright et al., 1990). 1.2.3.1 The iron-withholding strategy as a component of innate immunity Iron sequestration is recognized as an ancient host defense mechanism against invading pathogens (Beck et al., 2002) and it is widespread in occurrence. Upon infection, iron acquisition is critical for bacterial growth and pathogenicity (Bullen, 1981). However in the vertebrates, bacterial infection can drastically reduce plasma iron level (Lauffer, 1992) as the host withholds iron within the cells and tissues (Konijn and Hershko, 1977 ; Roeser, 1980; Brock, 1989). Some features of the ironwithholding defense system include constitutive components such as transferrin, lactoferrin and ovotransferrin, as well as processes which are induced at the time of microbial cell invasion. The suppression of iron efflux from macrophages hence, reduction in plasma iron and increased synthesis of ferritin by macrophages to accommodate iron from phagocytosed lactoferrin iron (Lauffer, R.B., 1992) is one such example. 6 Currently, the iron-withholding strategy is accepted as a new component of the innate immune system. (Singh, et al., 2002; Ganz, 2003). Singh et al. (2002) demonstrated that lactoferrin stimulates twitching, a specialized form of surface motility by chelating iron, causing the P. aeruginosa to wander around instead of forming clusters & biofilms. Conalbumin was also shown to block biofilm formation of P. aeruginosa through iron chelation, hence biofilm formation as well. Thus, iron deprivation inhibits the formation of resistant bacterial biofilms, prevents recalcitrant bacteria that survive initial defenses from forming squatters and favours the vulnerable unicellular forms that are better equipped to reach alternative iron sources (Singh et al., 2002) (Fig. 1.1). 7 4 hour 24 hour 3 days 7 days Fig. 1.1 Confocal microscopic images of GFP-labeled P. aeruginosa in biofilm flow cells perfused with lactoferrin-free (a-d) and lactoferrin-containing (20 µg/ml) (e-h) media. Images were obtained 4 h (a, e), 24 h (b, f), 3 days (c, g) and 7 days (d, f) after inoculating the flow cells. Images a, b, e and f are top views; scale bar, 10 µm. Images c, d, g and h are side views; scale bar, 50 µm. Results are representatives of 6 experiments. (Adapted from Singh et al., 2002). 8 1.3 The iron-binding proteins involved in infection and inflammation To achieve an iron-free physiological environment, mammals employ iron- binding proteins to reduce the level of extracellular iron to around 10-18 M (Bullen et al., 1978) so as to stall bacterial growth (Jamroz et al., 1993). At least two classes of iron-binding proteins, ferritin and transferrins, are present across phyla (Singh et al., 2002). 1.3.1 The transferrin family As a major iron transporter in the blood of vertebrates, transferrin absorbs iron in the gut, shuttles between peripheral sites of storage and uses, and maintains iron level sufficient to support cells having a particular demand for iron (Yoshiga et al., 1997; Jamroz et al., 1993). Transferrins are serum glycoproteins (extracellular), with a molecular weight of ~ 75-80 kDa. Each transferrin molecule is folded to give 2 globular domains. Each domain contains a specific binding site for a single Fe3+ (Caccova et al., 2002). Diferric iron is taken into cells by receptor-mediated endocytosis. Dissociation of iron from transferrin then occurs in an acidic endosome, after which the iron is transferred to the cytoplasm. Within cells, the iron is subsequently incorporated into metalloproteins or stored in the cytoplasm either within the iron storage protein, ferritin, or chelated to small molecules (Welch, 1992). At physiological pH, the affinity of transferrin for Fe3+ (Kd ~ 10-20 M) is very high. Lactoferrin, a member of the transferrin family, is a 78 kDa glycoprotein present in various secretions (eg. milk, tears, saliva and pancreatic juice). In the vertebrates, serum transferrin is an acute-phase protein as its concentration closely mirrors conditions of stress or infection, although its rise or fall varies with the infective microorganisms. Human lactoferrin is stored in specific granules of 9 polymorphonuclear granulocytes from which it is released following activation. It binds with high affinity to lipid A and may play an antibiotic role by depriving invading microorganisms of iron, which is required for their proliferation and (Yoshiga et al., 1997; Caccova et al., 2002). Owing to their bacteriostatic activity, members of the transferrin family (ovotransferrin and lactoferrin) have been considered major contributors to host iron sequestration. Transferrins have also been isolated from cockroach, mosquito, Bombyx mori, Drosophila and Manduca sexta at the genetic and protein level (Jamroz et al., 1993; Yoshiga et al., 1997; Yun et al., 1997; Yoshiga et al., 1999; Hueber et al., 1988). The involvement of transferrin in immune defense of mosquitoes has been shown by Yoshiga et al. (1997) as transferrin synthesis and secretion are increased upon exposure of mosquito cells (Aedes aegypti or Aedes albopictus) to bacteria. Inoculation of adult Drosophila with E. coli also led to dramatic increase in transferrin mRNA (Yoshiga et al., 1999). In the goldfish, transferrin serves as a primary activating molecule of macrophage antimicrobial response (Stafford and Belosevic, 2003). It was found that the products released by necrotic/damaged cells can enzymatically cleave transferrin, and the cleavage products of transferrin were able to induce nitric oxide response of macrophages. Addition of transferrin also significantly enhanced the killing response of the goldfish macrophages exposed to different pathogens or pathogen-associated molecular patterns. 1.3.2 The vertebrate ferritins Another important iron sequestration protein, ferritin, has been extensively investigated, showing pivotal roles in iron storage and detoxification. In the vertebrates, ferritin is mainly intracellular although trace levels of plasma ferritin can 10 be found in ng/L quantity. In higher vertebrates, ferritin has been indirectly linked to innate immune response since the synthesis of ferritin is regulated by proinflammatory cytokines at both transcriptional and translational levels (Torti et al., 1988; Konijn and Hershko, 1989; Roger et al., 1990; Huang et al., 1999). More recently, the ferroxidase sites of ferritin H subunit have been reported to be critical for direct DNA binding, suggestive of a new important role of ferritin in the protection of host cell genome by preventing DNA nicking due to free radical effects caused by free iron in the nucleus (Surguladze et al., 2004). An overview of the current understanding of both cytosolic and secreted ferritins in vertebrates is discussed below. 1.3.2.1 Cytosolic ferritins Ferritin is present in all types of mammalian cells, being most abundant in macrophages and hepatocytes. The structure of cytosolic ferritin from various organisms has been solved and they share a similar structure composed of 5 α-helices (Fig. 1.2A and B). In the native state, the ferritin complex is a hollow sphere (apoferritin) composed of 24 subunits (Fig. 1.2C), with very high iron binding capacity (4500 iron atoms). There are 24 subunits of two types, H and L (each of ~20 kDa), which exist in different ratios, from different tissues and in various physiological states (Nichol and Locke, 1999). With the completion of the human genome project, it is now known that there are at least 15 genes encoding ferritin Hchain subunits (FTHL1-4, FTHL7-8 and FTHL10-18) and 1 gene for ferritin L-chain subunit (FTL) (NCBI LocusLink, http://www.ncbi.nlm.nih.gov/LocusLink/list.cgi ). The human H- and L-subunits are ~ 55 % homologous and are coded by genes on various chromosomes. However, it is the H-chain that possesses ferroxidase activity. 11 (A) (B) (C) Fig. 1.2 Human H-chain (A) and human L-chain (B) both have 5 α-helices and the heavy chain ferritin subunits then assemble into a apoferritin complex of 24 subunits in 432 symmetry viewed down a four-fold axis. (C). The structures were obtained from the Protein Database. Human H-chain: pdb 2fha; human L-chain: pdb 1aew. The structure of the ferritin complex was adapted from Chasteen and Harrison, 1999. 12 Studies have revealed that Fe2+ enters the core of the apoferritin, after which it is oxidized to Fe3+ by the catalytic action of the amino acid side-chains of the H-chain. 1.3.2.2 Secreted ferritins Ferritin was demonstrated to exist in the serum with the development of a sensitive immunoradiometric assay (Addison et al., 1972; Jacobs et al., 1972; Slimes et al., 1974; Cook et al., 1974). The concentration of serum ferritin in normal subjects ranges between 12 and 250 µg/L (Jacobs et al., 1972; Slimes et al., 1974; Cook et al., 1974). The circumstantial evidence for ferritin circulating in the serum is usually closely related to reticuloendotheilial ferritin from clinical studies (Jacobs and Worwood, 1975). There appears to be structural differences between cytosolic and serum ferritin. Linder et al. (1996) have purified horse serum ferritin and compared it against horse spleen ferritin. Whereas intracellular ferritins are usually comprised of 24 subunits, each ~ 20 kDa, they reported that serum ferritin consists of more than 1 size of subunit and all were larger than intracellular ferritins. The iron content of serum ferritin was also much lower than that of cytosolic ferritin. Serum ferritin, and not its intracellular counterpart was also found to be glycosylated. Various studies have demonstrated the elusiveness of serum ferritin and several questions remain to be addressed. Firstly, the heterogeneity in ferritin molecules, some of which may be linked to its tissue of origin. Secondly, what are the underlying biological implications for the presence of dimers, trimers and larger polymers? Lastly, the existence of isoferritins from rat, human and horse tissues still remains to be explained. Until now, the source and nature of the trace level of plasma ferritin still remains ill-defined. It has been proposed that the presence of glycosylation indicates secretion of ferritin, possibly from phagocytic cells degrading 13 haemoglobin or direct release of cellular ferritin from damaged cell membranes (Cragg et al., 1981; Worwood, 1986). The only evidence for secretion of ferritin in mammals has been shown in rat hepatoma cells, where it was regulated by inflammatory cytokines and iron at the transcriptional level. Plasma ferritin concentration closely correlates with the iron status, which increases acutely in numerous physiological conditions, such as cancer, inflammation or infection (Linder et al., 1996; Tran et al., 1997). Several physiological functions for serum ferritin have been put forward over the years. These include: (i) a messenger with a hormonal effect on the mucosa of the small intestine, (ii) regulation of transferrin synthesis by hepatic parenchyma cells and (iii) scavenging and help in detoxifying ferrous iron leaking from damaged cells during infection (Jacobs and Worwood, 1975; Linder et al, 1996). 1.3.3 The invertebrate ferritins In invertebrates, ferritin has also been isolated from the haemolymph and in various tissues. Interestingly, insect ferritins, which are present in mg/L quantity, are mainly extracellular (Winzerling et al., 1995; Cappuro et al., 1996). An overview of the current understanding of both cytosolic and secreted ferritins in invertebrates is discussed below. 1.3.3.1 Cytosolic ferritins To date, invertebrate cytosolic ferritins have been found in the Calpodes ethlius, freshwater crayfish (Pacifastacus leniusculus), echinoderm and ticks (Nichol and Locke, 1989; Huang et al., 1996; Beck et al., 2002; Kopacek et al., 2003). In C. ethlius, ferritin was isolated from the midgut of the larvae. The holoferritin was stable 14 to heat at 75 oC or in the presence of SDS, proteinase K or Urea. Calpodes ferritin contains iron and is a glycoprotein having N-linked high-mannose oligosaccharides. There are 2 isoforms with a pI 6.5 – 7 and there are 2 major subunits of 24 and 31 kDa and 2 minor subunits of 26 and 28 kDa. The 24 kDa subunit is induced upon iron injections (Nichol and Locke, 1989). The ferritin of freashwater crayfish was purified from the hepatopancreas. It consists of 19 and 20 kDa subunits. It shows a closer relationship to vertebrate H-chains than to insect ferritins (Huang et al., 1996). In the echinoderm, a ferritin molecule was cloned from the coelomocyte cDNA library and it shows high homology to vertebrate ferritin. In vitro experiments showed that stimulated coelomocytes released iron into the culture supernatants and that the amount of iron in the supernatants decreased over time upon LPS or PMA treatments. There was also enhanced expression of ferritin mRNA after stimulation (Beck et al., 2002). This was perhaps the only study so far to have demonstrated the involvement of invertebrate cytosolic ferritin in innate immune defense. 1.3.3.2 Secreted ferritins The first report of secreted ferritin in invertebrates was from the Manduca sexta (Winzerling et al., 1995). Similar to C. ethlius cytosolic ferritin, Manduca haemolymph ferritin was also resistant to denaturation at 75 oC or to proteinase K. SDS-PAGE revealed the presence of 2 major subunits (24 and 30 kDa) and 2 minor subunits (28 and 32 kDa). The concentration of ferritin in the haemolymph was ~ 0.4 mg/ml, representing ~ 0.7 % of total haemolymph proteins. One of the Manduca ferritin subunits (Ms Fer) was subsequently cloned and it was found that it resembled more closely to the vertebrate L-chain (Pham et al., 1996). The Ms Fer transcript was 15 found to be expressed in the midgut, fat body and hemocytes, with highest expression in the midgut. Subsequently, secreted ferritins were isolated from A. aegypti, Musca domestica, D. melanogaster, C. ethlius and Galleria mellonaella (Dunkov et al., 1995; Capurro et al., 1996; Charlesworth et al., 1997; Nichol and Locke, 1999; Kim et al., 2001). In A. aegypti, the ferritin complex comprises of 24, 26 and 28 kDa subunits as well as small amount of 30 kDa subunits. The ferritin subunits were found to be present in larvae, pupae, and adult females, and they increased in animals exposed to excess iron (Dunkov et al., 1995). The Drosophila ferritin complex is both monomeric and dimeric and they are made up of 25, 26 and 28 kDa subunits. The deduced amino acid sequence of Drosophila ferritin subunit showed that it resembled closely to that of A. aegypti and it contained a signal sequence and a putative iron response element (Charlesworth et al., 1997). In C. ethlius, a 24 kDa nonglycosylated subunit and a 31 kDa glycosylated subunit constitute the haemolymph ferritin (Nichol and Locke, 1999). 