Measurement of salivary, urinary and fecal secretory IgA levels in children with partial or total IgA deficiency.

We measured salivary, urinary and fecal secretory IgA (sIgA) levels in 11 children with total IgA deficiency and in 6 children with partial IgA deficiency using an ELISA technique. This was based on flexible microplates coated with antisecretory component (SC) and peroxidase-conjugated anti-IgA as a second antibody. Selective IgA deficiency is diagnosed as a serum IgA concentration < or = 0.05 g/l; partial IgA deficiency is diagnosed as a serum concentration of IgA > 0.05 g/l but 2 SD below normal levels. No salivary or fecal sIgA, and only low levels of urinary sIgA, were detected in the selective IgA-deficient group. The partial IgA-deficient children presented with low levels of salivary, urinary and fecal sIgA. Fecal sIgA levels correlated with salivary sIgA levels (p < 0.01) but not with urinary sIgA levels (p > 0.05) in the IgA-deficient patients. We found that all the children with partial IgA deficiency, except one, had detectable, but low values of secretory IgA. Our data suggest that these patients also have a partial mucosal IgA deficiency.

Compensatory levels of salivary IgM anti-Streptococcus mutans antibodies in IgA-deficient patients.

The protective role of salivary IgA in dental caries has not been completely demonstrated, so, in order to elucidate this point, we evaluated 15 totally and partially IgA-deficient children in terms of the following variables: dental caries indexes, bacterial plaques, number of Streptococcus mutans and Lactobacillus in the saliva, and titers of IgA, IgG and IgM anti-Streptococcus mutans antibodies in the saliva. Age-matched healthy children served as the control group. IgA-deficient children showed caries scores lower than those of the healthy children; in addition, no statistical difference was found between amount of dental plaque and numbers of the bacteria in saliva. The totally IgA-deficient children presented IgM in levels much higher than the healthy children (p < 0.05). These data could indicate a compensation for the IgA deficiency by IgM.

Immunoassays of secretory IgA and secretory component.

1. A sandwich-type enzyme-linked immunosorbent assay (ELISA) is described for quantitation of secretory IgA (sIgA) in human serum, as well as an ELISA and a radioimmunoassay (RIA) for measurement of secretory component (SC) in human serum. Samples were reduced and alkylated prior to the measurement of SC. 2. Healthy individuals (N = 53) presented low levels of SC (median, 0.9 mg/l). The protein levels were significantly elevated when compared with the controls, in sera of women during the second (N = 31; median, 1.5 mg/l) and third (N = 35; median, 4.2 mg/l) trimesters of pregnancy and in sera of patients with alcoholic cirrhosis (N = 38; median, 2.4 mg/l) and acute viral hepatitis (N = 25; median, 2.4 mg/l). SC levels of women in the first trimester of pregnancy (N = 24; median, 0.5 mg/l) did not differ from the controls. 3. sIgA levels were also significantly elevated when sera of women in the third trimester of pregnancy (N = 41; median, 25.4 mg/l) and sera of patients with alcoholic cirrhosis (N = 32; median, 75.0 mg/l) or acute viral hepatitis (N = 38; median, 28.5 mg/l) were compared with controls (N = 49; median, 9.0 mg/l). Women in the first (N = 25; median, 7.7 mg/l) and second (N = 29; median, 10.2 mg/l) trimester of pregnancy did not present levels statistically different from the controls. 4. The results obtained for SC by RIA and ELISA were positively correlated (rs = 0.88; P less than 0.001). sIgA levels determined by ELISA were also positively correlated with the results of RIA-SC (rs = 0.77; P less than 0.001) or ELISA-SC (rs = 0.79; P less than 0.001). 5. The assays described are specific, relatively simple to perform, and can be useful for the study of the secretory immune system.

Human IgG but not IgM antibodies can protect mice from the challenge with live O6 Escherichia coli.

