Search

The allergic airway diseases chronic rhinosinusitis (CRS), allergic fungal rhinosinusitis (AFRS), asthma, allergic bronchopulmonary mycosis/aspergillosis (ABPM/A), and cystic fibrosis (CF) share a common immunological signature marked by T H 2 and TH 17 cell predominant immune responses, the production of IgE antibody, and atypical inflammatory cell infiltrate that includes eosinophils and other innate immune effector cells. Severe forms of these disorders have long been recognized as being related to hypersensitivity reactions to environmental fungi. Increasingly however,environmental fungi are assuming a more primary role in the etiology of these disorders, with airway mycosis, a type of non-invasive airway fungal infection,recognized as an essential driving factor in at least severe subsets of allergic airway diseases. In this review, we consider recent progress made in understanding the immune mechanisms that drive airway mycosis-related diseases, improvements in immune-based diagnostic strategies, and therapeutic approaches that target key immune pathways.

Panose (O-α-D-glucosyl-(16)-O-α-D-glucosyl-(14)-α-D-glucose) and maltose (O-α-D-glucosyl-(14)-α-D-glucose) coupled to bovine serum albumin (BSA) by means of an azophenyl linkage provided the synthetic antigens panosyl-BSA and maltosyl-BSA, used to examine the anti-hapten specificity of rabbit sera prepared against panosyl-BSA. The combining site specificity of cross-reactive antibodies in carrier-absorbed anti-panoside sera was investigated by quantitative hapten inhibition of maltosyl BSA-anti-panoside cross-precipitation and compared with data obtained by inhibition of homologous precipitation. Immunochemical data obtained demonstrate a heterogeneity in the combining site specificity of anti-hapten produced in response to panosyl-BSA and establish the occurrence of at least two distinct kinds of antibody directed against the α-(16)-α-(14) sequence of glucose units. One type of anti-panoside, inhibited more effectively by isomaltose than by maltose, shows isomaltose end group immunodominance. A second type of anti-panoside, cross-precipitated by maltosyl-BSA and more effectively inhibited by maltose than by isomaltose, shows maltose immunodominance.

Panosyl-BSA, a synthetic sugar-azoprotein conjugate consisting of the trisaccharide panose (glucosyl α-(1,6)-glucosyl-α(1,4) glucose) coupled to bovine serum albumin by an azphenyl linkage, cross precipitates antibody from types II and XX antipneumococcal horse sera. Hapten inhibition indicates that these cross reactions of synthetic antigen are mediated by the α-isomaltosyl moiety of the introduced panosyl haptenic grouping. While interaction of conjugate with horse anti-XX involves a fraction of antibody directed against an α-isomaltosyl end group, interaction with horse anti-II involves a fraction of antibody directed against the α-isomaltosyl end group. Purified fractions of antibody, cross reactive with panosly-BSA were specifically from antipneumococcal horse sera and their behavior with homologous antigen examined and compared with unfractionated serum.

Plasma glutathione peroxidase (GSHPx) (glutathione: H2O2 oxidoreductase) is a unique selenoglycoprotein. Treatment of this enzyme with glycopeptidase F partially deglycosylates it and establishes the presence of N-linked sugar moieties. Antibodies raised in a rabbit against the purified enzyme from plasma were found to be specific, noninhibitory, and capable of precipitating the enzymatic activity. The antibodies precipitated greater than 90% of the GSHPx activity of normal plasma, thus indicating that the selenoenzyme is the main if not the sole GSHPx activity of plasma. The antibodies did not precipitate RBC GSHPx. A slight cross-reactivity of the antibodies was found with rat plasma GSHPx. A GSHPx activity precipitation assay of normal plasma in the presence of selenium (Se)-deficient plasma indicates that no cross-reactive protein in the Se-deficient plasma interferes with the precipitation of the GSHPx activity from normal plasma. Thus, GSHPx protein as well as activity is deficient in plasma in the absence of Se. Antibodies against GSHPx either from RBCs or from plasma were used to specifically immunoprecipitate most of the GSHPx activity from RBCs or plasma, respectively, in healthy individuals to determine the amount of Se associated with the protein. GSHPx accounts for approximately 15% of the Se in RBCs and 12% of the Se in plasma. Thus, in normal individuals, these proteins account for only a fraction of plasma and RBC Se.

