ABSTRACT STRUCTURAL ANALYSIS OF THE POLYSACCHARIDES FROM THE ENVELOPES OF HETEROCYSTS AND SPORES OF A BLUE-GREEN ALGA BY Liliana A. Cardemil de Balboa Lindberg's combined gas chromatography-mass spectrometry tech- niques for the analysis of partially methylated alditol acetate sugar derivatives were used to study the structures of polysaccharides from the envelopes of heterocysts and spores of Anabaena cylindrica. Polysaccharides from both envelopes are highly branched. Glucose, mannose, galactose and xylose are at terminal positions, whereas glucose and mannose are at internal positions in these polymers. The molar percentages of the eleven partially methylated alditol acetate derivatives observed were approximately the same for the polysaccharides from the two envelopes. The side branches of these polysaccharides were removed by Smith degradation (periodate oxidation followed by reduction with sodium borohydride and mild acid hydrolysis) without measurable fragmenta— tion of the backbones. Gas chromatographic analysis of the partially methylated alditol acetate sugar derivatives showed that the back- bones of both envelope polysaccharides consist of glucose (61c) and mannose (Man) linked by (1+3) glycosidic bonds. Disaccharides, Liliana A. Cardemil de Balboa trisaccharides, and tetrasaccharides, obtained from the backbone polysaccharides by partial acid hydrolysis, were fractionated by column chromatography and separated by high voltage electrophoresis. Analysis of these oligosaccharides established that both backbone polysaccharides consist of repetitions of the structural unit Glc- Glc-Glc-Man, and that all linkages in the backbones are in the 8- configuration. All of the structural analytical data given in this dissertation suggest that the envelope polysaccharides from heterocyst and spore cells of Anabaena cylindrica are essentially identical. The polysaccharide from the envelope of spores can be extracted by boiling lipid-free spore envelopes with a concentrated solution of sodium azide. As shown by methylation analysis, extraction of the polysaccharide by boiling does not change its structure. An analysis of the amino acids present in lipid-free envelopes of heterocysts and spores shows that the amino-containing compounds are amino-acids. The results suggest strongly that a protein or peptide is a component of these envelopes. STRUCTURAL ANALYSIS OF THE POLYSACCHARIDES FROM THE ENVELOPES OF HETEROCYSTS AND SPORES OF A BLUE-GREEN ALGA BY «’\‘ ‘ LP (A: ’ . Liliana KZ’Cardemil de Balboa A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1975 To my husband Orlando and to my children Paulina and Orlandito ii ACKNOWLEDGEMENTS I wish to express my gratitude to Dr. C. Peter Wolk, who has given me freedom, advice and encouragement throughout these investigations. It has been a wonderful experience for me to work under his direction. I also want to extent my gratitude to Dr. Dietz Bauer, who helped me grow up in the field of Carbohydrate Chemistry. The helpful discussions with Dr. Paul Kindel, the advice and loan of laboratory equipment from Dr. Derek Lamport and the constructive criticism of Dr. Philip Filner are greatly appreciated. As members of my guidance committee, they have contributed to my scientific development. Many thanks to Maureen Meinert for English and editorial sugges- tions and to Andrew Mort for discussions on methods and equipment assistance. I thank both of you for your friendship. My sincere gratitude to my colleagues Taka Hirosawa, Paul Shaffer, Dr. Joseph Thomas and Dr. Robert Fisher for their friendship and encouragement. The assistance of Dr. Lloyd Wilson with the electrophoresis equip- ment and of Mr. Jack Harten with the gas chromatograph-mass spectrometer is warmly acknowledged. My warmest thought to my son Orlandito, who has been at home in Chile, waiting for his mother all of these years. iii The research reported here was supported by the U.S. Atomic Energy Commission and the U.S. Energy Research and Development Administration under Contract E(ll-l)-1338. iv TABLE OF CONTENTS LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . . . . LIST OF FIGURES O O O O O O O O O O O O O O O O O O O O O O 0 INTRODUCTION 0 O O I O O O O O O O O O O O O O O O O O O O O 0 LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . I. Elucication of the Glycosidic Linkages of Polysaccharides by Means of Methylation Analysis . . . . . . . . . . . . . . . . . . . . . Rationale of the Method . . . . . . . . . Earlier procedures . . . . . . . . Limitations of the classical methyla— tion procedures. . . . . . . . . Analysis of Methylated Polysaccharides. . Principal Contributions of Methylation Analysis to Knowledge of Polysaccharide Structure . . . . . . . . . . . . . . . II. Periodate Oxidation, and Elucication of the Linkages Between and Sequence of Sugars . . . . . Contributions of Periodate Oxidation to Knowledge of the Structure of Oligo- and Polysaccharides: The Classical Approach. . . . . . . . . . . . . . . . Limitations of the Classical Approach; New Procedures. . . . . . . . . . . . . l) Bromine oxidation of the dialde- hyde groups resulting from periodate oxidation. . . . . . 2) Sodium borohydride reduction of the dialdehyde groups result- ing from periodate oxidation . Examples of Use of Smith Degradation for Elucidation of the Structure of Polysaccharides . . . . . . . . . . . . III. Determination of the Sequence of Polysaccharide Molecules by Means of Partial Depolymerization . Partial Acid Hydrolysis . . . . . . . . V Page . viii . ix . l . 5 . 5 . S . 5 . 7 . 8 . 9 . 12 . 14 . 15 0 l6 . l6 . l7 . l9 . 19 Three Examples of Use of Partial Acid Hydrolysis for Determination of the Sequence of Monosaccharides in Polysaccharides . . . . . . . . . . . . . Acetolysis. . . . . . . . . . . . . . Hydrolysis with Specific Enzymes. . . . . . a. Glycosidases . . . . . . . . . . b. Glucanases . . . . . . . . . . . c. Contributions of hydrolytic enzymes to determination of the structure of polysaccharides and other complex carbohydrates. IV. Walls and Envelopes of Blue-Green Algae . . . . . . The walls of Vegetative Cells . . . . . . . The Envelopes of Heterocysts and Spores . . MATERIAL AND mops e e O I D C O O 0 e 0 e e e e e e e e e e e Part One - Methylation Analysis of the Envelopes of Heterocysts and Spores . . . . . . . . . . . . . . . . Preparation of Envelopes. . . . . . . . . . Sugar Composition Analysis. . . . . . . . . Methylation Analysis. . . . . . . . . . . . Gas Chromatography. . . . . . . . . . . . . Mass Spectrometry . . . . . . . . . . . . . Infrared Spectroscopy . . . . . . . . . . . Part Two - The Structure of the Backbone Polysaccharides Smith Degradation . . . . . . . . . . . . a. Periodate oxidation. . . . . . . b. Sodium borohydride reduction . . c. Mild acid hydrolysis . . . . . . Analysis of the Products of Smith Degradation . . . . . . . . . . . . . . . a. Fractionation. . . . . . . . . . b. Sugar composition analysis . . . c. Methylation analysis . . . . . . Partial Acid Hydrolysis of the Backbone Polysaccharide. . . . . . . . . . . . . . Paper Electrophoresis . . . . . . . . . . . Analyses of the Oligosaccharides. . . . . . Part Three - Additional Studies on Envelope Polymers . . Extraction of the Polysaccharide from Spore Envelopes . . . . . . . . . . . . . Amino Acid Analysis of Lipid-Free Envelopes of Heterocysts and Spores . . . vi Page 23 24 27 28 32 35 36 36 38 40 40 4O 40 43 45 46 47 47 47 47 48 48 49 49 SO 50 SO SO 51 54 54 SS Page RESULTS 0 O O O O O O O O O 0 O O I O O O O O O O O O 0 O O O O 5 7 Part One - Methylation Analysis of the Envelopes of Heterocysts and Spores . . . . . . . . . . . . . . . . 57 Part Two - The Structure of the Backbone Polysaccharides 70 Part Three - Additional Studies on Envelope Polymers . . 88 Extraction of the Polysaccharide from Spore Envelopes . . . . . . . . . . . . . 88 Amino Acid Analysis of Lipid-Free Envelopes of Heterocysts and Spores . . . 92 DISCUSSION. 0 O O O O C O O I O C O O O O O O O O O O O O O O O 96 REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . 105 vii Table II III IV VI LIST OF TABLES Page Analysis of the sugar composition of the envelopes of heterocysts and spores. . . . . . . . . . . . . . . . 58 Retention times relative to T-glucose for the par- tially methylated alditol acetate components derived from the polysaccharide of the heterocyst envelope . . . 64 Mole percent composition of heterocyst and spore envelope polysaccharides . . . . . . . . . . . . . . . . 69 Summary of the analyses performed on the oligosac- charides derived by partial acid hydrolysis of the heterocyst-polysaccharide backbone . . . . . . . . . . . 81 Sugar composition analysis of the polysaccharide extracted from the spore envelope, the residue envelopes, the intact, lipid-free spore envelopes and intact, lipid-free heterocyst envelopes. . . . . . . 89 Amino acid content as weight percent of the envelopes of heterocysts and spores. . . . . . . . . . . . . . . . 93 viii Figure 10 LIST OF FIGURES Procedures for isolation and purification of hetero- cysts and heterocyst envelopes . . . . . . . . . . . . . Infrared spectrum of a 10% w/v solution, in CCl4, of the methylated polysaccharide derived from the hetero- cyst envelope. . . . . . . . . . . . . . . . . . . . . . Gas chromatograms of the partially methylated alditol acetate derivatives obtained from the polysaccharide of the heterocyst envelope . . . . . . . . . . . . . . . Mass spectral fragmentogram obtained for each of the partially methylated alditol acetate components of the heterocyst envelope polysaccharide . . . . . . . . . Fractionation, by chromatography on a column of Bio- Gel P-2, of the products of Smith degradation of heterocyst envelopes . . . . . . . . . . . . . . . . . . Gas chromatograms (.2.2.4 column) of the partially methylated alditol acetate sugars derived from the backbones of the envelope polysaccharides of hetero- cysts and spores . . . . . . . . . . . . . . . . . . . . Fractionation, by chromatography on a column of Bio- Gel P-2, of the products of the partial acid hydrolysis of the backbone of the polysaccharide from heterocyst envelopes. . . . . . . . . . . . . . . . . . . . . . . . Paper electrOphoretograms of oligosaccharides obtained by partial acid hydrolysis of the backbones from heterocyst and spore envelope polysaccharides. . . . . . Gas chromatograms (.2.2.4 column) of the partially methylated alditol acetate sugars derived from each of the trisaccharides isolated, by paper electro- phoresis, from the backbone of the heterocyst-envelope polysaccharide . . . . . . . . . . . . . . . . . . . . . Reducing-end analyses of trisaccharides Tri-I and -IIe e e e e e e e e e e e e e e e e e e e e e e e e e 0 ix Page 42 6O 62 67 72 74 76 79 83 85 Figure 11 12 13 14 Page Enzymatic hydrolysis of disaccharides Di-I, -II, and -III and trisaccharide Tri—IV (Glc+Glc+Glc) isolated following partial acid hydrolysis of the backbones of the polysaccharides from the envelopes of heterocysts and spores . . . . . . . . . . . . . . . . . . . . . . . 87 Gas chromatograms (.2.2.4 column) of the partially methylated alditol acetate sugars derived from the extracted polysaccharide from spore envelopes and from the intact, lipid-free spore envelopes. . . . . . . 9l Amino acid profiles of the heterocyst envelope and spore envelope . . . . . . . . . . . . . . . . . . . . . 95 Possible structure of the polysaccharides from the envelopes of heterocysts and spores of Anabaena cylindrica O O O O O O O O O O .0 O O O O O O O O O O O O 102 INTRODUCTION Vegetative cells of certain filamentous blue-green algae can differentiate into two morphologically and physiologically distin- guishable types of cells: spores, which are perennating structures capable of germination; and heterocysts, which are important in algal nitrogen fixation (163). Deposition of a thick envelope exterior to the wall of a vegetative cell is the most conspicuous, and can be the first, morphological sign of differentiation. The envelope surround- ing a spore is entire, whereas the envelope surrounding a heterocyst is perforated by a large pore at each point of attachment to a vege- tative cell. Geitler suggested, on the basis of morphological observations, that the processes underlying differentiation of spore and hetero- cysts are evolutionarily (38) and ontogenically (39) related. Dif— ferently stated, these algae, Anabaena cylindrica Lemm., for example, are suitable for study of the relationships between alternative differentiations in a single organism. I have approached such a study by inquiring whether there are chemical similarities between the molecules forming the envelopes of heterocysts and spores. Both envelopes contain amino compounds, lipids and carbohydrate. The spore envelope contains six times more amino compounds, as per- cent of dry weight, than does the heterocyst envelope. Glucose, mannose, galactose and xylose constitute, and are present to l 2 approximately the same percentages in, the carbohydrate portions of the two envelopes (28). The two envelopes might, therefore, have the same polysaccharide molecule as their main component. Alternatively, the same sugar residues might form structurally different polysac- charides in the two envelopes. The polysaccharides from the two envelopes can be compared further by identifying the linkages between the sugar residues. Glycosidic linkages can be identified following methylation of free hydroxyl groups, e.g., by Hakomori's method (47). Carbon atoms involved in glycosidic bonds are unaffected. Reaction with methyl iodide in dimethyl sulfoxide, with dimethyl sulfinyl anion, generally allows complete methylation in a single reaction, without degradation of the polysaccharide. The methylated polysaccharide is then hydro- lyzed to methyl sugars, which are subsequently reduced and acetylated. The volatile methylated alditol acetate derivatives which result are identified by a combination of gas chromatography and mass spectrometry. Mass