We evaluated the ability of human anti-lipopolysaccharide (LPS) O6 immunoglobulin G (IgG) and IgM antibodies to protect mice challenged with Escherichia coli serotype O6:K2ac. Purified whole IgG, commercial gammaglobulin, whole IgM-effluent, pool of normal human serum (NHS), agammaglobulinaemic serum (test groups) or phosphate-buffered saline (control group) was injected into adult male 18 h before a challenge with viable O6 E. coli. The mortality rate was assessed over a period of 72 h. To determine the opsonic and phagocytic activity of the antibody isotypes, we incubated peritoneal macrophages from the control and test groups collected at different times after challenge with the live bacteria with acridine orange for fluorescent analysis. Tumour necrosis factor (TNF)-alpha and interleukin (IL)-6 were quantified in serum of both the test and control groups. All mice that received commercial gammaglobulin or NHS survived. Purified whole IgG (containing 1.1 mg/l of anti-LPS O6 IgG antibodies) protected 87.5% of the animals tested in this experiment, while whole IgM-enriched effluent with 1.5 mg/l of anti-LPS O6 IgM antibodies protected only 12.5%. The agamma serum showed no protective capacity compared with PBS (serving as control). The minimal concentration of anti-LPS O6 IgG antibodies able to protect 50% of animals was 0.137 mg/l of purified whole IgG. Whole IgM-enriched effluent showed no protective capacity independently of the concentration tested (0.048-17.0 mg/l of anti-LPS O6 IgM antibodies). Fluorescent analysis of peritoneal macrophages from animals pretreated with purified whole IgG showed no bacteria at 8 h after the challenge. By contrast, whole IgM effluent showed an increasing number of live bacteria at the same time. Mice that had received whole IgM effluent (1.5 mg/l of anti-LPS O6 IgM antibodies) before the challenge with LPS O6 presented 20.5 microg/l of IL-6 and 1.5 microg/l of TNF-alpha. Serum from animals pretreated with purified IgG did not present any detectable pro-inflammatory cytokine. Our findings suggest that IgG but not IgM antibodies protect animals from a challenge with E. coli O6 serotype.

Quantitation of salivary, urinary and faecal sIgA in children living in different conditions of antigenic exposure.

A sandwich-type ELISA was developed to quantify salivary, urinary and faecal secretory IgA (sIgA). The assay is based on binding of sIgA to microplates coated with anti-SC antibodies and reaction with peroxidase-labelled anti-IgA. The sensitivity of the technique was approximately 5 micrograms/L. Children, 1-6 years old (n = 142), were divided into two groups. Group 1 (n = 80) was composed of children living in a place with presumably low antigenic exposure conditions. Group 2 (n = 62) was composed of well-nourished (2A, n = 53) and malnourished children (2B, n = 9) living in a São Paulo slum with presumably high antigenic exposure. The subgroup 2A had salivary levels higher than group 1 and the ranges were similar to those found in the literature for older children and adults. The same subgroup presented a high incidence of undetectable faecal sIgA; their levels of urinary sIgA did not differ from group 1. The subgroup 2B did not have levels of salivary, urinary and faecal sIgA different from subgroup 2A. Our results suggest that environmental factors influence the ontogenesis of sIgA system.

Antibody titre and avidity in saliva and serum are not impaired in mildly to moderately undernourished children.

Avidities and titres of serum IgA, IgM, IgG, and salivary IgA antibodies for poliovirus type 3, beta-lactoglobulin and E. coli O antigens were analysed in well-nourished (n=58) and mildly to moderately undernourished (n = 18) 1-7-year-old children, living in slums in São Paulo, Brazil, with irregular water supply and no sewage system. The serum titres of antibodies in the undernourished children were not statistically different from the well-nourished ones, except for the IgA anti-beta-lactoglobulin titres which were lower in the youngest undernourished group. In respect to avidity of the serum antibodies, no statistical difference was observed, except that the youngest undernourished children presented avidities of serum IgA anti-E. coli O antigens higher than those of the well-nourished children of the same age range. With respect to salivary titres and avidities no difference was observed between well-nourished and undernourished children. The fact that the undernourished children were not deficient in antibody amount or quality, as measured in this study, suggests that in this respect their immune system was not impaired.

Differences within mono- and dizygotic twin-pairs in spectrotypes and clones of IgG2 antibodies to pneumococcal polysaccharide type 1 and C-polysaccharide after vaccination.

