Echinoderm immunity and the evolution of the complement system

Developmental and Comparative Immunology 23 (1999) 429±442 Echinoderm immunity and the evolution of the complement system Paul S. Gross a,c, Walid Z. Al-Sharif b, Lori A. Clow b, L. Courtney Smith a,b,* a Department of Biological Sciences, George Washington University, 2023 G St. NW, Washington, DC 20052 USA Institute of Biomedical Sciences Program in Genetics, George Washington University, Washington, DC 20052 USA c Life Sciences, University of Maryland, College Park, MD 20742 USA b Accepted 1 October 1998 Abstract Our understanding of in¯ammatory responses in humans has it roots in the comparative approach to immunology. In the late 1900's, research on echinoderms provided the initial evidence for the importance of phagocytic cells in reactions to foreign material. Studies of allograft rejection kinetics have shown that echinoderms have a non-adaptive, activation type of immune response. Coelomocytes mediate the cellular responses to immune challenges through phagocytosis, encapsulation, cytoxicity, and the production of antimicrobial agents. In addition, a variety of humoral factors found in the coelomic ¯uid, including lectins, agglutinins, and lysins, are important in host defense against pathogens and other foreign substances. Recently, a simple complement system has been identi®ed in the purple sea urchin that is homologous to the alternative pathway in vertebrates. The sea machine homologue of C3, is inducible by challenge with lipopolysaccharide, which is known to activate coelomocytes. Complement components have been identi®ed in all vertebrate classes, and now have been characterized in protochordates and echinoderms indicating the primordial nature of the complement system. Because it is thought that the complement system evolved from a few primordial genes by gene duplication and divergence, the origin of this system appears to have occurred within the common ancester of the deuterostomes. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Echinodermata; Complement; Alternative pathway; Evolution; Coelomocytes; Immunology 1. Historical background The comparative approach to immunology * Corresponding author. Tel.: +1-202-994-9211; fax: +1202-994-6100. E-mail address: [email protected] (L.C. Smith) had its origin in early studies of the causes and mechanisms of in¯ammation. Elie Metchniko€, the chief proponent of the cellular theory of immunity, asserted that in¯ammation was the most important phenomenon in pathology and proposed that the primary e€ectors of the immune response were circulating, amoeboid-like, phagocytic cells [1]. In sup- 0145-305X/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 5 - 3 0 5 X ( 9 9 ) 0 0 0 2 2 - 1 430 P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 port of his cellular theory of in¯ammation, Metchniko€ observed that under natural conditions, the bipinnaria larvae of the sea star, Astropecten pentacanthus, would respond to both physical injury and to infection by marine algae. These initial observations led Metchniko€ to perform the benchmark experiments in cellular immunology which included introducing rose prickles and glass rods into a bipinnaria larva and ®nding that mesodermal cells migrated to the injury site and encapsulated the prickle (for review see Ref. [1]). Furthermore, he demonstrated that this phenomenon also occurred when bacteria were introduced into the larva, noting that the bacteria were `devoured alive' by the phagocytic cells. In 1908, Metchniko€, along with Paul Ehrlich, was awarded the Nobel Prize for this groundbreaking work using echinoderms as a model, thus beginning the ®eld of comparative cellular immunology. 2. Allograft rejection For nearly half a century after Metchniko€, the comparative approach was largely ignored until William Hildemann became interested in how sessile animals on a crowded coral reef defended themselves against overgrowth and fusion. He and his colleagues investigated echinoderm immunity by employing allograft rejection to assess immune capabilities in these organisms. Sea stars (Protoreaster nodosus and Dermasterias imbricata ) and a sea cucumber (Cucumaria tricolor ) were used to demonstrate the capability of these animals to di€erentiate between self and allogeneic tissues [2,3]. Results indicated that they would reject ®rst set skin allografts in a chronic manner, that repeated allografts (2nd set and 3rd set) were rejected more quickly, and that autografts healed in and remained intact. Unfortunately, third party allografts were either not done or were inconclusive. Histological analyses of rejecting allografts had a signi®cant cellular in®ltrate that increased with time [2±4] and appeared to correspond to the phagocytes in the sea star larva that were observed by Metchniko€. These studies were the ®rst to demonstrate that echinoderms could respond to allogeneic tissue although the speci®city of the response was unclear. To clarify the non-self recognition capabilities of echinoderms and to characterize their rejection kinetics, primary, secondary and third party allografting experiments were performed on the sea urchin, Lytechinus pictus [5±7] and to a lesser extent on the sea urchin, Strongylocentrotus purpuratus [6]. Results indicated that sea urchins lacked speci®c non-self recognition and true immunological memory since 2nd set and 3rd party tissues were rejected at the same rate [6,8]. 3. Coelomocyte functions The defense capabilities of echinoderms were de®ned from allograft rejection experiments, but research has also concentrated on characterizing and understanding the functions of the coelomocytes, the cells that mediate these responses. Coelomocytes are found in the ¯uid ®lled coelom and, depending on the species, can be a mixture of several morphologically di€erent types. Sea urchins of the genus Strongylocentrotus have four types of coelomocytes which include phagocytes (present in all echinoderms, also called bladder, petal or ®liform amoebocytes), red spherule cells (also called eleocytes, morula cells and pigment cells), colorless spherule cells (also called white morula cells) and vibratile cells [9±12]. 3.1. Coelomocyte in®ltrates in injury and infection Coelomocytes have long been considered to be mediators of the immune response, in part, because of their presence surrounding injuries and infections in several echinoderm species and their appearance in grafted tissues. Skin infections become ringed with black or dark red tissues that are accumulations of red spherule cells in the sea urchin, Paracentrotus lividus [13] and in S. purpuratus (Smith, unpublished observation). In®ltrates of red spherule cells have also been documented as red spots on gonadal tissue of the sea urchin Strongylocentrotus intermedius P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 that surround metacercaria of a parasitic worm [14]. Cellular in®ltrates consisting of phagocytes and red spherule cells have also been noted around broken, infected, and regenerating spines of S. purpuratus [15]. Red pigmented cells accumulated at sites of surgical injuries during the initial phase of tissue transplantation in L. pictus [5] and histological analysis of allografts in the sea star, D. imbricata revealed mixed cellular in®ltrates in which cell densities increased during the chronic rejection process [3,4]. 