Scaml. J. Immunol. 35, 247-266, 1992


Invertebrate Immunity: Another Viewpoint Summary All vertebrates and invertebrates manifest self/non-self recognition. Any attempt to answer the question of adaptive significance of recognition must take into account the universality of receptor-mediated responses. These may take two forms: (I) rearranging, clonally distributed antigen-specific receptors that distinguish in the broadest sense between self and non-self, and non-self A from non-self B, latecomers on the evolutionary scene; (2) pattern recognition receptors, the earliest to evolve and still around, necessitating the requirement for induced second signals in T- and B-cell activation. Hither strategy need not force upon invertebrates the organization, struclure and adaptive functions of vertebrate immune systems. Thus, we can freely delve into the unique aspects ofthe primitive immune mechanisms of invertebrates. In contrast, using the opposite strategy which is still problematic, i.e. linking invertebrate and vertebrate defence, seems to give tis an approach to universality that might eventually reveal homologous kinship.

INTRODUCTION Invertebrates are exceedingly diverse. They include as many as two million species which belong to more than 20 phyla, from unicellular organisms up to the complex niulticellular protostomes and deuterostomes: thus it is not surprising to find diverse defence/immune responses whose effector mechanisms remain somewhat unexplored. This is in striking contrast to the mass of information on mammalian immune responses which has been derived essentially from the mouse, which belongs to one class of the phylum Vcrtebrata (clearly the reductionist or single laxon approach). The essential framework of immunology, the overwhelming burst of results since the 1960s, i.e. its theory, concept, technology, hasemanaicd primarily from this single animal (exclusive of the avian thymus and bursa of Fabricius paradigm). By both modern and classic analyses we suggest that certain immuno/dcfcncc responses (I) may be analogous in invertebrates and mice, revealing the general principle that immuno/defencc is a strategy that evolved to ensure survival after attacks by micro-organisms; (2) may be unique to invertebrates; (3) could be antecedents of vertebrate immunity; (4) are solely those ascribed to mice and men. Without question, the itnmune system is universal but the extent of interrelationships and the evolutionary origin is poorly understood. This holds not only for the protostome/deuterostomc lines, one means of classifying invertebrates, but relationships are even more debatable when the two great divisions Invertebrata/Vertebrata are considered. The immuno/defenee responses emphasized in this review are derived from protostomes and deuterostomes (perhaps an unfortunate classification with respect to immunity, since it is based on the embryologic formation ofthe mouth). Available evidence from results of immunodefcncc experiments suggests more kinship between tbe two groups than differences as the ontogcnetic scheme purports; erection of these two major groups may be irrelevant in the future with respect to immuno/defenee responses. We do not propose close kinship joining the invertebrates with vertebrates. This is problematic since invertebrates do not possess variable-region molecules (VRM) (e.g. Igs and TCRs). Development of these molecules was probably a crucial event in the evolution of jSrimordial vertebrates [1]. Their invertebrate precursors undoubtedly had a more than adequate defence system: witness the numerous extant species. Retaining this ancestral imtnunodcfcncc system (non-specific, innate, natural), the vertebrate line built upon it to evolve the modern immune system (specific, induced, adaptive). To ensure the perpetuation of the latter in the lace of countless numbers of antigens associated with rapidly mutating micro-organisms, this new VRM system was redeployed to trigger pre-existing effector 247



mechanisms of defence. Finally, there was probably still a further development, i.e. immunoglobulins, and antigen presentation by T-cel! receptors in relation to class I and class II MHC components. Although we may ponder the proposed function of a primordial VRM system, it is essential to define the animal group in which it may exist, perhaps a protochordate (despite the enormous distance, yet persistently proposed similarity based upon embryology), and how much of the system qualifies to be a true antecedent (Fig. I). Where have we gone sinee Metehinkoff's pioneer discovery [2]? In some respects perhaps phagocytosis

FIG. I. Hypothetical scheme explaining (he possible relationships and development of the primitive effector (defence system) and thai of a system characterized by [he presence of variable region molecules (VRM). Stage 1 indicates the presence of all components' effector systems as well as those of the VRM (immunoglobulin (Ig) and T-cell recepiors (TCR)|. The organism (progenitor) is ill-deftned (by dashed lines) and Ihe components are disorderly. Stage 2 presents the emergence of defined EFFECTOR SYSTFMS in a definite animal model (solid line) but still associated wilh an unknown animal model which begins to show order in the expression (spelling) of VRM components. Stage 3 reveals a definite organism ihai has retained ancient effector systems but now expresses also MHC and the two components of VRM (see also Fig. 2 which accounts for recognition in organisms thai possess ancient cflector systems (paitcrn recognition receptors) and ihose with clonally distributed recepiors (rearranging receptor genes).



