Maurice M. Rapport

Isolation of β-galactosidase and β-glucosidase from brain

Summary 1. β -Galactosidase ( β - D -galactoside galactohydrolase, EC 3.2.1.23) and β -glucosidase ( β - D -glucoside glucohydrolase, EC 3.2.1.21) were extracted from brain tissue and were separated from each other. β -Galactosidase from rat brain was purified 20-fold. The preparation catalyzed the hydrolysis of p -nitrophenyl- β -galactoside at a rate exceeding 100-fold that of p -nitrophenyl- β -glucoside. A similar preparation from calf brain was purified 40-fold but was less cleanly separated from glucosidase activity. A preparation of β -glucosidase from beef brain, purified 10-fold, catalyzed the hydrolysis of p -nitrophenyl- β -glucoside at a rate exceeding 50-fold that of p -nitrophenyl- β -galactoside. 2. The pH optimum, Km , and transferase activity of the enzymes as well as their competitive inhibition by galactono- and gluconolactones and by galactose were studied. Hydrolysis of p -nitrophenyl- β -lactoside to p -nitrophenol, glucose and galactose by the combined action of both enzymes was also investigated. 3. Attention is drawn to the use of these enzymes for the hydrolysis of glyco-phingolipids.

Immunochemical studies of organ and tumor lipids. IX. Configuration of the carbohydrate residues in cytolipin H.

Abstract The reaction between cytolipin H and antitumor serum is specifically inhibited by lactose. Cellobiose and d -galactose are ineffective. These observations provide strong evidence that the configuration of the glucose and galactose residues in cytolipin H have the configuration of the disaccharide 4-0-β- d -galactopyranosyl- d -glucopyranose (lactose).

The structure of plasmalogens IV. Lipids in normal and neoplastic tissues of man and in normal tissues of rabbit and rat

1. 1. Total lipid extracts of rabbit tissues (9 normal), rat tissues (9 normal, 1 cancerous), and human tissues (7 normal, 10 neoplastic) have been analyzed for their plasmalogen content by 2 independent methods: formation of the aldehyde p-nitrophenylhydrazone and addition of iodine under conditions specific for α,β-unsaturated ethers. 2. 2. The results agree with an average variation of less than 10%, indicating that almost all plasmalogens in mammalian organs have, as a common structural feature, an α,β-unsaturated ether linkage in a relatively long fatty chain. 3. 3. It is suggested that as the plasmalogen-lipid phosphorus ratio for a given organ is relatively constant, this ratio may be a more significant criterion than plasmalogen content for relating plasmalogen changes to organ function.

Immunochemical studies of organ and tumor lipids: XIII. Isolation of cytolipin K, a glycosphingolipid hapten present in human kidney☆☆☆

Abstract A neutral glycosphingolipid containing fatty acid-sphingosine (long chain base)-glucose-galactose-galactosamine in the molar proportions 1:1:1:2:1 has been isolated from normal human kidney by successive chromatographic separations on silicic acid, magnesium silicate, and silicic acid. The isolated substance, designated cytolipin K, was shown to have haptenic properties by screening it for reactivity with a large number of rabbit antisera prepared against diverse human tissues. Good reactions were observed in a small number of instances. The two most reactive antisera were obtained by combination immunization with lipids from a multiple myeloma-affected kidney and a pyelonephritic kidney, and with these antisera and optimal addition of auxiliary lipid, 0.003–0.005 μg. of cytolipin K was readily measured. It was shown that the immunological reactivity of cytolipin K was distinct with respect to five other lipid haptens present in human tissues, namely, cytolipin H, cytolipin G, galactocerebroside, cardiolipin, and Forssman hapten.

Lipid monolayers: Interaction of synthetic dihydroceramide lactosides with proteins

