Interest of immunomodulation as a mean to improve the preparation of polyclonal and monoclonal antibody reagents.

Immunomodulation by monoclonal antibodies (mAbs) was investigated in mice in order to improve the preparation of antibody reagents. Three different types of representative immunogens were chosen: a human soluble protein (secretory immunoglobulin A, SIgA), a bacterial polysaccharide from E. coli K1 and an envelope protein from the hepatitis B virus. These Ag are all of importance for diagnosis and exhibit different levels of immunogenicity. Antibody-mediated enhancement was observed against restricted and defined regions of each immunogen i.e.: the Fab epitopes of SIgA, the preS1 domain of the HBV envelope and associated cell wall components of the capsular PS. The epitopes which were enhanced appeared to be different from those recognized by the modulating mAb. Negative modulations were also observed. Moreover, new epitopes seemed to be generated. In both cases the level and direction of the modulation were irrespective of isotypy and affinity of the mAbs. Interestingly the positive modulatory effect was found to be correlated with an in vitro assay based on the binding of immune complex to antigen-presenting cells.

Human secretory component. II. Easy detection of abnormal amounts of combined secretory component in human sera.

A simple method, allowing easy detection of abnormally increased sIgA levels is described. It consists in quantitation of combined SC by gel double diffusion, using appropriate anti-SC immune sera. The technical conditions, locating the threshold of sensitivity of precipitation at about 25 microgram/ml, a value higher than that found in normal sera, were established. Comparison with other classical methods (SRID, ELISA and IHA) emphasizes the validity and simplicity of the technique which has shown convenient whenever a large number of sera have to be tested.

An ELISA method to measure total and specific human secretory IgA subclasses based on selective degradation by IgA1-protease.

We have taken advantage of the property of IgA1-proteases to selectively cleave the human IgA1 subclass into Fabalpha and Fcalpha-J chain-secretory component (Fcalpha-J-SC) fragments in order to design a novel ELISA method for measuring the two secretory IgA (S-IgA) subclasses in secretions. The assay is based on the loss of detection of S-IgA1 by a combination of peroxidase-labelled antibodies to secretory component and Fab following IgA1-protease treatment. The specificity is that of the protease and the sensitivity of the detection is 5 ng/ml. Moreover, the use of purified S-IgA1 and S-IgA2 controls is not necessary. The assay has been successfully applied to the analysis of colostral S-IgA antibodies (Abs) to HIV-1-gp160 from HIV-1 positive women. The major subclass of colostral S-IgA antibodies to gp160 was found to be of the alpha1 isotype but the specific activity of anti-HIV-gp160 S-IgA2 was, however, higher than that of S-IgA1.

Differences between the in vitro combinations of secretory component (SC) and immunoglobulin polymers (Ig) by enumeration of SC epitopes.

The enumeration of the SC epitopes has been established on 125I-labelled free and combined SC, by binding to anti-SC coated beads, then by addition of 3H-labelled anti-SC Fab' fragments of various specificities. The number of moles of Fab' fragments found on the beads increases in relation to the introduced amount. The extrapolation to an infinite concentration of added Fab' fragments gives the maximal theoretical accessible number of SC epitopes. The number of hidden epitopes (cryptotopes) is established by subtracting the total number found on sIgA, IgA-SC and IgM-SC from those found for free SC. These values are confirmed with Fab' fragments specific for the inaccessible determinant of SC. There are 4 cryptotopes in the case of sIgA, 3 for IgA1-SC, 2 for dimer IgA-SC and only 1 for IgM-SC (polyclonal or monoclonal). Thus the in vitro combinations of SC with polyclonal IgA dimers are different from the in vitro combinations with polyclonal or monoclonal IgM. The structural implications of these differences are discussed.

Enumeration of the antigenic determinants of human secretory component (SC) and secretory IgA (sIgA).

The number of antigenic determinants of human SC and sIgA was determined with specific anti-SC antibodies isolated from a pool of two hyperimmunized sheep. The Fab' antibody fragments were prepared and [125I] labelled, while pure SC and pure sIgA, isolated from colostrum, were labelled with 131I. The [125I] Fab' antibodies were added, in very large molar excess, to the [131I] antigens at various ratios. The Fab'-Ag complexes were separated from the antibody excess by gel-filtration. The highest Fab'/Ag molar ratios of the complexes, which correspond to the maximal number of accessible antigenic determinants, were calculated. We found at least 16 sites (16.6 +/- 1.36) on free SC while only 12 (12.27 +/- 0.51) were located on sIgA. These results confirm the existence of a significant number of hidden determinants of the SC subunit of sIgA and establish that about 1/4 of the SC molecule is implied in this binding.