as studied ever so intetiscly in invertebrates may still not be as illuminating as other aspects of invertebrate immunity. We must therefore await more results from novel experiments that may reveal unique features. Phagocytosis is an aneestral funetion which links a defensive mechanism with a means of obtaining nutrition in protozoans, but il ptays an ancillary role in modern functions in non-coelomate metazoans. In the more complex speeies, Ihe coclomatcs, techniques that include in vitro, bioehcmicat, immunochcmical and serological analyses have been especially advantageous in analysing invertebrate responses exclusive of phagoeytosis. As examples, the particular focus of this review will include the following: {1) cytotoxieity: (2) allogeneic/xenogencic recognition; (3) genetics of recognition in tunicatcs; (4) inducible antimicrobial peptides; (5) the prophcnoloxidase (proPo) activating system; (6) cytokincs; and (7) precursors of the Ig super family.

WHAT HAS S T R U C T U R E VERSUS F U N C T I O N REVEALED WITH RESPECT TO I M M U N E EVOLUTION? Homotogy. analogy, convergence, divergence, adaptive radiation We acknowledge the ferment, albeit limited (due to societal demands), in an interest in comparative immunology with particular reference to invertebrates. It is essential in our attempt to trace possible origins of vertebrate responses to reveal those genealogical relationships which have been conserved, i.e. structure and not function. Function may change discontinuously {see punctuation) with certain structures immediately taking on novel functions. Function has been dea!l with in an attempt to defme homology/analogy and convergence/divergence to account for functional invertebrate leucocytes [3-5]Whether we search for structure and developmental origins of coelomocytcs in protostomes seems unimportant for the moment since the available information is limited. However, there is evidence of the persistence ofan invertebrate structure that reflects morphological kinship to that of extinet/extant vertebrates. Here, we would like to disagree with Burnet [6-8] (his insistence that invertebrate immunologists may force invertebrate immuno/defence responses into vertebrate patterns), and divert attention momentarily in ihe face of new evidence. Using deuterostomes seems the likely starting point. Given unlimited leciinology for revealing conserved structure at all levels, e.g. cell, organ, but limited evidence, there are glimmers of what may be conserved structures [9]. Are there instances of, for example, a pharyngeal derivative among invertebrates that could be related lo a vertebrate thymus whose cells bear some structural resemblances which may possess the capacity to divide, as thymic cells are capable of doing in situ or in vitro, after stimulation by antigen? The pharyngeal region of the solitary tunicate and that of vertebrates seems a likely candidate. Cells from this region resemble lymphocytes and proliferate in response to allogeneic stimuli [9-16]. In those presenl-day relatives of extinct forms, the path ofan ancestral "immuno-stem cell", during evolution, could have taken several routes. This calls to mind three concepts: adaptive radiation, convergence and divergence. According to adaptive radiation, because of the constant competition for food and living space, each group of organisms tends to spread out and occupy as many different environmental niches as possible. With respect to the evolution of food getting and immunodefence, this principle is clear in protozoans and sponges [5]. Adaptive radiation assumes a process of evolution from a single aneestral species giving rise to several forms which occupy slightly different habitats. Such a situation is advantageous in evolution since it enables organisms and their inherent immunodefence eells to tap new sources of food or to escape from potential pathogenic micro-organisms by developing new strategies of immunodefence. Although present-day placenlal mammals are excellent examples of the benefits of adaptive radiation, with many groups occupying diverse environmental niches (e.g. terrestrial, aquatic), our information on mammalian immune systems has been limited mostly lo the laboratory mouse. The essence of adaptive radiation, in the immunological context, assumes the evolution from a single. ancestral form (immunodefence stem cell) of a variety of different immunocytes, eaeh of which is adapted and speeialized in some unique way to survive in a particular habitat. This is evident in the multitude of leucoeytic types which possess variable functions as those which we observe in invertebrates. Adaptive radiation which gives rise to several different types of descendants adapted by