spectrometry normally permits identification of the positions of the carbon atoms which are O-methylated and O—acetylated. Mass spectrometric data, when combined with the relative retention times obtained on different chromatographic columns, permit unambiguous identification of the methylated sugars (94). The analysis can be performed with 2 to 3 mg of polysaccharide material. I have applied these techniques to identify the glycosidic linkages of the polysaccharides of the envelopes of heterocysts and spores of A. cylindrica. The results indicate that 30% of the sugar residue are branched, and of these about one-third are doubly—branched. All four sugar components, glucose, mannose, galactose and xylose, 3 are found at terminal positions, whereas only glucose and mannose occupy internal portions of the molecule. The results of methylation analysis of the polysaccharides from the two types of cells were identical, within experimental error, suggesting that these polysac- charides are identical or nearly identical. However, because methyla— tion analysis does not elucidate the sequence of the sugar residues in a molecule, the polysaccharides could have had a different sequence of sugars. In order to determine the sequence of sugars in the polysac— charides, I have examined the products of the chemical degradation of the polysaccharides. I began with Smith degradation (44) because this procedure, which involves periodate oxidation, would not degrade sugars linked by C—3 which comprise about 47% of the sugar residues in the envelope polysaccharides. Sugar residues with vicinal hydroxyl groups, such as terminal sugars, or internal sugars linked by (1+2), (1+4) or (1+6) linkages, would, however, be oxidized. The dialdehyde sugar derivatives which were obtained by periodate oxidation (20,41,42) were then converted to polyalcohols by means of reduction with sodium borohydride. The linkages between the degraded and the non-oxidized sugars could then be cleaved selectively by mild acid hydrolysis (43). By this means I isolated the backbone of the polysaccharide of both heterocyst and spore envelopes. The sequence of sugars in the backbone polysaccharides was then determined by analysis of the products of partial acid hydrolysis. The anomeric configurations of the linkages in the backbone polysaccharides were identified enzymatically. 4 The backbones are identical or almost identical. They have the repeating structural unit Glc-Glc-Glc-Man. All of the sugar residues are linked by B(1+3) glycosidic bonds. The polysaccharide molecule was extracted by boiling spore envelopes with a 5M solution of sodium azide. The results of sugar composition analysis and methylation analysis indicate that the structure of the polymer is not changed by the extraction procedure. The results of analyses of the amino-containing compounds of the envelopes from the two types of cells are also reported. The purpose of these analyses was to identify the amino compounds in the envelopes, and to determine qualitative and quantitative similarities and/or differences between these compounds in the envelopes from the two types of cells. LITERATURE REVIEW I. Elucidation of the Glycosidic Linkages of Polysaccharides by Means of Methylation Analysis Rationale of the Method Earlier procedures. The most widely used procedure for identi— fication of glycosidic linkages in a polysaccharide proceeds by methyla- tion of the molecule. The procedure is based on the reaction of all free hydroxyl groups of the sugar residues with an alkylating agent in order to produce the corresponding O-methyl derivatives. Those sugar carbons lacking free hydroxyl groups because of the presence of glycosidic linkages will not be methylated. Methylation is followed by hydrolysis and by identification of the resultant partially O— methylated derivatives. Thus, if one finally obtains a 2,3,6—tri-O— methyl hexopyranose, it can be assumed that the linkage to this unit in the original polymer was through C-4 (Equation 1). The once popular methods, now largely superseded, for methyla- tion of polysaccharides were introduced by Haworth (53) and by Purdie and Irvine (117). a) Haworth's method. In this classical method, the alky— lating reagent is dimethyl sulfate in aqueous sodium hydroxide. b) Method of Purdie and Irvine. In this method, methyl iodide and silver oxide are the alkylating and base reagents, HonO 0H 0H methylo'ion H20C0H3 CHzocg-is CH203H3 0 0 CH3 CH3 CH3 CH3 CH3 CH3 n 1 hydrolysis CHZOCH3 n H H 0 CH3 CH3 2, 3, 6 - tri—O—methyl hexopyranose respectively. Polysaccharides are, in general, insoluble in organic solvents. It is therefore necessary, for this method, to methylate the polysaccharide partially by Haworth's method, and then to com- plete the methylation by dissolving the partially methylated poly— saccharide in methyl iodide. Alternatively, a mixture of methyl iodide and either methanol or acetone can be used. Methylation is then effected by adding the silver oxide catalyst. Another, but little used,\method for methylation involves treat- ing the polysaccharide with methyl iodide and methanol, in a solvent of liquid ammonia (103). Polysaccharides containing uronic acids present considerable difficulty in methylation because the uronic acid residues are degraded during the methylation procedure. One earlier method that 7 has been applied successfully involved conversion of the polysaccharide into the thallium salt, followed by methylation with methyl iodide and thallium hydroxide (33). ijrndal and co—workers (17) were able recently to methylate the capsular polysaccharide of Klebsiella type 52, which contains D-glucuronic acid, D-galactose and L-rhamnose in the ratio 1:3:2. In order to achieve methylation of the uronic-acid residues, they first reduced them to D-glucose with lithium aluminum deuteride. Subsequently, the polysaccharide was methylated by Hakomori's method which is described below. The glucuronic acid components thereupon appeared as glucose derivative deuterated at C-6. This derivative of glucose is easily identified by mass spectrometry. Limitations of the classical methylation procedures. The methods of Haworth and of Purdie and Irvine led to extensive degradation of the polysaccharide, as a consequence of the unfavorable effects of protic solvents (water). In addition, the polysaccharides were not completely methylated in a single reaction. It was necessary to remethylate the polysaccharide by the same or by a different pro- cedure. For these two reasons, a number of variations were introduced in these procedures so that they would methylate more efficiently: 1) One such variation is the method of Kuhn (146), which uses N/N— dimethyl formamide as solvent and methyl iodide as alkylating reagent in the presence of barium oxide. 2) Another variation is Hakomori's method (47), which has in recent years become the most widely used method. This method allows complete methylation of polysaccharides in a single step, without degradation. It uses dimethyl sulfoxide as solvent, dimethyl sulfinyl anion (prepared from sodium hydride and 8 dimethyl sulfoxide) as a base, and methyl iodide as alkylating reagent. This method has been modified by Sandford and Conrad (127). The methods of Kuhn and of Hakomori allow better methylation of polysaccharides because they use aprotic solvents (dimethyl formamide and dimethyl sulfoxide) which minimize the unfavorable effects of solvents during the alkylating step. They also permit the use of bases which are more efficient than hydroxide, in converting the hydroxyl groups of the polysaccharide to alkoxides, quantitatively, before addition of the alkylating reagent. Equation 2 shows the two steps of the reaction in which hydroxyl groups are methylated by the method of Hakomori. - + - + 3-SO-CH2 Na -————+ R-O Na CH3-SO-CH3 (Eq. 2) l) R-O-H + CH 2) R-o' Na+ + CH I ————+ R-O-CH + NaI 3 3 Analysis of Methylated Polysaccharides After a polysaccharide is fully methylated it is hydrolyzed to the individual monosaccharides. The depolymerization method employed should be one that minimizes demethylation or destruction of the sugar residues. Sulfuric acid has been used extensively for depoly- merization, although formic acid, hydrochloric acid, oxalic acid, acetyl bromide and methanolic hydrogen chloride have also been used. Talmadge et al. (139) used 2N trifluoroacetic acid for hydrolysis of the methylated polysaccharides of plant cell walls. Traditionally, the methylated sugars were separated from each other by means of column chromatography on cellulose, Celite, or a mixture of charcoal and Celite or by thin layer chromatography. 9 Mixtures of methylated monosaccharides can also be separated by paper chromatography and electrophoresis. More recently, ijrndal et a1. (14,15,16) developed a method for separation and analysis of the partially methylated sugars derived from permethylation and hydrolysis of oligo- and polysaccharides. In their procedure, par- tially methylated sugars are reduced and acetylated to form partially methylated alditol acetate sugar derivatives. These derivatives, being volatile, can be analyzed by a combination of gas chromatography and mass spectrometry (15,16). Each partially methylated alditol acetate derivative is identified on the basis of the pattern of fragments to which it gives rise in the mass spectrometer and on the basis of its retention time during gas chromatography. Mass-spectral and gas chromatographic analyses complement each other in this method, because the mass-spectral data permit identification of the linkages and distinguish between hexoses and pentoses, and between pyranosides and furanosides, whereas gas-chromatographic analysis permits identi- fication of specific isomeric sugars. Principal Contributions of Methylation Analysis to Knowledge of Polysaccharide Structure The general applicability of methylation analysis can be illus- trated best by means of examples of polysaccharides, the structural characteristics of which were defined by this procedure. 1. Glycogen was methylated and then hydrolyzed (54). The main methylated sugar was found to be 2,3,6-tri-O-methyl-a-D—glucopyranose, which indicated that the glucose units in glycogen were (1+4) linked. Subsequent analysis of permethylated glycogen showed that 2,3,4,6— tetra-O-methyl-D—glucopyranose and 2,3,-di-O-methyl-D-glucopyranose 10 were also present. The former of these two derivatives indicated the presence of terminal glucose (T-glucose) and the latter, the presence of (l+4)(l+6)-linked glucose (4,6-glucose). 2. Similarly, methylation analysis of starch (63) indicated the presence of 4-glucose. The presence, in addition, of 4,6-glucose proved that some glucose residues in starch have branches connected to C-6. 3. Methylation analysis combined with acetolysis demonstrated that cellulose is a B(l+4)-linked polysaccharide. The principal component was identified as 2,3,6-O-methyl-D-glucose (55,84). Under acetylating conditions, in the presence of an acid catalyst, cellulose is degraded to cellobiose octoacetate (56). The presence of cello- biose indicated that the anomeric configuration of the linkage is B. 4. A more recent contribution of methylation analysis to eluci- dation of the structure of a polysaccharide is that of Sandford and Conrad (127). They methylated the capsular polysaccharide of Aero— bacter aerogenes A3(Sl) by the method of Hakomori. The methylated sugars obtained after hydrolysis were separated by partition chroma- tography on a column of powdered cellulose, and analyzed by paper chromatography (127). This methylation analysis was subsequently complemented with partial acid hydrolysis and enzymatic cleavage of the linkages (23). By these means Conrad et a1. identified the presence of repeating unit in the polysaccharide molecule. r. ——6—D—G1cl—B—-iD——GluA—1—E—§-L—Fuc—— 4 B 1 [— D—Glc -11 ll 5. Tsusué and Fujita (141) have reported the presence in the blue-green alga, Tblgpothrix tennis, of a non-reducing glucofructan which accounts for one-fourth of the total carbohydrate content of the cells. Tsusué and Yamakawa (142) established the structure of this glucofructan by a combination of methylation analysis and specific enzymatic hydrolysis of the linkages. Methylation was achieved by the method of Hakomori, and the methylated sugars obtained after hydrolysis were trimethylsylilated and then analyzed by gas chromatography. The use of a- and B-glucosidases and a B- fructofuranosidase permitted identification of the anomeric configura- tions of some of the glycosidic linkages of the glucofructan. They proposed the following structure for these oligosaccharides: 17.2 4" Glc —— Fru ———2 Fru _ .1“ 6. A very important recent contribution of methylation analysis of polysaccharides to knowledge of cell wall structure has been that of Talmadge et al. (139), Bauer et a1. (9), and Keegstra et a1. (80). By supplementing the analytical method of Bjorndal et al. (15,16) with the use of purified hydrolytic enzymes, these authors have been able to identify, quantitate, and study the structures of the follow- ing macromolecular components of the walls of sycamore cells: neutral arabinan, neutral galactan, xyloglucan and acidic rhamnogalacturonan. 