Spectrotypes and clones of antibodies against pneumococcal capsular polysaccharide (Ps) type 1 and C-polysaccharide (C-Ps) were determined before and after immunization with a polyvalent pneumococcal Ps vaccine in 84 mono- or dizygotic twins. The method used was a micromodification of a rapid isoelectric focusing-affinity (IEF-affinity) immunoblot technique in agarose permitting characterization of isotype, light chain and Gm type. After vaccination the anti-type 1 Ps + anti-C-Ps clones were different in 75% of the monozygotic and 79% of the dizygotic twins. The anti-type 1 Ps clones differed among 72% of the monozygotic and 85% of the dizygotic twins (P > 0.05). Each twin had from zero to three clones producing IgG2 antibodies against type 1 Ps. A total of six different clones could be distinguished among all the twins. Vaccination enhanced the already actively secreting B-cell clones in 56 twins and newly recruited clones in 11 of the 84 twins; six among the 48 mono- and five among the 36 dizygotic twins. These new clones differed among the twins. Spectrotypes varied between all twins within the pairs. The fact that all twins differed in spectrotype is due to post-translational microheterogenity of the antibodies, events which are thus not genetically determined. The observation that even monozygotic twins possessed and responded with different clones within the pairs indicates that the V-region genes, which determine the final specificity of B cells, either differ from the original germ-line V region genes, e.g. owing to hypermutations or junctional diversity, or the rearranged germ-line genes occur accidentally although highly restricted.

Anti-Streptococcus mutans antibodies in saliva of children with different degrees of dental caries.

The aim of this study was to evaluate the relationship between the secretory immune system and dental caries. Forty-nine 3-5-year-old children with primary dentition were classified into three groups according to their caries indices: no caries (group I), one or two surfaces with caries lesions (group II) and rampant caries (group III). Lower numbers of mutans streptococci were found in group I in relation to groups II and III. Secretory IgA and anti-S. mutans IgA, IgM and IgG antibody levels were not significantly different among the groups. Western blotting analysis showed that some S. mutans proteins, including the 39, 59, 97 and 150 kDa molecular mass bands, were recognized by almost all the saliva samples. Antibodies against the 185 kDa band, known as antigen I/II, were present in all adults' saliva and in only one child in group III. The absence of antibodies to the 185 kDa band in children's saliva suggest a specific immunologic immaturity. Further prospective studies will be necessary to establish the possible effect of reactivity to this antigen on the S. mutans colonization in this age group.

Elevated levels and different repertoire profile of colostral anti-LPS antibodies may have a significant role in compensating newborn immunity.

A high prevalence of systemic infections caused by enterobacteria such as Escherichia coli is observed during the neonatal period. Lipopolysaccharide (LPS) is one of the major factors responsible for septic shock caused by these Gram-negative bacteria. We have recently demonstrated the presence of anti-LPS immunoglobulin (Ig)G antibodies in cord blood with a repertoire identical to that found in maternal serum. In the present study, we analyzed anti-LPS O111 antibody isotypes in maternal serum and colostrum from mothers and in cord serum from their respective full-term (n = 30) and preterm (n = 13) neonate infants. The main isotype found in serum samples from mothers of term infants was IgM (range between 28 and 54 mg/l), followed by IgA (1-2 mg/l) and IgG (2-3 mg/l). The range of IgG antibody concentrations in cord blood was between 2 and 3 mg/l, as a result of placental transfer. A novel observation in our study was that the LPS bands recognized by colostral antibodies were completely different from those recognized by IgG in serum. Colostral IgA antibodies recognized several bands not bound by serum IgG antibodies from the respective maternal serum, independently of the antibody quantity. In addition, we verified the pattern of LPS recognition by serum IgA and colostral IgA antibodies was identical, what suggested that the antibody isotype found in serum could probably be derived from differentiated IgA-positive cells which were homing to the mucosa through the mucosal homing mechanism. Identical pattern of recognition was obtained comparing the IgA and IgM isotypes in colostrum. Slight differences in the pattern of recognition were found between colostral and serum IgM antibodies. The fact that colostral antibodies recognize much more bands than serum antibodies may be important for the host to mount an effective immune response in the intestinal lumen, in order to prevent excessive absorption of LPS, reducing possible systemic effects caused by the molecule.