1.4 Model for host-pathogen competition for iron 1.4.1 The horseshoe crab Despite having only the innate immune recognition response, the horseshoe crab has managed to survive through ~400 millions of years in a harsh environment where a diversity of microorganisms flourishes. The hemolymph of the horseshoe crab contains mainly amebocytes (about 99 %) and plasma. The cell contains two types of dense granules, the large (L) granules and the small (S) granules (Fig 1.3A). The former is known to contain about 20 protein components including five clotting factors (Fig. 1.3B) and one anti-LPS factor, while the latter contains largely 16 (A) (B) Fig 1.3 (A) A hypothetical scenario to the coagulation-based clotting mechanism and containment of foreign invaders. The L granules are released more rapidly than the S granules and almost all the granules are exocytosed. (B) The LPS- and glucan-mediated pathways require the serine protease Factor C and Factor G respectively. The two pathways converge where proclotting enzyme is converted to the clotting enzyme, finally resulting in the formation of coagulin. Clot formation prevents entry of microorganisms into the limulus and prevents excessive loss of hemolymph. (Adapted from Hoffman et al., 1994) 17 tachyplesin and at least 6 other protein components (Iwanaga et al., 1994). On the other hand, the hemolymph plasma contains three major proteins: hemocyanin, limulin (lectin)/ C-reactive proteins, and α2-macroglobulin. Defense molecules found in the amebocytes and the hemolymph /plasma of the horseshoe crab are summarized in Table 2. The horseshoe crab is heavily dependent on the coagulation cascade, lectins and other defense factors for survival in the harsh environment. Since the study of innate immunity in the vertebrates has been hampered by its acquired immune system, the use of invertebrates such as the horseshoe crab as an experimental model would lend insights to our understanding of innate immunity in vertebrates. Without adaptive immune pathways, the horseshoe crab has proven itself to be an evolutionary success and this makes it an ideal organism for the study of invertebrate innate immunity in isolation, without interference from molecular interaction from the adaptive immune response. 18 19 1.4.2 Pseudomonas aeruginosa – a model pathogen for iron-piracy study Pseudomonas aeruginosa is the epitome of an opportunistic human pathogen, which is strongly involved in severe and often fatal infections in patients with cystic fibrosis, burns, ocular diseases, pneumonia, and other immunocompromised illnesses (Meyer et al., 1996). It is a Gram-negative aerobic rod, belonging to the bacterial family, Pseudomonadaceae (Fig. 1.4, see http://www.bact.wisc.edu/Bact330/lecturepseudomonas). P. aeruginosa is capable of producing a large variety of virulence determinants, such as cell-surface-associated factors (alginate, pilli, and LPS), extracellular factors (exotoxin A, exoenzymne S, proteases, cytotoxin, phospholipases, heat-stable haemolysins, pycocyanin and siderophores) (Ochsner et al., 1996), and is notorious for its resistance to antibiotics rendered by its outer membrane LPS. Moreover, its tendency to colonize surfaces in a biofilm form causes the cells to be impervious to therapeutic concentrations of antibiotics. Typically, Pseudomonas infection may compose of three distinct stages: (1) bacterial attachment and colonization; (2) local invasion; (3) disseminated systemic disease. To colonize, the fimbriae of Pseudomonas adhere to the epithelial cells of the upper respiratory tract and they may bind to specific galactose or mannose or sialic acid receptors on epithelial cells. A protease enzyme that degrades fibronectin to expose the underlying fimbrial receptors on the epithelial cell surface may assist the bacterium in colonization. During host invasion, the bacterial capsule or slime layer effectively protects cells from opsonization by antibodies, complement deposition, and phagocyte engulfment. The bacterium also produces 2 extracellular proteases (elastase and alkaline protease) to cleave proteins such as collagen, IgG, IgA, complement, fibronectin and fibrin. There are also reports of gamma interferon 20 (A) (B) Fig. 1.4 Gram staining of Pseudomonas aeruginosa (A) and colony morphology on agar plates (B). (Adapted from http://www.bact.wisc.edu/Bact330/lecturepseudomonas). 21 and tumor necrosis factor being inactivated by elastase and alkaline protease. Pseudomonas also produces pigments (pyocyanin, pyochelin and pyoverdin) to sequester iron from the environment under low-iron conditions for its growth in the host. Finally, an extracellular toxin, exotoxin A, which causes the ADP ribosylation of eukaryotic elongation factor 2, may exert some pathologic activity during the dissemination stage. There are various stimuli that influence the expression of virulence genes: temperature, pH, osmolarity, cell density, iron limitation, the level of oxygen and oxidative compounds, and nitrogen and phosphate concentrations (Ochsner et al., 1996). P. aeruginosa is an excellent model for the study of iron acquisition and metabolism. This is due to the fact that it requires oxygen and nitrogen compounds for respiration and these processes demand significant amounts of iron (Vasil and Ochsner, 1999). The control of iron-regulated genes in Pseudomonas is affected by the ferric uptake regulator (Fur) (Prince et al., 1993; Ochsner et al., 1995), which is an iron-responsive, DNA-binding repressor that binds as a dimer to the Fur-box in the promoter regions of iron-regulated genes (Hennecke, 1990). In the presence of low iron, Fe2+ dissociates from the Fur protein, which then releases from the Fur-box, allowing transcription to occur. Examples of Fur-regulated genes are pchR and pvdS, which encode a positive activator of pyochelin synthesis (Heinrichs and Poole, 1993) and a probable alternative sigma factor (Miyazaki et al., 1995), respectively. A genome-scale examination of iron-regulated genes in Pseudomonas has also been performed and interestingly, it seems that there were more genes induced under high iron conditions than in iron starved cells (Ochsner et al., 2002). This highlights the importance of iron regulation to the survival of this microorganism. 22 It is now known that the production of exotoxin A, pyoverdin and pyochelin are all negatively-regulated by iron (Vasil et al., 1986; Cox and Adams, 1985). Pseudomonas is able to produce large amounts of siderophores (pyoverdin and pyochelin) that act as powerful iron chelators for iron transport through the bacterial membranes via specific receptor proteins (Heinrichs et al., 1991) and have a TonBlike system for the translocation of iron through the cytoplasmic membrane. Within cells, iron is then released from the ferrisiderophores by a reductive mechanism before it reaches targets (Halle and Meyer, 1992). The prerequisite of iron for the in vivo survival of Pseudomonas within the host dictates that the bacteria must compete with the host for iron bound to host proteins, such as ferritin, transferrin, lactoferrin and haemoglobin (Meyer et al., 1996). So far, there are reports that pyoverdin and pyochelin are able to remove iron from transferrin and lactoferrin and promote P. aerginosa growth in media containing these iron-binding proteins or human serum (Takase et al., 2000). However, whether or not other iron-binding proteins in human serum are also targets of siderophores remains unknown. 1.5 The Rationale and Aims of this Thesis Until now, most research on insect ferritins has focused on their purification, cloning and characterization in iron homeostasis, neglecting the implications of invertebrate secreted ferritin (henceforth referred to as plasma ferritin) during infection. One supporting evidence of plasma ferritin in immune defense was demonstrated in Drosophila whereby secretion of ferritin in the haemolymph was upregulated almost instantly (~25 min) after LPS challenge (Vierstrate et al., 2003). After a 4 h LPS challenge, there seems to be no change in ferritin protein level in the haemolymph although an extra spot appeared in the 2-dimensional gel electrophoregram, which may suggest a post-translationally modified form. As 23 compared to other plasma iron-binding proteins, plasma ferritins have a larger capacity for iron and it is conceivable that as a rich nutrient pool, they would be obvious targets for bacteria to pirate iron from the host. Since the cytosolic counterpart may serve as a DNA protector, the plasma ferritin should be the first to ‘sense’ the need for DNA protection during infection. Insights from such studies will offer greater understanding of the functions of plasma ferritin in invertebrate iron control and innate immunity, and enlighten the enigmatic role of their mammalian analogue. In order to demonstrate the dynamic role of plasma ferritin on invertebrate defense during infection, we developed a gram negative infection model in the horseshoe crab to monitor the ferritin protein profile, the ferritin gene regulation and the control of plasma iron, using Pseudomonas aeruginosa as a model pathogen. 24 2. MATERIALS AND METHODS 2.1 Materials Horseshoe crabs, C. rotundicauda, were collected from Kranji estuary, Singapore. They were maintained in tanks and allowed to acclimatize overnight prior to experiments. Pseudomonas aeruginosa, ATCC 27853, were cultured in tryptone soya broth (TSB, Difco, Detroit, MI). Lipopolysaccharide, LPS, from E. coli O55:B5 was a product from Sigma. Unless otherwise stated, all molecular biology grade chemicals were from Sigma. 2.2 Infection of horseshoe crab and preparation of cell-free haemolymph /plasma The infection of horseshoe crab and preparation of plasma was performed as described by Wang et al. (Wang et al., 2003). Horseshoe crabs were intracardially administered with either 1 x 106 cfu of P. aeruginosa, or 10 µg of LPS per kg body weight or 200 µl of 0.05 M FeSO4.7H2O. Ng et al. (Ng et al., 2004) established the sub-lethal dose of P. aeruginosa at 106 cfu. Plasma was collected from the horseshoe crab after various time points (3, 6, 12, 24, 48, 72 h) of P. aeruginosa infection. The plasma was centrifuged at 150 x g to remove the amoebocytes, after which PMSF was added to a final concentration of 0.1 mM. As negative control, untreated and salineinjected horseshoe crab was employed. A 3 h time-point was chosen to compare the effects of LPS and iron on the induction of ferritin gene. The resulting plasma was used for subsequent iron assays and Western analysis. 2.3 Oxidative activity of the plasma during P. aeruginosa infection 2.3.1 Supercoil relaxation assay 25 The role of iron as a catalyst in the generation of free radicals is well accepted. The amount of free radicals (hence the oxidative activity) thus mirrors the amount of iron in the plasma. The oxidative activity of the naïve and infected plasma was examined via DNA backbone breakage using modifications of the procedure of Surguladze et al. (Surguladze et al., 2004). Supercoiled plasmid pGEM-T Easy DNA was used as substrate. Reaction mixtures of 20-30 µL contained 0.5 or 1 µg DNA, dissolved in 10 mM Hepes (pH 7.5), 50 mM NaCl, 2.5 mM MgCl2, 2.5 mM DTT. Serial 10-fold dilutions of plasma, glycerol and chelators [EDTA, K3Fe(CN)6 and K4Fe(CN)6] were added. The reactions were quenched by addition of 10 µL of 50% glycerol, 50 mM EDTA and 0.1% bromophenol blue. Following electrophoresis in 1 % agarose gel, the relative amount of supercoiled, linear and relaxed forms were measured by the intensity of digitized gel images. 2.4 Purification and identification of plasma ferritin 2.4.1 Purification and resolution of plasma ferritin Naïve horseshoe crab plasma was subjected to KBr differential density ultracentrifugation using modifications of the method of Kim et al. (Kim et al., 2001) and Dunkov et al. (Dunkov et al., 1995). Briefly, 2.64 g of KBr was added to 3 ml plasma and 3 ml of PBS (140 mM NaCl, 2.7 mM KCl, 1.8 mM Na2HPO4) in a Beckman Quickseal tube (12 ml). The mixture was overlaid with 6 ml of 0.9 % saline, and ultracentrifuged at 200,000 x g for 16 h at 4 oC. The resulting dark brown pellet was resuspended in 1.5 ml of 0.05 M sodium phosphate buffer, pH 6.5. Major protein contaminants were removed by exploiting the thermostability feature of ferritin. The protein mixture was heated at 75 oC for 15 min, cooled on ice for 5 min and centrifuged at 15,000 x g for 30 min at 4 oC. The supernatant was subjected to a 26 second round of heat denaturation. The resulting colourless protein mixture was concentrated 100 times using Microcon YM-30 (Millipore). After addition of SDS and glycerol to final concentrations of 1 % and 10 %, respectively, to the partially purified ferritin, it was resolved on SDS-PAGE (6 %) under non-reducing conditions to identify the presence of the ferritin complex. Ferritins from various organisms have been shown to be resistant to SDS (Nichol and Locke, 1989) and addition of SDS into the protein sample facilitates the disruption of other high molecular weight complexes into their subunits. The gel was immersed in Prussian Blue (1 % w/v potassium ferrocyanide in 0.1 M HCl) until the band-of-interest was stained. The ferritin protein band was excised and transferred into a benzoylated dialysis tubing (MWCO of 2 kDa, Sigma-Aldrich, Avg. flat width 32 mm). Five milliliters of 0.375 M Tris-HCl (pH 8.8) was added and the protein was eluted in 1 x SDS running buffer at 50 V for ~ 20 h. The eluted protein was subsequently concentrated with 5 volumes of acetone, boiled for 5 min and resolved in SDS-PAGE. 2.4.2 Two-dimensional gel electrophoresis The partially purified ferritin from naïve and infected plasma (pooled from 3 to 72 hpi) was resolved by two-dimensional gel electrophoresis. Briefly, 50 µg of the protein sample was solubilized in 9.8 M urea, 2 % CHAPS, 50 mM DTT, 0.5 % ampholyte (v/v) and absorbed into IPG strips, pH 5-8 (Biorad) via active rehydration overnight. Next, isoelectric focusing was performed at 300 V for 1 h, 1000 V for 1 h, 3000 V for 1 h, 12000 Vh and lastly 500 V for 99 h. The proteins were then reduced in 2 % DTT (w/v) for 15 min, alkylated in 2.5 % IAA for 15 min, sealed in 1 % agarose and subjected to SDS-PAGE. The protein spots were detected by silver staining. 27 Fig. 2.1 Schematic view of the partial purification and enrichment of horseshoe crab plasma ferritin. The plasma was first subjected to KBr ultra-centrifugation for 16 h, after which the resulting brown pellet was recovered. Major protein contaminants were removed by denaturation of the sample at 75 oC twice, and the final supernatant containing ferritin was enriched at least 100x by Microcon. 28 2.4.3 Edman degradation for N-terminal sequencing of ferritin To determine the N-terminal sequence, the partially purified proteins were transferred to PVDF membrane as described in Western analysis. The band-of-interest was excised and subjected to Edman degradation using ABI Procise 494 Protein Sequencer. 