various means to different environments may be termed divergent evolution. The opposite, convergent evolution, also occurs fairly frequently, i.e. two or more quite unrelated groups may, in becoming adapted to a similar environment, develop characteristics that are more or less similar. For example, wings have evolved not only in mammals, but in birds and in insects; so be it in immunocytes. In immunological terms this may translate into the evolution of haetnopoietic organs in mammals, birds and insects. This latter poinl of course calls to mind the related concepts of homologv and analogy. With respect to structure, homology refers to similarity in structure, in relationship to adjacent structures, in embryonic development and in nerve and blood supply. A seal's front flipper, a bat's wing, a cat's paw, a horse's front leg, etc. aresuperficially dissimilar and adapted for quite different functions. Slill they reflect a common genetic endowment and evolutionary relationship. The thymus seems an appropriate example as it appears in the various vertebrate elasses [17] and its possible pharyngeal antecedent in tunicates wbose cells are capable of replication both with and without antigenic stimulation [9] may be a precursor. In contrast, analogous structures are simply superficially similar serving a comparable function but having quite dilferent basic structures and developmental patterns. The presence of analogous structures does not imply an evolutionary relationship between the animals bearing them. For example, the wings of birds and the wings of butterflies are simply analogous; both enable their possessors to fly, but they have no developmental processes in common. Analogy holds true for functions of leucocytes in melozoans. To sequester infectious micro-organisms is the raison-d'etre, a eommon strategy developed for preservation of the species which need not require the same roots. With respect to molecular structure of cellular constituents clearly we can apply these terms at that level. For example, the haemoglobins of different animals, the cytochrome c's of vertebrates or the lactic dehydrogenases in birds and mammals are homologous proteins. The haemoglobins in different species have similar sequences of amino acids which reflect a common genetic pattern and evolutionary relationship. In contrast, haemoglobin and haemocyanin may be termed analogous molecules since they have similar functions (oxygen transport) but quite different molecular structures. With respect to immunology, sequence analyses have revealed a family of molecules (the Ig superfamily) that seem to reflect examples of both analogy and homology. For all of the members, the basic funetion is recognition. Those thai recognize non-self (TCR, Ig) are homologous but do they depart and become analogous to those whieh recognize self (?) [see Ig superfamily]. It is appropriate to mention an important component of evolutionary trees, i.e. how do we interpret the rate of morphological change in an evolving lineage? Despite the diversity among organisms, there is a measure of order and if imtnunology would beeome broader in its approach and not restricted essentially to one species we could observe these variations. For example, we know that there are two major lymphocyte types—T and B—first discovered by microscopy and later readily identifled as such by appropriate reagents. We could also identify members of the granulocyte family as neutrophils, eosinophils and basophils, and it is unlikely that we would incorrectly identify any member of the granulocyte family as a type oflymphocyte. What we are explaining is this; morphological gaps may separate even these closely related cell types, just as increasingly larger 'gaps' separate other mesodermally derived cells, e.g. those of the kidney, muscle and other types. There are rates which may explain morphological change, in an evolving lineage—let's say a lymphocyte. Gradualism, punctuation In a traditional model referred to as gradualism, most morphological change occurs within species as a result of genetic drift, directional selection and other processes that may affect the change in allele frequency. In an alternative model referred to as punctuation., we assume that most morphological change occurs rapidly during separation. The puncluational model is consistent with the observation that there are few intermediate forms between closely related species. With respect to the fossil record we have no information concerning immune responses except for what has been discovered by examining various living species (assumed lo be relatives of extinct species), the work of comparative immunologists. Perhaps a better approach would be to examine fewer species and proceed in depth, experimenting with at least two species. The available evidence strongly suggests that there is little difl'erence between certain observations, e.g. agglutinin production or graft rejection among many



invertebrates as presented in this review. To persist in an examination ofmany species may trap us and result in confusion as to what is being sought after coniparalivc immunology versus evolutionary immunology —especially if we view the two in the broadest context us a model, as is done for the great parent field, biology. Evolutionary versus comparative immunology According to Wake (1990), there is a distinction between evolutionary biology ;ind comparative biology (qua immunology) [18], Evolutionary biology operates in the historical context, and may utilize data from a diversity of sources (morphology, biochemistry, immunology, physiology, ecology, behaviour) in order to determine the history of lineages and mechanisms of change. Comparative biology, according to Ridley (1985), embraces design and history, or adaptation and phylogeny to explain the diversity of life [19], Enthusiastically supported by Rieppel (1987). comparative biology has been renewed because of new experimental techniques in functional morphology and the rebirth in systematic and evolutionary biology [20], Bartholomew (1987), who stresses comparisons between a minimum of two taxa [21], views comparisons as essential in studies of diversity to establish relations between and among phenomena to escape the 'chaos of unordered detail' (perhaps a typical example among invertebrate immunologists is reflected in the present situation with respect to analyses of humoral immune responses, especially the search lor and characterizalion ofagglutinins [22[). Finally, Wake in her admonitions warns against stating that the conclusions of good comparative biology indicate patterns of evolution without presenting a phylogcnctic context. When the comparisons do not allow the delineations of primitive and derived states, and thereby evolutionary trends, they are still useful as analogies.