3.2. Coelomocyte responses to foreign substances The survival of echinoderms in the microberich marine environment is dependent on their ability to defend themselves against microbial invasion. Combating infections must include rapid, ecient, and sometimes selective clearance of foreign invaders in order for an animal to survive. The capability of echinoderms to eliminate injected pathogens, foreign cells, and other types of particles has been well documented. Injections of bacteria into the coelomic cavity of the sea star D. imbricata [16] or into the sea urchins, Echinus esculentus [17], S. purpuratus, [18] or Strongylocentrotus droebachiensis [19] were eciently cleared in a few hours to a few days. In contrast to the sterile coelomic ¯uid of sea urchins, pre-injection analysis of coelomic ¯uid from the sea cucumber, Parastichopus californicus, revealed the presence of a natural bacteria ¯ora with as many as 108 bacteria per milliliter of coelomic ¯uid. Consequently, bacteria injected into the coelomic cavity of P. californicus were either not cleared or were cleared at varying rates which depended on whether the bacteria were isolated from the gut of the sea cucumber or were obtained from other sources [20]. Echinoderms are also capable of clearing xenogeneic cells, foreign non-cellular particles and proteins. Injection of cells from the sea urchin, Arbacia punctulata into the coelom of the sea star, Asterias vulgaris resulted in clumping of the sea star phagocytes, trapping or phagocytosis of the sea urchin cells, and rapid clearance of the injected cells from the coelomic ¯uid [21]. The sea cucumber, Holothuria polii eciently phago- 431 cytosed and encapsulated injected red blood cells (RBCs) [22], as did the sea urchin S. droebachiensis [23]. The sea urchin, L. pictus, eciently cleared T4 bacteriophage [24] and carmine particles injected into the coelomic cavity of P. californicus resulted in agglutination, encapsulation and brown body formation followed by the excretion of the brown bodies into the cloaca [20]. Even latex beads were quickly cleared from the coelomic cavity of the sea urchin, S. droebachiensis [23]. In addition to using repeated allografting to assess the speci®city of immune memory in echinoderms, repeated injections of foreign substances have also been performed. However, no di€erences have been noted in clearance rates of bacteria injected multiple times into the coelomic cavity of S. purpuratus, regardless of the interval between inoculations [18]. Similarly, accelerated clearance rates were not demonstrated when xenogeneic cells were injected a second time into A. vulgaris [21], nor after repeated injections of T4 bacteriophage into L. pictus [24]. These results corresponded with those obtained from allografting experiments, which had demonstrated the absence of speci®c immunological memory in echinoderms. The activities and functions of coelomocytes have also been studied in vitro, where the coelomic ¯uid from a number of echinoderms has been shown to be bactericidal. Coelomic ¯uid from P. lividus exhibited higher bactericidal activity in vitro when coelomocytes were present than after the cells had been removed [25]. Similar results were reported for coelomocytes from E. esculentus, which were found to be bactericidal against Pseudomonas sp. and had a wide range of antibacterial activity against both gram negative and gram positive marine bacteria [17,26]. Lysates of phagocytes and red spherule cells from P. lividus were bactericidal against both Vibrio spp. and Photobacterium spp. suggesting that lysozymes and echinochrome (the pigment of red spherule cells) produced by these cells might mediate the bactericidal activity [27]. In addition, echinochrome-A from E. esculentus had e€ective bactericidal activity against marine bacteria in vitro [28]. 432 P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 3.3. Cytotoxicity The phagocyte appears to be the cell type involved in cytotoxic reactions in mixtures of coelomocytes in vitro. When phagocytes from the sea urchin S. droebachiensis were co-cultured with phagocytes from either E. esculentus or Strongylocentrotus pallidus, 90% of the cells were killed, and in allogeneic mixtures of S. droebachiensis phagocytes, 70% of the cells were killed [29]. However, there was signi®cant variability in the amount of cell killing between di€erent combinations of cells, and Dales [30] was unable to repeat this result using coelomocytes from sea stars and di€erent species of sea urchins. This indicated that either that phagocytes do not have cytotoxic capabilities, or that this method of analyzing coelomocyte function had a number of technical diculties. On the other hand, others were able to isolate cytolytic granules from phagocytes of P. lividus, suggesting that these cells mediated their killing function through the release of this cytolytic material [31]. 3.4. Phagocytosis A subpopulation of coelomocytes from echinoderms are de®ned by their amoeboid behavior and their abilities to engulf foreign cells and particles. Phagocytes from two sea urchin species, Strongylocentrotus franciscanus and S. purpuratus, were noted to chemotax towards marine bacteria, with gram positive bacteria being phagocytosed more readily than gram negative bacteria [10]. Human and sheep RBCs were taken up within 30 min in vitro by phagocytes from the sea urchin, Strongylocentrotus nudus, and RBCs opsonized with coelomic ¯uid from animals that had been pre-injected with RBCs enhanced the phagocytic rate compared to nonopsonized red cells [32]. An important function of the phagocytic cells is degradation of phagocytosed material. Circulating phagocytes from the sea cucumber, H. polii, contain a rich selection of lysosomal enzymes, including acid and alkaline phosphotases, b-glucuronidase, aminopeptidase, acid and alkaline protease, and lipase [33]. The presence of lysozyme and acid phosphatase has also been documented in the sea cucumber P. californicus [20]. In addition, arylsulphatase, a lysosomal hydrolase known to play a role in in¯ammatory phenomena by catalyzing the hydrolysis of sulfate bonds, has been biochemically detected in coelomocyte lysate preparations of seven di€erent echinoderms [34] and a cDNA from S. purpuratus coelomocytes matched to arylsulfatase by expressed sequence tag (EST) analysis [35]. 