Formation of a hexagonal lattice structure by an R-form lipopolysaccharide of Klebsiella: comparative study of preparations extracted by the phenol-water and the phenol-chloroform-petroleum ether methods

An R-form lipopolysaccharide (LPS) extracted from Klebsiella strain LEN-111 (O3-:K1-) by the phenol-chloroform-petroleum ether method was compared with that extracted by the phenol-water method in the ability to form a hexagonal assembly. The LPS which was extracted by the phenol-water method and dialyzed against tap water to remove phenol showed ribbon-like structures, and it formed a hexagonal lattice structure with a lattice constant of 14.5 +/- 0.3 nm when it was precipitated by addition of two volumes of 10 mM MgCl2-ethanol. The LPS which was extracted by the phenol-chloroform-petroleum ether method and lyophilized consisted of ribbon-like structures and their fragments and it often formed small pieces of a hexagonal lattice, although the LPS before lyophilization did not form such a lattice. When the LPS extracted by the phenol-chloroform-petroleum ether method was precipitated by addition of two volumes of 10 mM MgCl2-ethanol, it formed essentially the same hexagonal lattice structure as that formed by the LPS extracted by the phenol-water method. From these results it is concluded that the ability of the LPS to form a hexagonal lattice structure does not depend upon the method of its extraction from bacterial cells.     

Electrophoretic analysis of lipopolysaccharide isolated from opaque and translucent colony variants of Vibrio vulnificus using various extraction methods

Lipopolysaccharides (LPS) from an opaque and a translucent colony variant of Vibrio vulnificus were isolated by several methods and the electrophoretic profiles were analysed. The phenol-water extraction method provided a better yield compared to the phenol-chloroform-petroleum ether method. In addition, two rapid micro-assays were used to isolate LPS for electrophoretic analysis. The electrophoretic pattern was the same for all LPS extracts and was similar in both variants. No high molecular weight bands, characteristic of smooth LPS, were detected in the LPS of this organism. This was in contrast to the smooth nature of the LPS of another member of the Vibrionaceae examined, V. cholerae. The result of this study showed no correlation between LPS and colony morphology in V. vulnificus.     

Chemical characterization of Campylobacter jejuni lipopolysaccharides containing N-acetylneuraminic acid and 2,3-diamino-2,3-dideoxy-D-glucose.

Lipopolysaccharides (LPS) of four nonencapsulated strains of the human enteric pathogen Campylobacter jejuni were chemically characterized. When applied to two of the strains, extraction by a modified phenol-chloroform-petroleum ether method (H. Brade and C. Galanos, Eur. J. Biochem. 122:233-237, 1982) gave better yields of LPS than did extraction by the conventional hot phenol-water technique. Constituents common to all LPS were D-glucose, D-galactose, L-glycero-D-manno-heptose, 3-deoxy-D-manno-2-octulosonic acid, D-glucuronic acid, D-galactosamine, and phosphorylethanolamine. Phosphate was present in a relatively high amount. In addition, the LPS of three strains contained N-acetylneuraminic acid, whereas the LPS of the strain lacking this component contained 3-amino-3,6-dideoxy-D-glucose. The lipid A component contained phosphate with D-glucosamine and 2,3-diamino-2,3-dideoxy-D-glucose as the major amino sugars. Ethanolamine-phosphate was present also. The major fatty acids were ester- and amide-bound 3-hydroxytetradecanoic and ester-bound hexadecanoic acids, with a minor amount of ester-bound tetradecanoic acid. This is the first report of N-acetylneuraminic acid in the oligosaccharide moiety and diaminoglucose in the lipid A of C. jejuni LPS.     

The effect of formalin-killing of Pasteurella multocida on the antigenicity and extractability of its lipopolysaccharide