2.4.4 Mass Spectrometric analysis of proteins The enriched plasma ferritin was stained with Coomassie Blue and excised. The gel pieces were washed with 50 mM NH4HCO3/50 % (v/v) acetonitrile and dehydrated with acetonitrile. The protein was then reduced with 10 mM DTT in 100 mM NH4HCO3 at 57 oC for 1 h, and S-alkylated with 55 mM IAA in 100 mM NH4HCO3 at room temperature for 1 h. In-gel digestion was carried out for 15 h with 12.5 ng/µl of trypsin at 37 oC. The digested protein fragments were extracted from the gel with 20 mM NH4HCO3, followed by 5 % formic acid in 50 % aqueous acetonitrile and finally 100 % acetonitrile. The protein was dried in a speed-vacuum and sent for MALDI-TOF and ES-Q-TOF analysis (Q-TOF 2, Micromass). The peptide mass fingerprint of the digested protein and raw MS/MS data from ES-Q-TOF was analysed by Mascot (http://www.matrixscience.com) against the MSDB (Mass Spectrometry protein sequence Database). Peptide sequences obtained by manual sequencing using ES-Q-TOF was analysed by MS-BLAST using Bork Group’s Advanced BLAST2 Search Service at EMBL (http://dove.embl- heidelberg.de/Blast2/msblast.htm) against the Swissprot database. 2.5 Cloning of horseshoe crab ferritin cDNA 29 In order to determine which form of ferritin (H- or L-chain) exist in the horseshoe crab as well as to understand the regulation of horseshoe crab ferritin during infection, the cloning of some of the full length ferritin subunits became necessary and was subsequently performed. 2.5.1 Degenerate RT-PCR of CrFer-H1 Degenerate primers were designed from the N-terminal sequence of the ferritin molecule: VQYDNDMKEP and two internal peptide fragments LLDYVNQR and DGLEALEDAMNLER. The horseshoe crab, C. rotundicauda cardiac total RNA was isolated by TRIZOLTM (Invitrogen) RNA extraction method. The plasma ferritin gene heavy chain form (CrFer-H) was amplified from naïve cardiac cDNA using NpFer: 5’-TAYGAYAAYGAYATGAARGARCC-3’, CGYTGYTTNACRTAATCAATTAG-3’ and RevpFer1: RevpFer2: 5’5’- TCNARRTTCATAGCRTCYTC-3’. 2.5.2 5’ and 3’ Rapid Amplification of cDNA ends (RACE) of CrFer-H1 After PCR amplification, two products of 240 bp and 340 bp were obtained and these were sequenced directly to obtain partial sequence information. The 5’ and 3’ ends of CrFer-H were isolated by 5’ and 3’ RACE using SMARTTM RACE cDNA Amplification Kit (Clontech). A 5’ RACE primer, 5RacepFer: 5’- TTGGCAAAACCCTTCCTGCCGACAGAGT-3’ was designed to obtain the 5’ UTR of CrFer-H, The remaining part of CrFer-H including its 3’UTR, was obtained by performing a 3’ RACE with 3RacepFer: 5’- ATACTCTTTGGACGACCGATGCATCAAC-3’ and a nested 3’RACE primer, 3RacenestedpFer: 5’-TATACATGAACATGGCGGCTCACTTTGG-3’. The final 30 RACE products were cloned into pGEM-T Easy vector (Promega) to determine the sequence of the insert using T7 and Sp6 primers. 2.5.3 PCR amplification and cloning of CrFer-H2 cDNA An EST clone of 600 bp, coding for ferritin-H, was isolated from a naïve cardiac cDNA library of the horseshoe crab (Ding et al., unpublished data). This clone, referred to as CrFer-H2, shows high homology to ferritin H chain polypeptide 1 from Branchiostoma lanceolatum (E-value of 3e-13) and contains the 3’ fragment of the horseshoe crab ferritin heavy chain. The 5’ end of CrFer-H was isolated by 5’ RACE using cardiac total RNA isolated by TRIZOLTM (Invitrogen) RNA extraction method. Using SMARTTM RACE cDNA Amplification Kit (Clontech), the 5’ fragment was obtained using the primer (5RaceCrFer): 5’- CCCATTCCATCATTCATAGCGCCAAGCT-3’. Table 3: Summary of primers used in the cloning of ferritin genes Gene Primer sequence CrFer-H1 5’-TAYGAYAAYGAYATGAARGARCC-3’ Primer name NpFer 5’-CGYTGYTTNACRTAATCAATTAG-3’ RevpFer1 5’-TCNARRTTCATAGCRTCYTC-3’ RevpFer2 5’-TTGGCAAAACCCTTCCTGCCGACAGAGT-3’ 5RacepFer 5’-ATACTCTTTGGACGACCGATGCATCAAC-3’ 3RacepFer 5’-TATACATGAACATGGCGGCTCACTTTGG-3’ 3RacenestedpFer CrFer-H2 5’-CCCATTCCATCATTCATAGCGCCAAGCT-3’ 5RaceCrFer 31 2.6 Regulation of ferritin during P. aeruginosa infection 2.6.1 Northern analysis of CrFer-H2 The tissue distribution of CrFer-H gene expression was examined in the hemocytes, hepatopancreas, heart, intestines, muscle and stomach, which were procured from either unchallenged (naïve) or FeSO4- or LPS-challenged horseshoe crab. The tissues were immediately frozen in liquid nitrogen prior to homogenization and extraction of total RNA. Ten microgram of total RNA was separated on 1.0 % formaldehyde-agarose gel, and blotted in a solution containing 20 X SSC (3 M NaCl, 0.3 M sodium citrate). Nucleic acid hybridization was carried out at 65 oC in a hybridization buffer containing 2X SSC, 0.1 % SDS, 5X Denhardt, 10 mg/ml of calf thymus DNA. The 5’ RACE fragment of CrFer-H (~ 550 bp) and 3’ RACE fragment of CrFer-H2 (~ 850 bp) were labeled with [α-32P] dCTP (Amersham) by random priming with RediprimeTM II (Amersham Biosciences). Either of these probes was used in the hybridization step. Membranes were washed for three cycles of 15-min each in 2X SSC/ 0.1% SDS at room temperature, 1X SSC/ 0.1% SDS at 37 oC and lastly 0.2X SSC/ 0.1% SDS at 42 oC. The hybridized membranes were exposed to Kodak films at -70 oC. 2.6.2 Western analysis After SDS-PAGE, the proteins were transferred in 25 mM Tris, 192 mM glycine, 20 % v/v methanol, to PVDF membrane (Bio-Rad). The membrane was blocked for 2 h at room temperature with 5 % skimmed milk, and hybridized with Manduca sexta anti-ferritin antibody (1:1000), followed by secondary anti-rabbit IgG conjugated with horseradish peroxidase (1: 10,000). The ferritin subunits were detected using Supersignal West Pico Chemiluminescence substrate (Pierce). 32 Polyclonal M. sexta anti-ferritin antibodies were raised in rabbits using ferritin purified from insect larval haemolymph. . 2.7 Strategy employed by P. aeruginosa to ‘steal’ host iron 2.7.1 Reaction between plasma ferritin and P. aeruginosa Two milliliters of naïve plasma was incubated for 24 h in 1 ml of TSB with or without 106 cfu of P. aeruginosa. The bacteria was pelleted at 10,000 x g for 5 min and the solution was clarified by centrifugation at 20,000 x g for 30 min. Plasma ferritin was then prepared and detected by Western analysis as previously described. 2.7.2 Quantification of total and labile iron pool in the plasma During infection, the in vivo host-pathogen competition for iron is highly dynamic and complex. The host attempts to withhold iron within cells and at the same time, the pathogen attempts to obtain host iron. In order to detect a possible change in the population of tightly- or weakly-associated iron during an in vivo infection, total iron and labile iron pool (LIP) of uninfected or infected plasma were determined in triplicates using modification of the procedure of Williams et al. (Williams et al., 1977). For total iron quantification, 50 µl of plasma or iron standard, (NH4)2(FeSO4)2 was incubated for 15 min at 37 oC with 250 µl of total iron reagent buffer, 10 g/L 2,4,6-tripyridyl-s-triazine [TPTZ] in 1 ml of 1 M HCl, 0.1 M sodium acetate buffer, pH 4.8, to which 6 M guanidine hydrochloride, and 5 g/L ascorbic acid were added just before analysis. The absorbance of each well of the 96-well plate was measured at 593 nm in an ELISA reader after being blanked with MilliQ water. The sample blank contained plasma and total iron reagent buffer without TPTZ, while the reagent blank contained MilliQ water and total iron reagent buffer. For LIP measurement, 50 µl of 33 plasma or iron standard was incubated with 250 µl of free iron reagent buffer (10 g/L 2,4,6-tripyridyl-s-triazine [TPTZ] in 1 ml of 1 M HCl, 0.1 M sodium acetate buffer, pH 4.8, 5 g/L ascorbic acid). The absorbance of each well of the 96-well plate was measured at 593 nm in an ELISA reader after being blanked with MilliQ water. The use of a low pH without 6 M guanidine hydrochloride would be expected to only cause dissociation of Fe3+ ions that are loosely-associated (labile). For instance, iron sequestered in the core of the holoferritin complex would only be released when the complex is disrupted by extreme pH of < 2 or > 10 (Caccova et al., 2002). Hence the use of two different conditions allows us to discriminate iron that is tightly-associated to proteins (such as being sequestered by ferritin) and the LIP. 2.7.3 Measurement of redox potential and pH It is well-known that the redox potential and pH directly control the binding of iron to iron-binding proteins (eg. transferrin) in human body fluids, hence its bioavailability for invading pathogens (Bullen et al., 2000). Pseudomonads are wellknown to employ siderophores as host iron-grabbers. To examine if P. aeruginosa employs other strategies to compete for host iron in vivo, the redox potential and pH of the naïve and infected plasma were measured. 2.8 Plasma ferritin-DNA interaction 2.8.1 Electrophoretic mobility shift of DNA by ferritin To verify the ability of ferritin to bind DNA, 10 or 50 µg ferritin purified from plasma samples of 3 h post-infection (hpi), 72 hpi or 3 h iron-load was reacted with 1 x 105 cpm/pmol of either one of two homologous DNA probes derived from the C. rotundicauda Factor C promoter (Wang et al., 2003): (i) LκBCom, a 65-mer 34 oligonucleotide: (5- GAAATTTTTCCTTCTTGTACATTGGAAAACGTTTTCACGTGACGTACTG ATTTGTCTGTCATGCA-3) or (ii) LDorThr, a 34-mer oligonucleotide (5 CATGCACGAGAAAAAAGCCGGGAAATCCATTAGA-3). The ferritin samples used in each DNA binding reaction was partially purified and enriched by KBr ultracentrifugation, heat denaturation and microcon concentration. Each binding reaction was incubated in 20 µl of binding buffer (10 mM HEPES, pH 7.9, 50 mM NaCl, mM EDTA, 2.5 mM DTT, 10 % glycerol and 0.05 % NP-40) at 25 oC for 30 min before electrophoresis on a native 4 % PAGE gel (with acrylamide and bisacrylamide ratio of 79:1). 2.8.2 Fluorescence measurement of ferritin complex-DNA interaction The environment whereby tryptophan residues reside in ferritin, in the absence or presence of a DNA probe was examined using luminescence spectrometer (Perkin Elmer, LS50B). Briefly, the 34-mer LDorThr probe was incubated in 50 mM sodium phosphate buffer, pH 6.5. The excitation wavelength was 280 nm using a slit width of 10 nm. The background fluorescence emission was scanned from 300 to 400 nm. Fluorescence emission was recorded for the naïve or infected (pooled 3-72 hpi) ferritin complex after incubation for 10 min with the same DNA probe in a final volume of 75 µl. 2.9 Statistical analysis The statistical analyses (P values) of the iron assays, redox potential and pH measurements of the horseshoe crab plasma over various time points of P. aeruginosa infection were performed by a Student’s t-test (Snedecor and Cochran,1967). 35 2.10 Homology modeling methods The predicted three-dimensional structure of CrFer-H1 and –H2 were homology modeled using SWISS-MODEL (http://swissmodel.expasy.org/ ). In a typical SWISS-MODEL First Approach Mode, there are basically 5 steps. The first step involves the search for an appropriate template (at least > 40 % homology) via a BLASTP programme. Next, the output sequence identity is checked against the target after which ProModII jobs are created. ProModII is also used to generate models and the final predicted structure is produced after energy minimization is calculated using Gromos96. 36 Fig. 2.2 Diagrammatic representation of the iron assay to measure total plasma iron and LIP. At pH 4.6, iron is released from proteins such transferrin and other weakly associated molecules. The released ferric or ferrous ions are reduced by ascorbic acid to yield ferrous ions, which in the presence of TPTZ, allows the measurement of the iron concentration at OD593. In the presence of a strong denaturant such as guanidine hydrochloride, all protein-bound iron will be released and this allows us to measure total plasma iron. 37 3. RESULTS 3.1 Iron contributes to the high oxidative activity of plasma. Transition metals, such as iron, are known to generate free radicals via Haber- Weiss reaction or the superoxide-driven Fenton reaction (Fig. 3.1A) that produces free radicals (Wentworth et al., 2000). These free radicals would be expected to accelerate host tissue injury if the iron is not tightly regulated. DNA nicking serves as a convenient and rapid assay to detect oxidative activity as a result of free radicals production, hence it is employed in the determination of oxidative activity of the limulus plasma. In accordance to the high iron content, naïve horseshoe crab plasma is highly oxidative and DNA nicking is observed when supercoiled plasmid DNA is incubated with 10x diluted plasma (Fig. 3.1B). DNA nicking also occurs via a free radical mechanism (Fig. 3.1C) as shown by the dose-dependent inhibitory effect of glycerol on DNA nicking. Increasing concentrations of EDTA rescued supercoiled plasmid DNA from being nicked, which suggests the involvement of metal ions (Fig. 3.1D). Since numerous transition metals can lead to generation of free radicals, we sought to determine the metal ion responsible for DNA nicking. Using potassium ferrocyanide and ferricyanide that react specifically with the ferric and ferrous iron respectively, we confirmed that iron is one of the major metal ions responsible for the generation of free radical (Fig. 3.1D). Surprisingly, DNA nicking was inhibited only at high concentrations of potassium ferrocyanide and ferricyanide and there was no dose-dependent inhibition with these chemicals. It was thus pertinent to determine if more free radicals would be generated during the early phase of bacterial infection and thus accelerate host tissue injury. Surprisingly, we did not detect any significant change in the oxidative effect of the plasma throughout 72 h of infection (Fig. 3.1E). 38 (A) (B) (C) 39 Fig. 3.1 Plasma iron regulates free radical-induced DNA nicking. Supercoiled plasmid pGEM-T Easy DNA was used as substrate. Reaction mixtures of 20-30 µL contained 0.5 or 1 µg DNA, dissolved in 10 mM Hepes (pH 7.5), 50 mM NaCl, 2.5 mM MgCl2, 2.5 mM DTT. Serial 10-fold dilutions of the plasma, glycerol and chelators [EDTA, K3Fe(CN)6 and K4Fe(CN)6] were added. Reactions were quenched by addition of 10 µL of 50 % glycerol, 50 mM EDTA and 0.1 % bromophenol blue. Following electrophoresis in 1 % agarose gel, the relative amount of supercoiled, linear and relaxed forms were measured by the intensity of digitized gel images. (A) The role of iron as a catalyst in the Haber-Weiss reaction. The reactive oxygen species (such as hydroxyl radicals) then react with biological macromolecules, causing damage in the process. (B) Concentration-dependent DNA nicking activity by naïve plasma and the effect of nicking buffer. Approximately 600 ng of plasmid DNA is degraded within 30 min of incubation at room temperature. (C) Effect of glycerol as a free radical scavenger in the highly oxidative plasma. The proportion of each topoisomer was calculated by normalising against total amount of plasmid DNA loaded in each well (as determined by ImageMaster). 