Conjugates consisting of methyl 2( p -azobenzamido)2-deoxy-α- d -glucopyranoside and methyl 2( p -azobenzamido)-2-β- d -glucopyranoside coupled to bovine serum albumin were used as synthetic antigens. Rabbits immunized with artificial antigen provided specific anti-methyl α-glucosaminide or anti-methyl β-glucosaminide anti-hapten sera. Quantitative hapten inhibition was employed to elucidate the specificity of anti-hapten combining sites and to evaluate some structural and configurational features contributing to the immunochemical interaction with the antibody. It is shown that the configuration at the anomeric carbon atom makes a major contribution to the immunochemical specificity of the anti-hapten antibodies against α- and β-methyl glycosides.

Interspecies cross-immunization of the horse with isolated donkey IgGa immunoglobulin gives rise to the production of specific antibodies. In addition to an isotypic specificity, horse anti-donkey IgGa antibodies also show a species specificity. Antisera obtained from a single horse during hyperimmunization were examined by immunodiffusion, quantitative precipitation, immunoadsorption and radio-immune binding. Identification of horse anti-donkey IgGa antibodies isolated by immunoadsorption showed IgGa and IgG(Ta) immunoglobulins present in each bleeding. While antibodies identified as IgGc and IgB immunoglobulins were not detected, IgGb, IgA and IgM antibodies varied in that they could be detected in some bleedings but not in others. Immunodiffusion and quantitative precipitation with donkey IgGa, its Fab, F(ab′), F(ab′)2, Fc and Fc′ fragments showed only intact donkey IgGa and its Fc fragment to precipitate with horse anti-donkey IgGa. Inhibition of solid phase radio-immune binding of 125I-Fab and 125I-Fc fragments however, established the presence of horse antibodies directed against antigenic determinants in the Fab as well as the Fc regions of donkey IgGa.

Fifty-four strains of Neisseria gonorrhoeae were examined for their ability to be agglutinated by various lectins. Wheat germ agglutinin, ricin, soybean lectin, and peanut agglutinin agglutinated all strains tested. Dolichos biflorus, Sophora japonica, Maclura pomifera, Ulex Europaeus, Lens culinaris, Canavalia ensiformis, Phaseolus lunatus, and Bandeiraea simplicifolia BS I and BS II lectins either failed to agglutinate or agglutinated some strains but not others. Agglutination of gonococci by wheat germ agglutinin, ricin, and soybean lectin is sugar specific and most effectively inhibited by ligands known to interact with lectin combining sites. Lectin reactive groupings appear to be independent of antigenic determinants conferring Gc serogroup specificity. Interactions with wheat germ agglutinin and ricin suggest the occurrence of multiple beta-linked N-acetyl-D-glucosamine (D-GlcNAc) and beta-D-galactosyl (beta-D-Gal) units as common structural features of gonococcal cell envelope polysaccharide. Interactions with soybean agglutinin, peanut agglutinin, and Dolichos biflorus lectins suggest that alpha-N-acetyl-D-galactosamine (alpha-D-GalNAc) units and beta-D-Gal linked to GalNAc and (or) GlcNAc may also occur as structural features of the polysaccharide components of the cell envelope of some gonococci.