The extraction of lipopolysaccharides (LPS) from formalin-killed (FK) Pasteurella multocida strain X-73 and from cells not exposed to formalin (NF) were compared by the Westphal and phenol-chloroform-petroleum ether (PCP) extraction procedures. The LPS was determined by: (1) serologic analyses with antiserum specific for LPS; (2) analyses for toxicity; and (3) chemical analyses for components expected to be in LPS (such as hexoses, heptoses, amino sugars, 3-deoxyoctulosonic acid, and fatty acids). Strain X-73, the strain most virulent for chickens, was markedly affected by formalin killing. Unlike many strains, which readily yield LPS into the aqueous phase when extracted with phenol at 68 degrees by the Westphal procedure, strain X-73 did so only with FK and not with NF cells. With the NF cells, LPS was extracted by EDTA from the precipitate obtained during the Westphal procedure. With the PCP procedure, LPS was extracted readily from NF cells, but not from FK cells. The change in extractability of LPS as a result of formalin-killing was the same for both the encapsulated form of X-73 and a nonencapsulated variant derived from it. Although formalin-killing affected the extractability of LPS, no antigenic differences could be detected by immunodiffusion. However, the chick-embryo toxicity of LPS extracted from NF cells was greater than that of LPS from FK cells.     

An unusual lipid A from a marine bacterium Chryseobacterium scophtalmum CIP 104199T

The hydrolysis of defatted cells of the marine bacterium Chryseobacterium scophtalmum CIP 104199T with 10% acetic acid (3 h, 100 degrees C) led to an unusual lipid A (LA) (yield 0.6%), obtained for the first time. Using chemical analysis, FAB MS, and NMR spectroscopy, it was shown to be D-glucosamine 1-phosphate acylated with (R)-3-hydroxy-15-methylhexadecanoic and (R)-3-hydroxy-13-methyltetradecanoic acids at the C2 and C3 atoms, respectively. It is similar to the monosaccharide biosynthetic precursor of lipopolysaccharide (LPS), so-called lipid X (LX). Unlike LX, LA can be isolated by the treatment of bacteria with organic solvents only after the preliminary acidic hydrolysis of the cells, which suggests that LA might be strongly, probably chemically, linked to other components of the outer membrane. However, LPS cannot be such a component, because extraction with phenol-water or phenol-chloroform-petroleum ether mixtures in high yields (5.34% and 0.5%, respectively) leads to preparations that do not contain 3-deoxy-D-manno-oct-2-ulopyranosonic acid, 3-hydroxyalkanoic acids, or LA.     

Characterization ofCampylobacter jejuni lipopolysaccharide

Lipopolysaccharides were isolated from dehydratedCampylobacter jejuni by combination of the phenol-chloroform-petroleum ether and phenol-water extraction techniques. Biochemical characterizations of lipopolysaccharide were performed on the two fractions of highest purity. Neutral sugar analyses detected galactose, glucose, trace amounts of mannose, and an unidentified deoxy-hexose. The primary amino sugars were galactosamine, glucosamine, and glucosamine-phosphate. Chemical analyses of other lipopolysaccharide components included phosphate, 3-deoxy-d-manno-octulosonic acid (KDO), and fatty acids. The predominant fatty acids were 3-hydroxytetradecanoic and hexadecanoic acids with lesser amounts of tetradecanoic acid. 3-Hydroxytetradecanoic acids were bound to lipid A by both amide and ester linkages.     

Structure of a D-glycero -D-manno-heptan from the lipopolysaccharide of Helicobacter pylori

A lipopolysaccharide (LPS) was isolated by hot phenol-water extraction from Helicobacter pylori strain D4 and found to contain no fucosylated poly-N-acetyllactosamine chain typical of most H. pylori strains studied but a homopolymer of D-glycero-D-manno-heptose (DD-Hep). The heptan attached to a core oligosaccharide was released by mild acid degradation of the LPS, and the following structure of the trisaccharide-repeating unit was established by chemical methods and 1H and 13C NMR spectroscopy: --> 2)-D-alpha-D-Hepp-(1 --> 3)-D-alpha-D-Hepp-(1 --> 3)-D-alpha-D-Hepp-(1 -->. 1H NMR spectroscopy performed on small amounts of the intact LPS revealed the presence of the same polysaccharide in LPS of H. pylori strains D2 and D5, but not strain D10.     

Processing of McCoy cell cultures infected with Chlamydia trachomatis: sequential isolation of chlamydial elementary bodies and lipopolysaccharide

Abstract Triton X-100 (TX100) enhances the liberation of chlamydial elementary bodies (EB) from host cells and dissolves the host cell membrane. In the presence of TX100 only differential centrifugation is needed to isolate reasonably pure EBs. The remaining high-speed supernatant still contains a large part of the chlamydial lipopolysaccharide (LPS), which can be isolated with the standard phenol-chloroform-petroleum ether extraction.     