40 (D) (E) Fig. 3.1 (D) Effect of metal chelators on oxidative activity of plasma using EDTA, potassium ferrocyanide & ferricyanide. There is DNA protection by all three chelators. There is possibly more Fe(II) than Fe(III) in the plasma. R, L and S annotate relaxed, linear and supercoiled forms of the pGEM-T Easy plasmid DNA. (E) Oxidative activity of plasma during P. aeruginosa infection of the horseshoe crab. There was no significance change in oxidative effect in the plasma over the course of infection. Reaction mixtures of 20-30 µL contained 0.5 or 1 µg supercoiled plasmid pGEM-T Easy DNA, dissolved in 10 mM Hepes (pH 7.5), 50 mM NaCl, 2.5 mM MgCl2, 2.5 mM DTT and 1 µl of 10 x diluted naïve or infected haemolymph. After 30 min incubation at room temperature, the reaction mixture was resolved by gel electrophoresis in 1 % agarose gel. The relative amounts of supercoiled, linear and relaxed forms were measured by the intensity of digitized gel images. Each bar is a result of at least three independent experiments. Statistical significance was calculated using a two-tailed Student’s T-Test. 41 3.2 Horseshoe crab plasma ferritin is made up of subunits of 21 kDa. After KBr ultracentrifugation, heat denaturation and microcon concentration, the enriched plasma ferritin is resolved in native SDS gel. The partially purified plasma ferritin complex exists as monomer and dimer in the native state, with an apparent molecular mass higher than that of horse ferritin (Fig. 3.2A). By KBr ultracentrifugation and heat treatment, significant enrichment of the horseshoe crab plasma ferritin was observed (Fig. 3.2B, lane 5). Both Prussian blue -stained bands from the non-reducing SDS-PAGE, when excised and electroeluted, were of 21 kDa size (Fig. 3.2B), suggesting that the horseshoe crab plasma ferritin complex comprises only one form of subunits of 21 kDa. The 21 kDa plasma ferritin subunit was recognized by Manduca sexta anti-ferritin antibody, reinforcing that the horseshoe crab plasma ferritin is made up of a single 21 kDa subunit. MS-BLAST using the Nterminal sequence VQYDNDMKEP and the partial sequences LIDYVNQR, DGLEALEDAMNLER (Fig. 3.2C) and further confirmed that the 21 kDa is a ferritin protein, with highest homology to Lymst yolk ferritin precursor (SwissProt P42578). 42 (i) (ii) 43 Fig. 3.2 Identification of limulus plasma ferritin complex and its subunits in the plasma. (A) The native state of limulus plasma ferritin was detected by Prussian blue staining. (B) The ferritin complex is made up of 21 kDa subunits. The native complex was electroeluted from SDS-PAGE and dissociated into its 21 kDa subunits. The ironstained band was excised and subjected to heat treatment, electroeluted, followed by silver stain. Plasma ferritin is significantly enriched after KBr differential density ultracentrifugation and a 21 kDa protein band can be specifically recognized by Manduca anti-ferritin antibody. Lane 1: protein marker; lane 2-4: Electroeluted proteins after native PAGE from limulus ferritin dimer, monomer and horse ferritin respectively, followed by silver staining; lane 5: Coomassie staining of the enriched ferritin; lane 6: Western analysis of limulus ferritin using M. sexta anti-ferritin antibody. (C) Protein sequencing of 21 kDa band. The N-terminal sequence and sequences of the trypsin-digested protein fragments are shown. 44 3.3 CrFer-H1 has 2 possible transcripts that are translated to plasma ferritin proteins To study the gene regulation of plasma ferritin in horseshoe crab, the corresponding gene was cloned from naïve cardiac cDNA. Initial degenerate primers designed from peptide sequences derived from protein sequencing yielded dominant bands (~ 240 and 330 bp) were used in RT-PCR with template mRNA of naïve heart, intestine and stomach. Two major cDNAs (Fig. 3.3A) of 240 and 330 bp were obtained. Direct DNA sequencing of these bands without cloning into pGEM-T vector using their corresponding gene-specific primers indicated that they possibly originated from the same gene and this is confirmed from the alignment of their deduced amino acids (Fig. 3.3B). RACE primers were then designed using the partial DNA sequence (Fig. 3.3C) and 5’RACE led to amplification of a single band (~ 500 bp), while 2 PCR products (~ 650 and 850 bp) were observed in the 3’RACE followed by a nested PCR (Fig. 3.3D). These RACE fragments were cloned in pGEM-T Easy vector and positive clones were screened by EcoRI digestion for subsequent determination of full length sequence (Fig. 3.3E). Interestingly, 2 cDNA sequences, henceforth designated as CrFer-H1a and –H1b due to their identical coding sequence but different transcript lengths, were obtained (Fig. 3.3F). The CrFer-H1 shows high homology to ferritin from Branchiostoma belcheri tsingtaunese (6e-36 and 3e-37, respectively). In accordance to results from protein sequencing, the CrFer-H1 protein contains the experimentally derived N-terminal sequence and two of three peptide fragments. However, there were some discrepancies between the experimentally derived peptide sequence ‘LIDYVNQR’ the deduced peptide sequence ‘LIDYVNKR’. This could possibly be due to polymorphism among ferritin H-subunit among individual horseshoe crabs since cDNA template for RT-PCR was prepared from total RNA from several individuals. The full length cDNAs of CrFer- 45 Naive Heart R1 No Intestine Stomach template R2 R1 R2 R1 R2 R1 R2 H2O bp 400 300 200 100 330 240 Fig. 3.3 (A) PCR products of the same size (~ 240 and 330 bp) were obtained from degenerate RT-PCR using naïve heart, intestine and stomach cDNA as template. R1 and R2 denote reverse degenerate primer 1 and 2. 46 (B) R1HT_Protein R2HT_Protein R1INT_Protein R2INT_Protein R2STC_Protein CrFerH Consensus .................X.......DNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI X DNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI .................X.......DNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI X DNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI .................X.......DNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI X DNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI ................FX.......XNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI FX NDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI .....G..........FX.......DXDMKEPKTGRYSLDDRCINAIQHQINEEMHASLI FX D DMKEPKTGRYSLDDRCINAIQHQINEEMHASLI MAAMMGKSLVLLVLTFFSTIETVRHDNDMKDSSMDRYILDNKCINGLQLQINEERHASLV F DNDMKD DRY LDNKCINGLQ QINEE HASLV VQYDNDMKEP dmk ry ld cin q qinee hasl 36 36 36 37 38 60 R1HT_Protein R2HT_Protein R1INT_Protein R2INT_Protein R2STC_Protein CrFerH Consensus YMNMAAHFGRNSVGRKGFAKFFKHSSD G YMNMAAHFGRNSVGRKGFAKFFKHSSDXGXXXCTN......................... YMNMAAHFGRNSVGRKGFAKFFKHSSDEERE..HAQKLIDYVNKRSGKVIAFDIKMP.G. Y MNMAAHFGRNSVGRKGFAKFFKHSSDEERE HAQKLIDYVNKRSGKVIAFDIKMP YMNMAAHFGRNSVGRKGFAKFFKHSSXQXVEXMHR......................... Y MNMAAHFGRNSVGRKGFAKFFKHSS Q E H YMNMAAHFGRNSVGRKGFAKFFKHSSDEERE..HAQKLIDYVNKRSGKVIAFDIKMP.R. Y MNMAAHFGRNSVGRKGFAKFFKHSSDEERE HAQKLIDYVNKRSGKVIAFDIKMP YMNMAAHFGRNSVGRKGFAKFFKHSSDEERE..HAQKLIDYVNKRSGKVIAFDIKMP.R. Y MNMAAHFGRNSVGRKGFAKFFKHSSDEERE HAQKLIDYVNKRSGKVIAFDIKMP YMHMASHFGSNAVGRKGFSKFFKHSSDEERE..HAQKLIDYINKRSGWVAAFDIKMPGKT Y MHMA HFG N VGRKGF KFFKHSSDEERE HAQKLIDYINKRSG V AFDIKMP LIDYVNQR ym ma hfg n vgrkgf kffkhss 71 92 71 93 94 118 R1HT_Protein R2HT_Protein R1INT_Protein R2INT_Protein R2STC_Protein CrFerH Consensus ............................................................ ..K.D.EV.EGWL..E....................P....................... K D P ............................................................ ..K.G.SG.RMAX............................................... K G ..K.G.SG.KDGX..S....................P....................... K G P IWKNGMEALQDALNLENHVNNKLHHLHQMADKICADPHLMNFLEGEFLTEQVESINELNT K G P 71 102 71 101 104 178 R1HT_Protein R2HT_Protein R1INT_Protein R2INT_Protein R2STC_Protein CrFerH Consensus ......................... ......................... ......................... ......................... ......................... FISQLGAMNDGMGEYLLDRELLEKS DGLEALEDAMNLER 71 102 71 101 104 203 Fig. 3.3 (B) Alignment of the deduced amino acid sequence of the 240 and 330 bp PCR products show that they may be encoded by the same gene. Amino acid sequences that harbor the sequence from protein sequencing are shown in red, with the corresponding sequence highlighted in green. 47 (C) Fig. 3.3 (C) Design of 5’ and 3’ RACE primers using the partial ferritin DNA sequence. Two 3’ RACE primers were designed to facilitate more specific amplification. Nucleotide sequences underlined indicate positions of the RACE primers. 48 (D) 5’RACE o 55 C o 3’RACE o o 60 C 65 C 55 C 60oC 65oC 3’RACE nested 55oC 60oC 65oC Nested PCR 66oC 1000 500 100 Fig. 3.3 (D) 5’ and 3’ RACE products of ferritin gene. A 5’ RACE product of ~ 500 bp and 2 3’ RACE products (~ 650 and 850 bp) were obtained using naïve heart cDNA as template (shown with arrows). Three different annealing temperatures were employed to determine the optimum annealing temperature for RACE. 49 (E) 5’RACE Clone no. 1 9 10 3’RACE 1 2 3 5 6 9 11 bp 9416 6557 4361 bp 2322 2027 1500 1000 500 564 100 Fig. 3.3 (E) Screening of positive clones by EcoRI digestion of pGEM-T Easy vector. Clones that harbor the insert of correct size of the 5’ and 3’ RACE products of CrFer-H1a and -H1b were subjected to DNA sequencing. 50 1 acgcgggggtttcataaaatataaagttctttgaatgagagccttctgtaccagtgggtg 61 caaaggctaattggcattgtatttgtttactaatttcagaaaatcgaatagtatttgctt 121 ttgtagctgttgacagtggcattcataccacagatactcagaggctgaaaccgtctaaag 181 1 ttgaatctctacgttgaaattgattttgaggtgtttttgagagtggcagccatgatggaa M M E 241 4 agagtgttcttgttagttgttctcgccctcggttctaccgtagtaggggtccagtatgac R V F L L V V L A L G S T V V G ↓V Q Y D 301 24 aatgacatgaaagaacctaaaaccgaccgatactctttggacgaccgatgcatcaacgcc N D M K E P K T D R Y S L D D R C I N A 361 44 attcagcatcagatcaatgaagaaatgcacgctagtctaatatacatgaacatggcggct I Q H Q I N E E M H A S L I Y M N M A A 421 64 cactttggcaggaactctgtcggcaggaagggttttgccaaattcttcaagcacagctcg H F G R N S V G R K G F A K F F K H S S 481 84 gacgaagaaagagagcatgcgcaaaagctaatcgattacgttaacaaacgaagtggcaaa D E E R E H A Q K L I D Y V N K R S G K 541 104 gtgattgcatttgatattaagatgccaggaaaggatgagtggaaggatggcttggaagca V I A F D I K M P G K D E W K D G L E A 601 124 ctggaagatgctatgaatttggagagacacgttaacaacaagttacaccatcttcatcac L E D A M N L E R H V N N K L H H L H H 661 144 atggctgataaaatttgcagtgacccacacttgatggactacattgagggagagtttctt M A D K I C S D P H L M D Y I E G E F L 721 164 acagaacaagtggaatcaatcaatgaatttaaaacctacattagccagcttggagccatg T E Q V E S I N E F K T Y I S Q L G A M 781 184 aacaatggtatgggagagtacctgtttgatcaccagttgctggagaaaagtagttaagaa N N G M G E Y L F D H Q L L E K S S * 1a 1b 841 841 aagaagctaaaacctgagtattcattacatcagttagttcctcagtttatttctttaata aagaagctaaaacctgaagtattcattacatcagttagttcctcagtttatttctttaat 1a 1b 901 901 cttcttttttattgtcacatgcttattcagtatgccaagtttggtcatgttttcttttta acttcttttttattgtcacatgcttattcagtatgccaagtttggtcatgttttcttttt 1a 1b 961 961 aatgaattttaacaaatatatgtggtttacaaatgggcaaaaaaaaaaaaaaaaaaaaaa aaatgaattttaacaaaatatatgtggtttacaaatggtttccattgggattttgttttt 1a 1b 1021 1021 aaaaa ttttgtagaaccttaggactatcaacagtattacttctgtagaagatgtaaggtataatg 1b 1b 1081 1141 cttgggatttttaattggattttcatgacttgtcctgggtataaacaaaaaactttttta tgagactaataaaatacagtaaaacccgaaaaaaaaaaaaaaaaaaaaaaaaaaaa Fig. 3.3 (F) Nucleotide sequence and deduced amino acid sequence of CrFer-H1a and -H1b. The RACE primers are shown in red and indicated by an arrow. Amino acids highlighted in yellow correspond to the peptide sequences obtained from Edman degradation and mass spectrometry. There is an IRE with the signature 5’-CAGTGC3’ and a C six nt upstream of the IRE loop (underlined) in the 5’UTR. Cleavage site: ↓; polyadenylation sites: double underlined; A+U destabilizing elements: bold and italicized; Phosphorylation sites as predicted by NetPhos 2.0 Server (http://www.cbs.dtu.dk/services/NetPhos/): bold and blue. 51 H1a and -H1b (GenBank Accession AY691511 and AY691509) are ~ 1025 bp and 1197 bp respectively. Both transcripts share a 334 bp 5’ UTR followed by an ORF of 606 bp. They differ only in their 3’ UTR where that of CrFer-H1a and –H1b are of 188 bp and 359 bp, respectively (Fig. 3.3F). The 3’ UTR of CrFer-H1b also contains a putative polyadenylation site, [A(T/A)TAAA], which ends 43 bases upstream from the poly(A) tail. The ORF of CrFer-H1 codes for 201 amino acids and the mature protein is predicted to be made up of 182 amino acids with a molecular size of 21.2 kDa. The predicted pI of CrFer-H1 is ~ 5.60. Computational analysis by NetPhos 2.0 Server predicted the presence of multiple possible phosphorylation sites on Ser and Tyr residues on CrFer-H1. 3.4 Another ferritin gene, CrFer-H2, codes for a secretory protein that is apparently absent in the plasma With a 3’ fragment (~ 650 bp) of a ferritin homologue obtained from an EST library, the 5’ end of another horseshoe crab secretory ferritin, designated as CrFerH2, was isolated (Fig. 3.4A and B). The full length CrFer-H2 cDNA (GenBank Accession AY691510) of 1159 bp, contains a 5’ UTR of 220 bp followed by an ORF of 612 bp. It is flanked by a 3’ UTR of 327 bp containing four putative polyadenylation sites [A(T/A)TAAA] which end 25 bases upstream from the poly(A) tail (Fig. 3.4C). The 3’ UTR also contains five AU-rich motifs (TATT or ATTTA). These sequences are present in many short-lived mRNAs such as oncogenes and cytokine mRNAs (Wentworth et al., 2000; Shaw and Kamen, 1986; Hennessey et al., 1989). Like ferritin mRNAs from various organisms, the CrFer gene contains a putative iron-response element (IRE) in the 5’ UTR. CrFer-H2 encodes a protein of 204 amino acids with a 22 amino acid signal peptide (MAAMMGKSLVLLVLTFFSTIET), which is cleaved to produce a mature 52 (A) (B) 3’ Fragment of CrFer-H from naïve cardiac cDNA library 1 2 3 4 5 6 5’RACE 55 60 65 bp 9416 6557 4361 2322 2027 500 564 100 Fig. 3.4 Cloning of CrFer-H2. (A) Screening of clones that harbor the 3’ fragment of CrFer-H2 after EcoRI digestion of pGEM-T Easy vector. The sequence obtained were then used to design a 5’ RACE primer to obtain the 5’ end of CrFer. (B) 5’ RACE product of CrFer-H2 at various annealing temperatures using naïve cardiac cDNA as template. A prominent DNA band of ~ 750 bp was obtained. 53 (C) 1 gtcgtatttaagaaactaattcgcgataagaagcgtactttgatcaacagtcttctgtgc 61 cagtgagtgcaaagactgagtaacaattatcaagcttttattcattgacttacataaact 121 atatcattactgttcttcgtcttttaatagtgaagttaaaactgcagttttccagcgata 181 1 aacgtcatattctcatcgcctcagcttaccttaacgtctatggctgccatgatgggaaaa M A A M M G K 241 8 tctcttgtcttattggtgctcactttcttctccacgatagagacggttaggcatgataat S L V L L V L T F F S T I E T ↓V R H D N 301 28 gatatgaaggattcttctatggatcggtatattttggacaataaatgcattaatggcctt D M K D S S M D R Y I L D N K C I N G L 361 48 caactacagatcaatgaagaaaggcatgctagtttggtgtacatgcatatggcttcccac Q L Q I N E E R H A S L V Y M H M A S H 421 68 ttcggtagcaacgctgttggtaggaagggcttcagcaagttttttaaacacagctcagat F G S N A V G R K G F S K F F K H S S D 481 88 gaggaaagggagcatgcgcagaaactaattgattatattaataaacgtagtggttgggtg E E R E H A Q K L I D Y I N K R S G W V 541 108 gctgcttttgacattaagatgccaggaaagacgatctggaagaatggcatggaagcactc A A F D I K M P G K T I W K N G M E A L 601 128 caagatgcactaaacctggagaatcatgtgaacaacaagttacatcacctccaccagatg Q D A L N L E N H V N N K L H H L H Q M 661 148 gctgataagatctgtgctgatcctcatttgatgaactttcttgaaggagagttcctcacc A D K I C A D P H L M N F L E G E F L T 721 168 gaacaagtggagtccatcaatgaactaaacaccttcatcagtcagcttggcgctatgaat E Q V E S I N E L N T F I S Q L G A M N 781 188 gatggaatgggagaatacctgcttgatcgtgagttgctcgaaaagagtaactaaattaaa D G M G E Y L L D R E L L E K S N * 841 901 961 1021 1081 1141 caccacaaaatagtcattttataacaatttttttttgtattagtctagtacaacttctgt actttcttcaactttggatatcacctagaacataagtatttttttatcaattttatcggt acatttcgtgttgcatttgtacagtagcttttgattaaaacattaatgagtttgttttat cttagagaagttcataatattaaggagttttgaagttgcattcttgacattaacaagaac ttgtttgtatttttgttttgtgattgaaataaatttctcttatattgcattaaaaaaaaa aaaaaaaaaaaaaaaaaaa Fig 3.4. (C) Nucleotide sequence and deduced amino acid sequence of CrFer-H2. The nucleotides boxed represent the primer employed for 5’ RACE. There is an IRE with the signature 5’-CAGTGC-3’ and a ‘C’ six nt upstream of the IRE loop (underlined) in the 5’UTR. Putative secretory signal as predicted by PSORT (http://psort.nibb.ac.jp/psort): ↓; polyadenylation sites: double underlined; A+U destabilizing elements: bold. Phosphorylation sites as predicted by NetPhos 2.0 Server (http://www.cbs.dtu.dk/services/NetPhos/): and blue. ferritin from Drosophila, Aedes mosquito, bold Manduca sexta and Lymst yolk ferritin (Fig. 54 polypeptide of 182 amino acids with a predicted mass of 21 kDa and an estimated pI of 5.70 (Fig. 3.4C). 