Abstract Appreciation of the chemical basis of specific lipid-protein interactions in biological systems requires more extensive study of nonspecific (as well as specific) interactions. Monolayers of glycosphingolipids were studied as these lipids are immunologically active and are known to form specific complexes with antibody. Rabbit γ-globulin, rabbit serum albumin, and bovine ribonuclease were injected under monolayers of synthetic N-myristoyl-, N-palmitoyl-, and N-stearoyl dihydrosphingosyl lactosides (C14-, C16-, and C18-dihydroceramide lactosides) at a definite film pressure. The interaction was followed by measuring the increase in surface pressure (π) as a function of time, of initial film pressure (πi), of temperature, and of protein concentration. The degree of interaction of these lipids with each of the proteins at 25°C. was in the order C16 > C18 ⪢ C14. In the interaction with rabbit γ-globulin, the surface pressure-temperature curve had a maximum at 25°C. for N-palmitoyl dihydroceramide lactoside and 35°C. for the stearoyl homolog. When different classes of lipids were compared, the order of interaction at 25°C. was: cholesterol > N-palmitoyl dihydroceramide lactoside > phosphatidyl ethanolamine > phosphatidyl choline. As expected, the increase in π was smaller at larger values of πi, but a maximum at low pressure (1 dyne/cm.) was found. With monolayers of N-palmitoyl dihydroceramide lactoside at an initial film pressure of 2 dynes/cm., the degree of interaction at 25°C. was in the order ribonuclease > γ-globulin > albumin. Curves relating degree of interaction to protein concentration showed discontinuities that were different for each system. The data agree with earlier studies on the relative reactivity of different classes of lipids and lend support to two concepts. First, the mechanism of penetration of a lipid film at low pressure by protein is one involving interaction between polar groups rather than simple diffusion of protein into the film. Second, the structure of water around the polar groups of both lipid monolayer and protein is an important factor in lipid-protein interaction.

IMMUNOCHEMICAL STUDIES OF ORGAN AND TUMOUR LIPIDS. XIV. GALACTOCEREBROSIDE DETERMINANTS IN THE MYELIN SHEATH OF THE CENTRAL NERVOUS SYSTEM.

THE organ specificity observed with some anti-brain sera, a well-established phenomenon usually associated with lipid antigens (BRANDT, GUTH and MULLER, 1926; WITEBSKY and STEINFELD, 1928), was recently shown to depend on the immunological activity of galactocerebrosides (JOFFE, RAPPORT and GRAF, 1963). The evidence was based on the finding that such antisera react with exceedingly small quantities of galactocerebroside (less than 0.005 p g ) but not with glucocerebroside or other lipid haptens such as cytolipin H (RAPPORT, GRAF, SKIPSKI and ALONZO, 1959), cytolipin K (RAPPORT, GRAF and SCHNEIDER, 1964), cytolipin G (GRAF, RAPPORT and BRANDT, 1961), Forssman hapten, and cardiolipin. The reactions of pure phrenosine and pure cerasine were indistinguishable. The reaction of galactocerebrosides with antibody is readily measured by the method of complement-fixation, but requires the presence of lecithin and cholesterol as auxiliary lipids (JOFFE et al., 1963). The dependence of the reaction on these auxiliary lipids appears to be a general phenomenon since it is similar for four different pure glycosphingolipid-antibody systems studied thus far, namely, cytolipin H, cytolipinK, galactocerebroside, and glucocerebroside (RAPPORT, 1963). The auxiliary lipid requirement needs to be evaluated from two different standpoints: first, the relative proportions of lecithin and cholesterol, and secondly, the weight of auxiliary lipid in relation to the weight of hapten. When this was done, with both intensity and sensitivity of reaction used as a guide, optimal reactions were obtained by using 100 times as much auxiliary lipid as glycosphingolipid; in this respect the four haptens listed above behave similarly. On the other hand, the proportions of lecithin and cholesterol did differ significantly for the four different hapten-antibody interactions, and the effect of changing these proportions was also seen, although to a much lesser degree, in studying different antisera with the same

Paper chromatography of saturated, unsaturated and hydroxy fatty acids.

Abstract 1. 1. Methods are presented for separating straight-chain saturated, unsaturated, and 2-hydroxy fatty acids from C 14 to C 24 by paper chromatography and for detecting the acids as bismuth sulfides. 2. 2. These methods are applied to the analysis of fatty acids isolated from bovine spleen glucocerebroside and bovine spinal cord galactocerebroside.

Immunochemical studies of organ and tumor lipids. XV. The reactivity of anti-lactose sera with cytolipin H.

Abstract A total of ninety antisera were prepared in rabbits against foreign globulinazophenyl lactose conjugates and studied for their content of precipitating antibody with phloroglucinol triazophenyl lactoside and for complement-fixation reactions with cytolipin H (ceramide lactoside). Of sixty antisera with a content of anti-lactose protein in excess of 100 μg/ml, sixteen contained complement-fixing antibody with avidity for cytolipin H comparable to that observed with antisera to human tumor tissue. These anti-lactose sera were therefore useful serological reagents for the measurement of cytolipin H. The correlation between antibody content determined by precipitation with a synthetic lactoside and by complement-fixation with cytolipin H was poor. Some anti-lactose sera contained potent inhibitors of complement-fixation. Differences in character between anti-lactose antibody and anti-tissue antibody were evident from 1. differences in the sensitivity to inhibition by lactose of the complement-fixation reaction with cytolipin H and 2. differences in the precipitability of the two types of antibody by the lactoside test antigen. Anti-lactose sera are more specific than anti-tissue sera in reactions with lipid haptens.