Lipopolysaccharide containing L-acofriose in the filamentous blue-green alga Anabaena variabilis

For the first time, an O-antigenic lipopolysaccharide (LPS) has been isolated from a filamentous blue-green alga (Anabaena variabilis). It was extractable with phenol-water, resulting in extraction of the bulk of the LPS into the phenol phase. The polysaccharide moiety of this LPS consists of l-rhamnose, its 3-O-methyl ether l-acofriose, d-mannose, d-glucose, and d-galactose. l-Glycero-d-mannoheptose and 2-keto-3-deoxyoctonate, the two characteristic sugar components of enteric LPS, and phosphate groups are absent from the A. variabilis O antigen. The only amino sugar present is d-glucosamine. Three hydroxy fatty acids were identified, namely, beta-hydroxymyristic, beta-hydroxypalmitic and beta-hydroxystearic acids, in addition to palmitic and unidentified fatty acid. The LPS of A. variabilis is localized in the outermost cell wall layer and behaves like a bacterial O antigen in serological tests. The passive hemagglutination yielded high titers with isolated LPS (pretreated by heat or by alkali) and rabbit antisera prepared against living or heat-killed cells. The position of the precipitation arcs after immunoelectrophoresis of the O antigen indicates the lack of charged groups. The water phase of the phenol-water extract contains, in high yield, a glucose polymer. It is serologically inactive as shown by the passive hemagglutination test and by agar-gel precipitation.     

Formation of a hexagonal lattice structure by an R-form lipopolysaccharide of Klebsiella sp.

We extracted an R-form lipopolysaccharide (LPS) by the phenol-water method from Klebsiella sp. strain LEN-111 (O3-:KI-) and followed the changes in ultrastructure of the LPS during the extraction procedure. When the LPS was obtained from the water phase of an extract by addition of 2 volumes of 10 mM MgCI2-ethanol, it consisted of membrane pieces with a hexagonal lattice structure with a lattice constant of 14 to 15 nm. The lattice structure of the LPS was disrupted into short rods with sodium dodecyl sulfate, but the same hexagonal lattice structure was again formed by precipitation with 2 volumes of 10 mM MgCI2-ethanol. The LPS preparation after two cycles of treatment by the phenol-water method, which contained no detectable amounts of proteins, kept an unaltered ability to form the hexagonal lattice structure. Extensive treatment with pronase and extraction with chloroform did not impair the ability of the LPS preparation to form the lattice structure. When the other salts, NaCI, CaCI2 or Zn(CH3COO)2, were used for precipitation of the LPS with ethanol in place of MgCI2, the LPS did not form the hexagonal lattice structure. However, if the LPS precipitated with NaCI-ethanol was converted to the magnesium salt form after it was electrodialyzed, it formed the same hexagonal lattice structure as the LPS precipitated with MgCI2-ethanol. From these results, it was concluded that the R-form LPS has the ability of in vitro self-assembly into a hexagonal lattice structure in the presence of Mg2+ without the help of other components such as proteins and free lipids from outer membrane.     

Rapid method for isolation and staining of bacterial lipopolysaccharide.

Classically bacterial lipopolysaccharide (LPS) purification and silver staining take several days. We designed a simple and fast method for LPS isolation which when combined with silver staining using Pharmacia PhastSystem both can be completed in few hours. The purity of LPS isolated by this simple method may not be comparable to that by the phenol-water method hence we recommend this rapid isolation and staining procedures for simple and fast study of LPS patterns in gels.     

Purification of rough-type lipopolysaccharides of Neisseria meningitidis from cells and outer membrane vesicles in spent media.

A procedure for the purification of Neisseria meningitidis lipopolysaccharide (LPS) from outer membrane vesicles (OMV) in spent growth media was developed. Five different LPS strains of group A N. meningitidis were grown in tryptic soy broth with vigorous aeration for 36-48 h, and centrifuged to collect both cells and supernatants. The amount of LPS in the OMV in the supernatants was higher or at least equal to that in the cells. The OMV in each supernatant were concentrated, pelleted by ultracentrifugation, and treated with 2% sodium deoxycholate to dissociate LPS from OMV. The LPS was then separated from capsular polysaccharides, proteins and phospholipids by gel filtration on Sephacryl S-300 column in 1% sodium deoxycholate, and precipitated from the column fractions in 70% ethanol. In addition, LPS was also extracted from cells with hot phenol-water, ultracentrifuged once after treatment with ribonuclease, and purified on Sephacryl S-300. When compared with an improved phenol-water extraction method, the LPS obtained from either OMV or cells by the above methods gave a 40-180% increase in yield. The LPS also had much higher activities in limulus amebocyte lysate assay, rabbit pyrogenic test, and enzyme-linked immunosorbent assay. The LPS purified from cells and from OMV were indistinguishable by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis.     

