Fish Immunoglobulin - A Sero-Diagnostician’s Perspective

Immunoglobulin M, the predominant isotype in fish, is generally of relatively low specificity. This major drawback for sero-diagnosis may be overcome by competitive Elisa techniques. Competing mammalian or avian antibodies of high affinity improve the resolution between specific and cross-reacting fish IgM. for Aquatic Animal Diseases are all direct (agent identification). Nonetheless some serological methods for (indirectly) diagnosing certain infections may soon be validated, rendering the use of serology for diagnostic purposes more widely acceptable (OIE, 1997). Present limitations are due not only to lack of experience but also to discrete features of the humoral immune response in fish. All land vertebrates have several kinds of Ig with different forms and functions (Turner and Owen, 1994), but findings suggest fish have only one kind, though further classes and subclasses are still being discussed (Wilson and Warr, 1992; Manning, 1994). The main plasma immunoglobulin in fish is a molecule generally called IgM, whose form with a high molecular weight resembles mammalian IgM in being a complex polymeric molecule (Wilson and Warr, 1992); but unfortunately, due to its inherent low specificity, IgM is the kind of immunoglobulin least appreciated in sero-diagnostics. Introduction The Manual of Standards for Diagnostic Tests of the OIE (Office International des Epizooties) includes prescribed and alternative tests for diseases of mammals, birds and bees in international trade. The agents of few of these diseases have to be identified. In most cases they can be screened or confirmed indirectly, mostly by detecting the humoral immune response of the host to the invading agent i.e. the specific immunoglobulins in the serum. Serological diagnosis ranges from simple qualitative slide agglutination to advanced quantitative ELISA-techniques. In recent years, the health screening of fish by detecting specific antibodies against fish pathogens has captured increasing interest, due partly to national and international legislation or policies on the monitoring of fish health (Dixon et al., 1994), but serological methods have not yet evolved to the point of providing a standard procedure. The methods presented in the OIE’s Diagnostic Manual Review Bull. Eur. Ass. Fish Pathol., 20(2) 2000, 61 Implications of the polymeric structure of IgM Affinity versus avidity ‘Affinity’ denotes the strength of binding between a monovalent hapten or a single determinant of an antigen and an antibody binding site. Actually, monomeric Igs have two binding sites, i.e. are divalent, so the multivalence of polymeric Igs can be deduced theoretically by multiplying the number of monomeric subunits by two. Practically, at least with larger antigens, the multivalence is somewhat reduced by steric hindrance. In distinction to affinity, the overall binding energy of an antibody is called its avidity. In the case of antigens with several epitopes, this is not just the sum, but the product, of the binding site affinities (‘bonus-effect’ of multivalence, Roitt, 1993). Thus, a polymeric IgM molecule of low affinity has high avidity and binds strongly to complex antigens with several, repetitive epitopes. This serves well as the defence of a lower vertebrate (e.g. fish) or as the first response of a higher vertebrate (see figure 1), which is phylogenetically or presently not yet able to produce antibodies of higher affinity. But this hampers serology since it favours cross-reactions. The low quality of individual fixation of (potentially also inadequate) epitopes, via the single bindingsites of Igs, is compensated by the quantity of bindings, causing the specificity of IgM to be lower than that of Igs with low molecular weight. Thus, a monovalent Ig of high affinity, whose one site is able to bind firmly, would be the most specific antibody but could not cross-link antigens and form precipitates or agglutinates. Monomeric, divalent IgG (mammals)/IgY (avians) is a compromise of evolution. Especially in agglutination methods used in mammalian and avian serology, attempts are made to destroy or reduce the activity of IgM (by heat, low pH, 2-mercaptoethanol or precipitation with rivanol) before testing the sera. Otherwise, due to the high agglutination potency of IgM, the impact of its potential crossreactivity would be over-proportional. Specificity of fish Ig Clearly, the affinity and diversity of antibodies in fish, as in all ectothermic vertebrates, is typically much lower (not only with respect to Ig classes) than in higher vertebrates (Warr, 1995). In teleosts and sharks there is no switch (of antibody producing B – lymphocytes) to low molecular weight Ig classes during an immune response which in mammals (and birds) is typically accompanied by the emergence of high affinity binding antibodies (Wilson and Warr, 1992, see also figure 1). Furthermore, in higher vertebrates, a single antigenic determinant can stimulate the immune system to produce several hundreds of different antibodies within one Ig class, but in fish, there is little variety of antibodies produced in response to a single defined hapten (Du Pasquier, 1982). Also, in poikilotherm vertebrates there is no affinity maturation of immunoglobulins like that in mammals. During an ongoing immune response in mammals the affinity for the particular antigen is sharpened in the B-cell blasts by somatic hypermutation before the cells differentiate to antibody-producing plasma cells (Pilström and Bengtén, 1996). At the same time a drop in antigen concentration increasingly favours the proliferation of highly affine B-cells (Roitt, 1993). The lack of affinity maturation in fish, despite presence of several (genetic) mechaBull. Eur. Ass. Fish Pathol., 20(2) 2000, 62 nisms to generate variability including somatic mutations (Pilström and Bengtén, 1996), may be due to the lack of suitable sites for clonal selection like germinal centres in the spleen of mammals (Wilson et al., 1992). Consequences for sero-diagnosis in fish A bias towards macroglobulins in agglutination assays is of little importance if mainly polymeric Igs are present, and since IgM has an excellent agglutinating potency, agglutination methods are fine if a moderate sensitivity is tolerable, i.e. for orienting titre monitoring after experimentally immunising fish. These methods are easy to carry out for the particulate antigen (Denzin, 1996). Soluble antigen can be coated on sheep erythrocytes for passive haemagglutination (principle: Hudson and Hay, 1976). As regards cross-reactivity: provided there is no infection eliciting cross-reacting antibodies in parallel to the immunisation regime, such titre monitoring is internally blanked for each individual, since titre changes are related to initial serum reactivity and, as a control, non-immunised animals are sampled at the same time. For diagnosis, more sensitive techniques are better. In complement fixation test, the reaction between antibodies and antigens can be made visible without depending on relatively large, visible agglutinates, but set-up of the test may be complicated for fish. Commonly used guinea-pig complement and rabbit haemolysin might have to be replaced by homologous components, since an incompatibility between fish and mammalian antibodies and complement has often been reported (Sakai, 1981; Saha et al., 1993). Generally, all more sensitive serological tests will reveal cross-reactivity, so a cut-off value has to be determined from the frequency distributions of concentrations of reacting antibodies in the sera of positive and negative controls. The lower the specificity of the Igs involved, the more the frequency distributions overlap, defining a wider cut-off range (see fig. 2). The cut-off value can be calculated (e.g. the mean activity in the negative control group plus twice the standard deviation, with a 2.5% risk (p = 0.025) of negative sera being falsely judged positive), but the value has to be chosen from the cut-off range according to the specific aims of sero-diagnosis as regards specificity and sensitivity. As seen from figure 2, the specificity and sensitivity of a test system are correlated inversely (Martin et al., 1987). Thus, a lack in specificity in a test speFig.2: Determination of cut-off values Fig.1: Humoral immune response in mammals with respect to antibody isotypes 0 1 2 3 4 5 6 7 IgG