12 II. Periodate Oxidation, and Elucidation of the Linkages Between and Sequence of Sugars Periodate oxidation has been valuable for elucidation of the structures of organic compounds, especially of carbohydrates. A great many publications concerned with carbohydrate structure mention the behavior of the compound studied towards periodate. Because of the large amount of pertinent literature, I will refer in this review only to the more important contributions of periodate oxidation to the elucidation of carbohydrate structure. Periodate reacts with polysaccharides to cleave the linkages between carbons carrying vicinal hydroxyl groups. Therefore, if a carbon of a pyranose or a furanose ring is involved in glycosidic linkage, that carbon will not be oxidized (25; see Equation 3), and therefore will not consume periodate. cuzon cuzos + - 3104' (T-Sugor consumes 2 mole of 10‘ , 4-Sugor consumes I mol, 3-Sugor none) 1 EHZOHO CHZOHO cnzor‘i) Osc-H och—O—gh/MW (Eq. 3) H 310; O H-ngH (formic acid released from T-Sugar) 13 Analytical data concerning the amount of periodate consumed and the amounts of products released after oxidation with periodate there- fore provide information about glycosidic linkages between sugar residues. In a polysaccharide composed of hexopyranose residues, the sugar at the reducing end will consume four moles of periodate and yield four moles of formic acid, if it is linked through C-6 (25,60; Equation 4a). If it is linked through any other carbon atom, the reducing-end sugar will also consume approximately 4 moles of periodate, but the products will be different. One mole of periodate is consumed by the aldehyde group of C-1, which is oxidized to formic acid; one is consumed by oxidation of the C-6 hydroxyl group with release of one mole of formaldehyde; and the two other moles of periodate are consumed by oxidation of the other carbon atoms which are not involved in a glycosidic linkage (25,68; Eq. 4—b,c,d). a H H 'ta“" gm) HO QHCOgH OH (M 01,004 Ho CHo+ HCHO "H o . The next step involves heterolysis of the bond between C-1 and the exocyclic O, to give a cyclic carbonion-oxonium ion (:) which most probably exists in the half-chain conformation (:). Reaction with water then forms the protonated reducing sugar (:), from which the reducing sugar @ is derived. 21 Were polysaccharides to be hydrolyzed in a similar fashion, hydrolysis of an internal linkage would require the reorientation of an entire chain so as to form the carbonium-oxonium ion (<:), Equation 5). In analogy to the observation that the introduction of groups of increasing size at C-5 increasingly reduces the hydrolysis rate, the rate of hydrolysis of the polysaccharide would be diminished by the introduction of a large, bulky terminal group of sugars (ll). Hydrolysis of the residue at the non-reducing end (terminal sugar) would not require reorientation. It has therefore been deduced that the non-reducing end (the terminal sugar residue) should be hydrolyzed more rapidly than other sugar residues in a polysaccharide. There is no direct experimental support for this viewpoint, but support has been obtained indirectly by determining the percentage of products formed after very limited hydrolysis. For instance, the amounts of D-glucose and of small oligosaccharides liberated by hydrolysis of a given weight of starch and cellulose are greater, and the amounts of intermediate-sized products less, than predicted on the basis of the assumption of completely random hydrolysis, for any degree or time of hydrolysis (8,10,11,101,ll7,l49). As an example, in the acid hydrolysis of potato starch, using dilute hydrochloric acid, D-glucose appears first (140,158). Thus, it appears that hydrolysis does not occur by random scission and that terminal linkages are hydrolyzed more rapidly than are other linkages. The rates of hydrolysis of polysaccharides with different types of linkages seem to parallel the rates of hydrolysis of the corresponding disaccharides. For example, amylose, a-(l+4)glucan, is hydrolyzed more readily than is a dextran containing mainly 22 B-D—(l+6)-glucosidic linkages (136), and correspondingly maltose is hydrolyzed more readily than isomaltose. However, other factors may be more important. For example, in guaran, a galactomannan.consist- ing of a backbone of B(l+4)-linked mannosyl residues, with galactosyl side chains d(l+6)-linked to alternate mannosyl residues (152), the a-D-(l+6) linkages of D—galactopyranose are cleaved more readily than are the B-D-(1+4) linkages of mannopyranose (11). The a-D-(l+6) linkages are more rapidly cleaved presumably because they are at terminal positions (11). The temperature and the concentration of the acid greatly influence the rate of hydrolysis. H0116 and Sejtli (65) have found that in acid hydrolysis of amylose there is, at the same total reducing power (the same chain length), more D-glucose formed with higher concentrations of acid (0.13N HCl at 100°) than with lower concentration of acid (0.06N HCl at 100°). The conditions of hydrolysis must, therefore, be carefully chosen if a maximal amount of structural data is to be obtained by partial acid hydrolysis. Column chromatography on cellulose, Celite, charcoal, or a mixture of charcoal and Celite can be used to separate the oligo— saccharide products of hydrolysis. Today, Bio-Gel (BioRad Labora- tories) columns seem to be the most suitable for this purpose. Information about the identities of, and ratios between, the oligo- saccharides formed can then be used to help to identify the structure of the polysaccharide. 23 Three Examples of Use of Partial Acid Hydrolysis for Determination of the Sequence of Monosac- charides in Polysaccharides l. Hydrolysis of rabbit liver glycogen with 0.05N sulfuric acid at 100°C for 8 hours yielded--in addition to D-glucose--maltose, isomaltose and maltotriose, demonstrating the presence of both (1+4) and (1+6) linkages in the original polymer (157). 2. The A3(Sl) polysaccharide from.Aerobacter aerogenes was hydrolyzed with 1N sulfuric acid at 100°C for variable time (23), and the resulting oligosaccharides separated by chromatography on columns of cellulose. The presence of cellobiose and of particular aldobiuronic, -triuronic, and -tetrauronic acids indicated that the polysaccharide is constituted of the repeating tetrasaccharide unit shown on page 10. 3. A two-step partial hydrolysis of galactoglucomannan from the leaves and stem tissue of red clover (19) indicated that the polysaccharide has a backbone of randomly distributed, B-(l+4)- linked glucosyl and mannosyl residues, to which galactosyl units are a-(l+6) linked. The first hydrolysis, performed with 25 mM oxalic acid, at 100°C for 6 h, cleaved the galactosyl branches and left the glucomannan backbone intact. The second hydrolysis with 0.1N H2504 for 6 h at 100°C, led to the release of four disaccharides (G+G, G+M, M+G, M+M), eight trisaccharides, and many tetra- and pentasaccharides, including homopolymers of glucose and of mannose. Because these results indicate that all permutation of sequences of glucose and mannose were present, it seems clear that the two types of monomer were arranged in a random sequence in the backbone of the original polymer. 24 Acetolysis Acetolysis can be defined as a chemical reaction leading to cleavage of O-R linkages with concomitant formation of an acetic acid ester. Lindberg and Lemieux have each proposed a different mechanism to explain the acetylation reaction leading to the cleavage of the O-R linkages of fully acetylated aldosides. Both mechanisms have been reviewed by Lemieux (89). Lindberg proposed that the attacking acetylium anion coordinates first with the oxygen atom of the ring (Equation 6(:)), to give an intermediate which undergoes ring- opening to yield an acyclic, resonance-stabilized carbonion ion @. The latter can either cyclize to anomeric form (@ or ©) or can react to give acyclic products (see Equation 6). HgCOAc szAc HgCOAc '0 0- Ac “—— Ac Ac AcO A° AcO (Eq. 6) on cm: on o @\ o Aquanmums 25 Lemieux (89) postulated that the attacking species coordinates first with the glycosidic oxygen atom, the RO—C-l bond ((E) and (b), Equation 7). This bond is weakened in the case of trans-C-l—C—2 systems by the participation of the acetoxy group on C—2, as shown in (E). The transition state (a) can then collapse either to give the anomers (:) or (:), or to form acyclic products. There are several procedures to achieve acetolysis. The most common procedure makes use of an acetic anhydride-sulfuric acid reagent. Other mixtures of acetic anhydride and inorganic acids have also been used: acetic anhydride-perchloric acid, acetic anhydride-zinc chloride, and acetic anhydride-trifluoroacetic acid- acetic acid. All of these procedures give rise to the attacking . . + . group, acetylium anion [(CH -CO) ; Equations 6 and 7; and see 3 reference 46]. @ © 6) Ac AC I2_ '86 H ZCOAc 0 R O—R 0 0R + '——0 "’ : __A£=__. _.__. 0A6 ‘_—" : 0 GAO o\cl 0\ I 1 (Eq. 7) Me Me 0-) | k 3 Fa u) / {3... Acychc Products 26 Acetolysis can complement acid hydrolysis. In the acetolysis reaction, (1+6) linkages are the most susceptible to degradation (76,159), whereas they are the least readily ruptured by acid hydrolysis. Different patterns of oligosaccharide fragments are therefore obtained from a complex polysaccharide by the two methods. Amylopectin can be mentioned as a case in point. Wolfrom and co- workers (158) isolated 1% 05 isomaltose octoacetate by acid hydrolysis of this polysaccharide, whereas no isomaltose could be isolated after acetolysis (140). Kocourek and Ballou (82) have used controlled acetolysis extensively to study the structure of yeast mannans. Their method comprises two steps: first acetolysis by acetic anhydride-acetic acid and sulfuric acid; and second, deacetylation in dry methanol with methanolic barium methoxide. Under these deacetylation condi— tions, the acetylated derivative of (1+6)-linked mannosyl residues does not lose the acetyl group that is linked either to the oxygen atom on C-5 of the mannosyl moiety according to Lindberg's hypothesis (see Equation 6(§)) or the glycosidic oxygen according to Lemieux's hypothesis (see Equation 7<§>)° The acetylated mannosyl derivatives which remain after the deacetylation reaction form acyclic products which are very easily hydrolyzed. Deacetylated oligosaccharides thereupon separate, almost immediately. These oligosaccharides con- tain (1+2), (1+3) and (1+4) linkages between mannosyl residues, and constitute the side branches of the original polymer. Recently, Rosenfeld and Ballou (125) have studied the kinetics Of the acetolysis reactions for several a- and B-linked disaccharides. {They found that the rates of acetolysis of B—linked disaccharides 27 decrease in the following order: (1+6)>>(1+3)>(l+2)>(1+4). For the a-linked disaccharides the rates decrease in the order: (l+6)>>(l+4)> (1+3)>(l+2). For the mannosyl-mannose disaccharides the order is o-(l+6)>>a-(l+3)>B-(l+4)>d-(l+2). Hydrolysis with Specific Enzymes Enzymes which cleave glycosidic linkages facilitate structural studies of polysaccharide molecules in two general ways. First, they can be used to degrade the polysaccharide to oligosaccharide frag- ments. Second, enzymes of known specificity can be used to determine the configuration of linkages. Two classes of carbohydrate hydrolases are known: glycosidases, which cleave only terminal sugars, and glycanases, which can cleave internal glycosidic linkages of polysaccharide and oligosaccharide molecules. The enzymes of both classes are named according to the sugar moieties in the glycoside and the anomeric configuration of the glycosidic bond hydrolyzed; for instance: o-D—galactosidase, B-(l+3)-endoglucanase. The glycosidases, also called glycosyl hydrolases, can be further divided into those which transfer glycosyl residues to suitable acceptors and those which only hydrolyze terminal sugars. Because of the great number of hydrolytic enzymes that have been reported which are active on carbohydrates, I will review here only those enzymes which are relevant to my work as well as certain others which have made contributions to research on the structure of polysaccharides. 28 a. Glycosidases o-Glucosidases. The a-glucosidases are a group of enzymes catalyzing the hydrolysis and/or transfer of a-D-glucosyl residues to suitable acceptors. They may be classified in three groups according to their mode of action and to their substrate speci- ficities (106): i) a-Glucosidases which are strictly hydrolytic enzymes. Examples are: l) the amyloglucosidases; 2) the a-glucosi— dases, highly purified from Saccharomyces, which hydrolyze a number of disaccharides having a-D—glucosidic linkages (52,99,115); 3) 0-D- glucosidases from alfalfa seeds (69) and barley malt (78); and 4) an a-D-(l+6)-glucosidase, from a rumen strain of Lactobacillus bifidus, which specifically cleaves (1+6) linkages (77). ii) a-Glucosidases, such as the maltases, which hydrolyze or, alternatively, transfer the a-glucosyl residue of maltose and d-glucosides to suitable acceptors. An additional example is the a-glucosidase from Aspergillus oryzae (112). This enzyme synthesizes nigerose (glucopyranosyl-a-(1+3)-glucose) by transferring the terminal glucose from.maltose to glucose. The enzyme is also able to form a trisaccharide identified as gluco- pyranosyl-a-(1+6)-glucopyranosyl-a-(1+3)-glucose. iii) a-Glucosidases which only transfer a-D—glucosyl groups to produce oligosaccharides and polysaccharides. Examples are the o-glucosidases from human heart, liver and skeletal muscle. These enzymes hydrolyze maltose, and transfer the glucosidic group to glycogen (61). 