Rapid preparation of samples for compositional sugar analysis of the "degraded polysaccharide" fraction of lipopolysaccharides from Vibrio cholerae

A simple and rapid method was devised for direct isolation and fractionation of the "degraded polysaccharide" (DPS) fraction of O-antigenic (or endotoxic) lipopolysaccharides (LPS) directly from heat-killed Vibrio cholerae (O1 and non-O1) cells without separating the LPS. Neither phenol-water extraction nor ultracentrifuge is needed in this method. V. cholerae NIH 41 was used as standard. The cells (3-5 g wet weight) were heated in 5% acetic acid at 100 C for 1.5 hr. The acetic acid extract obtained as the supernatant by centrifugation was evaporated to dryness in vacuo, and the resultant residue was dissolved in 10 ml of distilled water. The solution was mixed with 2 volumes of acetone, and the supernatant obtained by centrifugation was mixed with 5 volumes of acetone and centrifuged. Fraction Sed. II was recovered as the precipitate, while the supernatant was evaporated to dryness in vacuo, yielding fraction Sup. III. Sed. II had a sugar composition that was identical, at least qualitatively, to that of DPS isolated from LPS of the corresponding strain except for the absence of a fructose component in the case of V. cholerae NIH 41, while instead Sup. III from V. cholerae NIH 41 contained fructose.     

Peptides as requirement for immunotherapy of the guinea-pig line-10 tumor with endotoxins

Summary The transplantable line-10 hepatocellular carcinoma of guinea-pigs has been used as a model for the study of immunotherapy of malignant tumors. Cure rates of up to 100% have been obtained with ReGl-CM from 0 antigen-deficient (Re) mutant strains of Enterobacteriaceae, provided they were combined with mycobacterial trehalose dimycolate (cord factor, P3). Whereas highly endotoxic LPS extracts from all wild-type strains so far tested have failed to cause tumor regression, acid hydrolysis of such LPS samples led to residual fractions (RESI) that cross-reacted serologically with ReGl-CM samples (Chang and Nowotny, 1975) and provided cure rates up to 100%. RESI from Serratia marcescens was essentially nonpyrogenic and 100 times less lethal for chick embryos than potent endotoxins. Antigenic material associated with endotoxic extracts appears to be cryptic or sterically hindered from being effective in wild-type LPS but is exposed in ReGl and RESI samples.Reduction of the aminoacid content of ReGl-CM by microparticulate silica gel chromatography or by treatment with Triton X-100 significantly lowered the ability to bring about tumor regression without affecting endotoxicity. Antitumor activity could be restored by the addition of synthetic N-acetyl-muramyl-l-seryl-d-isoglutamine (MDP) or a nontoxic lipoid side fraction recovered during the isolation of ReGl-CM, which contained a small amount of peptidic substances. It is concluded that the addition of peptidic material, which may act as an adjuvant, to endotoxins is required to make them useful for immunotherapy of the weakly immunogenic line-10 tumor.Chemical procedures known to detoxify endotoxins while retaining adjuvanticity, such as succinylation and phthalylation, resulted in complete loss of endotoxicity and tumor-regressive potency of ReGl-CM. Transesterification with sodium methoxide led to a water-soluble phase, which cured 50% of tumor-bearing animals even though lethality and pyrogenicity were reduced by 100 times and 50 times, respectively. Thus there was no direct correlation between endotoxic potency and tumor-regressive activity. In addition, our findings indicate that a low level of toxicity may be required to obtain optimal levels of tumoricidal action.Abbreviations P3 trehalose dimycolate isolated by microparticulate silica gel chromatography (Ribi et al., 1974) - LPS lipopolysaccharide from wild-type gram-negative bacteria - ReGl endotoxic glycolipids from Re mutant gram-negative bacteria - CM chloroformmethanol - PW phenol-water - PCP phenol-chloroform-petroleum ether - ReGl-CM ReGl-PW, ReGl-PCP refer to ReGl extracted with CM, PW, or PCP, respectively - ACP acetone-precipitated by-product of ReGl-CM - B1, B2, B4 chromatographic fractions of ReGl-CM - lipid A hydrochloric acid hydrolyzate of LPS or ReGl - RESI organic solvent-insoluble fraction of lipid A (Chang and Nowotny, 1975) - KDO keto-3-deoxyoctonate - BSA bovine serum albumin - CWS cell wall skeleton of BCG (Bacillus Calmette-Guérin) - PPD purified protein derivative (tuberculoprotein) - TAP tuberculin-active peptide - MDP N-acetyl-muramyl-l-seryl-d-isoglutamine     