Human plasma glutathione peroxidase (GSHPx) has been shown to be a glycosylated selenoprotein distinct enzymatically, structurally, and antigenically from known cellular glutathione peroxidases. The extracellular location of the enzyme and the fact that it is glycosylated suggested that it is a secreted protein. Utilizing mutually non-cross-reactive antibodies to human cellular and plasma GSHPx, we conducted a search to determine the tissue of origin for plasma GSHPx. The cells screened were endothelial cells because they are the main source of extracellular superoxide dismutase, HL-60 cells (myeloid cell line) because they are the main source of extracellular H2O2, and Hep G2 cells (hepatic cell line) because they are the source of many plasma proteins. Human umbilical vein endothelial cells were metabolically labeled with either [35S]methionine or [75Se]selenious acid, and HL-60 cells and Hep G2 cells were metabolically labeled with [75Se]selenious acid. Proteins were immunopurified from the labeled cells and their media with either anti-red blood cell (RBC) GSHPx IgG or with anti-plasma GSHPx IgG. Utilizing anti-RBC GSHPx IgG, only the cellular form of the enzyme was precipitated from all the cells tested but not from their media. When anti-plasma GSHPx IgG was applied to the cells and their media, a selenoprotein was precipitated only from the media of Hep G2 cells. When Hep G2 cells were incubated in the presence of the carboxylic ionophore monensin, an intracellular selenoprotein could be detected using anti-plasma GSHPx IgG. The precipitation of the cellular form from all three cell types was partially inhibited by preincubation of the anti-RBC GSHPx IgG with purified RBC GSHPx while the precipitation of the selenoprotein from the medium of Hep G2 cells by anti-plasma GSHPx IgG was prevented by preincubation of the antibody with purified plasma GSHPx. We suggest that plasma GSHPx is synthesized by and secreted from hepatic cells. This is, to the best of our knowledge, the only known selenoprotein with a defined function that has been shown to be synthesized for secretion by mammalian cells.

Salmonella telaviv, Salmonella tranoroa, and Salmonella illinois were examined for their ability to interact with 15 purified lectins of known sugar specificity. The only interaction observed was between the lectin of Maclura pomifera and S. telaviv. M. pomifera lectin specifically agglutinated suspensions of S. telaviv and precipitated with its purified lipopolysaccharide and isolated lipid A free O polysaccharide. Quantitative inhibition assays showing methyl-alpha-D-galactopyranoside and N-acetyl-D-galactosamine to be potent inhibitors of Maclura lectin precipitation by S. telaviv O polysaccharide suggest that the interaction is mediated by D-galactose or N-acetyl-D-galactosamine units of bacterial polysaccharide structure, or both.

Strains of Neisseria gonorrhoeae were treated with pyocin 611 131 (pyocin 103) from Pseudomonas aeruginosa PA103, and isogenic resistant variants were isolated. The interaction of pyocin-sensitive and isogenic pyocin-resistant strains with wheat germ agglutinin (WGA) agglutinated all pyocin-sensitive, but not pyocin-resistant, strains. Binding of WGA to three pyocin-sensitive strains and their isogenic pyocin-resistant variants was examined quantitatively by using fluorescein-conjugated lectin. Pyocin-resistant strains maximally bound one-third to one-eighth the quantity of WGA bound by isogenic-sensitive strains. Linear Scatchard plots revealed homogeneous WGA-binding sites on three pyocin-sensitive and one pyocin-resistant strains. Biphasic Scatchard plots, obtained with two pyocin-resistant strains, show that WGA-binding sites in these strains are heterogeneous. The number of WGA-binding sites for pyocin-sensitive organisms ranged from 8 x 10(5) to 1 x 10(6) sites per coccus and from 1 x 10(5) to 3 x 10(5) sites per coccus for pyocin-resistant strains. The apparent association constant for WGA binding to pyocin-sensitive strains ranged from 3 x 10(6) to 6 x 10(6) liters/mol and from 6 x 10(6) to 1 x 10(7) liters/mol for pyocin-resistant strains. Gonococcal lipopolysaccharide was shown to serve as the pyocin 103 receptor by inhibition of pyocin activity. Lipopolysaccharide from a pyocin 103-resistant strain was not able to inhibit pyocin 103 activity. Pyocin 103 resistance was correlated with a structural alteration involving N-acetylglucosamine residues in gonococcal lipopolysaccharide. Based on interactions with wheat germ, soybean, and ricin lectins, a model of lipopolysaccharide structure in N. gonorrhoeae is presented.