4. Humoral factors 4.1. Cytolytic, bactericidal, and agglutating factors The capacity of humoral factors to damage target cells has been shown for a number of echinoderms. Hemolysins have been detected from coelomic ¯uid of the sea star, A. forbesi [36], from several sea urchin species [25,37], as well as from the sea cucumber, H. polii [38]. The hemolysin from P. lividus binds to erythrocytes, zymosan particles, lipopolysaccharide (LPS), and laminarin surfaces, but not to self or allogeneic cell membranes [39]. Factors in the coelomic ¯uid of the sea cucumber P. californicus, were variably e€ective at killing bacteria in vitro depending on the source of the bacteria [20]. Another type of humoral factor, agglutinins, appear to be involved in maintaining body integrity after injury and in encapsulating foreign invaders. Divalent cation-dependant hemagglutinins have been found in three sea urchin species [40], the sea star, Asterina pectinifera [41], and H. polii [38). The hemagglutinin in P. lividus, a heterotrimeric complex that binds rabbit RBCs, enhances the adhesive properties of autologous coelomocytes and may be involved in cell±cell and cell±matrix interactions such as clotting, wound repair, opsonization, and encapsulation [42]. Lectins are, in part, responsible for agglutination reactions by binding to speci®c carbohydrate structures on cells and on extracellular matrices. Lectins have been isolated from the coelomic ¯uid of several echinoderms and are thought to be important in identifying foreign P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 cells through opsonization. Two di€erent Ca2+dependent (or C-type) lectins have been identi®ed in the sea cucumber, Stichopus japonicus [43] and three lectins from the sea star A. pectinifera have di€erent binding abilities; one preferentially agglutinates rabbit RBCs, another binds human RBCs, and the third is a bacterial agglutinin [41]. Echinoidin, a C-type lectin from the sea urchin, Anthocidaris crassispina, also has an RGD sequence [44] that can mediate cell adhesion between mammalian cancer cells [45]. In fact, two ESTs with sequence similarities to the carbohydrate binding region of echinoidin have been identi®ed from a coelomocyte cDNA library from S. purpuratus [35]. The C-type lectin, CELIII, from the sea cucumber Cucumaria echinata, is another hemagglutinin that lyses rabbit and human RBCs by a novel pore-forming mechanism and may be toxic to foreign microbes [46]. In echinoderms, lectins play important roles in defense by performing opsonization and lytic functions as well as functioning in clot formation and wound repair. 4.2. Other humoral immune e€ectors The phenoloxidase system (PO) is important in immune defense in several groups of invertebrates. In arthropods, the PO system is well characterized and is part of an immune surveillance cascade (for review see Ref. [47]). In echinoderms, PO activity has been identi®ed in coelomic ¯uid and in circulating coelomocytes from certain echinoderms, including the sea star Asterias rubens and the sea urchin, Diadema antillarum [48,49]. A proposed model of activation, based on the arthropod system, involves the activation of a pro-enzyme by trypsin, to yield an intermediate which, in turn, is stimulated by calcium to form the active phenoloxidase tetramer [50,51]. Another common defense mechanism in plants and animals is the production of reactive oxygen intermediates, although at present the phagocytes from the sea urchin, S. nudus are the only echinoderm cells known to produce hydrogen peroxide in vitro when co-incubated with RBCs [32]. In vertebrates, a well studied family of immune 433 e€ector molecules are the cytokines, such as interleukin-1, which have wide spread e€ects on proliferation of immune cells and other stimulatory e€ects, resulting in rapid and ecient responses to immune challenges. Interleukin-1-like activity has been found in A. forbesi [52] and cytokine activity has also been observed in the sea star, Pisaster ochraceus [53]. Of interest is the heterodimer known as sea star factor from A. forbesi, which appears to have cytokine-like properties and may function to regulate the sea star immune response [54±56]. When minute quantities of sea star factor were injected into the coelomic cavity of A. forbesi, it functioned like a non-mitogenic lectin, stimulating rapid but temporary, localized, tight aggregations of circulating coelomocytes. Although none of the cytokine-like molecules from echinoderms have been sequenced and de®nitively shown to be interleukin homologues, the systemic responsiveness of echinoderms to minimal and localized immune challenges infers that some type of molecular signaling system exists in these animals. Based on the data reviewed above, echinoderms appeared to have a typical invertebrate type of immune system that functioned through non-adaptive mechanisms. Immune responses were based on coelomocyte activity (chemotaxis and phagocytosis) working in parallel with a variety of humoral factors (lectins, agglutinins, opsonins, bactericidal and probably fungicidal agents) that reacted directly with invading pathogens. No aspect of the echinoderm defense response was homologous to any subsystem of the vertebrate immune response. This changed when a homologue of a vertebrate complement component was identi®ed in the sea urchin, S. purpuratus [35]. 5. Molecular immunology of the purple sea urchin 5.1. Coelomocyte activation Coelomocytes mediate host responses to immune challenges through activities that include increased chemotaxis and phagocytosis, and the formation of cellular clots. These activities require signi®cant changes in cell shape which 434 P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 are mediated by changes in the actin cytoskeleton [57]. Pro®lin is a major actin binding and cytoskeletal modifying protein that plays a central role in either inhibiting or promoting the polymerization of actin monomers into ®laments [58] thereby modulating the shape of the cell cytoskeleton. An increase in pro®lin gene expression was noted in coelomocytes from S. purpuratus responding to injury [59] and to injections of LPS [60]. This suggested that a change in pro®lin message was an indication of coelomocyte activation during immune responses. This correlation was supported by studies of pro®lin expression during gastrulation, where increases in pro®lin transcripts occurred with the onset of migratory behavior of the primary mesenchyme cells and the formation of ®lopodia on secondary mesenchyme cells [61]. Consequently, titration of the pro®lin transcript content in coelomocytes was used to infer the activation status of the sea urchin immune response [59,60]. 