3.5 Common features of the 3 ferritin cDNAs There are multiple AU-rich motifs (TATT or ATTTA) in the 3’ UTR of CrFer-H1 and CrFer–H2. These sequences are present in many short-lived mRNAs such as oncogenes and cytokine mRNAs (Shaw and Kamen, 1986; Hennessey et al., 1989; Brown et al., 1996). Like ferritin mRNAs from various organisms, the horseshoe crab ferritin genes contain a putative IRE in the 5’ UTR. The three ferritin cDNAs also contain IRE at the 5’ UTR, which is predicted to fold into a typical stemloop structure for the binding of IRE-binding proteins under iron-depleted conditions (Fig. 3.5A). Structure prediction of CrFer-H1 and –H2 also revealed that they possess the typical structure of ferritins, with four long and one short α-helices (Fig. 3.5B). However, it appears that the N-terminal of CrFer-H2 is longer than that of CrFer-H1 and that it forms a loop which is located at the center of the four α-helices. Phylogenetic analysis revealed that both ferritin cDNAs are clustered with secreted ferritin from Drosophila, Aedes mosquito, Manduca sexta and Lymst yolk ferritin (Fig. 3.5C). To date, reports have documented that vertebrate ferritin genes do not have secretory signals and are mainly intracellular. The invertebrate ferritins are more divergent than the vertebrate ferritins. Sequence alignment of CrFer-H1 and –H2 shows that they share ~ 72 % identity. Similar to ferritin-H from other organisms, the critical residues for ferroxidase activity are present in CrFer-H1 and -2 (Fig. 3.5D). In the vertebrate ferritin-H, the ferroxidase centre is made up of Tyr25, Tyr28, and Tyr30, and polynuclear Fe-complex formation involves Glu58 and His61 (Brown et al., 1996; Lind et al., 1998). The proposed structure of the human H-ferritin 55 CrFer-H1 A C C G T CrFer-H2 G G Chicken-H Crayfish CrFer1_2 CrFerH Echinoderm Human-H Human-L Mouse-L Rana-H Rat-H Rat-L Snail-S Consensus TCCTGCTTCAACAGTG TTG ACGGAAC TCCTGCTTCAACAGTGCTTGGACGGAAC. TCCGGCTTCGCCAGTGTGTGAACGAGCT. T CC GCTTC CAGTGT TGAACG GCCTTCTGTACCAGTGGGTGCAAAGGCT. CCT CT A CAGTG TG A G GTCTTCTGTGCCAGTGAGTGCAAAGACT. CT CT CAGTG TG A GA TTGTGCGTTCGCAGTGTCGAAACCAAGC. T TGC T CAGTGT AAC A C TCCTGCTTCAACAGTGTTTGGACGGAAC. T CCTGCTTCAACAGTGTTTG ACGGAAC TCTTGCTTCAACAGTGTTTGAC..GAACA T C TGCTTCAACAGTGTTTGA GAAC TCTTGCTTCAACAGTGTTTGAACGGAAC. T C TGCTTCAACAGTGTTTGAACGGAAC TCTTGCTTCAACAGTGTGTGAACGAGCT. T C TGCTTCAACAGTGT TGAACG TCCTGCTTCAACAGTGCTTGAACGGAAC. T CCTGCTTCAACAGTG TTGAACGGAAC TCTTGCTTCAACAGTGTTTGGACGGAAC. T C TGCTTCAACAGTGTTTG ACGGAAC TCTTGCTGCGTCAGTGAACGTACAGAAA. T C TGCT C CAGTG G AC GAA c cagtg 28 28 28 28 28 28 27 28 28 28 28 28 Fig. 3.5 (A) Predicted secondary structures of CrFer IRE at the 5’ UTR and the alignment of CrFer IRE sequence with that from other organisms. Both IREs are predicted to fold into a stem-loop structure with the consensus sequence ‘CAGTGN’ and a ‘C’ 6 nt upstream using default parameters in both Genebee (http://www.genebee.msu.su/ ) and RNAfold software (http://rna.tbi.univie.ac.at/cgibin/RNAfold.cgi ). 56 CrFer-H1 CrFer-H2 N N C C Fig. 3.5 (B) Both CrFer-H1 and –H2 are predicted to possess the typical 5 αhelices of ferritins. Structure of CrFer-H1 and –H2 are predicted by homology modeling using Swiss-Model. Templates for CrFer-H1 are pdb 1fha and 2fha (48.1 % identity for both), while templates for CrFer-H2 are pdb 1rcg and 1mfrO (42.1 % and 49.5 % identity respectively). 57 Secreted Cytosolic Vertebrate L-chain Vertebrate H-chain Fig. 3.5 (C) Phylogenetic analysis of CrFer-H1 and –H2. Both CrFer-H proteins form a cluster with secreted ferritin from Drosophila, Aedes mosquito, Manduca sexta and Lymst. 58 Fig. 3.5 (D) Central region of subunits of human H-ferritin. The dinuclear ferroxidase center with 2 Fe3+ modeled on those found for Fe3+ in E. coli ferritin A subunit. (Adapted from Chasteen and Harrison, 1999). 59 CrFer-H1 CrFer-H2 Consensus ...MMERVFLLVVLALGSTVVGVQYDNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI MMER FLLVVLAL STV GVQ DNDMKEPK DRY LDDRCINAIQHQINEEMHASLI MAAMMGKSLVLLVL FFSTIE V HDNDMKD MDRYILDNKCINGLQ QINEE HASLV MAAMMGKSLVLLVLTFFSTIETVRHDNDMKDSSMDRYILDNKCINGLQLQINEERHASLV mm l vl st v dndmk dry ld cin q qinee hasl 57 60 CrFer-H1 CrFer-H2 Consensus YMNMAAHFGRN VGRKGFAKFFKHSSDEEREHAQKLIDYVNKRSGKV AFDIKMPGKDEW YMNMAAHFGRNSVGRKGFAKFFKHSSDEEREHAQKLIDYVNKRSGKVIAFDIKMPGKDEW YMHMASHFGSNAVGRKGFSKFFKHSSDEEREHAQKLIDYINKRSGWVAAFDIKMPGKTIW Y MHMA HFG NAVGRKGF KFFKHSSDEEREHAQKLIDYINKRSG VAAFDIKMPGK W ym ma hfg n vgrkgf kffkhssdeerehaqklidy nkrsg v afdikmpgk w 117 120 CrFer-H1 CrFer-H2 Consensus KDGLEALEDAMNLE HVNNKLHHLHHMADKIC DPHLMDYIEGEFLTEQVESINEFKTYI KDGLEALEDAMNLERHVNNKLHHLHHMADKICSDPHLMDYIEGEFLTEQVESINEFKTYI KNGMEALQDALNLENHVNNKLHHLHQMADKICADPHLMNFLEGEFLTEQVESINELNTFI K NGMEALQDALNLENHVNNKLHHLHQMADKICADPHLMNFLEGEFLTEQVESINEL TFI k g eal da nle hvnnklhhlh madkic dphlm egeflteqvesine t i 177 180 CrFer-H1 CrFer-H2 Consensus SQLGAMNNGMGEYLFDHQLLEKS SQLGAMNNGMGEYLFDHQLLEKS SQLGAMNDGMGEYLLDRELLEKS S QLGAMNDGMGEYLLDRELLEKS sqlgamn gmgeyl d lleks 200 203 CrFer2 protein CrFerH protein N-terminal Fragment 1 Fragment 2 Fragment 3 Consensus ...MMERVFLLVVLALGSTVVGVQYDNDMKEPKTDRYSLDDRCINAIQHQINEEMHASLI V DNDMK MAAMMGKSLVLLVLTFFSTIETVRHDNDMKDSSMDRYILDNKCINGLQLQINEERHASLV V DNDMK ......................VQYDNDMKEP............................ V DNDMK ............................................................ ............................................................ ............................................................ 57 60 10 0 0 0 CrFer2 protein CrFerH protein N-terminal Fragment 1 Fragment 2 Fragment 3 Consensus YMNMAAHFGRNSVGRKGFAKFFKHSSDEEREHAQKLIDYVNKRSGKVIAFDIKMPGKDEW LIDY N R YMHMASHFGSNAVGRKGFSKFFKHSSDEEREHAQKLIDYINKRSGWVAAFDIKMPGKTIW LIDY N R ............................................................ ...................................LIDYVNQR................. LIDY N R ............................................................ ............................................................ 117 120 10 8 0 0 CrFer2 protein CrFerH protein N-terminal Fragment 1 Fragment 2 Fragment 3 Consensus KDGLEALEDAMNLERHVNNKLHHLHHMADKICSDPHLMDYIEGEFLTEQVESINEFKTYI GLEALEDA NLER KNGMEALQDALNLENHVNNKLHHLHQMADKICADPHLMNFLEGEFLTEQVESINELNTFI GMEALQDA NLE ............................................................ ............................................................ .DGLEALEDAMNLER............................................. GLEALEDA NLER ..GLEPSEDPGNLER............................................. GLEP EDP NLER 177 180 10 8 14 13 CrFer2 protein CrFerH protein N-terminal Fragment 1 Fragment 2 Fragment 3 Consensus SQLGAMNNGMGEYLFDHQLLEKS SQLGAMNDGMGEYLLDRELLEKS ....................... ....................... ....................... ....................... 200 203 10 8 14 13 Fig. 3.5 (E) CrFer-H1 and –H2 share ~ 72 % identity (top) and there are likely to be other isoforms of plasma ferritin (bottom). CrFer-H1 and –H2 contain the critical residues for ferroxidase activity (boxed in red). Alignment of CrFer-H1, -H2 and the peptide sequences from protein sequencing reveal that there are other unidentified isoforms in the plasma. 60 ferroxidase site is illustrated in Fig. 3.5D (Chasteen and Harrison, 1999).There are also likely to be other ferritin isoforms in the plasma besides CrFer-H1 and –H2 since there are peptide sequences that do not match the deduced amino acid sequence of CrFer-H1 and –H2 (Fig. 3.5E). 3.6 CrFer-H2 is ubiquitously expressed and its transcription is responsive to LPS and bacterial infection. High homology between CrFer-H1 and –H2 (~ 70 % at the DNA level) indicates that both would not be distinguishable by Northern analysis. To understand the regulation of plasma ferritin gene expression during infection, transcript levels of CrFer-H2 were investigated during LPS and P. aeruginosa challenge. The basal level of the 1 kb CrFer-H2 transcript is detectable in naïve hepatopancreas, muscle and stomach (Fig. 3.6A). The expression is most prominent in the heart and intestine. In contrast, hemocytes do not express CrFer-H2 gene. When challenged for 3 h with LPS, the CrFer-H2 gene is upregulated in the hepatopancreas, heart, intestine, muscles and stomach by 22-, 12-, 37-, 19- and 40- fold, respectively (Fig. 3.6B). However, iron-loading did not result in any significant change in transcription of CrFer-H2 in all tissues (Fig. 3.6B), although the plasma ferritin level was increased by 5-fold. P. aeruginosa induced CrFer-H2 transcription in hepatopancreas as early as 3 hpi. The transcript (1.2 kb) level peaked at 12 hpi in the hepatopancreas and stomach (Fig. 3.6C and D). Both CrFer-H2 transcripts expressed in the heart increased gradually over the 72 h time course of infection. Interestingly, the intestine and stomach did not show significant upregulation in the 1 kb transcript. It is likely that the 1.2 kb and 1 kb CrFer-H2 transcripts, which are detected in the heart and stomach tissues, are generated by tissue-specific choice of two out of the four possible polyadenylation sites in the CrFer-H2 cDNA (Fig 3.5A). Alternative splicing and 61 (A) Naive 3 h LPS 3 h FeSO4 A HP HT I M S A HP HT I M S A HP HT I M S kb CrFer-H2 1 Ribosomal L3 (B) Induction of CrFer-H transcript to LPS or FeSO4 stimuli 45 40 35 Fold 30 25 20 Fe 15 LPS 10 5 0 Ame HP HT INT MUS STC Fig 3.6. (A) Northern analysis to study differential expression of CrFer-H2 in various naïve, 3h LPS-induced and 3h FeSO4 induced tissues of the limulus. CrFer-H2 is expressed as a 1 kb transcript in the hepatopancreas (HP), heart (HT), intestine (I), muscle (M) and stomach (S) but not in the amoebocytes (A). Upregulation of this transscript is observed in heart, intestine, muscle and stomach. Iron loading of the limulus did not induce transcription of CrFer-H2 in any tissue. (B) Quantitative analyses of CrFer-H expression using ImageMaster software. The fold induction in the LPS and iron-induced tissues were estimated by taking the basal value of the naïve tissues as 1. 62 (C) (D) Fig 3.6. (C) Northern analysis to study kinetics of CrFer-H2 expression Both alternative splicing and polyadenylation has yielded ferritin-H transcriptinofvarious tissues after infection with P. aeruginosa and (D) the change in fold of CrFer-H2 normalized against actin 3. Total RNA from at least three individuals was pooled for each sample. Transcripts of larger size were generated in heart, intestine and stomach tissues during saline or bacterial challenge but not in the naïve tissues. 63 multiple polyadenylation have yielded various sizes of ferritin transcripts in Drosophila (Lind et al., 1998). 3.7 LPS and iron-loading can regulate ferritin protein synthesis during Pseudomonas infection. To investigate the regulation of plasma ferritin level during bacterial challenge or iron-loading, Western analyses using M. sexta anti-ferritin antibody were performed. There is a 4-fold increase in the plasma ferritin at 3 hpi, after which the plasma ferritin completely disappeared reproducibly between 6-48 hpi, and was only detected again at 72 hpi (Fig. 3.7). The plasma ferritin is also upregulated ~ 2-fold by iron loading for 3 h. 3.8 P. aeruginosa ‘steals’ host iron by degrading plasma ferritin. The disappearance of plasma ferritin at between 6 to 48 hpi (Fig. 3.7) suggests that it could either have been degraded by P. aeruginosa or it has translocated into specific cells. This prompted us to determine, in a cell-free environment, whether plasma ferritin was degraded by P. aeruginosa. Ferritin could still be detected when naïve plasma was incubated with TSB for 24 h (Fig. 3.8A). However, the ferritin was not detectable when P. aeruginosa was incubated with plasma and TSB, suggesting that P. aeruginosa has degraded plasma ferritin in this cell-free system. In order to monitor the iron status during Pseudomonas infection, we determined iron that is tightly-associated to proteins and the labile iron pool (LIP). The proportion of these two populations of iron may shed some light on the possible change in the type of proteins that associate with and regulate the level of plasma iron. At a concentration of ~ 40 mg/ml in the plasma, iron is at least 8 to 10-fold higher than that in human plasma. Over a time course of 72 h of P. aeruginosa infection, the total 64 Fig. 3.7 Regulation of limulus plasma ferritin at the protein level during P. aeruginosa infection. Plasma ferritin protein level during Pseudomonas infection as detected by Western blot using Manduca anti-ferritin antibody. There is drastic increase in ferritin protein in the plasma after 3 h, after which the protein level decreased to an undetectable level. In contrast, iron loading of the horseshoe crab after 3 h only resulted in s smaller magnitude of ferritin increment. The level of ferritin was highest after 72 h. 65 plasma iron level was maintained within a narrow range of 35.55 to 42.44 µg/ml (Fig. 3.8B). In the naïve state, the LIP ranged from 7.32 to 20.68 µg/ml. Thus, most of the extracellular iron is tightly-associated with proteins, constituting 50-70 % of the total plasma iron. There was a significant increase in LIP at 3 to 12 hpi (Fig. 3.8B), indicating a shift towards more labile iron. Interestingly, there were large individual variations in the population of tightly-associated iron and LIP during P. aeruginosa infection, while the total plasma iron level was maintained throughout infection. In vitro experiments have demonstrated that maintenance of pH and redox potential of normal human plasma is crucial for iron-withholding power against strains of E. coli and Klebsiella pneumoniae (Bullen et al., 2000). Our measurements of redox potential and pH of the horseshoe crab plasma revealed that these two critical parameters were only altered after 24 hpi (Fig. 3.8C). It is noteworthy that during this period of time, viable bacteria can still be detected in the horseshoe crab. The pH of the plasma remained consistently above 7.3 throughout 72 hpi. The redox potential was stabilized at -15 mV after which it decreased slightly to -30 mV. This suggests that P. aeruginosa does not lower host plasma redox potential or pH during its iron piracy from the host. 66 67 Fig. 3.8 Strategy employed by P. aeruginosa to obtain host iron. (A) P. aeruginosa can degrade ferritin in vitro. Plasma ferritin protein was incubated with or without Pseudomonas for 24 h, followed by its detection via Western blot using Manduca anti-ferritin antibody. (B) Pseudomonas infection does not result in hypoferraemia of the limulus. Total plasma iron was maintained throughout the course of infection, while LIP increased at 3 and 12 hpi, possibly as the pathogen attempts to ‘steal’ host iron. (C) Pseudomonas does not lower plasma redox potential or pH to ‘steal’ host iron. Each time-point is obtained from at least four individuals. Statistical significance was calculated using a two-tailed Student’s T-Test. *: significant; **: highly significant. A significant difference with P < 0.05 and P < 0.01 is represented by one and two * respectively. Data for bacteria viability was adapted from Ng et al., 2004. 