A modified fast micro method in agarose for isotype, allotype, light chain and idiotype-specific analysis of antibody clonotypes to bacterial virulence antigens.

A modified fast micro method for spectrotypic/clonotypic analysis of human IgG1-4 antibodies against bacterial virulence antigens of polysaccharide or protein nature is described. Serum samples of as small volumes as 0.5 microliter were isoelectrically focused in micro agarose gels made in plexiglass matrices and blotted using immunoaffinity-mediated capillary blotting onto nitrocellulose membranes previously coated with antigen. The bands of the antigen-specific antibodies were identified with respect to isotypes, light chain types, allotypes or idiotypes by incubating the nitrocellulose membranes with mouse monoclonal anti-human IgG subclass antisera and then with alkaline phosphatase-conjugated rabbit anti-mouse immunoglobulins. The method was applied for characterization of human monoclonals against tetanus toxoid (TT) and for the analysis of variable patterns of clonotypes in IgG subclass-deficient patients. The usefulness of the technique was also demonstrated by comparing the variable specificity and reactivity of different commercial monoclonals against human IgG subclasses. This method is fast, specific, sensitive, uses little material, is simple and reproducible.

Placental transfer of IgG antibodies against Haemophilus influenzae type b capsular polysaccharide in Brazilian term and preterm newborns.

Placental transfer of antibodies to polysaccharide antigens is still a controversial subject. The incidence of invasive Haemophilus influenzae type b (Hib) infections is high in countries where the vaccine has not been included in routine immunization schedules. In the present work, we proposed to evaluate the natural immune response to Hib capsular polysaccharide in term and preterm Brazilian newborns and their respective mothers. Although the means, medians, and ranges of antibody titres in paired maternal and cord sera from preterm neonates were similar, the maternal levels were slightly higher than the cord levels and a poor correlation between these levels was verified. Term neonates showed similar antibody levels to those of their respective mothers and a very significant correlation between these levels was observed.

Allergic and immunologic parameters in patients with Fanconi's anemia.

BACKGROUND: Fanconi's anemia (FA) is a rare recessive chromosomal instability disorder, characterized by progressive bone marrow failure and congenital defects. Patients with FA present with recurrent infections, particularly those of the respiratory tract. OBJECTIVE: The aim of the present study was to evaluate whether patients with FA have altered antibody-mediated immune responses. METHODS: A group of 12 patients with FA, 5-32 years old (6 males) was studied. Serum levels of IgG, IgM, IgA and IgG subclasses, isohemagglutinin titers and specific IgG antibodies to poliovirus and measles were determined using standard methods. Immediate skin tests to common inhalant allergens were performed, and total and specific serum IgE was quantitated using a fluoroenzymatic assay (Uni-CAP, Pharmacia). Antipneumococcal antibodies were measured by ELISA before and 4-8 weeks after immunization with pneumococcal vaccine (Pneumo 23, Pasteur Mérieux Connaught). Responses to serotypes 1, 3, 5, 6B, 9V and 14, which are the most prevalent in our country, were studied. RESULTS: Ten patients had elevated IgE levels in sera, and 7 of them had detectable specific IgE and positive immediate skin tests. An inadequate response to pneumococcal vaccination was found in 2 of the 12 patients. Isohemagglutinin titers and levels of IgG, IgM, IgA and IgG subclasses and antipoliovirus and antimeasles antibodies were within the normal limits for age in all patients. Two patients had undetectable IgG4 levels (below 5 mg/dl). CONCLUSIONS: The results indicate that a proportion of patients with FA (2/12) in our study had inadequate responses to pneumococcal vaccination. No other significant abnormalities of the immune system were found in these patients.

Evaluation of serum levels of IgG subclasses and anti-ribosyl-ribitolphosphate IgG and IgG2 in children with Haemophilus influenzae b meningitis.