29 B-Glucosidases. The best known and purified B-glucosidase enzyme is that which is obtained from almonds. It can be extracted from crushed, defatted almonds with water, and precipitated with alcohol (144). The enzyme can be purified first by zinc sulfate which precipitates many COntaminants of the enzyme, while the enzyme itself remains soluble. The B-glucosidase is then purified further by precipitation with tannin (144). Helferich and Kleinschmidt (59), using (diethylamino)ethyl Sephadex, were able to purify the enzyme 33-fold. The purified B-glucosidase preparation retains B-galacto- sidase activity. For a long time it was believed that the same enzyme was responsible for both activities. Today, however, on the basis of a difference in behavior of the two B-hexosidases during adsorption on polysterene B-D—glucoside (57,58), it has been con- cluded that the two hexosides are hydrolyzed by distinct enzymes (58). It is probable that the two enzymes have very similar proper- ties, and normally exist as a firm aggregate (131). Another B-glucosidase that may be mentioned is the inducible B-glucosidase of Saccharomyces (81), which has an absolute requirement for a correct position of the hydroxyl group at C-3 and C-4. B-Glucosidases with differing thermal stabilities have been found in Neurospora (98). Aspergillus niger forms several 8- glucosidases which are specific for B-(l+3)- or B—(l+4)-linked glucans (83,138). u-Mannosidases. a-Mannosidases have been found in several mammalian tissues, including liver (106), and in seeds and kernels of almonds (106). The main substrates for the almond a-mannosidase 30 are the a-D-mannopyranosides, a-D-lyxopyranosides (lyxose is a pentose in which the hydroxyl groups at C-2, C-3 and C-4 have the same configuration as in mannose), and heptopyranosides having the configuration of D—mannose at C-2, C-3 and C—4 (116). a-Mannosidase from jack beans has been very useful in studying the sequence and anomeric configuration of sugars in chains of various complex carbohydrates (93). B-Mannosidases. These enzymes occur in yeast, snails, insects, oysters, and in the pancreas and epididymis of rats. Reese and Shibata (123) have found B-mannosidase and a-galactosidase activities associated with a B-mannanase from a number of fungi grown on glucomannans and galactomannans of higher plants. Two of these fungi, Penicillium verruculosum and Penicillium ochro-chloron, produce high yields of these enzymes. The B-mannosidase, which has an absolute requirement for the B-configuration, is responsible for the hydrolysis of mannobiose and mannotriose, substrates which are not hydrolyzed by the B-mannanase enzyme. Reduced mannobiose is highly resistant to the enzyme, whereas reduced mannotriose is readily hydrolyzed. Recently a B-mannosidase has been isolated from the fruiting body of the mushroom Polyporus sulfureus. The purified enzyme is free of a-mannosidase activity (148) and can therefore be used for identification of the anomeric configuration of mannosyl residues in oligosaccharides. a-Galactosidases. a-Galactosidases are found in animals, including humans, where it is present in the thyroid, kidney, and spleen (40), and in the following plants, from which it has been 31 studied in detail: Vicia faba (26), watermelon (5), alfalfa (24), plantain (24), coffee (114). Yeast and basidiomycetes (92) are good sources of the enzyme. Li and co-workers (91) have partially purified an a-galactosidase from Diplococcus pneumoniae. The enzyme was shown to be a sulfhydryl enzyme. Intracellular a-galactosidase can be induced in E. coli and Aerobacter aerogenes (64). Most of the a-galactosidases show a high degree of specificity towards the alkyl and aryl-a—D-galactopyranosides and towards o-D- galactopyranosyl oligosaccharides of the raffinose family. The facility with which a-D-galactopyranosyl residues of oligosaccharides can be hydrolyzed decreases with increasing size of the oligosaccharides. B-Galactosidase. The B-galactosidase from E. coli is the most extensively purified and characterized enzyme of this type. This enzyme has been a valuable aid in the determination of the mechanism and specificity of glycosidase action as well as for the elucidation of the genetic regulation of enzyme synthesis. Wallenfels et al. (146) crystallized the B-galactosidase from E. coli ML 309 by extraction of lysed cells, removal of nucleic acids, and fractiona- tion of the extracts with alcohol and ammonium sulfate. The crystal- lized enzyme has a molecular weight of 518,000 (129). The enzyme is a tetramer, each subunit of which has a molecular weight of 130,000 daltons (134). All of the B-galactosidases studied have been shown to be very specific for B-D—galactosidic linkage. Inversion of the B-linkage to the a-D-configuration, or changes in the configuration of the 32 ring (e.g., conversion of the pyranose ring to furanose form) prevents enzyme action. B-Xylosidases. Some xylan-fermenting bacteria produce B-xylosidases, which catalyze the hydrolysis of a series of xylose— containing oligo- and polysaccharides, such as xylobiose, xylotriose and xylan (67). B-D—Xylosidase has also been found in the subcellular particles of rat liver, in the chick embryo and in cartilage (106). b. Glucanases a—Glucanases. The a-amylase from human salivary glands hydrolyzes a-(l+4)-glucosidic linkages in polyglucans such as amylopectin, glycogen and dextrins. The linkage that is cleaved by the enzyme is selected at random, with the exception that linkages close to the terminal ends of the polysaccharide are cleaved rela— tively slowly (12). In contrast, the B-amylase isolated and purified from potatoes (6) preferentially cleaves the penultimate glycosidic bond of its substrate. Neither of these enzymes can cleave the branch points of glycogen and amylopectin. d—(1+3)-Glucanase is an exo—splitting enzyme produced by Aspergillus nidulans. Because of the presence of its substrate in the fungus, it is normally a constitutive enzyme (165). The enzyme degrades a—(1+3) glucan to glucose, but is inactive on nigeran, a glucan with mixed o-(l+3) and a-(1+4) linkages. B-Glucanases a) Cellulases. The best known B-glucanases are the cellulases, which are B-(l+4)-glucanases. Cellulases have been 33 obtained from the snail, Helix pomatia (135), and from the fungi Aspergillus orgzae, Streptomyces sp. and Myrothecium verrucaria (126). Most of the cellulases cleave terminal as well as internal residues, and yield great amounts of cellobiose. The enzyme from Streptomyces sp. also produces much of cellotriose. Reese et al. (124) have found that the activity of cellulase falls off sharply as the substrate decreases in length from six to two sugar residues. The partially purified cellulase from Streptomyces sp. QMB 814, unlike other known cellulases, is free of oligosaccharidase activity. The formation of cellotriose and cellobiose by this enzyme is there- fore due to its random endosaccharidase activity. b) B-Glucanases other than cellulases. Glucanases are known with specificity for B-(1+2), B-(l+6) and B-(1+3) linkages. B-(l+2)-Glucanases isolated from the crown gall-inducing bacterium Agrobacterium tumefaciens (Wisconsin A-6) and related organisms (100) cleave B-(l+2) linkages at random. B-(l+2)-Glucanase can be induced in Penicillium sp. and Aspergillus quadricinctus when they are grown on B-(l+2) glucan. These enzymes break the internal linkages of B-(l+2) glucan releasing sophorose (glucopyranosyl-B-(l+2)-glucose) and other B-(l+2)-linked oligosaccharides (121). 8-(l+6)-Glucanases cleave lutean and pustulan, the best known B-(l+6) glucans. Lutean is produced as the malonic ester by certain fungi, including Penicillium luteum and P. aculeatum (3). Pustulan is found in the lichen Umbilicaria pustulata (95). The enzyme yields B-(l+6) oligosaccharides of varying chain lengths. 34 B-(l+3)-Glucanases cleave B-(1+3) glucosidic linkages from B-(1+3) glucans such as laminaran. The B-(1+3) exoglucanase from Basidiomycete QM806 (120) degrades B-(1+3) glucans and related oligosaccharides from the terminal (non-reducing) end of the chain. It also degrades B-(l+3) glucans having B-(l+6)-glucosidic side groups on every third unit of the chain, to glucose and gentobiose (glucosyl-B-(1+6)-glucose, ref. 75). The B-(l+3)-endoglucanases from RhizoPus arrhizus (120) and Bacillus circulan (121) act on B-(l+3) glucans in a random fashion to give a series of oligosac- charides. The Rhizopus enzyme also cleaves B-(1+4) oligosaccharides. All B-(l+3) oligosaccharides are susceptible to this enzyme, with the exception of laminaribiose (123). Among the enzymes which can cleave glucans containing a mixture of B-(l+3) and B-(l+4) linkages are the lichenases. Most of the lichenases, however, are cellulases, B-(1+3) endoglucanases or a mixture of both. A specific lichenase which has no action on cellulose or on B-(l+3) glucan has been prepared from a commercial bacterial amylase (Novo Terapeutisk Lab., Copenhagen, Denmark) and from a diastase (N°57, Rohn Haas Co., Philadelphia, Pennsylvania) (122). The predominant product released from lichenin by these enzymes is B-(1+3) cellobiosyl-glucose. The enzymes appear to have an affinity for a trimeric portion of the lichenin molecule, and split the adjacent B-(1+4) linkage (122). 35 c. Contributions of hydrolytic enzymes to determination of the structure of polysaccharides and other complex carbohydrates 1. The a-mannosidase from jack beans, which has been puri- fied extensively by Li and co-workers (93), has been used to determine the anomeric configuration of mannosyl residues attached to several animal proteins. The enzyme releases mannose from ovomucoid, oroso- mucoid, and ovoalbumin suggesting the presence of o-mannosidic linkages in these proteins (90). 2. By sequentially applying a purified a-galactosidase from figs, Hakomori and co-workers (48) sequenced the carbohydrate moiety of two ceremide trihexosides (glycolipids) from human erythrocytes and hamster fibroblasts. The enzymatic hydrolysis was complemented with methylation analysis. The structures were found to be galacto- pyranosyl-a-(1+4)-galactopyranosyl-B-(l+4)-gluc0pyranosyl ceremide. 3. Bauer et a1. (9) have also combined enzymatic hydrolysis with methylation analysis in order to sequence a hemicellulosic xyloglucan extracted from walls of cultured sycamore cells. They used a partially purified endoglucanase, obtained from the fungus Trichoderma viride, which cleaves 8-(l+4)-glucosidic linkages. The oligosaccharide fragments released by the enzyme were subjected to methylation analysis. It was concluded that the xyloglucan has a repeating heptasaccharide unit which consists of 4 residues of B-(1+4) glucose and 3 residues of terminal xylose. 4. The fine structure of the glycogen of the blue-green alga Anacystis nidulans has been investigated recently (149). After 36 selective hydrolysis of all of the o-(l+6) linkages by an isoamylase from Pseudomonas amyloderamose, the resulting mixture of linear chains was subjected to gel permeation chromatography on Acrylex P-lO. The average chain length was 8 sugar residues. These results indicated that the structure of the blue-green algal glycogen is more similar to that of phytoglycogen from sweet corn than to bac- terial glycogen which has a chain length of 14 sugar residues as determined by the same methods. IV. Walls and Envelgpes of Blue-Green Algae The Walls of Vegetative Cells The cells of blue-green algae, like the cells of gram negative bacteria, have a multilayered cell wall (2,37,79,105,107), the total thickness of which is 40 to 55 nm. The layers of this wall, from inside to outside, have been named by Jost (79) as LI, LII' LIII' and LIV' Layer L is an electron transparent layer, the width of which I varies from cell to cell and from one region to another within the same cell (2). Allen (2) has suggested that this layer is an artifact, formed by retraction of the protoplast during the prepara— tion of the sample for electron microscopy. The next layer, L , is electron opaque. It is the lysozyme- II sensitive layer (74,96) and is therefore presumed to be constituted of peptidoglycan (murein). Frank et a1. (37), who first analyzed the composition of the peptidoglycan-containing layer of the wall, found that muramic acid and glucosamine are present in the 1:1 ratio common to the backbones of bacterial peptidoglycans. Diaminopimelic acid, 37 alanine, glutamic acid, and aspartic acid, the amino acid components of this layer, are also present in the peptidoglycan from bacterial walls. Layer L is an electron transparent layer which appears to III be made up of a parallel array of fibrils of 6 to 9 mm diameter (50). The fibrils are sensitive to proteolytic enzymes but resistant to other hydrolytic agents, suggesting that they are composed of proteins (49). The LIV layer, approximately 75-80 A thick, is the external layer of the wall. Under the electron microscope it has the appear— ance of either a single dark line (30) or of two dark lines spaced apart (79). A layer of similar appearance and location has been observed in gram negative bacteria and identified as the lipopoly— saccharide layer (107). A lipopolysaccharide of very similar compo- sition has been extracted from the walls of the blue-green alga Anacystis nidulans with phenol-water (151). It contains mannose, glucose, galactose, rhamnose, 2-keto-3-deoxyoctonic acid and 2~amino~ 2-deoxyheptose, as does lipopolysaccharide from bacterial walls. Weckesser et al. (150) have identified for the first time an O- antigenic component of a lipopolysaccharide extracted from a blue- green alga (Anabaena variabilis). The polysaccharide moiety of this O-antigen consists of L-rhamnose, its 3-O-methyl ester L-acofriose, D-mannose, D-glucose, and D-galactose. The only amino sugar present is D—glucosamine. L-Glycero-D-mannoheptose and 2-keto-3-deoxyoctonic acid, present in the O-antigen of the enteric bacteria, are absent from the lipopolysaccharide of A. variabilis. However, the O-antigen behaves, in serological tests, like a bacterial O—antigen. 