Human secretory component. IV: Antigenic regions involved in in vitro binding to dimeric IgA.

The aim of this report was to identify the region(s) of the secretory component (SC) molecule involved in in vitro binding to dimeric IgA. Inhibition of the SC binding was tested by Fab' antibody fragments directed against the accessible (A) and inaccessible (I) regions of SC. Antibodies directed against the main 38.5-kDa trypsin fragment of SC, and antibodies from two immune sera with a wide anti-SC spectrum, were also used. The specificity and activity of the five antibody preparations were established by double-diffusion in gel and by their SIgA combining capacity. Inhibition curves were established by RIA using constant amounts of 125I SC, dimeric IgA and increasing quantities of the various Fab' antibodies. These results indicated involvement of a larger part than the I region of the SC molecule in combination with dimeric IgA, perhaps including a second (minor?) site of binding on the accessible parts in addition to the major region located on I.

Both Fc alpha domains of human IgA are involved in in vitro interaction between secretory component and dimeric IgA.

The sites of interaction between 125I labelled human secretory component (SC) and dimeric IgA were located by studying the inhibitory effect of various antibodies to IgA. Several Fab' fragments were isolated from three sera of hyperimmunized rabbits. The specificity of these different antibody preparations, as determined by a RIA inhibition test or by ELISA, showed that two were directed against both domains of Fc alpha, two against C alpha 2, two against C alpha 3 and one against Fd alpha. A monoclonal antibody against C alpha 3 was also used. The results indicate that both the C alpha 2 and C alpha 3 domains are equally and independently involved in the interaction between SC and dimeric IgA.

IgM reassociation in the absence of J-chain.

Human monoclonal IgM pentamers with different biophysical properties (euglobulin, pseudoglobulin or cryoglobulin) were reduced and reassociated in the absence of J-chain. Reassembly occurred for 50-82% of the monomers. The reassociated molecules consisted of covalent oligomers and pentamers. The deficiency of J-chain (estimated to be less than 0.17% of normal) was shown by alkaline-urea overloaded gel electrophoresis followed by silver staining. The addition of exogenous J-chain, from polymeric IgA or IgM, did not significantly modify the reassembly ratio. Thus J-chain does not seem to be an absolute requirement for IgM polymerization.

High-level ability of secretory IgA to block HIV type 1 transcytosis: contrasting secretory IgA and IgG responses to glycoprotein 160.

The IgG and secretory IgA (S-IgA) responses to the HIV-1 envelope (gp160 antigen) were analyzed in the colostrum (Col) and in the cervicovaginal fluid (CVF) of HIV-l-infected women. We show IgG antibodies (Abs) to the recombinant gp160 to be predominant as compared with the corresponding S-IgA isotype. The low level of the S-IgA response cannot be related to a general disturbance of the mucosal-associated Iymphoid tissue (MALT) because the level of a current Ab to a caries-associated antigen from Streptococcus sobrinus was in the normal range in these secretions. The major subclass of IgA to gp160 was of the alpha1 isotype both in Col and in CVF. However, the specific activities of S-IgA1 and S-IgA2 were different when expressed as the ratio of the anti-gp160 related to total Ig of each subclass. Indeed, the specific activity of the S-IgA2 was predominant over S-IgA1 in the Col, whereas the reciprocal results were found in CVF, showing a subcompartmentalization of these secretions. The ability of S-IgA and IgG to block one of the pathways involved in the HIV-1 penetration across mucosa, i.e., transcytosis through epithelial cells, was evaluated using a functional in vitro assay. Both S-IgA and IgG Abs impaired virus transcytosis, irrespective of the level of antigp160 specific activities. However, specific S-IgA was more efficient than IgG. These features suggest that mucosal specific S-IgA to HIV-1 could be relevant in decreasing infectivity of HIV-1 in corporal fluids.

Molecular analysis of the modulatory factors of the response to HBsAg in mice as an approach to HBV vaccine enhancement.

Mice were injected with immune complexes containing the recombinant hepatitis B surface antigen (HBsAg) vaccine (S + preS2) bound to different monoclonal antibodies (mAbs), in order to determine whether an enhancement of the response to a human vaccine could be obtained and observed. Enhancement and indifference were observed, as well as a decrease in immunogenicity. No relationship could be established between any effect and affinity or isotype of the bound mAbs. The preS2 region was rendered more immunogenic when an IgG2a mAb was bound to the S region of the HBsAg. The response to the S region was not modulated, whereas immunogenicity of the preS2 colinear region was decreased by antibody shielding. The mAb which was the most efficient as an enhancer of the antibody response also increased binding of the complexed immunogen to antigen presenting cells. The binding of a human mAb to the sole S region, but not to the preS2 region, should be tested as a potentiating agent of the anti-preS2 human immune response.