5.2. Homologues of complement components A random primed, directionally cloned cDNA library was constructed from coelomocytes that had been analyzed for their level of activation after injection of LPS [35,60]. This library was used in an EST study to identify genes that were transcribed in activated cells and that encoded immune e€ector proteins [35]. Two ESTs of particular interest appeared similar to members of the vertebrate complement cascade. The ®rst, EST064, matched to the thioester protein family which is de®ned by the presence of a highly reactive carbonyl that is formed between cysteine and glutamine residues within the conserved thioester site that is capable of forming covalent bonds with hydroxyl or amino groups on target proteins. The thioester protein family includes complement components C3, C4, C5 and a2-macroglobulin (a2M). Although homologues of the thioester proteins have been identi®ed in many vertebrates, homologues of a2M have also been characterized in invertebrates including arthropods and molluscs [62]. The second clone, EST152, matched to several short consensus repeats (SCRs) or complement control protein modules. SCRs are small domains that are typical of proteins involved in the vertebrate complement system including complement receptors, factor B (Bf) and C2, and a number of complement regulatory proteins, but have also been found in noncomplement proteins such as factor C from the horseshoe crab Limulus [63], factor XIIIb, a blood clotting protein in vertebrates, [64] and the interleukin 2 receptor [65]. These two ESTs, being expressed by the immune cells of the sea urchin, S. purpuratus, suggested that a simple complement system was present in this invertebrate and was composed of, at least, a thioester containing protein and a complement receptor or regulatory protein. This was the ®rst molecular evidence that an echinoderm had an element within its immune system that was homologous to a system found only in vertebrates; evidence that inferred homology of immune systems within the deuterostomes [35]. The existence of an echinoderm complement system had been hinted at previously, in that phagocytes from S. droebachiensis showed augmented phagocytosis of RBCs opsonized by human C3, suggesting that coelomocytes had a receptor for C3b or C3bi [23,66±69]. Furthermore, inhibitors of complement opsonization decreased or blocked phagocytosis of RBCs by sea urchin phagocytes [70]. Others investigating the hemolytic activity in coelomic ¯uid from A. forbesi and H. polii also suggested that a complement-like system functioned in echinoderms [36,71]. However, comparisons of the holes produced by the hemolytic factor from two species of sea cucumber and those resulting from mammalian complement showed signi®cant di€erences by electron microscopy, indicating that they were most likely produced by di€erent mechanisms [39,72]. These early results suggested that echinoderms might have an alternative pathway of complement that lacked a terminal pathway. However, the lectin pathway of complement activation [73,74] or a lectin opsonin was not considered, even though certain sugars were shown to block the reaction reaction [70]. P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 5.3. SpC3, a homologue of vertebrate C3 The ®rst de®nitive identi®cation of a complement component in an invertebrate came when the sequence of EST064 was completed. Analysis of the deduced amino acid sequence indicated that the encoded protein, SpC3, was a new member of the thioester complement protein family, and was a homologue of the vertebrate complement component C3 [75]. The homology to vertebrate C3 was based on several conserved regions identi®ed from the alignment between SpC3 and members of the thioester protein family and included the thioester site, ba junction (no ag junction), C3 convertase site, two factor I cleavage sites, and cysteines in conserved positions including those involved in forming the interchain disul®de bridge [Fig. 1(A)]. On reducing gels, SpC3 revealed two chains with sizes similar to those in other C3 proteins. The gene (Sp064) is expressed speci®cally by the coelomocytes, the protein is present in coelomocytes and in coelomic ¯uid. The phylogenetic analysis of the thioester family of proteins indicated that SpC3 is the most ancient member, implying that the alternative pathway of the complement system was present in the common ancestor of the deuterostomes rather than the common ancestor of the vertebrates, as had been previously assumed [75]. 5.4. SpC3 is induced in coelomic ¯uid by LPS Since the SpC3 protein was present in coelomic ¯uid of S. purpuratus [75], we were interested to know if the amount of SpC3 protein would change after challenge with LPS. Sea urchins used for this experiment were maintained in the controlled environment of a large marine aquarium for approximately 1.5 yr before they were challenged. We determined that these animals had become immunoquiescent since the concentration of SpC3 in the coelomic ¯uid had fallen to undetectable levels and could serve as excellent controls for studies of immune activation. Using Western blots, animals receiving a single injection of LPS showed detectable SpC3 in circulation 15 min post-injection (Fig. 2). 435 SpC3 concentration peaked at approximately 2 hr, gradually declining after 4 hr, and returning to control levels by 24 hr post-injection [Fig. 2(A)]. Animals receiving injections of ®lter sterilized sea water did not show a sharp peak in SpC3 concentration at 2 hr, but instead showed a slow increase in circulating SpC3, which became identical, at 24 hr, to animals receiving LPS [Fig. 2(B)]. Animals receiving no injections showed no discernible SpC3 in the coelomic ¯uid [Fig. 2(C)]. The densometric scans of the Western blots show these changes more clearly [Fig. 2(D)]. The rapid appearance of SpC3 in coelomic ¯uid after LPS injection may have been due to the rapid secretion of stored SpC3 and/or to the induction of SpC3 gene transcription, followed by translation of the message and secretion of the newly produced protein. The appearance of SpC3 in coelomic ¯uid was transient perhaps because of the minimal challenge posed by a single injection of LPS (see legend to Fig. 2). After ®ve injections of LPS, SpC3 appears in circulation in much greater amounts and is present for at least 90 days (Clow, Gross, Shih and Smith, unpublished). 5.5. SpBf, a homologue of vertebrate factor B The completed analysis of cDNAs that included the EST152 sequence indicated that the encoded protein was a homologue of vertebrate Bf, and was called SpBf [Fig. 1(B)] [76]. Like other members of the Bf/C2 family, SpBf has a mosaic structure which includes ®ve SCRs, a von Willebrand factor (vWF) domain, and a serine protease domain. The gene encoding SpBf (Sp152) is expressed speci®cally in coelomocytes. The alignment between SpBf and vertebrate Bf/ C2 proteins revealed amino acids in conserved positions for serine protease activity, a conserved factor D cleavage site, and Mg2+ binding sites that, in vertebrate Bf proteins, function in interactions with C3b during the formation of C3 convertase. The presence of ®ve SCRs in SpBf was unusual since all other Bf/C2 proteins have three SCR domains. Sequence alignments and phylogenetic analyses of SCR domains from all the Bf/C2 proteins were used to identify similarities among these domains and to predict any 436 P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 