Immunochemical Studies of Organ and Tumor Lipids

Cytolipin H is a glycosphingolipid whose haptenic function was first demonstrated as a result of studies of the antigenic components of a human epidermoid carcinoma (1, 2). Analysis of the isolated molecule showed that it contained fatty acid, a lipid base, glucose, and galactose in equimolar proportions and thus had a composition similar to that of a substance isolated from ox spleen (3). Subsequently it was shown that cytolipin H isolated from ox spleen was immunochemically indistinguishable from that isolated from human tumor, despite small differences in degree of unsaturation and specific rotation (4). Further studies have now revealed other structural features, namely, that the two monosaccharide residues are bound as a disaccharide with the configuration of lactose (5), that the fatty acid residue is represented by chains ranging from C-14 to C-24 with a distribution that differs for each preparation, and that the lipid base is principally sphingosine (6). Further study of this molecule is necessary in order to develop more fully the relation between structure and immunological activity, but the supply of sphingolipid has been extremely limited. On one hand, large amounts of human carcinoma tissue are not available, and on the other, the method of isolation from ox spleen described by Klenk and Rennkamp (3) is very indirect, since their primary objective was the isolation of gangliosides rather than cytolipin H. We have therefore restudied the method of isolation with a view to developing an efficient procedure that would yield a pure product in gram quantities. The method developed involves the following steps: (a) dehydration of the tissue, (b) extraction of lipids and removal of water-soluble impurities from them, (c) preparation of an ether-acetone-insoluble fraction, (d) preparation of an acetic acid-acetone-insoluble fraction, (e) chromatography on Florisil, and (‘j) chromatography on silicic acid. In this way it was possible to isolate over 1.5 g of cytolipin H from 40 kg of ox spleen, or approximately 75 7* of that estimated to be present in this tissue.

IMMUNOLOGICAL REACTIONS OF MYELIN IN VITRO

Early in 1962, Joffe, Rapport and Graf found that, contrary to previous reports,1’ * galactocerebroside would indeed satisfy the requirements for a solution to a long-recognized problem, namely, the identity of the lipid antigen that accounts for the organ specificity observed with some antibrain sera. This study was published in January, 19633 and showed that both naturally-occurring and synthetic galactocerebrosides had approximately equal activity, measured by complementfixation, with an antibrain serum that reacted strongly with brain lipid extracts of 10 animal species, but not with lipid extracts of other organs. Pure galactocerebroside did not react, and the strong reactions shown were only obtained in the presence of auxiliary lipids, in this case, a mixture of equal parts by weight of lecithin and cholesterol. The best reactions (judged by both sensitivity and intensity) were seen when 100 parts of the auxiliary lipid mixture were used for each part of galactocerebroside hapten. At the six unit level of complement-fixation the endpoint of 50 per cent hemolysis (indicating consumption of five complement units) was obtained with about 0.005 p g of galactocerebroside. Specificity was established by showing that other pure lipids of analogous structure (glucocerebroside, cytolipin H), similarly formulated with auxiliary lipid, were inactive at concentrations over 200 times greater. The magnitude of the auxiliary lipid effect was undoubtedly the decisive feature that led to the erroneous interpretations in the earlier reports: on the one hand, fairly pure cerebroside fractions showed little or no activity, whereas the fractions containing many lipids, among which were both lecithin and cholesterol, but only a trace of cerebroside displayed intense reactivity. Hence it was concluded that galactocerebroside was not the key substance. Since we find that optimal reactivity is obtained with ratios of auxiliary to essential lipid of 100 to 1, these phenomena no longer present any difficulty in interpretation. However, the auxiliary lipid phenomenon still presents a tremendous obstacle in attempting to relate the quantitative reactivity of antibody with artificial formulations of test antigens to this reactivity with similar antigenic determinants as they are found in nature. Are they, indeed, present in sufficient concentration, in available sites, in a form that permits them to combine with specific antibody? The availability of pure myelin prepared by Norton and Autilio permitted us to direct our attention to some of these questions. Myelin is quite insoluble in aqueous medium, but the complement-fixation method is singularly geared to cope with this difficulty in a routine way. Preliminary tests showed that the myelin had no anticomplementary effect, thus permitting the sensitivity of the complement-fixation method to be increased to three complement units. FIGURE 1 shows the isofixation curves4 obtained with an antibovine brain serum that contains antigalactocerebroside antibody using two different preparations of myelin, one “light” and the other a mixture of “light” and “heavy.” These curves show the quantities of antigen and