Characterisation of the main molecular forms of human fecal immunoglobulin A.

The molecular composition of fecal IgA is poorly documented, although it is of theoretical and practical importance to determine the different forms of IgA in faeces. Two main molecular forms were isolated by successive steps of ion-exchange chromatography and gel-filtration. The first consisted of secretory IgA dimers dissociating into slightly lower molecular mass forms under the influence of the electric field during electrophoresis. The other contained cleaved-IgA complexed with alpha 1-antitrypsin, that is considered to be a serum origin marker. These results confirm that secretory IgA are relatively resistant to digestive enzymes in vivo, and suggest that alpha 1-antitrypsin-bound fragments originate from serum IgA monomers. Analysis of the proportions of these forms may be of value in the investigation of gut diseases.

Immune complexes as immunizing agents to increase the number of monoclonal antibody producing hybrids and to deviate the response to poorly immunogenic epitopes.

The use of immune complexes (IC) in an antibody excess, as immunizing agent, led to a large increase in the mouse polyclonal response to human SIgA. This enhanced response, as compared to SIgA alone, was analysed with mouse polyclonal anti-alpha chain antibodies (Ab). A kinetic study showed an early rise (between days 14 and 21) of the antibody response against the discontinuous epitopes of SIgA while the anti-IgA response increase was delayed. Induction of hybridomas with an IC consisting in SIgA containing an excess of anti-alpha chain Ab, increased 10-fold the number of positive wells. Moreover, two of these MAb were specific for weakly immunogenic epitopes. One recognized only SIgA (anti-C), i.e. the association between the alpha chain and the secretory component (SC), while the other mainly combined to IgA dimers (anti-P). Both these MAb will be useful tools for structural studies and for the dosage of secretory Ab.

[Use of serum s-IgA detection in liver pathology (author's transl)]. / Intérêt de la détection des IgA sécrétoires dans le sérum en pathologie hépatique.

Secretory IgA (sIgA) were searched in 60 sera of healthy blood donors and in 1 590 sera of subjects having various diseases. 20 percent of these subjects showed an increased amount of sIgA in their sera. The only subjects presenting a constant increase (sometimes more than 20 fold the normal amount) were people with liver diseases. Quantitation of sIgA, in relation with the determination of the IgA/transferrin ratio (IgA/T) in sera, showed an important difference between Laennec's cirrhosis on one hand and virus hepatitis or post-hepatitic cirrhosis on the other. In Laennec's cirrhosis a moderate increase in sIgA went with a strong elevation of the IgA/T ratio, the latter being proportional to the degree of evolution of the disease. In virus hepatitis, the sIgA amount was largely increased while the IgA/T ratio remained at a normal value.

Unexpectedly high levels of some presumably protective secretory immunoglobulin A antibodies to dental plaque bacteria in salivas of both caries-resistant and caries-susceptible subjects.

The role of salivary antibodies in protection against cariogenic bacteria is actually a matter of debate. Correlation between caries experience and naturally occurring antibodies was extensively investigated. Comparison of salivary antibodies from 21 caries-resistant and 22 caries-susceptible subjects was carried out by using a new quantitative method. Secretory immunoglobulin A (S-IgA) antibodies to Streptococcus sobrinus and Streptococcus sanguis cells were detected in all salivas and at similar levels in both groups. When assayed with two major antigens from S. sobrinus, i.e., protein antigen I/II and cell wall carbohydrates, only specific activities of antibodies to the protein component were increased (P < 0.01), but this occurred unexpectedly in the caries-susceptible group. Western blot (immunoblot) analysis with the culture supernatant and cell wall proteins from S. sobrinus showed the same antibody specificity in both groups. No selective increase of the protease-resistant S-IgA2 subclass was found, and avidities of antibodies to both antigen I/II and cell wall carbohydrates were similar. Our results demonstrate that naturally induced S-IgA antibodies against S. sanguis, S. sobrinus, and the major antigens of the latter are not sufficient to inhibit caries development.

Secretory component-binding properties of normal serum IgM.