Fig. 1. Schematic representations of SpC3 and SpBf (75, 76). A. The 210 kD SpC3 protein (total length=1699 amino acids) is composed of two chains, a (130 kD) and b (80 kD). The ba junction, thioester site and cleavage sites for a putative factor I and C3 convertase are positioned to scale. Four of the ®ve consensus N-linked glycosylation sites (CHO) appear to carry sugars (Al-Sharif and Smith, unpublished) and are shown. The positions of the putative interchain disul®de bond is indicated by dotted lines and is based on conserved positions of cysteines in SpC3 that correspond to cysteines involved in disul®de bonding in other C3 proteins (90; see Figure 3 in Ref. [75]). To receive an alignment of all the thioester family proteins including SpC3 by e-mail, send a request to [email protected] with the message; GET ALIGN:DS31395.DAT. B. The SpBf protein is a mosaic protein typical of the Bf/C2 protein family. It is composed of ®ve SCRs, a von Willebrand Factor domain, and a serine protease domain. The conserved cleavage site for a putative factor D is indicated in addition to the amino acids involved in the protease function. To receive an alignment of the Bf/C2 protein family including SpBf by e-mail, send the request to [email protected] with the message; GET ALIGN:DS33817.DAT. functional signi®cance of ®ve SCRs in SpBf. Based on known functional importance of individual SCRs in Bf for binding to C3b in vertebrates [77±79], results suggested that the three SCRs closest to the vWF domain in SpBf might be more important for binding than the two Nterminal SCRs. Furthermore, SpBf was found to be the most primitive member of the Bf/C2 protein family, and having ®ve SCRs appeared to be the primitive state for the Bf/C2 protein family [76]. 5.6. The feed-back loop of the alternative pathway Together, SpC3 and SpBf are among the ®rst complement proteins to be described from an invertebrate, and they appear to be the central components of a primitive complement system that is homologous to the alternative pathway. Previously, Lachmann [80] proposed a sequence of evolutionary steps that began with an `archeocomplement system' which functioned essentially the same as the alternative pathway, and culminated in the higher vertebrate complement system. The archeo-complement system consisted of a C3-like protein with a thioester site, and a factor B-like protein containing SCRs and a serine protease domain and would be capable of autoampli®cation through a positive feedback loop, generating large numbers of activated C3 molecules that would quickly opsonize a pathogen (or any other surface lacking protection against complement attack). A model of how the sea urchin complement system might function, is based on the components that have been identi®ed in the sea urchin thus far. It is feasible that P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 437 Fig. 2. Western blot analysis of SpC3 in coelomic ¯uid from S. purpuratus after challenge with LPS. Immunoquiescent sea urchins were injected with 2 mg LPS/ml coelomic ¯uid (A), 2 ml sterile sea water/ml coelomic ¯uid (B), or received a needle injury only (C). The volume of coelomic ¯uid was estimated according to Smith et al. [60]. Coelomic ¯uid (300 ml) was withdrawn into a syringe (pre-loaded with 300 ml of ice cold Ca2+/Mg2+-free sea water with 30 mM EDTA and 50 mM imidazole) by inserting the needle through the peristomeum into the coelomic cavity. Samples were taken 15 min prior to injection (tÿ0.25), 15 min post injection (t+0.25), 2 hr post injection (t2), 4 hr post injection (t4), 8 hr post injection (t8), and 24 hr post injection (t24). Coelomocytes were pelleted immediately and ¯uid samples were stored at ÿ708C until use. To prevent general immune activation of all the animals from LPS contamination in the water, sea urchins were maintained in small individual aquaria with aeration, at 168C, for the duration of the experiment. Coelomic ¯uid was mixed 2:1 with lysis bu€er (lysis bu€er=4% SDS, 20% b-mercaptoethanol, 20% glycerol, 0.1 M Tris), loaded onto SDS-polyacrylamide gels (4.5% stacking, 8% separating) and run under reducing conditions. Proteins were electroblotted to nitrocellulose (BioRad), blocked with 5% milk in Tris bu€ered saline, incubated with rabbit anti-SpC3-a' chain antibody, followed by goat±anti-rabbit HRP antibody (Pierce) and visualized on ®lm by ECL (Pierce). The primary antibody was produced according to Al-Sharif et al. [75] and directed to a peptide (SGGDGGEQNAAVKVRDDFRETWFFD) that corresponds to the N-terminus of the a chain after cleavage at the putative C3 convertase site. The control lane (C) on all blots represents coelomic ¯uid from a single animal that was maintained in a sea water system open to the Paci®c Ocean. D. Relative densitometric scans of bands from the western blots corrected to the signal from the control (C) lane. W=LPS challenged animal, Q=SSW challenged animal, .=needle injury only animal. the echinoderm complement system might function like the proposed archeo-complement system in which activation would be initiated either spontaneously by the `tick-over' mechanism of SpC3 [81], or by the lectin pathway that has been identi®ed in vertebrates and ascidians [73,74,82,83] and probably functions in echinoderms as well. Activated SpC3, with an exposed thioester in the ¯uid phase would bind to an appropriate unprotected surface such as a microbe. Next, SpBf would bind to SpC3, rendering the bound SpBf susceptible to cleavage and activation by a putative factor D. The resulting complex would then function as a C3 convertase through the activity of the serine protease domain on SpBf. More SpC3 would be activated by the convertase which might also bind to the pathogen forming more convertase creating the positive feedback loop. Eventually, the pathogen would become covered with complement complexes, stimulating ecient phagocytosis or encapsulation by phagocytes. In addition, a factor I-like homologue would be a necessary inhibitor/regulator to cleave and degrade activated SpC3 in the presence of certain cofactors on cell surfaces and in circulation (for review of complement regulation, see Ref. [84]). This type of opsonization system would be remarkably ecient, and therefore, more valuable in host defense than simple opsonins. At present there is no direct evidence for echinoderm homologues of factor D, factor I (and cofactors), or a complement receptor, however, our preliminary data suggests that the sea urchin complement system may be activated and regulated by mechanisms that are similar to those that function in the 438 P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 Fig. 3. Phylogeny of Complement in Deuterostomes. The appearance of complement components in deuterostome phylogeny suggests origins in the echinoderms with duplications of genes and pathways in more advanced animals. The information was compiled from the following references which are also cited within the table: [74±76, 83, 89±107]. No citations are included for the classical, alternative, and terminal pathway components in higher vertebrates. alternative pathway in vertebrates (Al-Sharif and Smith, unpublished). 