Self and non-self revisited in relation to recognition Investigations of defence systems in invertebrates have led to the conclusion that there are no immunoglobulins nor T-cell receptors. There arc proteins, related to the immunoglobulin superfamily molecules which mediate the discrimination between self and non-self [23]. Experimental immunologists may take two courses. One may stress the more practical, oriented towards the response to natural disease-causing micro-organisms, a tact which largely girded the field in the historical context. And then there came the ferment ofthe 1960s, which was more theoretical. It embodied Ihe use ofany antigen (non-self) to challenge the immune system, revealing the immense plasticity and capabilities. This heralded the T and B system, the receptor concept, networks, and the 'assistance league' of specific functions and interactions associated with the less than central but essential actors in the immune response scenario (macrophages, granulocytes). Invertebrate responses need lo catch up with this approach and, at the same time, define unique responses, those which arc analogous and which reveal certain concepts: cooperation, suppression, tolerance. Recognition seems to be the key point, whether immune or not; nevertheless, a recent analysis seems a bit reductional in at least two points. A first point is purely semantic. The term 'anticipatory' (and its contrary non-:inticipatory) seems inappropriate, especially since it compounds the already existing difficulties and tedium in defining the terminology that cloaks immune responses [24], Second, this term has, to our knowledge, not appeared in the literature, which could cloud still further the bewildering assortment of jargon. Invertebrates can 'anlicipate' a contact with pathogens and they can respond to such a contact by spontaneously recognizing the so-called heterophilic antigens [25. 26] or chemical radicals oflen present at the surface of bacteria such as LPS orzymosan. Asa probe which supports the anticipatory character of this response, it has often been shown that a first contact with a pathogen induces a stimulation ofthe invertebrate defence system. Its function may not be entirely a defence task, bul it is trephocytic or symbiotic and may also play a role in morphogenesis [27]. The main diflcrence between anticipatory and non-anticipatory responses is how specific the recognizing 'paratopes' are: whereas in invertebrates this is a 'broad-spectrum" recognizing system, which means that relatively simple molecules like terminal sugar units build up the 'epitope'. in vertebrates the recognized area is extremely individualized and is a highly specific epitope. The same



holds for the recognizing systems: invertebrates possess simple lectin-like recognition sites, soluble or acquired, found on membranes without a T-ccil-like recognition system, whereas vertebrates, by contrast, have a highly specialized recognition system against non-self components including those which are infectious.