Pneumococcal type 37 capsular polysaccharide was obtained free of contaminants by affinity chromatography on Con-A, wheat germ agglutinin, Maclura pomifera lectin and HOPC-8 mouse myeloma protein affinity columns. The immunochemical reactivity of native and periodate oxidized borohydride reduced type 37 polysaccharide antigen with polyclonal rabbit and monoclonal mouse anti-Pn37 hybridoma antibodies was studied by quantitative precipitation. Quantitative hapten inhibition studies, employing the isomeric series of alpha- and beta-(1----2), (1----3), (1----4) and (1----6)-linked glucobioses as competitive inhibitors of antibody precipitation establish a specificity for anti-Pn37 antibody directed at least in part, against the Glc beta(1----2) Glc (sophorosyl) unit. A high mol. wt, D-glucose containing polysaccharide antigen, cross-reactive with rabbit anti-Pn37 is reported which was found to occur in the culture medium of 7 of 19 strains of Actinomyces examined.

1. 1. Antigenic analysis employing rabbit antisera specific for horse immunoglobulins established the occurrence of IgG, IgG(T), IgM and γ 1 -component in the normal serum of the donkey, mule, hinny and zebra. 2. 2. With the rabbit reagents employed, the corresponding immunoglobulins of each equine specie and the interspecific hybrids were immunologically indistinguishable. 3. 3. Antigenic analysis of sera obtained from horse umbilical cord blood showed only IgM to be present; IgG(T), γ 1 -component and IgG were not detected. 4. 4. IgG-globulins of the donkey, mule, hinny and zebra were isolated from normal serum by chromatography on columns of DEAE-cellulose and their extinction and sedimentation coefficients determined.

Antiserum to crude lactoperoxidase was employed in agar diffusion and immunoelectrophoretic analyses of fractions obtained in the isolation and purification of lactoperoxidase from cow's milk by ion-exchange chromatography and gel filtration. The presence of seven antigenic components could be detected in a crude enzyme preparation by immunological analysis and their fate monitored during the purification procedure. Purified lactoperoxidase, prepared by ion-exchange chromatography and gel filtration, was found by immunological analysis to consist of a single antigenic component.

The production of specific anti-enzyme in response to immunization with purified lactoperoxidase could be demonstrated by assays of peroxidase activity on the supernatant fluids and precipitates obtained from quantitative precipitin determinations. Spectrophotometric analysis of specific precipitates formed in antibody excess for a distinctive antigen marker showed all the hemoprotein added as antigen to be quantitatively precipitated by antibody. Lactoperoxidase isolated from the milk of ruminants such as the goat and sheep were examined for their cross reactivity with anti-bovine lactoperoxidase and found to be immunochemically indistinguishable from cow's milk lactoperoxidase in their immunodiffusion and quantitative precipitin behavior. Bovine lactoperoxidase is immunologically distinct from bovine transferrin, ceruloplasmin, red protein, and lactollin as well as dog myeloperoxidase. Exposure of lactoperoxidase to proteolytic enzymes or 8 M urea does not result in any change detectable by immunodiffusion. The diffusion coefficient of bovine milk lactoperoxidase was estimated by immunodiffusion to be 5.2 × 10 −7 cm sec − .

1. 1. Antisera prepared against equine IgG globulin, its Fab, Fc fragments and L chains were employed in immunodiffusion analysis to examine their cross reactivity with canine IgGa,b globulins and a canine IgGd paraprotein. 2. 2. Interspecies antigenic similarities were found and cross reactive groupings assigned to various parts of the canine γG-globulin molecule. 3. 3. Immunodiffusion and cross absorption studies with anti-equine IgG, showed canine IgGa,b and IgGd to share some antigenic determinants in common with equine IgG and IgB absent from equine IgG(T).