6. The evolution of complement in deuterostomes The theory that the complement cascades in higher vertebrates evolved from a few primordial genes through gene duplication and subsequent divergence of function [85] is based on similarities in protein sequence and function (e.g., the thioester protein family; the C2/Bf family; members of the terminal pathway), clustered organization of some complement genes (regulators of complement activation cluster, C2 and Bf linkage in mammals) and parallels between pathways (e.g., the alternative and classical pathways; the lectin pathway and the C1 complex) [86,87]. Complement systems of varying complexity have been identi®ed in all vertebrate classes [88] and with the identi®cation of C3 homologues in the sea urchin [75] and in a tunicate [89], this expands the range of animals in which comp- lement has been found to include the entire lineage of deuterostomes. The number of complement pathways in an organism and the complexity of each pathway correlates with phylogeny; more complex systems being found in more advanced deuterostomes (Fig. 3). All deuterostomes have, at minimum, a simple alternative pathway with ancient opsonin function, and most deuterostomes can activate the complement system through the lectin pathway. In the elasmobranchs, where immunoglobulins are ®rst identi®ed, the number of components, pathways (the classical and terminal pathways appear), and functional capabilities (antibody e€ector system) of the complement system expands signi®cantly. In higher vertebrates (teleosts and tetrapods), the alternative and classical pathways are expanded further, adding more components to complete the system (Fig. 3). Analysis of the complement system in animals throughout the deuterostome lineage can be imagined as `snap shots of complement evolution' in which several of the evolutionary steps have been P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 preserved, making investigations of the molecular evolution of the complement system accessible in living organisms, rather than being lost to extinction. The phylogenetic analyses of SpC3 and SpBf indicate that these echinoderm proteins are the most primitive of the thioester and Bf/C2 families of complement components, respectively [75,76]. The sea urchin system conforms to the `archeocomplement system' suggested by Lachmann [80] as the simplest recognizable complement system. Thus, the sea urchin complement system may bear similarities to the system that functioned in the ancestral deuterostome and that gave rise to the complement cascades in the higher deuterostomes [75]. Future investigations of the complement components in echinoderms may provide keys to the primitive beginnings of the deuterostome complement system, which may allow us to identify homologous functions in higher vertebrates and to piece together how the multiple subsystems in the mammalian immune response came to function together. Acknowledgements The authors would like to sincerely thank Drs. John D. Lambris and J. Oriol Sunyer for the anti-peptide antiserum that was used in the SpC3 induction study. References [1] Metchniko€ E. Lectures on the comparative pathology of in¯ammation: delivered at the Pasteur Institute in 1891. (Starling FA, Starling EH, MD., Trans.) Kegan Paul, Trench, Trubner & Co., Ltd., London 1893. [2] Hildemann WH, Dix TG. Transplantation reactions of tropical Australian echinoderms. Transplantation 1972;15:624±33. [3] Karp RD, Hildemann WH. Speci®c allograft reactivity in the sea star Dermasterias imbricata. Transplantation 1976;22:434±9. [4] Varadarajan J, Karp RD. Histological versus morphological assessment of graft rejection in invertebrates. Transpl 1983;35:629±31. [5] Co€aro KA, Hinegardner RT. Immune response in the sea urchin Lytechinus pictus. Science 1977;197:1389±90. 439 [6] Co€aro KA. Transplantation immunity in the sea urchin. Doctoral dissertation, University of California, Santa Cruz, CA, 1979. [7] Co€aro KA. Memory and speci®city in the sea urchin Lytechinus pictus. In: Manning MJ, editor. Phylogeny of immunological memory. New York, NY: Elsevier/ North-Holland Biomedical Press, 1980. p. 77±80. [8] Smith LC, Davidson EH. The echinoid immune system and the phylogenetic occurrence of immune mechanisms in deuterostomes. Immunology Today 1992;13:356±62. [9] Boolootian RA, Giese AC. Coelomic corpuscles of echinoderms. Bio Bull 1958;15:53±63. [10] Johnson PT. The coelomic elements of sea urchins (Strongylocentrotus ). III. In vitro reaction to bacteria. J Invert Pathol 1969;13:42±62. [11] Karp RD, Co€aro KA. Cellular defense systems of the Echinodermata. In: Manning MJ, editor. Phylogeny of immunological memory. Amsterdam, The Netherlands: Elsevier/North-Holland, 1980. p. 257±82. [12] Smith VJ. The echinoderms. In: Ratcli€e NA, Rowley AF, editors. Invertebrate blood cells. New York, NY: Academic Press, 1981. p. 513±62. [13] HoÈbaus E. Coelomocytes in normal and pathologically altered body walls of sea urchins. In: Jangoux M, editor. Proceedings of the European colloquium on echinoderms. Rotterdam, The Netherlands: A.A. Balkema, 1979. p. 247±9. [14] Shimizu M. Histopathological investigation of the spotted gonad disease in the sea urchin, Strongylocentrotus intermedius. J Invert Pathol 1994;63:182±7. [15] Johnson PT, Chapman FA. Abnormal epithelial growth in sea urchin spines (Strongylocentrotus franciscanus ). J Invert Pathol 1970;16:116±22. [16] Kaneshiro ES, Karp RD. The ultrastructure of coelomocytes of the sea star Dermasterias imbricata. Biol Bull 1980;159:295±310. [17] Wardlaw AC, Unkles SE. Bactericidal activity of coelomic ¯uid form the sea urchin Echinus esculentus. J Invert Pathol 1978;32:25±34. [18] Yui M, Bayne C. Echinoderm immunology: bacterial clearance by the sea urchin Strongylocentrotus purpuratus. Biol Bull 1983;165:473±85. [19] Plytycz B, Seljelid R. Bacterial clearance by the sea urchin, Strongylocentrotus droebachiensis. Dev Comp Immunol 1993;17:283±9. [20] Dybas L, Frankboner PV. Holothurian survival strategies: mechanisms for the maintenance of bacteriostatic environment in the coelomic cavity of the sea cucumber, Parastichopus californicus. Dev Comp Immunol 1986;10:311±30. [21] Reinisch C, Bang FB. Cell recognition of the sea star (Asterias vulgaris ) to the injection of amoebocytes of the sea urchin (Arbacia punctulata ). Cell Immunol 1971;2:496±503. [22] Canicatti C, D'Ancona G. Cellular aspects of 440 [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 Holothuria polii immune response. J Invert Pathol 1989;53:152±8. Bertheussen K. Endocytosis by echinoid phagocytes in vitro. I. Recognition of foreign matter. Dev Comp Immunol 1981;5:241±50. Co€aro KA. Clearance of bacteriophage T4 in the sea urchin Lytechinus pictus. J Invert Pathol 