Elucidation of the structure of the Pasteurella haemolytica serotype T10 lipopolysaccharide O-antigen by n.m.r. spectroscopy

The aqueous-phase lipopolysaccharide isolated from Pasteurella haemolytica serotype T10 cells by the phenol-water extraction method was found to be S-type lipopolysaccharide which possessed O-antigenic polysaccharide chains composed only of D-galactose residues. Structural analysis of the O-polysaccharide, using a combination of 1D and 2D 1H- and 13C-n.m.r. methods, led to the identification of the disaccharide repeating-unit as [----3)-alpha-D-Galp-(1----3)-beta-D-Galf-(1----]n. The serological cross-reactivity between P. haemolytica serotypes T4 and T10 can now be related to the structural similarity of the antigenic LPS O-polysaccharides.     

Structural elucidation of the lipopolysaccharide core region of the O-chain-deficient mutant strain A28 from Pseudomonas aeruginosa serotype 06 (International Antigenic Typing Scheme).

The lipopolysaccharide (LPS) of the Pseudomonas aeruginosa serotype 06 rough-type mutant A28 was isolated by a modified phenol-chloroform-petroleum ether extraction method. Deoxycholate-polyacrylamide gel electrophoresis indicated a single band with mobility similar to that of the complete core region of the wild-type parent serotype 06 (International Antigenic Typing Scheme) strain. Compositional analysis of the LPS indicated that the core oligosaccharide was composed of D-glucose (three units), L-rhamnose (one unit), 2-amino-2-deoxy-D-galactose (one unit), L-glycero-D-manno-heptose (two units), 3-deoxy-D-manno-octulosonic acid (two units), L-alanine (one unit), and phosphate (two units). Under the mild conditions of hydrolysis with methanolic hydrogen chloride, a 7-O-carbamoyl substituent was observed on the second heptose residue. The glycan structure of the LPS was determined by employing one- and two-dimensional nuclear magnetic resonance spectroscopy and mass spectrometry-based methods with a backbone oligosaccharide that was obtained from the LPS by deacylation, dephosphorylation, and reduction of the terminal glucosamine. On the basis of the results of the present study and our earlier work with the P. aeruginosa 06-derived core-defective mutant R5 (H. Masoud, E. Altman, J. C. Richards, and J. S. Lam, Biochemistry, 33:10568-10578, 1994), a structural model for the complete core oligosaccharide is proposed.     

Toxins of Bacteroides fragilis and Bacteroides thetaiotaomicron rods as stimulators of adhesion molecule expression on the surface of vascular endothelial cells

Lipopolysaccharides from four Bacteroides fragilis strains: one nonenterotoxigenic (NTBF) and three enterotoxigenic (ETBF), and from three B. thetaiotaomicron strains were extracted by hot phenol-water method and purified. B. fragilis enterotoxin was prepared according to the procedure of van Tassell et al. (1992). The influence of the examined toxins on the expression of adhesion molecules: ICAM-1, VCAM-1 and E-selectin on HMEC-1 (human dermal microvascular endothelial cells) was assayed in ELISA test with monoclonal antibodies. Four concentrations of toxins were applied: 0.01, 0.1, 1.0 and 10.0 (micrograms/ml). Endothelial cells were activated for 24 hours (ICAM-1 and VCAM-1 expression) and for 4 hours (E-selectin expression). The coloured product of immunoenzymatic reaction was measured by reading the absorbance at wavelength 492 nm. Two controls were performed in each experiment: with resting HMEC-1 and E. coli O55:B5 LPS (Sigma, USA). Bacteroides fragilis and B. thetaiotaomicron lipopolysaccharides stimulated three adhesion molecules under investigation. Their activity was comparable, but weaker than the activity of E. coli O55:B5 LPS. ICAM-1 was the most stimulated molecule. B. fragilis enterotoxin induced two adhesion molecules: VCAM-1 and E-selectin demonstrating weaker stimulatory activity than E. coli LPS. Stimulation of adhesion molecules on vascular endothelial cells should be considered to be a biological activity of B. fragilis and B. thetaiotaomicron endotoxins and B. fragilis enterotoxin.     