38 The walls of many blue-green algae are surrounded by mucilage. The mucilage from several species of blue-green algae has been analyzed chemically (13,28,66,102,110,lll). The polysaccharide portion of the mucilage of Anabaena cylindrica contains 47% glucose, 25% mannose, 6% galactose, 21% xylose and traces of fucose. Muci- lages from some other species seem to have uronic acids as components (66,110,111), although these are absent from Anabaena cylindrica. Certain mucilages of blue-green algae are diffluent whereas other mucilages are relatively denser. However, the chemical difference between the two is not known. The Envelopes of Heterocysts and Spores External to the 4—layered wall, a thick envelope is present in some filamentous blue-green algae in the specialized cells known as heterocysts and spores. The heterocyst envelope, as seen with the electron microscope, consists of a peripheral "fibrous" region, a thick homogeneous layer underlying it, and a laminated layer which borders the cell wall (87). These envelope layers appear to surround the heterocyst com- pletely except at their junction to vegetative cells. All around this junction, the laminated layer becomes very thick and, together with the homogeneous layer, forms a thick bottleneck-like structure surrounding the region of attachment (87). No sharp boundary sepa- rates the homogeneous layer from the fibrillar layer; these two layers may be made up of the same material (87,32). The laminated layer of the envelope consists of four glyco- lipids (85,156) which are characteristic of heterocysts (104,107, 39 164). The ehcmical structure of these lipids was elucidated by Nichols and co-workers (18) and by Lambein and Wolk (85). The sugar moieties of the glycolipids are glucose and galactose. Because the lipid content of the enveloPe corresponds to the laminated layer (156), the carbohydrate portion must correspond to the fibrous and homogeneous layers. ,The composition of the carbohydrate portion was studied by Dunn and Wolk (28), who reported that it contains 73% glucose, 21% mannose, 4% xylose and 3% galactose. The spore envelope surrounds the spore completely and, unlike the heterocyst envelope, is not organized in layers. However, electron transparent and electron dense regions can be clearly distinguished in the envelopes of spores of Cylindrospermum sp. (22). Dunn and WOlk (28) also performed a chemical analysis on the envelOpes of spores of Anabaena cylindrica. They found 41% carbo- hydrate, 24% amino compounds, 1% lipids and 2% ash. The composition of the carbohydrate portion was reported as 76% glucose, 17% mannose, 4% xylose, and 3% galactose, a composition which is very similar to the composition of the carbohydrate portion of the heterocyst envelope. MATERIAL AND METHODS Part One - Methylation Analysis of the Envelopes of Heterocysts and Spores Preparation of Envelopes Anabaena cylindrica was grown in photosynthetic fermentors as described previously (155). Filaments used for analysis of hetero- cyst envelopes were collected twice weekly, with subsequent additions of fresh medium to the fermentors, whereas spores were collected after 3 weeks of culture without intervening addition of fresh medium. Filaments were concentrated by centrifugation at 48,200 x g (Sorvall RC-ZB centrifuge with KSB attachment, Ivan Sorvall, Inc., Norwalk, Conn.). Vegetative cells were broken by cavitation, at setting no. 3 of a Model S-125 Sonifier (Heat Systems Co., Melville, N.Y.). Pub- lished techniques were used for the isolation of heterocysts (161) and spores (164). Envelopes were isolated as described by Dunn and WOlk (28). The isolated envelopes were extracted with a mixture of chloroform:methanol (2:1, v/v), lyophilized, and maintained in vacuum desiccator at 60°C until used. A summary of the procedure for iso— lating heterocysts and heterocyst envelopes is shown in Figure 1. Sugar Composition Analysis Sugar composition was determined by gas chromatography of the alditol acetate derivatives as follows: 2 mg of envelopes were 40 41 Figure 1. Procedures for isolation and purification of hetero- cysts (161) and heterocyst envelopes (28). a. Sonication of intact filaments using a model S-125 Sonifier. b. Heterocysts (H) and vegetative cells hnfi are separated from spores (S) by centrifugation in a cesium chloride (CsCl) solu- tion of density (p) 1.4, at 16,300 g for 20 minutes. c. The heterocysts are separated from vegetative cells by centri- fugation in a CsCl solution of density 1.3 at 16,300 g for 20 minutes. d. The purified heterocysts are suspended in distilled water and broken in a cooler Bflhler cell-mill for 60-90 minutes with use of 0.5 mm glass beads. e. Envelopes (E) and intact heterocysts are separated from membranes (M) and cell walls by suspension in and centrifuga— tion in 1M NaCl at 48,200 g for 20 minutes. f. The envelopes are separated from intact heterocysts by cen- trifugation in a CsCl solution of density 1.4 at 16,300 g for 20 minutes. 9. Lipids are extracted from heterocyst envelopes with a mixture of chloroform and methanol (2:1, v/v). h. Purified, lipid-free heterocyst envelopes are obtained. The procedures for isolating spores (164) differed from the above procedures in the following ways: after sonication of the fila- ments, spores were separated from cell fragments by centrifugation in a CsCl solution of density 1.55 at 16,300 g for 20 minutes. A second centrifugation at 16,300 g for 20 minutes, in a CsCl solu- tion of density 1.44, separates spores (which sediment) from heterocysts. Spore envelopes are obtained in the same way that heterocyst envelopes are obtained except that because 60-90 minutes of treatment in the cell mill breaks essentially 100% of the spores (28) there is no need to remove intact spores from the resulting suspension of fragments. 42 M l Sonication n P L4 ®+@/s~~\ Supernatant 1cm: P l.3 pi" Pellet ® Supernatant m V peuet cell mill Efi E NaCl IM Solution /Spin \ Ea/ 4% £34 CsCl /°l.4 S in R" \ a E \H/ (I Pellet Supernatant ‘CM‘ 'MOOH 2 l Envelopes Lipid Free Figure 1 43 hydrolyzed with 1 ml of 2N trifluoroacetic acid (TFA) in sealed vials at 121°C for 60 minutes. The sugars were converted to alditol acetates (l). The final acetic anhydride solution containing the mixture of alditol acetate was injected directly into the .2.2.4 column that is described below. Chromatography was performed with a Perkin Elmer 900 gas chromatograph (Perkin-Elmer Corp., Norwalk, Conn.) temperature programmed at l.5°C/min from 130°-180°C, with a helium flow rate of 60 ml/min. Methylation Analysis Three milligrams of dried heterocyst envelopes and 5 mg of dried spore envelopes were methylated by Hakomori's method (47) as described by Sandford and Conrad (127). The methylated envelopes were extracted with a mixture of chloroform:methanol (1:1, v/v). The residual material was dialyzed against water, dried, remethylated, and extracted with chloroform/ methanol. The chloroform/methanol extracts were concentrated to a volume of 2 ml and separated from water-soluble reagents of low molecular weight either by dialysis against several changes of dis- tilled water for 48 h or by passage through a Sephadex LHZO column (1.3 cm i.d. x 70 cm; Pharmacia Fine Chemicals, Inc., Piscataway, N.J.). In the latter case, fractions of 1.5 ml were collected, and those which were positive to the anthrone reagent (27) were pooled and dried in air at 60°C. The dried methylated polysaccharide was hydrolyzed with 1 ml of 2N TFA for 75-80 minutes at 121°C in sealed tubes, and the partially methylated aldoses were converted to partially methylated 44 alditol acetate sugar derivatives (1). For mass spectrometry the partially methylated aldoses were reduced with NaBD (Merck, Sharp 4 and Dohme of Canada Limited, Montreal, Canada) instead of NaBH4. The presence of deuterium in the anomeric carbon provides a dis- tinguishable difference between the fragmentation patterns derived from C-3 and C-4 linked hexoses (94). The qualitative identification of each partially methylated alditol acetate component shown by gas chromatography was made from the following information: (i) the position of methoxy groups in the methylated alditol acetate as obtained by mass spectra analysis (94); (ii) the sugar composition of the unmethylated envelopes (l); and (iii) the relative retention times of the components on 3 chromatographic columns (see below). The three columns were cali- brated, using known standards, against retention times relative to terminal glucose (94,139) for the same columns and the same tempera— ture programs. Standard partially methylated alditol acetate derivatives were prepared from sycamore cell walls (139), xyloglucan (9), maltose (Pfanstiehl Laboratories, Inc., Waukegan, 111.), arabinogalactan (Pfaltz and Bauer, Inc., Flushing, N.Y.) and yeast mannan (Sigma Chemical Co., St. Louis, Mo.). The positions of peaks derived from the envelope polysaccharides were then compared by linear interpolation between the positions of the standards, with published retention times of partially methylated alditol acetate derivatives (94,139). I denote each partially methylated alditol acetate derivative by indicating the linkages (in addition to C-1) by which the corresponding glycosyl residue of the original sample was connected to other sugars. Thus, I write 2,3-mannose instead of 45 1,2,3,5-tetra-O—acetyl-4,6-di-O-methyl mannitol. Partially methylated alditol acetate derivatives which are obtained from sugar residues that occupy non-reducing terminal positions in the polysaccharide, that is, which are glycosidically linked to other sugar residues only at C-1, are denoted by T- (e.g., T-glucose = 1,5 di-O—acetyl 2,3,4,6-tetra-O-methyl glucitol). Quantitative analysis was performed by integration (Autolab Computing Integrator, System IV, Spectra-Physics, Santa Clara, Calif.) of the gas chromatographic peaks. So as to express the data as mole percent of the recovered carbohydrate, each peak area was divided by the molecular weight of the appropriate, partially methylated alditol acetate derivative and the sum of these ratios was normalized to 100. Gas Chromatography Gas chromatography, when not combined with mass spectrometry, was performed in the Perkin Elmer 900 Gas Chromatograph with helium as carrier. Acetic anhydride solutions containing a mixture of partially methylated alditol acetate derivatives were injected directly into the chromatographic columns. Three different glass, gas-chromatographic columns (2 mm i.d. x 1.22 m) were used. They contained, on a stationary phase on Gas Chrom Q (80-100 mesh): (i) three percent ECNSS-M, a nitrile silicone-polyester co-polymer (Applied Science Laboratories, Inc., Ann Arbor, Mich.); chromatography on this column was performed isothermally at 155°C with a helium flow rate of 60 ml/min; (ii) two-tenths percent poly(ethylene glycol adipate), 0.2% poly(ethylene glycol succinate), and 0.4% silicon GE-XE60; this column, denoted 46 .2.2.4, was temperature programmed at 1°C/min from 110° to 180°C, with a helium flow rate of 60 ml/min; (iii) three percent silicone polymer OV-225 containing methyl, phenyl and cyanopropyl groups (Applied Science Laboratories, Inc.). Chromatography was performed isothermally at 170°C, with a helium flow rate of 60 ml/min. The latter program, which resolved branched sugars and also provided sharp peaks with approximately equal base widths, was the only non- isothermal program which provided closely reproducible relative peak areas. Accurate quantification of the branched sugars is important, because the total amount of branches should balance the terminal sugars. Mass Spectrometry Combined gas chromatography-mass spectrometry (GC-MS) was per- formed in an LKB-9000 GC-MS (LKB Instruments, Inc., Rockville, Md.) interfaced with a PDP 8/I computer. Glass, tubular columns of 2 mm i.d. and 1.22 m length containing either 3% OV-225 on Gas Chrom Q (100-120 mesh) or 3% SP-240l, a trifluoropropyl silicone (Supelco, Inc., Bellefonte, Pa.) on Supelcoport (100-120 mesh) were used. The .2.2.4 column previously described was also used in order to identify 3-mannose as 3-hexose. Oven temperature was rapidly increased from 110° to 170°C immediately after the 3-mannose peak was recorded. In all of the GC-MS chromatography, the carrier gas was helium at a flow rate of 25-30 ml/min. The temperature program for the OV-225 and .2.2.4 columns was as described above. The SP-240l column was temperature programmed at 3°/min from 160° to 230°C. The ion source was at 240°C, and the scanning voltage was 70 e.v. Background was subtracted, and the data prepared in the form of bar 47 graphs or fragmentograms, by means of a computerized data system (137). Infrared Spectroscopy Infrared spectra were recorded on a Perkin-Elmer 621 Spectro- photometer using a 0.025 mm KBr demountable cell. The sample for infrared spectroscopy was a 10% solution of permethylated heterocyst or spore envelopes in CC14. Part Two - The Structure of the Backbone Polysaccharides Smith Degradation a) Periodate oxidation. Sixty milligrams of lipid-free hetero- cyst or spore envelopes (obtained as described in Part One of Material and Methods) were oxidized at 5°C, in complete darkness, using 160 mg of sodium metaperiodate dissolved in 50 ml of sodium acetate buffer (0.1M, pH 5). The rate of the reaction was determined by measuring the absorption at 223 nm, the absorption maximum for periodate (45), with a DB-G spectrophotometer (Beckman Instruments, Inc., Irvine, Calif.). A solution containing 160 mg of sodium metaperiodate in 50 ml of sodium acetate buffer was used as blank. For these measure- ments, 10 ul aliquots of the reaction mixture (freed of particulate matter by centrifugation) were taken at regular time intervals and diluted to 2.5 ml with distilled water. The reaction was run for 9 days or, after the first occasion, for 6 days. Periodate remaining in the reaction mixture was destroyed by the addition of a few drops 48 of a 0.1M solution of sodium arsenite (NaAsOZ), after which no reaction could be obtained with KI-starch. The reaction mixture was then dialyzed against several changes of distilled water, and the volume reduced to 60 ml with a rotary evaporator at 40°C. b) Sodium borohydride reduction. Ninety milligrams of NaBH4 were stirred with 60 ml of the above reaction mixture for 10 h at room temperature (ca. 23°C), after which 1N HCl was added dropwise to destroy excess NaBH4. c) Mild acid hydrolysis. The pH of the solution was adjusted to 0.5 with HCl and the solution stirred for 6-8 h at room tempera- ture. After hydrolysis, the solution was centrifuged at 1000 g and a small pellet (weight, about 2 mg) was separated from a clear yellow supernatant liquid. The pellet was washed with 0.5 ml of 0.05N acetic acid 2 or 3 times by centrifugation. The combined supernatant liquids, but not the pellet, reacted positively to the anthrone reagent (27), which shows the presence of sugars. The pellet was then discarded. The combined supernatant fluids were concentrated to 5 ml in a rotary evaporator, and passed through anion and cation exchange columns in order to remove C1-, Na+, and other ions present. The anion exchange resin was AG3-X4, 100-200 mesh, chloride form (BioRad Laboratories, Richmond, Calif.), converted to hydroxyl form with 0.5N NaOH solution. The cation exchange resin was Dowex 50W-X8, 100-200 mesh, hydrogen form (BioRad Laboratories). These columns were packed in glass tubes 1.2 cm i.d. and 15 cm long. Samples were eluted with distilled water, and fractions of 2 ml were collected. 