Our aim was to investigate why serum IgM is poorly transferred into secretions in normal subjects. Indeed, the low IgM level in secretions contrasts with the capacity of monoclonal IgM to bind to secretory component (SC), but it is not well established to what extent normal serum IgM can do so. The mean SC affinity was studied with a polyclonal IgM preparation from 250 normal subjects and with a representative pool of 100 different monoclonal IgM. The SC-binding percentages varied as a function of the IgM/SC molar ratio according to a common hyperbolic curve, with similar association constants: Ka = 4.19 +/- 2.61 x 10(7) M-1 (polyclonal pool) and Ka = 5.80 +/- 2.73 x 10(7) (monoclonal pool). It thus appears that the large difference in IgM concentrations between blood and secretions cannot be due to an SC-binding defect of serum IgM, but is probably explained by its low diffusion from blood to the extravascular compartment.

Nonimmune macromolecular complexes of Ig in human gut lumen. Probable enhancement of antibody functions.

Protein Fv, a human sialoprotein recently described in the stools of patients suffering from liver diseases, binds the variable domain of H chains without impairing Ag binding. Normal subjects are shown here to secrete protein Fv under a hidden form, saturated with luminal Ig. In feces, the nonimmune complexes are essentially of 1800 and 800 kDa M(r); they contain protein Fv molecules bound with F(ab')2 fragments produced by cleavage of Secretory IgA during colonic transit. The 800-kDa complexes correspond to 6 molecules of F(ab')2 fragments bound to a sole protein Fv dimer. This was established by comparison with an in vitro-made complex having a valency of 6. Investigation of the role of protein Fv shows that in vitro addition of this free molecule to antivirus or to anti-Salmonella typhi antibodies allows or augments agglutination of the corresponding pathogens. This property is of major interest for secretory antibodies because it favors their role in Ag conveyance in the mucus stream. It seems therefore that protein Fv is a novel key factor of the immune defense in gut.

Secretory component is bound to the paraproteins in sera of IgA and IgM gammopathies.

Secretory immunoglobulins A (SIgA) and M (SIgM) were investigated in 20 sera containing high levels of monoclonal polymeric IgM or IgA. In the sera of patients suffering from Waldenström's macroglobulinemia (WM), the level of SIgA was found to be low, whereas that of SIgM was extremely high. Reciprocally, in the multiple myeloma (MM) patients, SIgA were increased and SIgM were dramatically decreased. Electrophoretic analysis showed these SIgA and SIgM to have the same monoclonal pattern as the corresponding paraprotein. Hence these molecules must originate from the malignant clone. The most likely mechanism involved is an intravascular formation of the secretory-like immunoglobulins. Free secretory component (SC) could diffuse passively from the digestive lumen and bind the circulating myeloma polymeric immunoglobulins. Such a possibility of in vivo binding of free SC to IgM and IgA polymers leads to a reconsideration of the secretory origin of SIgM and SIgA in normal human serum.

Non-immune VH-binding specificity of human protein Fv.

The specificity of human F(ab)-binding Protein Fv (previously called Protein F), a sialoprotein released into the digestive tract mainly during hepatitis, was investigated with fragments or chains of monoclonal immunoglobulins. Protein Fv bound an unreduced H-chain dimer of a monoclonal human IgA2m(1) molecule but neither the corresponding L-chain dimer, nor several Bence-Jones molecules. Using enzymatic subfragments of F(ab)mu, or F(ab')2 gamma, a significant binding was observed with Fv fragments (VH + VL), while Fb fragments (CH1 + CL) were inactive. Taken altogether, these results prove that the structure recognized by Protein Fv is located in the VH domain. This structure probably involves discontinuous epitopes linked by a disulphide bond, which are destroyed by combined reduction and dissociation. Protein Fv does not interfere with the antigen-binding site, since there was no reciprocal inhibition with the antigen-antibody reaction.

Non-immune binding of human protein Fv to immunoglobulins from various mammalian and non-mammalian species.

Reactivity of the secretory protein Fv with immunoglobulins (Ig) from various species of vertebrates was investigated. Binding was observed throughout all taxonomic classes: mammalian, avian, reptilian, amphibian and fish. Contrasting with this wide spectrum, no significant binding was found with most mammalian ungulates, such as horse (Perissodactyl), cow, sheep and goat (Artiodactyls). Nevertheless, disruption of the hydrogen bonds of Ig allowed these non-reactive molecules to bind. Such a conserved reactivity during evolution, and our previous data on the effect of the cleavage of the intra-chain disulphide bonds, suggest that protein Fv recognizes a discontinuous framework structure involving both the FR1 and FR3 regions in the variable domain of the heavy chain of Ig.