In 40 children with Haemophilus influenzae b (Hib) meningitis, we determined serum levels (mg/dl) of IgG subclasses using the radial immunodiffusion method; 67.8 per cent of these children were less than 24 months old. In 14 children of the sample we measured serum IgG and IgG2 anti-ribosyl-ribitolphosphate (anti-PRP) (by enzyme-linked immunosorbent assay, ELISA) in the acute and convalescent phases of the disease. Lower IgG2 levels than those of the control group were obtained in all age ranges: 3-12 months, 1-2 years (p < 0.01), and 2-5 years (p < 0.001). IgG4 was also present in lower levels in patients of all age ranges (p < 0.05, p < 0.001, and p < 0.01 respectively). Serum levels of IgG anti-PRP and IgG2 anti-PRP measured were very low in the acute phase of the disease in all age ranges and there was no notable increase in levels during the convalescent phase of the disease. This result indicates that children less than 24 months old do not produce sufficient levels of IgG and IgG2 anti-PRP even after Hib meningitis.

Placental transfer of IgG and IgG subclass antibodies anti-purified Escherichia coli LPS O16, O6 and O111.

We evaluated 22 paired maternal and cord sera regarding the presence of IgG and IgG subclasses against purified Escherichia coli LPS O6, O16 and O111 employing ELISA for titre and avidity analysis, isoelectric focusing associated with affinity-blotting for spectrotypic analysis, and the Western-blotting technique for recognition of the various bands in lipopolysaccharide (LPS). Levels of anti-LPS IgG antibodies in cord sera were equivalent to their respective maternal sera, showing a significant correlation (P < 0.0001). IgG1 antibody levels were higher in cord sera than in maternal sera (P < 0.005 for anti-O111, P < 0.05 for anti-O16 and P < 0.02 for anti-O6). Cord IgG2 antibody levels were not different from the maternal levels (P > 0.1). The levels of IgG3 and IgG4 were undetectable. The avidity of anti-O6 and anti-O111 IgG in 10 cord sera showed an extremely significant correlation with maternal antibody avidity (P < 0.0001). Identical patterns of recognition were found in the paired samples analysed by Western blotting. Most of the serum samples recognized the O-repetitive chains and also the region corresponding to core and lipid A. Although the antibody spectrotypes varied among individuals, paired cord and maternal serum samples showed identical patterns. Our findings suggest the occurrence of placental transfer of IgG antibodies against LPS O6, O16 and O111, mainly involving the IgG1 or IgG2 subclasses.

Immunoassays of secretory IgA and secretory component

1. A sandwich-type enzyme-linked immunosorbent (ELISA) is described for quantitation of secretory IgA (sIgA) in human serum, as well as an ELISA and a radioimmunoassay (RIA) for measurement of secretory component (SC) is human serum. Samples were reduced and alkylated prior to the measurement of SC. 2. Healthy individuals (N = 53) presented low levels of SC (median, 0.9 mg/l). The protein levels were significantly elevated when compared with the controls, in sera of women during the second (N = 31; median, 1.5 mg/l) and third (N = 35; median, 2.4 mg/l) and acute viral hepatitis (N = 25; median 2.4 mg/l). SC levels of women in the first trimester of pregnancy (N = 24; median, 0.5 mg/l) did not differ from the controls. 3. sIgA levels were also significantly elevated when sera of women in the third trimester of pregnancy (N = 41; median, 25.4 mg/l) and sera of patients with alcoholic cirrhosis (N = 32; median, 75.0 mg/l) or acute viral hepatitis (N = 38; median, 28.5 mg/l) were compared with controls (N = 49; median, 9.0 mg/l). women in the first (N = 25; median, 7.7 mg/l) and second (N = 29; median, 10.2 mg/l trimester of pregnancy did not present levels statistically different from the controls. 4. The results obtained for SC by RIA and ELISA were positively correlated (rs = 0.88; P < 0.001). sIgA levels determined by ELISA were also positively correlated with the results of RIA-SC(rs = 0.77;P<0.001) or ELISA-SC (rs = 0.79; P < 0.001). 5. The assays described are specific, relatively simple to perform, and can be useful for the study of the secretory system