49 Fractions reacting positively with anthrone were pooled, and concen- trated to a volume of 5 ml. Analysis of the Products of Smith Degradation a) Fractionation. A l or 2—ml aliquot of the above solution was passed through a Bio-Gel P-2 column, 400 mesh (BioRad Labora- tories; 1.3 cm i.d. x 100 cm long), maintained at 50°C, in order to separate mono- or oligosaccharides present and of molecular weight less than 1800 daltons. The sample was eluted with distilled water (7 ml/h). Fractions of 1.5 ml were collected and tested for the presence of sugars with the anthrone reagent. The fractions that reacted positively with the reagent were pooled. A solution con— taining 2 mg, each, of glucose, maltose, raffinose, and stachyose was used as standard in order to define the fractions corresponding to particular molecular weights. The void volume of the column was determined with arabinogalactan (Pfaltz and Bauer, Inc., Flushing, N.Y.). Because all of the sugar content of the reaction mixture came in the void volume of the column, the steps using ion exchange resins and gel filtration were subsequently replaced with dialysis against distilled water. The fractions collected from the Bio-Gel P-2 column, following pooling, or--alternatively--dia1yzed portions, were dried in a rotary evaporator. The dried material was washed with absolute methanol and evaporated to complete dryness at 60°C several times in order to remove the borate ions present as a result of the NaBH reduction. The dried polysaccharide was resuspended in 4 2 ml of distilled water and lyophilized. 50 b) Sugar composition analyeis. Sugar composition analysis of this polysaccharide material was performed by gas chromatography of the alditol acetate derivatives using the ".2.2.4" column (1). c) Methylation analysis. Five milligrams of polysaccharide obtained by Smith degradation was methylated (47,127) and analyzed (94) by the procedures described in Part One of Material and Methods. Gas chromatographic and mass spectral analyses were performed using the .2.2.4 and SP-240l columns, respectively. Partial Acid Hydrolysis of the Backbone Polysaccharide Fifteen milligrams of backbone were hydrolyzed with 1 ml of either 0.5N TFA (trifluoroacetic acid) or 0.5N H2804 at 100°C for 30 min in sealed vials. The hydrolysis was stopped by neutralization of the solution with 5N NaOH. The neutralized solution was passed through the Bio-Gel P-2 column as described above, except that l-ml fractions were collected. The tubes corresponding to oligosaccharide fractions of the same molecular weight were pooled, dried at 60°C, and stored in vacuo at 60°C. Paper Electrophoresis Different oligosaccharides of the same molecular weight were separated on Whatman 3MM paper by means of an E-C 453 Pressure Plate Paper Electrophoresis Apparatus (Milton Roy Company, St. Petersburg, Florida), at 20 V/cm. The electrophoresis apparatus was connected to a Lauda K-2/R water circulating pump (Lauda Instruments Division, Brinkmann Instruments, Inc., Westbury, N.Y.) maintained at 5°C. The buffer was 0.2M sodium borate, pH 10. For qualitative analyses, the 51 paper was cut in strips 6.3 cm wide x 60 cm long. For quantitative analysis and recovery of oligosaccharides, the paper was cut in sheets 20 cm wide and 60 cm long. The paper was equilibrated with buffer solution for l h before application of samples. Twenty-five microliters of each sample were applied on the 6.3 x 60 cm strips and 250 pl on the 20 x 60 cm sheets. The duration of electrophoresis was 6 h for the disaccharides, 10 h for the trisaccharides, and 14 h for the tetrasaccharides. Oligosaccharides were detected on the paper strips by means of silver nitrate-alcoholic sodium hydroxide (166). Oligosaccharides localized by this means (reaction at the margins of the 20 x 60 cm sheets) were eluted with distilled water for 24 h. The solution containing an oligosaccharide was concentrated to 1-2 ml and passed through the anion and cation exchange columns described previously. The solution was then dried and washed several times with methanol at 60°C for removal of borate ions. Analyses of the Oligosaccharides Four different analyses were performed in order to identify and to sequence oligosaccharides: (i) Sugar composition analysis, by gas chromatography of the alditol-acetate derivatives, using the .2.2.4 column (1); (ii) Methylation analysis, by gas chromatography of the par- tially methylated alditol acetate derivatives, using the same column (94,139); (iii) Reducing-end analysis, performed by reduction of the oligosaccharide with 1-2 mg of NaB3H (185 mCi/mmol, New England 4 52 Nuclear, Boston, Mass.) in 1 m1 of 1N NH40H solution. After 1 h of reduction the excess of NaB3H4 was destroyed with glacial acetic acid and the oligosaccharide dried, and washed and dried several times with absolute methanol at 60°C. The tritiated oligosaccharide was then hydrolyzed with 1 m1 of 2N TFA for l h at 121°C in a sealed vial. The products of hydrolysis were reduced with NaBH4 in 1N NH4OH. The resulting sugar alcohols were separated by paper electro- phoresis in 0.05M sodium borate solution, pH 9.2, at 20 V/cm (166), for 6-8 h. Twenty-five microliters of the sample solution (concen- tration ca. 1 mg/ml) were applied to 6.3 x 60 cm strips of paper (Whatman 3MM) previously equilibrated for l h with the buffer solu- tion. The sugar alcohols were detected with silver nitrate-alcoholic sodium hydroxide reagent modified by the addition of 4% pentaerythritol to the final 0.5N NaOH alcoholic spray (166). The strip of paper was developed at the margins and the remainder of the paper was scanned for radioactivity in a model 7200 Radiochromatogram Scanner (Packard Instrument Co., Downers Grove, 111.). (iv) Analysis of the anomeric configuration of (1+3) linkages, performed by enzymatic hydrolysis of the three disaccharides, and the trisaccharide G1c+Glc+Glc, obtained by partial acid hydrolysis of the backbones from the polysaccharides of heterocyst and spore envelopes. The enzymes and reaction conditions used were: (a) a-Glucosidase partially purified from Saccharomgces cerevesiae (51; obtained from Sigma Chemical Co., St. Louis, Mo.). This enzyme was used in a 0.1M sodium phosphate solution, pH 6.8, at 37°C. The enzyme was shown to be active on maltose (Pfanstiehl S3 Laboratories, Inc., waukegan, Ill.) and maltotriose (Aldrich Chemical, Inc., Milwaukee, Wisc.), and inactive on cellobiose (Eastman Organic Chemicals, Rochester, N.Y.). (b) B-Glucosidase purified from almonds (62; obtained from Sigma Chemical Co.), used in 0.1M sodium acetate, pH 5, at 37°C. The enzyme was shown to be active on cellobiose and inactive on maltose and maltotriose. (c) a-Mannosidase purified from jack bean meal (93). The enzyme was used in a 0.05M solution of sodium acetate, pH 4.5, at 37°C. The enzyme was shown to be active on p-nitrophenyl—a—D- mannopyranoside and mannosyl-a-(l+2)-mannose and inactive on p-nitro- phenyl-B-D-mannopyranoside. The enzyme and the nitrophenyl manno- pyranosides were a generous gift of Dr. Yu-Teh Li (Tulane University, New Orleans, La.). The mannosyl-mannose was generously provided by Dr. Clinton E. Ballou (University of California, Berkeley, Calif.). (d) B-Mannosidase purifed from Polgporus sulfureus (148), also from Dr. Li. This enzyme was used in a 0.05M solution of glycine-HCl, pH 3, at 37°C. The enzyme was shown to be active on p—nitrophenyl-B-D—mannopyranoside and inactive on p-nitrophenyl—a-D— mannopyranoside and mannosyl-a-(1+2)-mannose. (e,f) B-Mannosidase-containing B-mannanases 5339Q from Penicillium ochro—chloron and S339K from Penicillium verruculosum (123), obtained as a generous gift from Dr. E. T. Reese (U.S. Army Natick Laboratories, Natick, Mass.). These enzymes were used in 0.05M sodium acetate solution, pH 4.5, at 50°C. The enzymes were shown to be active on p—nitrophenyl-B-D-mannopyranoside and inactive on p-nitrophenyl-a-D—mannopyranoside. 54 (g) B(1+3)Endoglucanase Sl76N partially purified from Rhizopus arrhizus QM1032 (121), also from Dr. Reese. The enzyme was used at 50°C, in 0.01M sodium acetate solution, pH 5. One-half milligram of enzyme (except for the mannosidases, which were provided by Dr. Li as solutions) and 1 mg of Glc+Glc, Glc+Man, Man+Glc or Glc+Glc+Glc, obtained by partial acid hydrolysis of back- bone, were dissolved in 1 ml of the appropriate buffer. The sample was shaken for 24 h in a water bath at the temperature indicated above. The course of the enzyme reaction was assayed as follows. After 0, 3, 6 and 24 h of incubation, lO-ul aliquots of the reaction mixture were diluted with 0.99 ml of distilled water and tested by means of the Nelson test (21) for reducing-end equivalent released. Laminaribiose(glucosyl-B-(1+3)-glucose) and nigerose (glucosyl- G-(l+3)-glucose) were generously provided by Dr. Reese, and malto- pentaose by Dr. A. Kivilaan (Michigan State University). Part Three - Additional Studies on Envelope Polymers Extraction of the Polysaccharide from Spore Envelopes The polysaccharide was extracted by boiling 45 mg of lipid-free spore envelopes with a SM solution of sodium azide (NaN3) for about 5 minutes. The suspension was cooled, and centrifuged at 1000 x g. In this way, a pellet was separated from a clear yellow supernatant fluid. The pellet was washed 2-3 times with 1 ml of water, centri- fuged and the water added to the clear yellow supernatant fluid. 55 The combined supernatant fluids were dialyzed for 24 h against 3 changes of distilled water, concentrated to 5 ml in a rotary evapora- tor, and lyophilized. The pellet was dried at 60°C in a vacuum desiccator. Both the pellet and the extracted lyophilized material were maintained under vacuum at 60°C for further analysis. One milligram of the pellet material and 1 mg of the extracted polysaccharide were analyzed for sugar composition by gas chromatography of their alditol acetate derivatives (1). One milligram of hetero— cyst and spore envelopes (lipid free) were also analyzed for sugar composition, as control samples. Three milligrams of the extracted polysaccharide were methylated (47) and analyzed (94), as has been previously described in Part One of Material and Methods. The .2.2.4 column was used for all of the gas chromatographic analyses. Amino Acid Analysis of Lipid—Free Envelopes of Heterocysts and Spores Three milligrams of heterocyst and spore envelopes (lipid free) were hydrolyzed with 3N, and a separate 3 mg with 6N, HCl in sealed vials flushed with nitrogen gas. The hydrolysis with 3N HCl was performed for 4 h at 108°C in order to detect amino sugars present in the envelopes. The hydrolysis with 6N HCl was performed for 18 h at 108°C in order to achieve optimal release of amino acids. After hydrolysis the solutions were dried at 50°C under nitrogen, and the residues resuspended in citrate buffer, pH 2.87. Amino acid analysis was performed with an automatic Technicon Amino Acid Analyzer (Technicon Instruments Corporation, Tarrytown, N.Y.) by the 56 accelerated method using Chromobeads C—2 (128). The peaks were integrated by the Autolab System IV Computing Integrator. Each amino acid was identified on the basis of its retention time rela- tive to the internal standard, norleucine. Correction factors cor- responding to the color reactions of the individual amino acids with ninhydrin were applied automatically. RESULTS Part One - Methylation Analyeis of the Envelopes of Heterocysts and Spores Both heterocyst and spore envelopes have the same sugar components and in the same amounts (Table I). My results are very close to those reported by Dunn and Wolk (28), except that the amount of galactose found here is higher (6.3 :_1.6% as compared with 3 i'l% reported previously). The methylated polysaccharide was separated into chloroform/ methanol soluble and chloroform/methanol insoluble portions. The soluble portion accounted for two-thirds of the total polysaccharide content of the envelopes as estimated by the anthrone test. All of the sugar content of the insoluble portion became soluble in chloroform/ methanol when remethylated, and in terms of methylation analysis was identical to the portion soluble in chloroform/methanol after the first methylation. Infrared spectroscopy of the methylated polysac- charide showed no peak at 3 um (Figure 2). The ECNSS-M and OV-225 columns resolve 9 of the 11 partially methylated alditol acetate components of the envelopes (Figure 3a,c). The .2.2.4 column partially resolves two peaks at relative retention times of 1.33 and 1.35 (peaks #4a and 4b, Figure 3b; Table II). The SP-2401 column partially resolves two other peaks in the region of terminal (T-) hexoses (peaks #2a and 2b, Figure 3d). 57 58 «v.2. eds an H. as 2 H 23 $828 3 H as 3 H as $2.52 3 H 3 3 H as 398.8 3 h em 3 H as Hoax 28m seems: 8528329. 8285. .6 5:52.58 .1. 222 _ Ems .mmoOHa>cm mo mcofluouommum ucaHOMMHo mono» Bonn oo>flumo .coflumfi>oo pusocmum.fl asaop some a we ucaouom mace zoom .mouomm was mumwooumumo mo momOHa>co on» NO cofluflmomfioo swoon any mo mwmwamsm .H magma 59 Figure 2. Infrared spectrum of a 10% w/v solution, in CCl , of the methylated polysaccharide derived from the heterocyst envelope. The absence of absorption peaks in the region 2800- 3400 A in comparison with the spectrum of a 10% solution of methanol in chloroform, shows that less than 2% of the hydroxyl groups which were originally free in the polysaccharide remain unmethylated. 