68 3,9 Ferritin switches from a DNA-binding to non-DNA-binding conformer during infection The H-subunit of human ferritin has been shown to bind DNA without significant preference for base composition, sequence, or the nature of the DNA ends (Surguladze et al., 2004). Interestingly, in this study, only the naive ferritin complex binds to both DNA probes to form a complex, resulting in their gel mobility shift (Fig. 3.9A and B). In contrast, no ferritin-DNA complex was formed with plasma ferritin purified from 3 and 72 hpi. Neither did the ferritin from 3 h iron-loaded plasma bind to the DNA even though the ferritin protein level increased under these conditions (Fig. 3.7). Moreover, naïve ferritin-DNA complex interaction also led to a blue shift from 343 nm to 340 nm with a drastic increase in fluorescence emission intensity when excited with a wavelength 280 nm (Fig. 3.9B). In contrast, there was no significant difference in the fluorescence emission intensity from the infected ferritin complex alone or in the presence of DNA. This indicates that the environment where tryptophan residues reside in the naïve ferritin complex may be influenced when the molecule interacts with DNA. 3.10 Uninfected and infected plasma ferritin complexes contain different ferritin isoforms. The inability of infected plasma ferritin to bind DNA suggests that during infection, the plasma ferritin has undergone a change in its physicochemical properties. The uninfected and infected plasma ferritins were resolved into their respective subunits by two-dimensional gel electrophoresis. Figure 3.10 shows 3 distinct spots between pI 6.2-6.4, which make up the 21 kDa band from the uninfected plasma ferritin. Interestingly, the infected plasma ferritin also contains these 3 isoform in addition to 2 new members at pI ~ 6.7 and 7.8 (Fig. 3.10). 69 70 Fig. 3.9 Uninfected plasma ferritin but not infected plasma ferritin can bind to DNA in a sequence-independent manner. (A) Using LDorThr and LkBCom as probes, plasma ferritin from naïve, 3, 6 and 72 hpi individuals, as well as 3 h iron-loaded individuals were incubated and run on 4 % PAGE gel. Only plasma ferritin from uninfected horseshoe crabs can bind DNA in a sequenceindependent manner. LDorThr probe: 5 CATGCACGAGAAAAAAGCCGGGAAATCCATTAGA-3; LkBCom probe: 5– GAAATTTTTCCTTCTTGTACATTGGAAAACGTTTTCACGTGACGTACTGATT TGTCTGTCATGCA-3. (B) Fluorescence measurement of ferritin complex-DNA interaction. Naïve ferritin alone leads to emission at 343 nm while naïve ferritin in the presence of DNA leads to emission at 340 nm; blue shift with increased intensity. This implies that the tryptophan residues in the naïve ferritin complex may become more buried into the hydrophobic region of the complex during DNA interaction. Only slight increase in intensity for infected ferritin alone or with DNA and no shift was observed. This indicates negligible infected ferritin complex-DNA interaction. Excitation = 280 nm; Slit width = 10 nm, scan speed = 50 scans/min. 71 Infected ferritin Naive ferritin Fig. 3.10 Uninfected and infected plasma ferritins consist of different 21 kDa isoforms. Uninfected plasma ferritin is made up of 3 isoforms, which are also present in infected plasma ferritin, in the pI range of 6.2-6.4 (boxed). However, 2 extra spots (~ pI 6.7 and 7.8) emerge in the infected protein (red arrows). A streak in the infected protein in the pI range 5.5-6.0 was also observed. In both cases, ~ 80 µg of partially purified protein was resolved using 2 dimensional gel electrophoresis and the spots were detected by silver staining. 72 4. DISCUSSION 4.1 Plasma ferritin is directly involved in innate immune response. 4.1.1 The horseshoe crab plasma ferritin evades degradation by P. aeruginosa to prevent iron loss. Pathogen-mediated degradation of iron-binding proteins such as transferrin followed by release of iron has been reported. P. aeruginosa can use transferrin as a source of iron via pyoverdin without pH change (Wolz et al., 1994). Hissen et al. (2004) also demonstrated that proteolytic degradation of transferrin by Asperigillus fumigatus may play a secondary role in iron acquisition for its survival in the serum. Most recently, Larson et al. (2004) have shown that the Gram-negative diplococcus Neisseria meningitidis can trigger rapid redistribution and degradation of cytosolic ferritin within infected epithelial cells. Using indirect immunofluorescence microscopy, it was observed that cytosolic ferritin is aggregated and recruited to intracellularly where the invading meningococci was located. In meningococci infected cells, the half-life of ferritin was 5.3 h compared to 20.1 h for uninfected cells. The effect of meningococci was reversed by supplementation of ascorbic acid and lysosomal protease inhibitor, leupeptin which retarded meningococcal replication. It was proposed that accelerated ferritin degradation occurs as a response to an iron starvation state induced by meningococci infection and that ferritin degradation provides the intracellular pathogen with a critical source of iron. As mentioned in the Introduction section, the use of P. aeruginosa as the model pathogen in this study is justified by the fact that it requires oxygen and nitrogen compounds for respiration and these processes demand significant amounts of iron (Vasil and Ochsner, 1999). Besides, the relevance of P. aeruginosa in human clinical diseases makes it interesting to be employed in the infection of the limulus so as to shed some light on the role of plasma ferritin in iron homeostasis against host iron piracy in an iron-rich 73 environment. P. aeruginosa has been successfully cultured from mud samples obtained from the natural habitat of the limulus (data not shown) and this indicates that it is potentially a natural pathogen for the limulus. This validates the use of this pathogen for infection of the limulus in this study. Our observation of the absence of plasma ferritin from 6 to 48 hpi was highly reproducible (Fig. 3.7) and it indicates that plasma ferritin has either been degraded by P. aeruginosa or that it had undergone a change in localization from extracellular to intracellular. This is supported by the upregulation of ferritin gene expression upon LPS or P. aeruginosa challenge (Fig. 3.6). Here, we present evidence that P. aeruginosa, which is limited to extracellular invasion (Kharazami, 1991; Pederson, 1992; Doring, 1997), is able to degrade plasma ferritin possibly to obtain the host iron for its growth and survival (Fig. 3.8A). Indeed, the large iron capacity of plasma ferritin as compared to other iron-binding proteins would make it a very attractive candidate for host iron piracy. However, a count of the viable bacteria that remains during the course of infection showed that less than 200 cfu of bacteria could be detected in the plasma at 3 hpi and that by 18 hpi, P. aeruginosa was completely cleared in almost all infected individuals (Ng et al., 2004). Hence this would disfavour the hypothesis that disappearance of plasma ferritin was due to degradation by Pseudomonas in vivo from 6 to 48 hpi. In fact, it would suggest that the ferritin in the uninfected horseshoe crab plasma is able to translocate into specific cells and bind DNA with no sequence preference (Fig. 3.9). During infection, the containment of plasma ferritin intracellularly serves as a mechanism to evade P. aeruginosa degradation, hence withholding the iron as the immediate defense mechanism. In the course of recovery, two additional isoforms of ferritin purified from the infected plasma (Fig. 3.10) probably contributes to the inability of the ferritin complex to 74 interact with DNA. The perculiar shift in the DNA-binding nature of naïve and infected plasma ferritins may suggest that both conformations may possibly play mutually exclusive roles in the host during bacterial infection. 4.1.2 Regulation of plasma ferritin may contribute to iron homeostasis and constant free radical level. The horseshoe crab plasma iron concentration is at least 8-10 fold (in the 235 µM range) higher than that reported in humans. This indicates both the physiological importance of ferritin to this organism and yet, at the same time the ferritin molecule presents itself as a “bait”, hence imposing potential danger to the host since it serves as a nutrient pool of iron for an invading pathogen. Iron is crucial for the normal functioning of the adaptive immune system. The horseshoe crab does not exhibit hypoferraemia when challenged with P. aeruginosa since adaptive immunity is nonexistent in the horseshoe crab. It is noteworthy that the horseshoe crab is able to overcome a dose of P. aeruginosa that would have been lethal to mice (Stieritz and Holder, 1975; Soothill, 1992). Interestingly, the oxidative activity of the plasma does not exhibit significant changes during the course of infection by a pathogen that depends heavily on host iron for growth and proliferation. In the Haber-Weiss reaction, iron acts as a catalyst to convert the substrate superoxide anion and hydrogen peroxide to oxygen, hydroxide ion and hydroxyl radicals. The constant plasma oxidative activity may be due to the substrates being limiting in the Haber-Weiss reaction. Alternatively, it may be due to upregulation of stress-related proteins that may contribute to suppress free radical formation. Nevertheless, we show that the tight control of iron may be a factor that determines the level of free radical in the plasma during infection. 75 4.2 A dynamic role of ferritin during Pseudomonas infection. At 3 hpi of P. aeruginosa, the increase in plasma ferritin level coincides with an increase in the labile iron pool (LIP). This may reflect the immediate response of the horseshoe crab to synthesize more plasma ferritin to sequester iron away from the pathogen, involving both transcriptional and translational up-regulation of ferritin. Simultaneously, it is likely that plasma ferritin undergoes translocation into cells, possibly via cell-specific receptors to evade P. aeruginosa degradation. Although ferritin receptors have not been reported in the horseshoe crab, the presence of ferritin receptors on rat hepatocytes has been demonstrated (Mach et al., 1983). Thus, based on the assumption that ferritin receptors exist in the horseshoe crab, we propose the following model for the dynamic role of ferritin during Pseudomonas infection (Fig. 4.1). In the uninfected animals, most iron is stored in plasma ferritin complex, while the LIP represents ~ 15 % of the total iron content in the plasma. There is basal level of ferritin in plasma, and the ferritin population may consist of homo- and/or heteropolymers of ferritin, which may have a dynamic distribution in the plasma, cytoplasm and nucleus. The ferritin complex may translocate into specific cells via specific receptors and into the nucleus where it may bind DNA. Indeed, ferritin translocation into cell nuclei has been reported by Thompson et al. (2002) and ferritin complex-DNA interaction has been demonstrated by Surguladze et al. (2004). The low intracellular iron level results in repression of IRE on the CrFer-H transcripts by IRE-binding proteins. During infection, even though ferritin expression is induced, the plasma ferritin is restricted within the intracellular space to evade degradation by Pseudomonas. This allows the host to withhold iron within cells, leading to intracellular iron sequestration and hence intracellular iron level rises. As a result of 76 increased intracellular iron level, IRE-binding protein repression is relieved and more ferritin protein is synthesized. The presence of LPS on the bacteria cells also induces the synthesis of more CrFer-H transcripts. In the course of recovery, the two CrFer-H proteins, which may have undergone some post-translational modifications (such as phosphorylation), and other isoferritins may form the.infected plasma ferritin, which is not able to bind DNA. The presence of isoferritins is supported by the mass spectrometric analysis where peptide fragments that do not correspond to the deduced amino acid sequence of CrFer-H1 and –H2 were evident. Thus, there may exist many more isoferritins in the limulus that await to be identified and cloned. The uninfected and infected ferritins may play differential roles. The naïve and infected ferritin complexes are assembled with varying ferritin isoforms, leading to differential roles of both forms of ferritin complexes during infection. The dynamic interplay between the synthesis of specific isoforms during infection and the assembly of unique ferritin complexes may hence protect the host against (i) iron piracy by pathogen and (ii) free radical-induced damage. 77 78 Fig. 4.1 Proposed model for the dynamic role of plasma ferritin during infection. (1) In the uninfected state, there is basal level of ferritin in plasma. The ferritin population may consist of homo- and/or heteropolymers of ferritin, which may be found in the plasma, cytoplasm or nuclei. In the nucleus, ferritin may bind DNA with no sequence preference. (2) The low intracellular iron level results in repression of IRE on the CrFer-H transcripts by IRP. (3) During infection, there is containment of plasma ferritin intracellularly to evade Pseudomonas degradation. (4) The host withholds iron into cells, leading to intracellular iron sequestration and hence intracellular iron level rises. (5) As a result of increased intracellular iron level, IRP repression is relieved and more ferritin protein is synthesized. (6) The presence of LPS on the bacteria cells also induced the synthesis of more CrFer-H transcripts. (7) There may be differential role in the uninfected and infected ferritins, which may be assembled using ferritin of varying isoforms and/or post-translational modifcations (such as phosphorylation), leading to specific roles of both forms of ferritin during infection. The dynamic interaction between the synthesis of specific isoforms, post-translational modification during infection and the assembly of unique ferritin complexes may hence protect the host against (i) iron piracy by pathogen and (ii) free radical-induced damage. 