Bovine tissues were analyzed for the presence of lactoperoxidase by immunodiffusion analysis. A protein antigenically identical to lactoperoxidase was shown to be present in the sublingual, submaxillary, and parotid glands. An antigenically related but not identical protein was present in pig salivary glands. As the enzyme could not be detected in any of the other tissues employed, it could be concluded that if present, it represented less than 0.04% of the extracted protein. Most notable among the tissues which did not contain the enzyme, are the thyroid, uterus, and lymph node. A method is presented for the isolation of lactoperoxidase from the submaxillary gland. The enzyme is a major protein of this tissue, comprising between 1 and 2% of the wet weight of the tissue. The relationship of lactoperoxidase to peroxidase activity reported in other studies of the salivary gland is discussed. It is clear that the “salivary peroxidase” and “iodide peroxidase” can be attributed to lactoperoxidase.

Rabbit antiserum to a BSA-cellobiose conjugate was examined by immunodiffusion and quantitative hapten inhibition employing the synthetic polyvalent hapten phloroglucinol triazophenyl-s- d -cellobioside as antigen. The isomeric series of s-linked glucobioses and mixed oligosaccharides possessing either terminal nonreducing or internal (subterminal) cellobiose were assayed for their ability to inhibit precipitation. Anticonjugate was more effectively inhibited by oligosaccharides possessing a terminal non-reducing cellobiose unit than by compounds with subterminal cellobiose groups. antibody could readily distinguish the s-(1,4) from the s-(1,3) and s-(1,6) glucosidic linkages. The portion of horse anti-VIII, precipitable by BSA-Cello was examined by hapten inhibition and p- nitrophenyl -s- d -cellobioside , cellotriose and cellotetraose found to be equipotent inhibitors of precipitation suggesting a specificity directed against the s- d -cellobiosyl or cellotriose unit.

The immunochemical behavior and enzymic activity of α-amylase from Aspergillus oryzae were examined before and after treatment of enzyme with N -bromosuccinimide or H 2 O 2 in the presence of Cu ++ . Under conditions resulting in the loss of 60–90% of the enzymic activity, preparations still reactive with antibody were obtained which showed a decrease in capacity per unit weight in precipitating antibody. Under certain conditions, 60% of the amylase activity could be destroyed by N -bromosuccinimide without any detectable change in immunochemical behavior. Immunochemical reactivity and enzymic activity thus show a difference in susceptibility toward these oxidizing reagents. Amylase-antiamylase specific precipitates formed with 7S antibody or bivalent 5S antibody fragments and soluble complex formed with univalent 3.8S antibody fragments showed only a small fraction of the activity of uncombined enzyme when starch was used as substrate. Similar findings were obtained when maltotriose or maltotetraose were substituted for starch in the assay of residual activity. These findings suggest that decreased activity shown by insoluble complexes is not due to aggregate formation but rather to the location of antibody attachment with respect to active sites. A Lineweaver-Burk treatment of kinetic data with antibody complexes is consistent with this interpretation.

The trisaccharide panose, 6- O - α -d-glucopyranosyl-4- O - α -d-glucopyranosyl-d-glucose, coupled to bovine serum albumin (BSA) via an azophenyl linkage was used as an artificial antigen to immunize rabbits. Antisera rendered specific for the introduced haptenic grouping by absorption with whole calf serum were studied by immunodiffusion, quantitative precipitation and hapten inhibition. Anti-hapten showed a specificity directed against the introduced p -phenylazo- s -panoside group. Hapten inhibition assays employing various di- and triglucosides, including two structural isomers of panose, established the relative importance of an α-configuration and a nonreducing terminal (1,6)-(1,4) ordered sequence of glucoside bonds for optimal interaction with antibody. A heterogeneity in the specificity of antibodies formed in response to s-panosyl-BSA was evidenced from cross reactions obtained between anti-panoside and various sugar azoprotein conjugates.