A method for purification of bacterial R-type lipopolysaccharides (lipooligosaccharides)

A new gel filtration method was developed for purification of R-type lipopolysaccharides (lipooligosaccharides) from some nonenteric gram-negative bacteria, including Neisseria meningitidis, Haemophilus influenzae, and Bordetella pertussis. These wild-type lipooligosaccharides are poorly extractable by the phenol-chloroform-ether extraction method of C. Galanos, O. Luderitz, and O. Westphal [1969) Eur. J. Biochem. 9, 245-249) and therefore a new procedure was developed for their isolation. The lipooligosaccharides (LOS) were first extracted by hot phenol-water, treated with RNase, then disaggregated in deoxycholic acid, and purified by gel filtration on Sephadex G-75. By comparison the conventional hot phenol-water purification method using repeated ultracentrifugations yielded less LOS. The yield of LOS by gel filtration was 30 to 108% higher and the purity was better.     

Procedure for isolation of bacterial lipopolysaccharides from both smooth and rough Pseudomonas aeruginosa and Salmonella typhimurium strains.

Lipopolysaccharide (LPS) is a major component of the outer membrane of gram-negative bacteria. It is now well established that within a single organism, size heterogeneity of this molecule can exist. We have developed a LPS isolation procedure which is effective in extracting both smooth and rough LPS in high yields (51 to 81% of the LPS present in whole cells as quantitated by using hydroxy fatty acid, heptose, and 2-keto-3-deoxyoctonate yields) and with a high degree of purity. The contamination by protein (0.1% by weight of LPS), nucleic acids (1%), lipids (2 to 5%), and other bacterial products was low. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the LPS demonstrated the presence of a high degree of size heterogeneity in the isolated smooth LPS as well as the presence of significant amounts of rough-type LPS. The Pseudomonas aeruginosa LPS interacted well with a monoclonal antibody in a variety of immunochemical analyses. The usefulness of the procedure was demonstrated by comparing LPS preparations obtained from wild-type and mutant strains of P. aeruginosa and Salmonella typhimurium. For example, it was shown that the LPS of an antibiotic supersusceptible mutant Z61 of P. aeruginosa, which was previously characterized as identical to wild type with respect to the ratio of smooth to rough LPS molecules isolated by the phenol-water procedure, actually contained only a small proportion of O-antigenic side chains.     

Chemical and ultrastructural differences in endotoxic glycolipids from Salmonella minnesota Re mutant extracted with various solvent systems

Endotoxic glycolipids ( ReGl ) extracted from the whole cells (WC) and cell walls of heptose-less Re mutant of Salmonella minnesota with hot phenol-water (PW), phenol-chloroform-petroleum ether (PCP), and chloroform-methanol (CM) were analyzed chemically and examined with an electron microscope. ReGl -PW-(WC) contained mannose and proteins as contaminants, ReGl -PCP(WC) consisted of an excess amount of amino compound (cadaverine), and ReGl -PCP(WC) consisted of proteins and cadaverine, in addition to the ReGl constituents. The ultrastructure of ReGl -PW(WC) resembled onions when stained with uranyl formate, and was spherical when stained with uranyl acetate, sodium phosphotungstate and ammonium molybdate, whereas those of ReGl -PCP(WC) and ReGl -CM(WC) were shaped like ribbons. However, the shadowed ultrastructures of ReGl -PCP(WC) and ReGl -CM(WC) showed small pieces of flat and wide sheets, respectively. ReGl -PCP(WC) re-extracted by the PW method was found to be converted into the onion-like structure which was similar to that of ReGl -PW(WC), while an intermediate form (fingerprint-like) was observed after re-extraction of ReGl -PW(WC) with PCP. It was strongly suggested that the ultrastructural arrangement of ReGl was dependent on the solvent systems used for extraction.