60 2-68 essences; com. 88 88 coon 8% 80¢ aouoqlosqv m e n Om 8:235 593.263 Figure 2 61 Figure 3. Gas chromatograms of the partially methylated alditol acetate derivatives obtained from the polysaccharide of the heterocyst envelope. a. The ECNSS-M column was used for determinations of relative retention times. b. The .2.2.4 column was used for separation of 3—glucose and 3-mannose. c. The OV—225 column was used isothermally (llustrated) for determinations of relative retention times, and was temperature-programmed for quantification of peak areas. d. The SP-2401 column was used for separation of T-glucose and T-mannose. The programs used for the different columns are described in Material and Methods (Part One). Peaks 1-9 are identified in Table II. Peak 10, O-acetyl inositol, the internal standard, comes at 180°C, and is therefore not seen in a and c, which were performed isothermally at 155° and 170°C, respectively. The unidentified peak between peaks 1 and 2 in the chromatograph taken with .2.2.4 column lacks the m/e = 43 fragment, charac- teristic of partially methylated alditol acetate derivatives, in the mass spectrometer. 62 h) EONSS - M G) ISOTHERHAL l 1 l l 1 A 4 A L 5 l0 '5 20 23 30 35 40 45 50 85 6° ” 7° TIME (M500...) T m (a g) 12A 'mmemmms MOORAIUEO OV - 225 lSOTHEIMAL _L 4 J l l A | J 1 5 IOIS 20253035045505.06810 TllEtmiwtes) H) 89-240! TEMPERATURE WED C) L l l L l 1 l O 5 l0 IS 20 25 30 35 TIME (situates) Figure 3 Table II. 63 Retention times relative to T-glucose for the partially methylated alditol acetate components derived from the polysaccharide of the heterocyst envelope. The .2.2.4 column was programmed at l°/min from 110° to 180°C. The ECNSS-M and OV-225 columns were used isothermally at 155° and 170°C, respectively. Only the .2.2.4 column resolved 3-glucose from 3-mannose. T-Glucose and T- mannose, which have the same relative retention time in these columns, were partially resolved by the SP- 2401 column. 64 TABLE II Retention times relative to T-glucose for the methyl-alditol- acetate components of the polysaccha ride of the heterocyst envelope Column Peak! Sugar ldentlficatlon ECNSS-M .2.2.4 0V-225 l T-Xyl 0.66 0.55 0.56 23 T-Glc 1.00 LN 1.01 2h T-Man 1.00 1.00 Ll!) 3 PM 1.28 1.15 1.19 4a 3-Glc 1.98 1.33 1.82 4b 3-Man 1.98 1.35 1.82 5 4-Glc 2.50 1.50 2.30 6 2,3-Man 3.20 1.57 2.94 7 3,4-Glc 4.00 1.74 3.36 8 3, 6-Glc 5.11 l. 85 4.27 9 2,3,4-Glc 5.60 1.89 5.06 65 Peaks #1 to 9 appear in the same order on the four columns (Figure 3a-d). The combined GC-MS analysis performed with the OV-225 and SP-2401 columns (Figure 4) identify these peaks as: T-pentose (peak #1), T-hexose (peak #2a), T-hexose (peak #2b, SP-2401 column), T-hexose (peak #3), 3-hexose (peaks #4a and 4b), 4-hexose (peak #5), 2,3-hexose (peak #8), and 2,3,4-hexose (peak #9). The molecular fragments of each of these partially methylated alditol acetate sugars are as reported by Lindberg (94) (Figure 4). In the fragmentograms of Figure 4, 3-hexose and 4-hexose are easily distinguished because the methyl sugars had been reduced with NaBD4: 161 and 234, whereas the 3-hexose (peak #4) gives fragments of m/e the 4-hexose (peak #5) gives fragments of m/e 162 and 233. Gas chromatographic analysis identified the T-sugars. Peak #1 (T-pentose) is T-xylose, as shown by chromatographic identity with T-xylose present in xyloglucan and sycamore cell walls. Peak #2a (T-hexose) is T-glucose, as shown by chromatographic identity with T-glucose from maltose. Peak #2b (T-hexose) from column SP-2401 is T-mannose. It has the same retention time as T-mannose present in yeast mannan, and is chromatographically resolvable from T-glucose and T-galactose present in standard polysaccharides. Peak #3 (T- hexose) is T-galactose, as shown by chromatographic identity with T-galactose present in arabinogalactan and xyloglucan. Peak #5 (4-hexose) is 4-glucose, as shown by chromatographic identity with 4—glucose present in maltose, sycamore cell walls and xyloglucan. Each of the other partially methylated alditol acetate deriva- tives was identified by determining its retention times relative to T—glucose, as described in Material and Methods, Part One, using 66 Figure 4. Mass spectral fragmentogram obtained for each of the partially methylated alditol acetate components of the hetero- cyst envelope polysaccharide. Gas chromatography was performed on the SP-2401 column. Ordinates: peak heights at specific values of m/e. The abscissa: scan number, at 105 per scan. 67 333 lalsl‘lalnlnlalilalilsl IlAlLlLi“ 306 AlllelALlilLl Lilalllililajllwin 305 11.1.1.1.1.1.1.11.111.1 .11111.l.l.1 111AlAlLlllltllAAll‘lllAlAlLlAlAJ AlLJALIJLIAlliLlAlAl AIIAILJ‘ALAIAA 4111;InLL].IMIIIALLIJILIAIILLLLLLL 254 235 141111111115 llll‘llAlAlAILLAJAJAl 205 LJAIAIAJLIII IAILJAIAJAJAIAIAJAIAJ ISO AlAlAlLlJlllLl—LIAIAIAA l89 AlnljlnlllnLLlnlnljlllj ll‘lLLs'A.1__ LlAlLL451All lllkAM‘l‘lll‘l‘lAl lSl Jllllll‘lAll ll? 46 45 AAA: 4“; - 1 , sgkvem egg @©... (9 TI! AKS ® 50 IOO ISO SCAN NUMBER Figure 4 68 columns ECNSS-M, .2.2.4 and OV-225 (isothermal program, Figure 3c). The relative retention times for all of the partially methylated alditol acetate components of the heterocyst envelope are listed in Table II. The relative retention times are identical, within experi- mental error, to those reported by Talmadge et al. (139) and Lindberg (94) for the same partially methylated alditol acetate derivatives and the same columns. The 3-mannose peak (Peak #4b, Figure 3b) is resolved only by the .2.2.4 column. The mass spectrum of this peak is characteristic of a 3-hexose. Its relative retention time, 1.35, differs by only 0.02 units from the relative retention time of 3-glucose (Table II). The mole percent compositions of the partially methylated alditol acetate components of the heterocyst and spore envelope polysaccharides, determined by quantification of peak areas using the OV-225 column, as described in Material and Methods, Part One, are presented in Table III. The compositions of the polysaccharides from the envelopes of the two cell types are equal, within experi- mental error. The amounts of T-xylose and T-galactose, and the sum of T— mannose, 3-mannose and 2,3,-mannose present in the methylated poly- saccharide (Table III) accOunt for the total amounts of xylose, galactose and mannose present in the envelopes as given by sugar composition analysis (Table I). For every mole of the doubly branched sugar, 2,3,4-glucose, there must be two moles of T-sugars. Hence, its mole percent was added twice in the balance sheet, Table III, in which branches and terminals are compared. 69 Table III. Mole percent composition of heterocyst and spore envelope polysaccharides. Each value presented is the mean + standard deviation, derived from methylation analysIs of three different envelope preparations. TABLE III Mole percent composition of heterocyst and spore envelope polysaccharides (from methylation analyslsl Sugar Mole Percent Mole Percent Heterocyst Spore T-Xylose 3.6 1 0.6 4.5 1 0.7 T-Glucose 22.0 _+. 1.0 39.9% 23.8 1 1.0 43.3% T-Mannose 6.2 1 1.0 Terminals 6.8 t 1.5 Terminals T-Galactose 8.1 i 1.0 8.2 t 1.2 3-Glucose 10.6 1 0.6 10.7 1 0.7 3-Mannose 7.0 i: 0.6 30'” 7.5 1 0.8 29'” 4-Glucose 12.9 1 0.6 ”"3" 11.1 1 0.8 ”m" 2,3-Mannose 5.8 i 1.0 6.2 t 1.0 3,4-Glucose 5.3 i 0.8 29.9% 4.9 i 0.9 27.4% 3,6-Glucose 6.0 :1: 1.2 +12.8 6.6 t 1.0 _+3_._7_ 2,3,4-Glucose 12.8 1- 1.3 42.7% 9.7 t 1.0 37.1% Branches Branches 70 Part Two - The Structure of the Backbone Polysaccharides Periodate oxidation is complete at 6 days, after which time no further decrease in absorption at 223 nm is observed. The entire sugar content of the reaction mixture passes through the Bio-Gel P-2 column with the void volume (Figure 5). Periodate oxidation of 60 mg of dried heterocyst envelopes containing 55 mg of polysaccharide (based on sugar composition analysis of 1 mg of envelope material) yields 30 mg (ca. 55%) of a white lyophilized polysaccharide. The recovery from spore envelopes is lower: only 20 mg (ca. 49%) of lyophilized material is obtained from 60 mg of dried spore envelopes, containing 41 mg of polysaccharide. Glucose and mannose, in molar ratio approximately 3:1 (73.7 :_1.7%:26.3 :_l.7% and 73.0 i_1.4%:27.0 :_l.4% in material derived from the envelopes of heterocysts and spores, respectively) were found in these polysaccharides. Methyla— tion analysis (Figure 6) showed the presence of only two sugars (1+3)-linked glucose (3-Glc) and (1+3)-linked mannose (3-Man). Because terminal sugars were not detected, branches accounted for less than 1% of the sugar present, I therefore refer to these poly- saccharides as "backbones" of their respective envelope polysaccharide. The products of partial acid hydrolysis of the backbones, frac- tionated by means of a Bio-Gel P-2 column, corresponded in their elution volume to mono-, di-, tri-, tetra-, and pentasaccharides (Figure 7). Thirty percent of the total sugar congent was found in the void volume of the column. When re-hydrolyzed under the same conditions, the residual 30% was broken down to mono-, di-, tri- and tetrasaccharides (24%) and to higher oligosaccharides, the great majority of which were of molecular weight less than 1800 daltons. 71 Figure 5. Fractionation, by chromatography on a column of Bio-Gel P-2, of the products of Smith degradation of heterocyst envelopes. Fractions of 1.5 ml were collected and assayed for sugar by the anthrone test. The arrows show the peak fractions in which, in a prior run, glucose, maltose, maltotriose, stachyose and arabinogalactan (void volume) were eluted. Mg/ml 20 LE) l.O 0.5 72 Void Volume i O B b 0 £ 0 0 O a g ‘6 g... *— Trlsaccharide <— Disaccharide ‘— Monosoccharlde RIJJIIJIJJJLIIIIIJIAI ILIAIIIJIIILIIIL 20 3O 4O 5O Fraction Number Figure 5 73 Figure 6. Gas chromatograms (.2.2.4 column) of the partially methylated alditol acetate sugars derived from the backbones of the envelOpe polysaccharides of heterocysts (solid line) and spores (dashed line). Inositol was added as standard. 74 8353 ms: Aum .UN 0. cos-m 6:85 Bonn 40 Figure 6 75 Figure 7. Fractionation, by chromatography on a column of Bio-Gel P-2, of the products of the partial acid hydrolysis of the backbone of the polysaccharide from heterocyst envelopes. Fractions of 1.0 ml were collected and assayed for sugar with the anthrone reagent. The arrows show the peak fractions in which, in a prior run, glucose, maltose, maltotriose, stachyose, maltopentaose, and arabinogalactan (void volume) were eluted. 76 epuouooosouow ——+ epgaouooosgo —> OleOHOODSQJL ——D “31101100050119; —-b OleOQOODSDTUOd —-§ 0 E 2 o 2 ———o of. _._ a '5 W > N .'.. Iw/Bw Figure 7 20 3O 4O 50 60 7O 80 90 Fraction Number IO 77 Di-,tri—, and tetrasaccharides having the same electrophoretic mobilities (Figure 8) were obtained in very close to the same ratio from the backbones of spore and heterocyst envelope polysaccharides. Paper electrophoresis separated the disaccharide fractions from the envelopes of both types of cells into three disaccharides (Figure 8), Di-I, -II and -III, having, respectively, the following electro- 1 (V/cm)-1. The phoretic mobilities: 0.243, 0.312 and 0.382 um s’ trisaccharide fractions were both separated into four trisaccharides (Figure 8), Tri-I, -II, -III and -IV, having electrophoretic mobili- ties of 0.125, 0.160, 0.201, and 0.243 pm 3.1 (V/cm)-1. The two tetrasaccharide fractions were also separated into four components, Te-I, -II, -III, and -IV, having respective electrophoretic mobili- ties of 0.079, 0.109, 0.139, and 0.188 pm 5.1 (V/cm)-1. The disaccharides were identified by sugar composition analysis (Table IV) and by reducing-end analysis as: Glc+Man (Di-I), Man+Glc (Di—II) and Glc+Glc (Di-III). The trisaccharides were subjected to sugar composition analysis (Table IV), methylation analysis (Figure 9) and, where necessary, reducing-end analysis (Figure 10). Based on these analyses, the trisaccharides were identified as: Glc+Glc+Man (Tri-I), G1C+Man+Glc (Tri-II), Man+Glc+Glc (Tri-III), and Glc+Glc+Glc (Tri-IV). The tetrasaccharides were analyzed only for their sugar composition. All contain glucose and mannose, in molar ratio 3:1. The disaccharides Glc+Man (Di-I, Figure 11a) and Glc+G1c (Di-III, Figure 11c) were hydrolyzed by B-glucosidase but not by a-glucosidase. Similarly, the trisaccharide Glc+Glc+Glc (Tri-IV) was completely hydrolyzed by B-glucosidase, but not measurably hydrolyzed by a- glucosidase (Figure 11d). Man+Glc (Di-II) was hydrolyzed by 78 Figure 8. Paper electrophoretograms of oligosaccharides obtained by partial acid hydrolysis of the backbones from hetero- cyst and spore envelope polysaccharides. OR indicates the origin. In order of increasing electrophoretic mobility are found disac- charides Di-I, -II, and -III; trisaccharides Tri-I, -II, -III, and -IV; and tetrasaccharides Te-I, —II, ~III, and -IV. Magnifi- cation: X 0.33. .- K O 79 Figure 8 Table IV. 80 Summary of the analyses performed on the oligosaccharides derived by partial acid hydrolysis of the heterocyst- polysaccharide backbone. Ratios between the amounts of the different oligosaccharides of a given chain length were calculated, following elution, from gas chromato- graphic determinations of the amounts of their component sugars. Very similar ratios were obtained by use of the anthrone reagent, with oligosaccharides derived from the backbone polysaccharides of both heterocyst and spores. 81 2 "$52288 8m: :5 o; >_ 0.— oA :. oh o4 __ 8. a; .9— 2%.. 8:52.58 .88 828888.58 o; 85185186. >_ E 3 81881821: E E 8185...: E E 8818518. __ E E 8182. __ a 3888 E 8218616.? _ E E 82.66. _ E 82 88 e538 885122 c s: 82. 1.5.8885 s Passage”. 3 3 “Emma 8:52.58 .88 s 825858 .88 a SEESEE 8288885 82888895 889.88 >_ REE 82 Figure 9. Gas chromatograms (.2.2.4 column) of the partially methylated alditol acetate sugars derived from each of the trisac- charides isolated, by paper electrophoresis, from the backbone of the heterocyst-envelope polysaccharide. Peak #1 is T-Glc (in Tri-I, Tri-II and Tri-IV) or T-Man (in Tri-III); peak #2 is 3-Glc; peak #3 is 3-Man; and peak #4 is inositol, added as standard. 