Sirishinha, Stitaya (University of Rochester School of Medicine and Dentistry, Rochester, N.Y.), and Peter Z. Allen . Immunochemical studies on α-amylase. III. Immunochemical relationships among amylases from various microorganisms. J. Bacteriol. 90: 1120–1128. 1965.—Immunochemical relationships among amylases obtained from a selected group of microorganisms were examined, and a cross-reaction was detected between the α-amylases of Bacillus stearothermophilus and B. subtilis . Immunodiffusion and quantitative precipitin studies, as well as cross-neutralization tests, indicate that B. stearothermophilus α-amylase reacts with a portion of antibody present in antisera to crystalline B. subtilis α-amylase. Amylases from these two species thus have some aspects of structure in common. Limited data obtained by immunodiffusion suggest that groupings which confer cross-reactivity to the B. stearothermophilus enzyme are lost after exposure to mercaptoethanol in the presence of ethylenediamine-tetraacetate, followed by treatment with iodoacetamide. With the antisera employed and within the concentration range examined, no immunochemical cross-reaction was observed among amylases from Aspergillus oryzae, B. subtilis, B. polymyxa, B. macerans, Pseudomonas saccharophila , and Euglena sanguinis . Immunoelectrophoresis of partially purified B. stearothermophilus α-amylase by use of antiserum to the crude enzyme, together with localization of amylase activity in immunoelectrophoretic plates, suggests that B. stearothermophilus α-amylase is antigenic in the rabbit.

Investigation of bovine lacrimal and harderian glands revealed the presence of the enzyme lactoperoxidase, which was isolated and purified. A nonheme, iron-containing protein was identified at the same time. Both proteins are present in milk, mammary glands, and salivary glands. Their roles are discussed: The lactoperoxidase may be important in controlling bacterial flora.

Intracellular recording and horseradish peroxidase (HRP) injection techniques were employed to examine the projections of superficial layer [stratum griseum superficiale (SGS) and stratum opticum (SO)] superior collicular (SC) neurons in the hamster that sent axon collaterals into the deep laminae (those ventral to the SO) of this structure. Sixty-nine neurons were studied, selected from a sample of over 185 HRP-filled superficial layer cells on the basis of having heavily stained axons. Of the 69 cells included in the study, 43.4% (n = 30) sent at least one axon collateral to the deep laminae. Not all cell types in the superficial layers contributed equally to this interlaminar projection: 78.6% (n = 11) of the recovered wide-field vertical cells, 55.0% (n = 11) of the narrow-field vertical cells, 16.7% (n = 2) of the stellate cells, 40.0% (n = 2) of the marginal cells, 18.2% (n = 2) of the horizontal cells, and 28.6% (n = 2) of neurons we could not classify on the basis of their somadendritic morphology projected to the deep layers. Within a given cell class, there were no significant morphological or physiological differences between the neurons that possessed deep axon collaterals and those that did not. The deep axon collaterals of most of the interlaminar projection neurons were restricted to the stratum griseum intermediate (SGI). In this layer, the largest segment of the axon arbor was located lateral to a projection line that was orthogonal to the SC surface and that passed through the soma of the cell in question. These results, along with those of a previous study (Mooney et al., 1984), which demonstrated that the dendrites of deep layer cells may extend through the SO and into the SGS, indicate that there is an extensive anatomical substrate by which sensory information may be communicated from superficial to deep layer SC neurons.

Selenium deficiency can develop in patients receiving intravenous hyperalimentation. This depletion can result in a decrease in both cellular and plasma glutathione peroxidase. Previous investigations have revealed that with repletion of selenium, there is a rapid increase in the recovery of plasma glutathione peroxidase and a more slow recovery in red blood cell glutathione peroxidase. This suggests that the plasma enzyme is not present as a result of hemolysis of red blood cells.