1978;32:384±5. Stabili L, Pagliara P, Metrangolo M, Canicatti C. Comparative aspects of Echinoidea cytolysins: the cytolytic activity of Spherechinus granularis (Echinoidea) coelomic ¯uid. Comp Biochem Physiol 1992;101A:553± 6. Service M, Wardlaw AC. Bacteridical activity of coelomic ¯uid of the sea urchin, Echinus esculentus, on di€erent marine bacteria. J Mar Biol Ass UK 1985;65:133±9. Gerardi G, Lassegues M, Canicatti C. Cellular distribution of sea urchin antibacterial activity. Biol Cell 1990;70:153±7. Service M, Wardlaw AC. Echinochrome-A as a bactericidal substance in the coelomic ¯uid of Echinus esculentus (L.). Comp Bioehcm Physiol 1984;79B:161±5. Bertheussen K. The cytotoxic reaction in allogeneic mixtures of echinoid phagocytes. Exp Cell Res 1979;120:373±81. Dales RP. Phagocyte interactions in echinoid and asteroid echinoderms. J Marine Biol Assoc UK 1992;72:473±82. Pagliara P, Canicatti C. Isolation of cytolytic granules from sea urchin amoebocytes. Eur J Cell Biol 1993;60:179±84. Ito T, Matsutani T, Mori K, Nomura T. Phagocytosis and hydrogen peroxide production by phagocytes of the sea urchin Strongylocentrotus nudus. Dev Comp Immunol 1992;16:287±94. Canicatti C. Lysosomal enzyme pattern in Holothuria polii coelomocytes. J Invert Pathol 1990;56:70±4. Canicatti C, Miglietta A. Arylsulphatase in echinoderm immunocompetent cells. Histochem J 1989;21:419±24. Smith LC, Chang L, Britten RJ, Davidson EH. Sea urchin genes expressed in activated coelomocytes are identi®ed by expressed sequence tags. Complement homologues and other putative immune response genes suggest immune system homology within the deuterostomes. J Immunol 1996;156:593±602. Leonard LA, Strandberg JD, Winkelstein JA. Complement-like activity in the sea star Asterias forbesi. Dev Comp Immunol 1990;14:9±30. Ryoyama K. Studies on the biological properties of coelomic ¯uid of sea urchin. I. Naturally occurring hemolysin in sea urchin. Biochim Biophys Acta 1973;320:157±65. Canicatti C, Parrinello N. Hemaglutinin and hemolysin level in coelomic ¯uid from Holothuria polii (Echinodermata) following sheep erythrocyte injection. Biol Bull 1985;168:175±82. Canicatti C. Binding properties of Paracentrotus lividus [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] (Echinoidea) hemolysin. Comp Biochem Physiol 1991;98A:463±8. Ryoyama K. Studies on the biological properties of coelomic ¯uid of sea urchin. II. Naturally occurring hemagglutinin in sea urchin. Biol Bull 1974;146:404±14. Kamiya H, Muramoto K, Goto R, Sakai M. Lectins in the hemolymph of a star®sh, Asterina pectinifera: puri®cation and characterization. Dev Comp Immunol 1992;16:243±50. Canicatti C, Pagliara P, Stabili L. Sea urcin coelomic ¯uid agglutinin mediates coelomocyte adhesion. Eur J Cell Biol 1992;58:291±5. Matsui T, Ozeki Y, Suzuki M, Hino A, Titani K. Puri®cation and characterization of two Ca2+-dependent lectins from coelomic plasma of sea cucumber, Stichopus japonicus. J Biochem 1994;116:1127±33. Giga YA, Ikai I, Takahashi K. The complete amino acid sequence of echinoidin, a lectin from the coelomic ¯uid of the sea urchin Anthocidaris crassispina. J Biol Chem 1987;262:6197±203. Ozeki Y, Matsui T, Titani K. Cell adhesive activity of two animal lectins through di€erent recognition mechanisms. Fed Eur Biochem Soc 1991;289:145±7. Hatakeyama T, Nagatomo H, Yamasaki N. Interaction of the hemolytic lectin CEL-III from the marine invertebrate Cucumaria echinata with the erythrocyte membrane. J Biol Chem 1995;270:3560±4. SoÈderhaÈll K. Prophenoloxidase activating system and melanizationÐa recognition mechanism of arthropods? A review. Dev Comp Immunol 1982;6:601±11. Smith VJ, SoÈderhaÈll K. A comparison of phenoloxidase activity in the blood of marine invertebrates. Dev Comp Immunol 1991;15:251±61. Roch P, Canicatti C, Sammarco S. Tetrameric structure of the active phenoloxidase evidenced in the coelomocytes of the echinoderm Holothuria tubulosa. Comp Biochem Physiol 1992;102B:349±55. Ashida M, Iwama R, Iwahana H, Yoshida H. Control and function of prophenoloxidase activating system. In: Payne CC, Burges HD, editors. Proceedings of the third international colloquium on invertebrate pathology. Brighton, UK: University of Sussex, 1982. p. A81± 86. SoÈderhaÈll K, Smith VJ. Prophenoloxidaes-activating cascade as a recognition and defense system in arthropods. In: Gupta AP, editor. Hemolytic and humoral immunity in arthropods. New York, NY: John Wiley, 1986. p. 251±85. Beck G, Habicht GS. Isolation and characterization of a primitive interleukin-1-like protein from and invertebrate, Asterias forbesi. PNAS 1986;83:7429±33. Burke RD, Watkins RF. Stimulation of star®sh coelomocytes by interleukin-1. Biochem Biophys Res Commun 1991;180:579±84. Prendergast RA, Suzuki M. Invertebrate protein stimulating mediators of delayed hypersensitivity. Nature 1970;227:277±9. P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 [55] Prendergast RA, Liu SH. Isolation and characterization of sea star factor. Scand J Immunol 1976;5:873±89. [56] Kerlin RL, Cerbra JJ, Weinstein PD, Prendergast RA. Sea star factor blocks development of T-dependent antibody secreting clones by preventing lymphokine secretion. Cell Immunol 1994;156:62±76. [57] Edds KT. Morphological and cytoskeletal transformation in sea urchin coelomocytes. In: Cohen WD, editor. Blood cells of marine invertebrates: experimental systems in cell biology and comparative physiology. New York, NY: A.R. Liss, 1985. p. 53±74. [58] Sohn RH, Goldschmidt-Clermont PJ. Pro®lin: at the crossroads of signal transduction and the actin cytoskeleton. Bioessays 1994;16:465±72. [59] Smith LC, Britten RJ, Davidson EH. SpCoel1, a sea urchin pro®lin gene expressed speci®cally in coelomocytes in response to injury. Mol Bio Cell 1992;3:403±14. [60] Smith LC, Britten RJ, Davidson EH. Lipopolysaccharide activates the sea urchin immune system. Dev Comp Immunol 1995;19:217±24. [61] Smith LC, Harrington MG, Britten RJ, Davidson EH. The sea urchin pro®lin gene is expressed in mesenchyme cells during gastrulation. Dev Biol 1994;164:463±74. [62] Armstrong PB, Quigley JP. Alpha 2-macroglobulin: an evolutionary conserved arm of the innate immune system. Dev Comp Immunol 1999;23(4±5):381±98. [63] Muta T, Miyata T, Misumi Y, Tokunaga F, Nakamura T, Toh Y, Idehara Y, Iwanaga S. Limulus factor C: an endotoxin-sensitive serine protease zymogen with a mosaic structure of complement-like, peidermal growth factor-like, and lectin-like domains. J Biol Chem 1991;266:6554±61. [64] Nonaka M, Matsuda Y, Shiroishi T, Moriwadi K, Nonaka M, Natsuume-Sakai S. Molecular cloning of the b subunit of mouse coagulation factor XIII and assignment of the gene to chromosome 1: close evolutionary relationship to complement factor H. Genomics 1993;15:535±42. [65] Leonard WJ, Depper JM, Kanehisa M, Kronke M, Pe€er NJ, Svetlik PB, Sullivan M, Greene WC. Structure of the human interleukin-2 receptor gene. Science 1985;230:633±9. [66] Kaplan G, Bertheussen K. The morphology of echinoid phagocytes and mouse peritoneal macrophages during phagocytosis in vitro. Scand J Immunol 1977;6:1289± 96. [67] Bertheussen K. Endocytosis by echinoid phagocytes in vitro, II. Mechanisms of endocytosis. Dev Comp Immunol 1981;5:557±64. [68] Bertheussen K. Receptors for complement on echinoid phagocytes. II. Puri®ed human complement mediates echinoid phagocytosis. Dev Comp Immunol 1982;6:635±42. [69] Bertheussen K, Seljelid R. Receptors for complement on echinoid phagocytes. I. The opsonic e€ect of vertebrate sera on echinoid phagocytosis. Dev Comp Immunol 1982;6:423±31. 