79 4.3 Insights into the role of plasma ferritin: from horseshoe crab to mammalian plasma ferritin Whereas human, horse and rat plasma ferritins have been isolated, no corresponding gene sequence of these plasma ferritins have been reported (Southern and Baker, 1982; Linder et al., 1996). In contrast, the limulus possesses at least two or more secreted ferritins. The subtle structural differences between CrFer-H1 and -H2 raise the question for the requirement of isoferritins in the limulus. How many genes encode for limulus H-chain ferritin? Is there a redundancy among isoferritins or do they have different enzymatic activities for mutually exclusive functions? Do homoor heteropolymers of H-chain ferritin exist and what are their biological significance? To date, heteropolymers have only been known to be built from H- and L-chain ferritin subunits, and the H- : L-chain ratio varies during inflammation. Even more interestingly is the finding that there exist several putative phosphorylation sites on limulus H-chain ferritin subunits. Phosphorylation and dephosphorylation is a mechanism that has been widely employed in various signaling as well as DNA binding molecules. If phosphorylation of ferritin subunits does occur, does it confer extra roles to plasma ferritin such as in DNA protection? Future investigation on limulus isoferritins may help to unravel more mysteries behind plasma ferritin such as their in vivo composition (uninfected and infected states) and the structural requirements for the polymer formation. The origin of vertebrate plasma ferritin still remains controversial, except for an individual report that rat hepatoma cells secrete ferritin upon iron and cytokine stimulation (Tran et al., 1997). Nevertheless, it has been widely acknowledged that fluctuation in plasma ferritin of higher invertebrates is a strong indication of physiological changes, such as inflammation and cancer (Tran et al., 1997). During infection, it is speculated that plasma ferritin scavenges and helps detoxify Fe2+ 80 leakage from damaged cells (Linder et al., 1996; Tran et al., 1997). Surprisingly, searching the recently completed rat and human genome sequences for ferritin-H and -L gene homologs did not reveal any candidate genes that may encode secretory ferritin. This casts doubts on the authenticity of plasma ferritin in higher vertebrates. More research is therefore needed to unravel the origin of vertebrate plasma ferritin. Is it secreted via an unknown mechanism or released during cell damage/lysis? Does this indicate that it is only present in invertebrates where it functions as an iron transporter like transferrin in innate immunity? Does the much higher level of plasma iron in invertebrates dictate the requirement of more iron-binding proteins for immediate deployment against pathogens’ attempt to usurp host iron? As iron is a double-edged sword (acting both as a regulator of iron in host and as a potential nutrient source for the invading pathogen), the regulation and control of iron requires the orchestrated cooperation amongst numerous molecules revolving around this important process of host-pathogen interaction. Our study on horseshoe crab plasma ferritin has clearly served to provide an overview of the dynamic host-pathogen interaction in the context of plasma iron homeostasis and innate immunity. This is clearly lacking in many studies whereby iron-withholding by the host and iron-piracy by the pathogen are investigated separately. 81 5. CONCLUSION AND FUTURE PERSPECTIVES In summary, this project demonstrates that there exists various isoforms of horseshoe crab plasma ferritin which may assemble into homo- or heteropolymers for iron sequestration during a microbial infection. The involvement of the ferritin proteins in innate immune defense can be supported by the observation that the CrFerH1 and –H2 genes are highly induced in various tissues upon LPS or bacterial challenge. The existence of several isoforms of CrFer-H complicates the study of ferritin genes expression upon infection since it is likely that each isoform may have a unique expression profile when induced by LPS. Another complication stems from the understanding of ferritin protein regulation during infection. The use of the Manduca sexta anti-ferritin antibody has probably only allowed us to examine the translational regulation of some, but not all ferritin isoforms in the context of innate immunity. It would be ideal to examine the protein profiles of each individual isoform to have greater understanding of their respective role during infection. Nevertheless, we have demonstrated for the first time, several novel findings. Firstly, the phenomenal loss of plasma ferritin during infection suggests that it transits into specific cells where it is retained within cells during P. aeruginosa infection. To date, such a detailed examination of plasma ferritin protein level during bacterial infection (over 72 h time course) has not been reported in other invertebrates. Secondly, P. aeruginosa is able to degrade plasma ferritin in vitro to obtain iron for its growth in the host. Studies on P. aeruginosa have only demonstrated the ability of this pathogen to degrade transferrin and it would be interesting to determine whether the pathogen employs the same protease in degradation of iron-binding proteins. Lastly, the differential roles of uninfected and infected ferritin complexes in DNA binding suggests that they may play specific roles during infection. 82 Future work will focus on the identification of a plasma ferritin receptor or other interacting partners via yeast-2-hybrid screening. Yuan et al. (2004) have demonstrated the interaction of granulocyte colony-stimulating factor receptor (GCSFR) and H-ferritin using yeast two-hybrid screens, GST pull-down experiments & immunoprecipitation studies in vitro and in vivo. They observed that G-CSFR interacts with H-ferritin and dissociates after 30 min upon G-CSF induction and then reassembles at 45 min. Their work suggests that LIP may be released from the dissociated H-ferritin, and then induce formation of intracellular reactive oxygen species in the bone-marrow hematopoietic cells. It has also been reported that human H-kininogen is a ferritin-binding protein (Torti et al., 1998). H-kininogen is a multifunctional protein; inhibits cysteine proteases, plays a role in contact activation of the coagulation cascade, and is the precursor of the potent proinflammatory peptide bradykinin. H-kininogen as a ferritin-binding protein may link ferritin in the complex chain of interactions by which H-kininogen mediates its multiple effects in contact activation and inflammation. Autoantibodies to serum ferritin (FBP1 and FBP2) in bovine have also been found (Orino et al., 2004). FBP1 and FBP2 are IgM and IgG respectively. It will be interesting to identify novel interacting partners of plasma ferritin in the horseshoe crab. It is also likely that various isoforms of ferritin, which constitute the uninfected and infected forms of plasma ferritin complexes confer specific innate immune functions in the horseshoe crab during P. aeruginosa infection. Changes in ferritin subunits during inflammation and infection have been observed in humans but this phenomenon has not been explored in detail with so far. Thus, investigating the transcript and protein profiles of the various isoforms may shed light on the role of each individual subunit. There may also exist other iron-binding proteins in the 83 horseshoe crab plasma that are involved in innate immune defense. One such example is transferrin which has been found to be present in organisms across phyla, even in invertebrates such as M. sexta and A. aegypit (Winzerling et al., 1995; Lowenberger C., 2001). A detailed examination of the dynamics of other ferritin subunits and other iron-binding proteins in the horseshoe crab will help build a more complete map of iron-mediated host defense during bacterial invasion. 84 References Addison, G.M., Beamish, M.R., and Hales C.N. (1972) An immunoradiometric assay for ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. J. Clin. Pathol. 25: 326-329. Andrews, S.C., Robinson, A.K., Rodriguez-Quinones, F. (2003) Bacteria iron homeostasis. FEMS Microbiol. Rev. 27: 215-237. Beck, G., Ellis, T.W., Habicht, G.S., Schluter, S.F., Marchalonis, J.J. (2002) Evolution of the acute phase response: iron release by echinoderm (Asterias forbesi) coelomocytes, and cloning of an echinoderm ferritin molecule. Dev. Comp. Immunol. 26: 11-26. Bone, R.C., Grodzin, C.J., Balk, R.A. (1997) Sepsis: A new hypothesis for pathogenesis of the disease process. Chest 112: 235-243. Brock, J.H. and Mainou-Fowler, T. (1983) The role of iron and transferrin in lymphocyte transformation. Immunol. Today 4: 347-351. Brock, J. (1989) Iron and cells of the immune system. In: Bullen, J.J. and Griffiths, E, editors. Iron and infection: Molecular, Physiological and Clinical Aspects, 2nd edition. Wiley, Chichester, Pp 289-327. Brock, J.H. (1999) Iron and the immune system. In: de Sousa M, Brock J, editors. Iron in immunity, cancer and inflammation, Brooklyn, NY: John Wiley. Pp 111-22. 85 Brown, C., Lagnado, C., Goodall, G. (1996) A cytokine mRNA-destabilizing element that is structurally and functionally distinct from A+U-rich elements. Proc. Natl. Acad. Sci. USA 93: 13721-13736. Bullen, J.J., Spalding, P.B., Ward, C.G., Roger, S.H.J. (1992) The role of Eh, pH and iron in the bactericidal power of human plasma. FEMS Microbiol. Lett. 94: 47-52. Bullen, J.J., Rogers, H.J., Griffiths, E. (1978) Role of iron in bacterial infection. Curr. Top. Microbiol. Immunol. 80: 1-35. Bullen, J.J. (1981) The significance of iron in infection. Rev. Infect. Dis. 3: 1127-1138. Bullen, J., Griffiths, E., Rogers, H., Ward, G. (2000) Sepsis: the critical role of iron. Microbes and Infection. 2: 409-415. Caccavo, D. et al. (2002) Antimicrobial and immunoregulatory functions of lactoferrin and its potential therapeutic application. J. Endotoxin Res. 8: 403-417. Capurro, M.L., Iughetti, P., Ribolla, P.E.M., Bianchi, A.G. (1996) Musca domestica Hemolymph Ferritin. Archives of Insect Biochemistry and Physiology 12: 197-207. Charlesworth, A., Georgieva, T., Gospodpv, I., Law, J.H., Dunkov, B.C., Ralcheva, N., Barillas-Mury, C., Ralchev, K. and Kafatos, F.C. (1997) Isolation and properties of Drosophila melanogaster ferritin. Molecular cloning of a cDNA that encodes one 86 subunit, and localization of the gene on the third chromosome. Eur. J. Biochem. 247: 470-475. Chasteen, N.D. and Harrison P.M. (1999) Mineralization in ferritin: An efficient means of iron storage. J. Struc. Biol. 126: 182-194. Cook, J.D., Lipschitz, D.A., Miles, L.E.M. (1974) Serum ferritin as a measure of iron stores in normal subjects. Am. J. Clin. Nutri. 27: 681-687. Cox, C.D. and Adams, P. (1985) Siderophore activity of pyoverdin for Pseudomonas aeruginosa. Infect. Immun. 48: 130-138. Cragg, S.J., Wagstaff, M. and Worwood, M. (1981) Detection of a glycosylated subunit in human serum ferritin. Biochem. J. 199: 565-571. Crichton, R.R., Wilmet, S., Legssyer, R., Ward, R.J. (2002) Molecular and cellular mechanisms of iron homeostasis and toxicity in mammalian cells. J. Inorg. Biochem. 91: 9-18. Dellinger, R.P., Opal, S.M., Rotrosen, D. (1997) From the bench to the bedside: The future of sepsis research. Chest 11: 744-753. Doring, G.(1997) Cystic fibrosis respiratory infections: interactions between bacteria and host defence. Monaldi Arch Chest Dis. 52:363-6 87 Dunkov, B.C., Zhang, D., Choumarov, K., Winzerling, J.J., and Law, J.H. Isolation and Characterization of Mosquito Ferritin and Cloning of a cDNA that Encodes One Subunit. (1995) Archives of Insect Biochem. Physiol. 29: 293-307. Dunkov, B.C. and Georgieva, T. (1999) Organization of the Ferritin Genes in Drosophila melanogasteer.DNA Cell Biol. 18: 937-944. Epstein, J., Eichbaum, Q., Sheriff, S., Ezekowitz, R. (1996). The collectins in innate immunity. Curr. Opin. Immunol. 8: 29-35. Fearson, D., Locksley, R. (1996). The instructive role of innate immunity in the acquired immune response. Science 272: 50-53. Franc, N.C., White, K. (2000) Innate recognition systems in insect immunity and development: new approaches in Drosophila. Microb. Infect. 2: 243-250. Foukas, L.C., Katsoulas, H.L. Paraskevopoulou, N.m Metheniti, A., Lambropoulou, M. and Marmaras, V.J. Phagocytosis of Escherichia coli by Insect Hemocytes Requires Both Activation of the Ras/Mitogen-activated Protein Kinase Signal Transduction Pathway for Attachment and 3 Integrin for Internalization. J. Biol. Chem. 273: 14813-14818. 88 Halle, F. and Meyer, J.M. (1992) Iron release from ferrisiderophores. A multi-step mechanism involving a NADH/FMN oxidoreductase and a chemical reduction by FMNH2. Eur. J. Biochem. 209: 621-627. Hennecke, H. (1990) Regulation of bacterial gene expression by metal-protein complexes. Mol. Microbiol. 4: 1621-1628. Hennessey, S., Frazier, B., Kim, D., Deckworth, T., Baumgartel, D., Rotwein, P., Frazier, W. (1989) Complete thrombospondin mRNA sequence includes potential regulatory sites in the 3' untranslated region. J. Cell Biol. 108: 729-741. Heinrichs, D.W. and Poole, K. (19930 Cloning and sequence analysis of a gene (pchR) encoding an AraC family activator of pyochelin and ferripyochelin receptor synthesis in Pseudomonas aeruginosa. J. Bacteriol. 175: 5882-5889. Henrichs, D.E., Young, L. and Poole, K. (1991) Pyochelin-mediated iron transport in Pseudomonas aeruginosa: involvement of a high-molecular-mass outer membrane protein. Infect. Immun. 59: 3680-3684. Hissen, A.H.T., Chow, J.M.T., Pinto, L.J., and M.M. Moore. (2004) Survival of Asperigillus fumigatus in Serum Involves removal of Iron from Transferrin: the Role of Siderophores. Infec. Immun. 72: 1402-1408. 89 Huang, T.S., Law, J.H. and Soderhall, K. (1996) Purification and cDNA cloning of ferritin from the Hepatopancreas of the freshwater crayfish Pacifastacus leniusculus. Eur. J. Biochem. 236: 450-456. Huang, T.S., Melefors, O., Lind, M., Soderhall, K. (1999) An atypical iron-responsive element (IRE) within crayfish ferritin mRNA and an iron regulatory protein 1 (IRP1)like protein from crayfish hepatopancreas. Insect Biochem. Mol. Biol. 29: 1-9. Huebers, H.A., Heubers, E., Finch, C.A., Webb, B.A., Truman, J.W., Riddifore, L.M., Martin, A.W. and MAssover, W.H. (1988) Iron binding proteins and their roles in the tobacco hornworm, Manduca sexta (L.). J. Comp. Physiol. 158: 291-300. Iwanaga, S. Kawabata, S. ans Muta, T. (1998). New Types of clotting factors and defense molecules found in horseshoe crab hemolymph: their structure and functions. J. Biochem. (Tokyo) 123: 1-15. Jacobs, A., Miller, F. and Worwood, M. (1972) Ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. Br. Med. J. 4: 206-208. Jacobs, A.and Worwood, M. (1975) Ferritin in serum. Clinical and biochemical implications. Eng. J. Med. 292: 951-956. Jamroz, R.C., Gasdaska, J.R., Bradfield, J.Y., and Law, J.H. (1993) Transferrin in a cockroach: molecular cloning, characterization and suppression by juvenile hormines. Proc. Natl. Acad. Sci. USA 90: 1320-1324. 