Hyperimmune horse serum from a single animal (horse 46) immunized with group B (strain B-11) meningococcal vaccine provides a standardized, readily available diagnostic reagent used in primary isolation medium and for serogrouping of meningococci. Identification of the heavy-chain isotypes of specific anticapsular polysaccharide and anti-lipopolysaccharide isolated from horse 46 serum revealed a differential distribution in the occurrence of immunoglobulin classes. Meningococcal anticapsular antibodies of horse 46 serum were restricted predominately to the immunoglobulin M (IgM) class, with only trace amounts of IgGa present, whereas anti-lipopolysaccharide concomitantly produced showed a heterogeneity in its heavy-chain isotypes, consisting of IgM, IgGa, IgGb, moderate amounts of IgB, and a small amount of IgA.

Immunodiffusion analysis of papain digestion products, heavy and light chains of horse IgG-globulins with several rabbit and anti-horse IgG sera, have permitted the demonstration of five antigenic specificities (Fc 1 , Fc 2 , L, L sp and Fab sp ) associated with these equine antigens. Reactivity with anti-Fc 1 is shown by both F′c and Fc fragments, while anti-Fc 2 reactivity is shown only by Fc fragment. Absorption of anti-Fab serum with L chain Fc fragment provides a reagent (anti-Fab sp ) which precipitates only with Fab fragment, IgG-globulin or reduced and alkylated IgG. Upon exposure to detergent, reduced and alkylated, horse IgG loses its precipitability by anti-Fab sp , while reactivity with anti-L or anti-Fc reagents remains unaffected; reactivity is recoverable, however, following removal of detergent. Antisera to horse L chain show two kinds of specificity, one type (anti-L) directed against L chain determinants of IgG and Fab fragment, while another fraction of antibodies (anti-L sp ) is directed against determinants present on purified, reduced and alkylated horse L chains. Anti-L sp reagent does not precipitate with native IgG or Fab fragment, but does not react with free L chains and urea or detergent or heat-treated IgG-globulins.

Purified levan is antigenic in man and subcutaneous injection of 1 mg. leads to the production of precipitins and skin sensitivity. Apple amylopectin, maize glycogen, and synthetic polyglucose did not stimulate antibody production in small numbers of individuals injected. Quantitative precipitin studies carried out with several levan preparations employing human antilevan show variations in immunochemical behavior indicating structural differences among levans. Evidence that precipitins formed in response to levan injection are antilevan antibodies was obtained by analysis of specific precipitates formed in the region of antibody excess for ketosugar. Injection of levan giving rise to production of antilevan failed in one individual to give a non-specific anamnestic rise in either antidextran or antiblood group A precipitins. Laminarin, a neutral polysaccharide, was found to give precipitation with normal human sera.

Laminaribiose ( 3-o-β- d -glucopyranosyl- d -glucose ) was coupled to bovine serum albumin (BSA) by an azophenyl linkage to provide the synthetic antigen BSA-p-phenylazo-β-laminaribioside. The behavior of antisera prepared in rabbits immunized with the β-laminaribiosyl conjugate was examined by immunodiffusion, quantitative precipitation and hapten inhibition. Anticonjugate absorbed with carrier protein showed the greatest reactivity with the homologous β-laminaribiosyl-BSA antigen, but also showed some cross precipitation with β-cellobiosyl, β-sophorosyl and β-gentiobiosyl-BSA conjugates. Glucobioses linked through the β-(1 → 2), β-(1 → 3), β-(1 → 4) and β-(1 → 6) positions, as well as laminaridextrins and tri and tetrasaccharides of β-linked glucose possessing a laminaribiose moiety either at a nonreducing end location or at a subterminal location, were assayed for their ability to inhibit antilaminaribioside precipitation. Hapten inhibition data showed anticonjugate a possess a high degree of specificity directed against the terminal nonreducing β-laminaribiosyl end group.