441 [70] Bertheussen K. Complement-like activity in sea urchin coelomic ¯uid. Dev Comp Immunol 1983;7:21±31. [71] Parrinello N, Rindone D, Canicatti C. Naturally occurring hemagglutinins in the coelomic ¯uid of Holothruia polii delle Chiaje (Echinodermata). Dev Comp Immunol 1979;3:45±54. [72] Canicatti C. Membrane damage by coelomic ¯uid from Holothuria polii (Echinodermata). Experientia 1987;43:611±4. [73] Turner MW. The lectin pathway of complement activation. Res Immunol 1996;147:110±5. [74] Matsushita M, Endo Y, Nonaka M, Fujita T. Complement-related serine proteases in tunicates and vertebrates. Curr Opin Immunol 1998;10:29±35. [75] Al-Sharif WZ, Sunyer JO, Lambris JD, Smith LC. Sea urchin coelomocytes speci®cally express a homologue of complement component C3. J Immunol 1998;160:2983± 97. [76] Smith LC, Shih C-S, Dachenhausen SG. Coelomocytes speci®cally express SpBf, a homologue of factor B, the second component in the sea urchin complement system. J Immunol 1999;161:6784±9. [77] Horiuchi T, Kim S, Matsumoto M, Watanabe I, Fujita S, Volanakis JE. Human complement factor B: cDNA cloning, nucleotide sequencing, phenotype conversion by site-directed mutagenesis and expression. Mol Immunol 1993;30:1587±92. [78] Hourcade DE, Wagner LM, Oglesby TJ. Analysis of the short consensus repeats of human complement factor B by site-directed mutagenesis. J Biol Chem 1995;270:19716±22. [79] Xu Y, Volanakis JE. Contribution of the complement control protein modules of C2 in C4b binding assessed by analysis of C2/Bf chimeras. J Immunol 1997;158:5958±65. [80] Lachmann PJ. An evolutionary view of the complement system. Behring Inst Mitt 1979;63:25±37. [81] Lachmann PJ, Halbwachs L. The in¯uence of C3b inactivator (KAF) concentration on the ability of serum to support complement activation. Clin Exp Immunol 1975;21:109±14. [82] Sato T, Endo Y, Matsushita M, Fujita T. Molecular characterization of a novel serine protease involved in activation of the complement system by mannose-binding protein. Int Immunol 1994;6:665±9. [83] Ji X, Azumi K, Sasaki M, Nonaka M. Ancient origin of the complement lectin pathway revealed by molecular cloning of mannan binding protein-associated serine protease from a urochordate, the Japanese ascidian, Halocynthia roretzi. Proc Natl Acad Sci 1997;94:6340± 5. [84] Liszewski MK, Farries TC, Lublin DM, Rooney IA, Atkinson JP. Control of the Complement System. Adv Immunol 1996;61:201±83. [85] Bentley DR. Structural superfamilies of the complement system. Expl Clin Immunogenet 1988;5:69±80. [86] Campbell RD, Law SK, Reid KBM, Sim RB. 442 [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] P.S. Gross et al. / Developmental and Comparative Immunology 23 (1999) 429±442 Structure, organization, and regulation of the complement genes. Ann Rev Immunol 1988;6:161±95. Reid KMB, Campbell RD. Structure and organization of complement genes. Immunol Med 1993;20:89±125. Lambris JD, Mavroidis M, Sunyer JO. Phylogeny of third component of complement, C3. In: Erdei A, editor. New aspects of complement structure and function. Austin, Texas: R.G. Landes & Co, 1994. p. 15±34. Nonaka M, Azumki K. Opsonic complement system of the solitary ascidian, Halocynthia roretzi. Dev Comp Immunol 1999;23(4±5):431±7. Pancer Z, Gershon H, Rinkevich B. Cloning of a urochordate cDNA featuring mammalian short consensus repeats (SCR) of complement-control protein superfamily. Comp Biochem Physiol B 1995;111:625±32. Raftos D, Green P, Mahajan D, Nair S, Pearch S, Hutchinson A. An opsonic collectin-like protein from tunicates. Dev Comp Immunol 1997;21:89 [personal communication]. Hanley PJ, Hook JW, Raftos DA, Gooley AA, Trent R, Raison RL. Hag®sh humoral defense protein exhibits structural and functional homology with mammalian complement components. PNAS 1992;89:7910±4. Ishiguro H, Kobayashi K, Suzuki M, Titani K, Tomonaga S, Kurosawa Y. Isolation of a hag®sh gene that encodes a complement component. EMBO J 1992;11:829±37. Fugii T, Nakamura T, Tomonaga S. Component C3 of hag®sh complement has a unique structure: identi®cation of native C3 and its degradation products. Mol Immunol 1995;32:633±42. Fugii T, Nakamura T, Sekizawa A, Tomonaga S. Isolation and characterization of a protein from hag®sh serum that is homologous to the third component of the mammalian complement system. J Immunol 1992;149:117±23. Nonaka M, Takahashi K. Complete complementary DNA sequence of the third component of complement of lamprey. Implication for the evolution of thioester containing proteins. J Immunol 1992;148:3290±5. [97] Nonaka M, Takahashi M, Sasaki M. Molecular cloning of a lamprey homologue of the mammalian MHS class III gene, complement factor B. J Immunol 1994;152:2263±9. [98] Culbreath L, Smith SL, Obenauf SD. Alternative complement pathway activity in nurse shark serum. Amer Zool 1991;31:131A. [99] Smith SL. Nurse shark complement Ð in retrospect and prospect. Dev Comp Immunol 1997;21:144. [100] Ross GD, Jensen JA. The ®rst component (Cln) of the complement system of the nurse shark (Ginglymostoma cirratum ). 1. Hemolytic characteristics of partially puri®ed C1n. J Immunol 1973;110:175±82. [101] Jensen JA, Festa E, Smith DS, Cayer M. The complement system of the nurse shark: hemolytic and comparative characteristics. Science 1981;214:566±9. [102] Hyder-Smith S, Jensen JA. The second component (C2n) of the nurse shark complement system: puri®cation, physico-chemical characterization and functional comparison with guinea pig. Dev Comp Immunol 1986;10:191±206. [103] Dodds AW, Smith SL, Levine RP, Willis AC. Isolation and initial characterisation of complement components C3 and C4 of the nurse shark and the channel cat®sh. Dev Comp Immunol 1998;22:207±16. [104] Jensen RA, Fuller L, Iglesias E. The terminal components of the nurse shark C-system. J Immunol 1973;111:306±7. [105] Turner MW. Mannose-binding lectin: the pluripotent molecule of the innnate immune system. Immunol Today 1996;17:532±40. [106] Sato T, Endo Y, Matsushita M, Fujita T. Molecular characterization of a novel serine protease involved in activation of the complement system by mannose-binding protein. Int Immunol 1994;6:665±9. [107] Thiel S, Vorup-Jensen T, Stover CM, Schwaeble W, Laursen SB, Poulsen K, Willis AC, Eggleton P, Hansen S, Holmskov U, Reid KB, Jensenius JC. A second serine protease associated with mannan-binding lectin that activates complement. Nature 1997;386:506±10.