90 Kharazami, A. (1991) Mechanisms involved in the evasion of the host defence by Pseudomonas aeruginosa. Immunol Lett. 30: 201-205. Kim, B.S., Lee, C.S., Yun, C.Y., Yeo, S.M., Park, W.M., Kim, H.R. (2001) Characterization and immunological analysis of ferritin from the haemolymph of Galleria mellonella. Comp. Biochem. Physiol. Part A 129: 501-509. Konijn, A., Hershko, C. (1977) Ferritin synthesis in inflammation. I. Pathogenesis of impaired iron release. Br. J. Haematol. 37: 7-19. Konijn, A., Hershko, C. (1989) The anemia of inflammation and chronic disease. In: de Sousa M, Brock J, editors. Iron in immunity, cancer and inflammation, Brooklyn, NY: John Wiley. Pp 111-22. Kopacek, P., Zdychova, J., Yoshiga, T., Weise, C., Rudenko, N. and Law, J.H. (2003) Molecular cloning, expression and isolation of ferritins from two tick species – Ornithodoros moubata and Ixodes ricinus. Insect Biochem. Mol. Biol. 33: 103-113. Larson, JA, Howie HL, So M. (2004) Neisseria meningitidis accelerates ferritin degradation in host epithelial cells to yield an essential iron source. Mol. Microbiol. 53: 807-820. Lauffer, R.B. (1992) Iron and human disease. CRC Press, Inc. Pp 162-172; 180-198. 91 Lind, M.I., Ekengren, S., Melefors, O., Soderhall, K. (1998) Drosophila ferritin mRNA: alternative splicing regulates the presence of the iron-responsive element. FEBS Lett. 436: 476-482. Linder, M.C., Schaffer, K.J., Hazegh-Azam, M., Zhou, C.Y.J., Tran, T.N., and Nagel, G.M. (1996) Serum ferritin: does it differ from tissue ferritin? J. Gastroenterol. Hepatol. 11: 1033-1036. Litwin, C.M. and Calderwood, S.B. (1993) Role of iron in regulation of virulence genes. Clin. Microbiol. Rev. 6: 137-149. Lowenberger C. (2001) Innate immune response of Aedes aegypti. Insect Biochem Mol Biol.31:219-29. Mach, U., Powell, L.W., and Halliday, J.W. (1983) Detection and isolation of a hepatic membrane receptor for ferritin. J. Biol. Chem. 258: 4672-4675. Medzhitov, R., Janeway, C.A.J. (1997) Innate immunity: impact on the adaptive immune response. Curr. Opin. Immunol. 9: 4-9. Meyer, J., Neely, A., Stintzi, A., Georges, C. and Holder, I.A. (1996) Pyoverdin is Essential for Virulence of Pseudomonas aeruginosa. Infect. Immun. 64: 518-523. 92 Miyazaki, H., Kato, H., Nakazawa, T., Tsuda, M. (1995) A positive regulatory gene, pvdS, for expression of pyoverdin biosynthesis genes in Pseudomonas aeruginosa PAO. Mol. Gen. Genet. 248: 17-24. Ng, P.M.L., Jin, Z., Tan, S.S.H., Ho, B., Ding, J.L. (2004) C-reactive protein: a predominant LPS-binding acute phase protein responsive to Pseudomonas infection. J. Endotoxin Res.10: 163-174. Nichol, H.and Locke, M. (1989) The characterization of ferritin in an insect. Biochem. 19: 587-602. Nichol, H. and Locke, M. (1999) Secreted ferritin subunits are of two kinds in insects. Molecular cloning of cDNAs encoding two major subunits of secreted ferritin from Calpodes ethlius. Insect Biochem. Mol. Biol. 29: 999-1013. Ochsner, U.A., Johnson, Z., Lamont, I.L., Cunliffe, H.E. and Vasil, M.L. Exotoxin A production in Pseudomonas aeruginosa requires the iron-regulated pvdS gene encoding an alternative sigma factor. (1996) Mol. Microbiol. 21: 1019-1028. Ochsner, U.A., Wilderman, P.J., Vasil. A.I. and Vasil, M.L. (2002) GeneChip expression analysis of the iron starvation response in Pseudomonas aeruginosa: Identification of novel pyoverdine biosynthesis genes. Mol. Microbiol. 45: 1277-1287. Ochsner, U.A., Vasil, A.I., and Vasil, M.L. (1995) Role of the ferric uptake regulator (FUR) of Pseudomonas aeruginosa in the regulation of siderophores and exotoxin A 93 expression: purification and activity on iron-regulated promoters. J. Bacteriol. 177: 7194-7201. Orino et al. (2004) Characterization of bovine serum ferritin-binding proteins. Comp. Biochem. Physiol. A 137: 375-381. Pederson, S.S. (1992) Lung infection with alginate-producing, mucoid Pseudomonas aeruginosa in cystic fibrosis. APMIS Suppl. 28: 1-79. Pham, D.Q., Zhang, D., Hufnagel, D.H. and Winzerling, J.J. (1996) Manduca sexta haemolymph cDNA sequence and mRNA expression. Gene 172: 255-259. Prince, R.W., Cox, C.D.m and Vasil, M.L. (1993) Coordinate regulation of siderophore and exotoxin A production: molecular cloning and sequencing of the Pseudomonas aeruginosa fur gene. J. Bacteriol. 175: 2589-2598. Soderhall, E. and Cerenius, L. (1998). Role of the prophenoloxidase-activating system in invertebrate immunity. Curr. Opin. Immunol. 10: 23-28. Roeser, H. (1980) Iron metabolism in inflammation and malignant disease. In: Jacobs A, Worwood M, editors. Iron biochemistry in biochemistry and medicine II, London: Academic Press. Pp 605-614. 94 Rogers, J.T., Bridges, K.R., Durmowicz, G.P., Glass, J., Auron, P.E., Munro, H.N. (1990) Translational control during the acute phase response. Ferritin synthesis in response to interleukin-1. J. Biol. Chem. 9: 14572-14583. Salzet, M. (2001). Vertebrate innate immunity resembles a mosaic of invertebrate immune responses. Trends in Immunol. 22: 285-288. Shaw, G., Kamen, R. (1986) A conserved AU sequence from the 3′ untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46: 659-668. Singh, P.K., Parsek, M.R., Greenberg, E.P., and Welsh, M.J. (2002) A component of innate immunity prevents bacterial biofilm development. Nature 417: 552-555. Slimes, M.A., Addiego, J.E. Jr, Dallman, P.R. (1974) Ferritin in serum: diagnosis of iron deficiency and iron overload in infants and children. Blood 43: 581-590. Snedecor, G.W. and Cochran, W.G. (1967) Shortcut and npn-parametric methods. In: Cochran WG, ed. Statistical Methods, Iowa: Iowa State University Press pp 121-134. Soothill, J.S. (1992) Treatment of experimental infections of mice with bacteriophages. J. Med. Microbiol. 37: 258-261. Southern, L.L. and Baker, D.H. (1982) Iron status of the chick as affected by Eimeria acervulina infection and by variable iron ingestion. J. Nutri. 122: 2353-2362. 95 Stafford, J.L. and Belosevic, M. (2003) Transferrin and the innate immune response of fish: identification of a novel mechanism of macrophage activation. Dev. Comp. Immunol. 27: 539-554. Stieritz, D.D., Holder, I.A. (1975) Experimental studies of the pathogenesis of infections due to Pseudomonas aeruginosa: description of a burned mouse model. J. Infect. Dis. 131: 688-691. Surguladze, N., Thompson, K.M., Beard, J.L., Connor, J.R. and Fried, M.G. (2004) Interactions and Reactions of Ferritin with DNA. J. Biol. Chem. 279(15):14694-702. Takase, H., Nitanai, H., Hoshino, K. and Otani, T. (2000) Impact of Siderophore Production on Pseudomonas aeruginosa Infections in Immunocompromised Mice. Infect. Immun. 68: 1834-1839. Thompson, K.J., Fried, M.G., Ye, Z., Boyer, P. and Connor, J.R. (2002) Regulation, mechanisms and proposed function of ferritin translocation to cell nuclei. J. Cell Sci. 115: 2165-2177. Torti, S., Kwak, E., Miller, S., Miller, L., Ringold, C., Myambo, K., Young, A., Torti, F. (1988) The molecular cloning and characterization of murine ferritin heavy chain, a tumor necrosis factor-inducible gene. J. Biol. Chem. 263: 12638-12644. Torti, S.V. and Torti, F.M. (1998) Human H-kininogen is a ferritin-binding protein. J. Biol. Chem. 273: 13630-13635. 96 Touati, D. (2000) Iron and oxidative stress in bacteria. Archives of Biochem. Biophy. 373: 1-6. Tran, T.N., Eubanks, S.K., Schaffer, K.J., Zhou,, C.Y.J., and Linder, M.C. (1997) Secretion of Ferritin by Rat Hepatoma Cells and Its regulation by Inflammatory Cytokines and Iron. Blood 90: 4979-4986. Truty, J., Malpe, R., and Linder, M.A. (2001) Iron Prevents Ferritin Turnover in Hepatic Cells. J. Biol. Chem. 276: 48775-48780. Vasil, M.L., Chamberlain, C. and Grant, C.C.R. (1986) Molecular studies of Pseudomonas exotoxin A gene. Infect. Immun. 52: 538-548. Vasil, M.L. and Ochsner, U.A. (1999) The response of Pseudomonas aeruginosa to iron: genetics, biochemistry and virulence. Mol. Microbiol. 34: 399-413. Vierstraete, E., Verleyen, P., Beggerman, G., D’Hertog, W., Van den Bergh, G., Arckens, L., De Loof, A. and Schoofs, L. (2003) A proteomic approach for the analysis of instantly released wound and immune proteins in Drosophila melanogaster haemolymph. Proc. Natl. Acad. Sci. USA. 101: 470-475. Wang, L., Ho, B. and Ding, J.L. (2003) Transcriptional regulation of Limulus Factor C. Repression of an NFκb Motif Modulates its Responsiveness to Bacterial Lipopolysaccharides. J. Biol. Chem. 278: 49428-49437. 97 Ward, C.G., Bullen, J.J., in Bullen, J.J., Griffiths, E. (Eds.), Iron and infection, Molecular, Physiological, and Clinical Aspects, 2nd edition. Wiley, Chichester, 1999, pp 369-450. Welch, S. (1992) Transferrin. The Iron Carrier. CRC Press, Boca Raton. Wentworth, A.D., Jones, L.H., Wenworth, Jr, P., Janda, K.D., and Lerner, R.A. (2000) Antibodies have the intrinsic capacity to destroy antigens. Proc. Natl. Acad. Sci. USA 97: 10930-10935. Williams, H.L., Johnson, D.J., and Haut, M.J. (1977) Simulataneous spectrophotometry of Fe2+ and Cu2+ in serum denatured with guanidine hydrochloride. Clin. Chem. 23: 237-240. Winzerling, J.J., Nez, P., Porath, J., Law, J.H. (1995) Rapid and efficient isolation of transferrin and ferritin from Manduca sexta. Insect Biochem Mol Biol. 25:217-224. Wolz, C., K. Hohloch, A. Ocaktan, K. Poole, R.W. Evans, N. Rochel, A.M. AlbrechtGary, M.A. Abdallah, and G. Doring. (1994) Iron release from transferrin by pyoverdin and elastase from Pseudomonas aeruginosa. Infect. Immun. 62: 4021-4027. Worwood, M. (1986) Serum ferritin. Clinical Sci. 70: 215-220. 98 Wright, S., Ramos, R., Tobias, P., Ulevitch, R., Mathison, J. (1990). CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding proteins. Science 249: 1431-1433. Wright, S. (1991). Multiple receptors for endotoxin. Curr. Opin. Immunol. 3: 83-90. Yoshiga, T., Hernandez, V.P., Fallon, A.M. and Law, J.H. (1997) Mosquito tranferrin, an acute-phase protein that is up-regulated upon infection. Proc. Natl. Acad. Sci. USA 94: 12337-12342. Yoshiga, T., Georgieva, T., Dunkov, B.C., Harizanova, N., Ralchev, K. and Law, J.H. (1999) Drosophila melanogaster transferrin. Cloning, deduced protein sequence, expression during the life cycle, gene localization and up-regulation on bacterial infection. Eur. J. Biochem. 260: 414-420. Yuan et al. (2004) Regulation of LIP level and ROS formation through interaction of H-ferritin with G-CSF receptor. JMB 339: 131-144. Yun, E.Y., Kang, S.W., Hwang, J.S., Goo, T.W., Kim, S.H., Jin, B.R., Kwon, O., Kim, K.Y. (1997) Molecular Cloning and Characterization of a cDNA Encoding a Transferrin Homolog from Bombyx mori. Biol. Chem. 360: 1455-1459. 99 [...]... check Between 6-48 h of infection, the host plasma ferritin evades Pseudomonas- mediated degradation by transiting from extracellular to intracellular space, during which different ferritin isoforms constitute the ferritin complex Our data show that the host recovers its level of plasma ferritin by 72 h Furthermore, we demonstrate that contrary to the naïve ferritin, which binds the host DNA sequence-independently... α-helices and the heavy chain ferritin subunits then assemble into a apoferritin complex of 24 subunits in 432 symmetry viewed down a four-fold axis (C) The structures were obtained from the Protein Database Human H-chain: pdb 2fha; human L-chain: pdb 1aew The structure of the ferritin complex was adapted from Chasteen and Harrison, 1999 12 Studies have revealed that Fe2+ enters the core of the apoferritin,... addressed Firstly, the heterogeneity in ferritin molecules, some of which may be linked to its tissue of origin Secondly, what are the underlying biological implications for the presence of dimers, trimers and larger polymers? Lastly, the existence of isoferritins from rat, human and horse tissues still remains to be explained Until now, the source and nature of the trace level of plasma ferritin still... using Pseudomonas aeruginosa as a model pathogen We demonstrate the interesting phenomenon that on one hand, Pseudomonas attempts to degrade the host ferritin in order to usurp the host iron for its survival On the other hand, the host maintains iron homeostasis by tightly regulating its level of plasma ferritin, plasma redox potential and pH that keeps the plasma free radicals in check Between 6-48 h of. .. it has been isolated from plasma in trace amount Plasma ferritins serve as extracellular iron storage molecules and loss of plasma iron to pathogen is detrimental to the host during infection Interestingly, the horseshoe crab plasma iron level is 8-10-fold higher than human plasma In this study, horseshoe crab plasma ferritin complex was purified, characterized and its dynamic role in innate immune... ethlius, ferritin was isolated from the midgut of the larvae The holoferritin was stable 14 to heat at 75 oC or in the presence of SDS, proteinase K or Urea Calpodes ferritin contains iron and is a glycoprotein having N-linked high-mannose oligosaccharides There are 2 isoforms with a pI 6.5 – 7 and there are 2 major subunits of 24 and 31 kDa and 2 minor subunits of 26 and 28 kDa The 24 kDa subunit is induced... comprised of 24 subunits, each ~ 20 kDa, they reported that serum ferritin consists of more than 1 size of subunit and all were larger than intracellular ferritins The iron content of serum ferritin was also much lower than that of cytosolic ferritin Serum ferritin, and not its intracellular counterpart was also found to be glycosylated Various studies have demonstrated the elusiveness of serum ferritin. .. molecules found in the amebocytes and the hemolymph /plasma of the horseshoe crab are summarized in Table 2 The horseshoe crab is heavily dependent on the coagulation cascade, lectins and other defense factors for survival in the harsh environment Since the study of innate immunity in the vertebrates has been hampered by its acquired immune system, the use of invertebrates such as the horseshoe crab... colonize surfaces in a biofilm form causes the cells to be impervious to therapeutic concentrations of antibiotics Typically, Pseudomonas infection may compose of three distinct stages: (1) bacterial attachment and colonization; (2) local invasion; (3) disseminated systemic disease To colonize, the fimbriae of Pseudomonas adhere to the epithelial cells of the upper respiratory tract and they may bind to specific... into the culture supernatants and that the amount of iron in the supernatants decreased over time upon LPS or PMA treatments There was also enhanced expression of ferritin mRNA after stimulation (Beck et al., 2002) This was perhaps the only study so far to have demonstrated the involvement of invertebrate cytosolic ferritin in innate immune defense 1.3.3.2 Secreted ferritins The first report of secreted .. .DYNAMIC ROLE OF PLASMA FERRITIN DURING PSEUDOMONAS INFECTION: INSIGHTS FROM THE LIMULUS ONG SEK TONG DERRICK (Bachelor of Science (Hon)) A THESIS SUBMITTED TO THE FOR THE DEGREE OF MASTER OF. .. free radical level 75 A dynamic role of ferritin during Pseudomonas infection 76 iv 4.3 Insights into the role of plasma ferritin: from horseshoe crab to mammalian plasma ferritin 80 CONCLUSION... during P aeruginosa infection of the horseshoe crab Identification of limulus plasma ferritin complex and its subunits in the plasma (A) The native state of limulus plasma ferritin was detected