CYTOTOXICITY Antibody is the vertebrate's invention. However, for various reasons the cytotoxic cell immunity with limited TCR-MHC repertoire could have developed at an early stage of invertebrate evolution since it is still found in some invertebrates [I, 28-31], Williams (1987) has proposed that the Ig superfamily probably evolved from a heterophilic recognition system which underwent extensive diversification [32]. This recognition system probably developed into an immune system derived from a cytotoxicity system which involved programmed cell death. Since natural cell cytotoxicity may be gaining favour as a possible link between invertebrate and vertebrate cell-mediated responses, this view seems not farfetched [31]. With the demonstration of graft rejection in invertebrates, a response accompanied by specificity and memory, the idea that the TCR a/^-MHC co-evolution may have begun in certain invertebrates becomes an interesting idea worthy of more intense scrutiny. In a more expanded form, there are two explanations with respect to the evolution of effector mechanisms and the non-clonal origins ofthe immune system as presented by Janeway [33]. The effector mechanism may have existed before specific antigen receptors evolved, but they were added to preexisting effector mechanisms; the result was to enhance the precision and broaden the scope of recognition that could trigger responses. Succumbing to teleology, there was the need to evolve the clonal distribution of receptors which could account for the'recall' phenomenon in the immune system, i.e. anamnesis or memory (Fig. 2). Assuming that there is a lack ofthc rearranging receptor gene families, which could permit clonal selection, one question can be quickly posed: how was, and even now how is, effector function regulated in primitive organisms? According to Janeway, 'primitive effector cells bear receptors that allow recognition of certain pathogen-associated molecular patterns that are not found in the host.' He proposes the term pattern recof^nition receptors [33]. There are two interpretations with respect to the mechanisms of how pattern recognition receptors might operate. Although Janeway assures that pattern recognition receptors arc still an important component ofthe vertebrate immune system, they are not clonally distributed on lymphocytes and do not therefore confer on them the 'recall' or immune memory response, generally accepted as "the hallmark ofthe specific immune response' [33], According to his proposal, Janeway suggests three types of pattern recognition receptors which serve to activate effector functions of primitive immune systems before the evolution of rearranging gene families: (1) uncharactcrized receptors which allow NK cells to distinguish between target cells: (2) the antigen-presenting ceil receptors acted upon by adjuvants to induce second signals; (3) several T-cell surfaee molecules of unknown specificity that can induce T-cell activation upon cross-linking. Cytotoxieity as observed in vitro constitutes a more refined model than experimentally produced transplants (that are rejected by cytotoxic eells) to test the extent of reeognitive potentialities of invertebrate leucocytes. This would extend the analyses of host mechanisms devoted almost exclusively to graft rejection responses to those against parasites [34] and neopiasia [34 36]. Valembois et al. [27] have examined the cells involved in cytotoxic reactions in sipunculids by using both micro.scopic and biochemical techniques, revealing at least two categories. One leucocyte type is responsible for specific recognition and it. in turn, induces a second leucocyte to kill target cells in a non-specific manner. Only the killer eells have been observed by microscopy [38]; they are small hyalocytes or granulocytes with few granules and numerous ribosomes. As seen by both electron microscopy and time-lapse microcinematography. contact between killer cells and target cells is essential before eytolysis can occur [39]. Other characteristics of the response include its inhibition by EGTA. cytochalasin B, vinblastinc. colchicine and various drugs which are inhibitors of polypeptide synthesis; the effector molecule is a protein. Investigations of enzyme activity assume that a phospholipase derived from leucocyte killers provokes the lysis of target cell membranes [25].



Non-clonal pattern receptors Clonal, rearranging receptors

4Non-Clonal Response to Pattern


Non-Clonal Response to Pattern


Specific, Clonal Response to Ligand "^ + Co-Stimulator FIG. 2. The development of the immune system. Panel A illustnites Ihe proposed nature of lymphocyles in pre-vertebrates. Such cells have several different pattern recognition receptors distributed non-cionally. All such cells can respond to a particular pattern. leading to a protective immune response to a micro-organism expressing that pattern. Panels B and C illustrate the proposed naiure of the vertebrale immune system, m which clonally distributed receptors deriving from rearranging receptor genes have been added to cells that retain the pattern recognition receptors of their pre-vertebrate ancestors. In panel B, all such cells retain the abilily to respond to a particular pattern, leading lo protective immunity against microbes expressing that pattern. This type of response does not lead to clonal expansion and. thus, cannot lead to immunological tnemory. In panel C, the response to a specific ligand. recognized by a clonally distributed receptor encoded in rearranging receptor genes, generates specific immunity, clonal expansion, and immunological memory, provided the appropriate costimulatory activity is also present. Both types of response may be important in protective immunity. {From Ref, 33.) in earthworms, large leucocytes are not involved in in vitro cytotoxic reactions. After harvesting coelomocytes and filtering them on a nylon wool column, chloragocytes and large macrophage-like cells adhered to nylon fibres. The only active cells which passed through the columns were small, nonadhcrcnt leucocytes which were cytotoxic [25], More recently. Suzuki and Cooper have shown that coelomocytes of the earthworms Eisenia fetida :tnd Lumbricu.s terrestris demonstrated significant spontaneous cytotoxicity against allogeneic and xenogcneic target cells (manuscript in preparation). In a 24-h trypan blue assay, allogeneic killing using Lumbricus ceWs was significant aia P value of 0.04, but



using Eisenia, another species, the level of killing is greater (/*< 0.0002), Cytotoxicity between the two earthworm species (i.e. xenogcneic) was significantly higher (/'

Invertebrate immunity: another viewpoint.

All vertebrates and invertebrates manifest self/non-self recognition. Any attempt to answer the question of adaptive significance of recognition must ...
7MB Sizes 0 Downloads 0 Views