83 TR! -I CD CD ©fl LL (1.11.11 i ‘1 LL01» TRl-III (2) ® 1 LLRETWJ; ® 1 ® l J l l _ _l 40 60 TIME (minutes) Figure 9 *:=—@ @ 84 Figure 10. Reducing-end analyses of trisaccharides Tri-I and -II. Bands of glucitol and mannitol are shown at the base of each electrophoretogram. NaB3H4 has reduced a mannosyl residue in Tri-I and a glucosyl residue in Tri-II. Tri-III and Tri-IV were not sub- jected to reducing-end analysis because Tri—III lacks 3—Man (Figure 5) and Tri-IV lacks mannose, so that both trisaccharides must have glucose at the reducing-end. 2 Mannitol ‘i We” Mannitol Glucitol 85 TRISACCHARIDE I Glucitol Distance (cm) TRISACCHARIDE II l 1 (5 Distance (cm) Figure 10 - 'WI3XIO4 E o. o —- 2xl04 .3 .2 8 o .9 8 - lo4 a: r-—fi .55 Origin “—7 Ii . F a 3 x lo‘ 8 - 2 3: l0‘ ‘3. Z‘ :8 ‘6 8 - 4 -- IO '8 a: Origin ? l J_°_. LA I 86 Figure 11. Enzymatic hydrolysis of disaccharides Di-I, -II, and -III and trisaccharide Tri-IV (Glc+Glc+Glc) isolated following partial acid hydrolysis of the backbones of the polysaccharides from the envelopes of heterocysts (circles) and spores (triangles). The appearance of reducing-end groups, as percent of total sugar residues present, is shown as a function of time of hydrolysis of (a) Glc+Man (Di-I), (C) G1c+G1c (Di-III) and (d) Glc+Glc+Glc (Tri-IV) by B-glucosidase from almonds (0,4) and by a-glucosidase from yeast (0, A); and as a function of time of hydrolysis of (b) Man+Glc (Di-II) by B-mannosidase from Polyporus sulfureus (O, A) and by a-mannosidase from jack beans (0, A). 87 3:: me...» am ON 0. o. n 0 nor a 1 e\ l \O l 3 i\ 1 :8 8W ilk 1 QC“) 2: «9:: ms; nu ON 0. o. m — -i q q fl ON Om 0 0m \. A 0 V n q d d T 0 ON O¢ so. wlllohum 0. Om wowed wowed Figure 11 88 B-mannosidase, but not by a—mannosidase (Figure 11b). B-Mannosidase- containing B-mannanase from both Penicillium ochro-chloron and Penicillium verruculosum also readily cleave the linkage of this disaccharide. In confirmation of these results it was found (i) that Di-III has the same electrophoretic mobility as laminaribiose and is separable from nigerose by electrophoresis and (ii) that ca. 93% of the backbone from heterocyst envelope polysaccharide is hydrolyzed to mono—, di-, and trisaccharides in 1.5 h by the B(1+3) endoglucanase from Rhizopus arrhizus. Part Three - Additional Studies on Envelope Polymers Extraction of the Polysaccharide from Spore Envelopes Twenty-five milligrams (83% of the total amount of polysaccharide that the envelope contains and 55% of the total envelope material) was extracted from 45 mg of lipid—free spore envelopes by boiling with a 5M solution of NaN3 for 5 min. Table V shows the sugar composition of the extracted polysac— charide (Table V a) as well as of the residue after extraction (Table V b). The sugar compositions, determined at the same time, of spore (Table V c) and heterocyst (Table V d) envelopes are also given for the sake of comparison. Figure 12 shows a gas chromatogram of the partially methylated alditol acetate components derived from the polysaccharide extracted from the envelopes of spores. The gas chromatogram is superimposed 89 mi. 9.2. H62 “.3 m .8 N .3 c .3 o .8 88:5 CNN fimw Emu o .mm 32:55. o .8 Tu o .v o .N 888.23 m .8 o .m m .e o .m 083x o u I. a 8:52.58 .5 222 > 555 AU: mamOHm>cc umxoououoc cumulowmwa .uomucw one AOV mamoao>co 020mm OOHMIUHmfiH .uomusw Ono .Anv mamOHw>cm mopwmou as» .Amv amoaobom 020mm on» 80H: oouomuuxm ooflumoooommaom on» no mwmhamcm cowuflmomfioo Homom .> canoe 90 Figure 12. Gas chromatograms (.2.2.4 column) of the partially methylated alditol acetate sugars derived from the extracted poly- saccharide from spore envelopes (solid line) and from the intact, lipid-free spore envelopes (dashed line). 91 85: Om 05 cm Om Co v 5523 EN. N. ON 0. Figure 12 92 over another obtained for the partially methylated alditol acetate components derived from the lipid-free spore envelope. The original material and the extracted polysaccharide have the same components, in approximately the same proportions. Amino Acid Analysis of Lipid-Free Envelopes of Heterocysts and Spores The total amounts of amino acid present in 3 mg of heterocyst and spore envelopes are given in Table VI. These amounts were esti- mated by adding the weights of all of the amino acids present in the envelopes. These values were determined by conversion, from.moles to gramsJ of the results of computation performed by the Autolab Computer Integrator. No amino sugars were found in the envelope of either type of cell; 0.01% of amino sugars would have been detectable. Amino acid composition profiles for both types of envelope are shown in Figure 13. Both contain the same amino acids except that ornithine is present only in the spore envelope. The heterocyst envelope con- tains small amounts of methionine which is absent from spore envelopes, and more arginine than does the spore envelope. 93 w .2 22 82 .82 2 me .o 8 .m 2815...: .5 88.88 2: c. 3.852. 8.8 8.8 8 .5 88.88 8 55 _> 555 mauomm pom mumwoosouoz mo momon>cm 0:0 :0 voooucm unmfio3 no 0500500 peas ocfiE< .H> OHQMB 94 Figure 13. Amino acid profiles of the heterocyst envelope and spore envelope . § :2 I x\\\\\\\\\\\\\\\\ g M spore \\\\\\\\\\\\\\\‘ g \\\\\\\\\\‘ g. T \\\\\\\\\\\\\\\\\' Leu Ileu K\\\\\\\\V Ala \\\\\\\\\\‘ 5 \\\\\\\\\\\\\\‘ g \\\\\\\\\\ a; X\\\\\\‘ \\\\\\\\\\\\\\\\\\\\‘ Asp Thr l0+~ 8 so, % aIOW Figure 13 DISCUSSION The results of methylation analysis of the polysaccharides from the envelopes of heterocysts and spores led me to the following conclusions: (i) Xylose, galactose, mannose and glucose occupy terminal positions in the polysaccharides, whereas the internal portion of the molecules is composed of mannose and glucose. (ii) Forty-five to forty-eight percent of the sugars, including all of the branched sugars and a majority of the linear sugars, in the polysaccharides have C-3 involved in a glycosidic bond. (iii) The polysaccharides are highly branched.. About 30% of the sugar residues are branched (Table III), and there is an unusually large amount (10-13%) of a doubly branched sugar, 2,3,4-G1c. Four types of evidence, taken together, demonstrate that the poly— saccharides are completely methylated, so that the high frequency of appearance of branched and doubly branched sugars is not an arficat of undermethylation. First, the chromatograms (Figure 3b and 3d) show no peaks corresponding to unmethylated alditol acetate deriva- tives. If the polysaccharides had been only partially methylated, small peaks of O-acetyl-mannitol, -galactitol and —glucitol would have been expected, with retention times of from 1 to 3 min less than O-acetyl-inositol. Second, the infrared spectrum (Figure 2) of the methylated polysaccharide shows no absorption, attributable 96 97 to free hydroxyl groups, at 3 pm. Third, the results of quantitative analysis are closely reproducible (Table III). Fourth, the amount of terminal sugars balances the amount of branched sugars (Table III). (iv) Because both polysaccharides have the same components and in approximately the same proportions, the polysaccharide molecules from the envelopes of heterocysts and spores may be identical or almost identical. However, because methylation analysis gives only the linkages between, and not the sequence of the sugars, there remains the possibility of a different sequence of sugars in the polysaccharides from the two cell types. Smith degradation and partial acid hydrolysis of the products of that degradation, dis- cussed below, allowed me to determine the sequence of the internal portions (backbones) of the polysaccharides from heterocyst and spore envelopes. The structures of the two backbones will be com- pared later in this discussion. The only sugar-containing products of Smith degradation come in the void volume of the Bio-Gel P-2 column (Figure 5). Therefore, they have a molecular weight in excess of 1800 daltons, i.e., contain at least 11 sugar residues. When these products were subjected to methylation analysis, terminal sugars were undetectable, i.e., present at a frequency of less than 2% of the total sugar content (Figure 6). The Nelson test showed the presence of 0.2% reducing- end sugar residues. Due to the limited solubility of those products, this test may have underestimated their content of reducing sugars. I conclude that the polysaccharide products of Smith degradation of the envelope polysaccharides contain between 100 and 500 sugar residues. Because branched sugars were also not detectable (Figure 98 6), the polysaccharide products constitute the linear, internal portions, or backbones, of the highly branched envelope polysaccharides from which they were derived. Only glucose and mannose, essentially all 3—1inked (Figure 6), are present in these backbone polysaccharides. The observed ratio of glucose to mannose was 2.8:1 for the backbone polysaccharide derived from the heterocyst envelopes and 2.7:1 for the corresponding backbone from spores. These ratios are in fairly good agreement with the ratios (2.7:1 and 2.3:1, respectively) between total 3-1inked glucose and total 3-linked mannose from the intact polysaccharides from heterocyst and spore envelopes, as given by methylation analysis (Table III), and are consistent with a 3:1 ratio of glucose to mannose in the backbone polysaccharides. Because 4-Glc is broken down by the Smith degradation procedure, and because oligosaccharides are not recovered (Figures 5 and 6), it is clear that 4-Glc cannot be a frequent constituent of the back- bone of the original polysaccharide. The absence of branched sugars (Figure 6) following Smith degradation implies that in the intact polysaccharides, the branches consist of T-sugar linked to the back- bone either directly or through the 4-G1c residues. Although it is unknown whether sequences of seriate 4-Glc residues are present in the intact envelope polysaccharides, the stoichiometry of 4-Glc to total 3—linked mannose (Table III) suggests that a single 4—G1c residue may be a constituent in a basic subunit (see below) of the intact polysaccharides, and would be present in a side branch. As shown by methylation analysis of the products of Smith degra— dation, all T-sugar and 4-Glc residues have been degraded. On the 99 other hand, there appears to have been negligible degradation of other portions of the original envelope polysaccharides. Thus, ca. 53% of the original polysaccharide from heterocyst envelopes is comprised of T-sugars plus 4-Glc (Table III), whereas ca. 45% of the polysaccharide originally present is lost during the Smith degradation procedure. The lower recovery of backbone from spores, as percent of envelope material, is consistent with the higher content of amino compounds in their envelopes (28; see Table VI, page 92). The presence of both glucose and mannose in oligosaccharides derived from partial acid hydrolysis implies that the backbones do not consist of separate glucan and mannan chains, and the fact that no di-, tri-, or tetrasaccharide contains two mannosyl residues implies that the backbones do not consist of a random distribution of glucosyl and mannosyl moieties. The presence of a single mannosyl residue within each tetrasaccharide implies that the backbones consist of a repeating unit Glc-Glc-Glc-Man, an interpretation which is consistent with the structures and relative frequencies of di— and trisaccharides which were observed (Table IV). Because Glc+Man, Glc+G1c and Glc+Glc+Glc are completely hydrolyzed by B-glucosidase and not measurably by a—glucosidase, and because Man+Glc is hydrolyzed by B-mannosidase but not by a-mannosidase, I conclude that all of the sugars in the backbone are B-linked. I have found no differences between the envelope polysaccharides from heterocysts and spores. As I have shown here, the backbone of both consists of B(l+3)-linked glucose and mannose with a repeating sequence of Glc—Glc—Glc-Man. Moreover, all of their other glycosidic 100 linkages are present in the same frequency, within experimental error (Table III). Although it remains to be determined whether or not the branches are linked with the same anomeric configurations to the same sugar residues in the backbones of the two polysaccharides, the evidence available to date seems sufficient to suggest, as a working hypothesis, that the envelope polysaccharides from the two types of differentiated cells are essentially identical. I propose that, if the two polysaccharides are essentially identical, their structure resembles, to a first approximation, the structure presented in Figure 14. Only a portion of the mannosyl residues in the backbones is substituted, and certain of the terminal sugars are present in amounts not stoichiometric with the backbone sugars (Table III). Therefore, even if the sequence of substitutions of glucosyl residues in the backbone were correct, the "basic" structural unit illustrated would not be, in the strict sense, a "repeating" subunit. The results of sugar composition analysis and methylation analysis of the polysaccharide extracted from the spore envelope (Table V a, Figure 12) indicate that the structure of the polysac- charide remains unchanged upon extraction. Extraction does not fractionate the total polysaccharide into two dissimilar portions, except that the extracted portion may be relatively richer in galactose. Thus, the structure proposed for the complete polysac- charide (Figure 14) is presumably valid also for the extracted polysaccharide. Whether or not these polysaccharides are produced in small amounts by vegetative cells is unknown. The vegetative cells of our cultures, in contrast to those grown in more concentrated inorganic 101 Figure 14. Possible structure of the polysaccharides from the envelopes of heterocysts and spores of Anabaena cylindrica. Linkages to terminal sugars (T-S) indicated by dashed arrows can be present in only certain of the "basic" units illustrated. On the basis of approximate stoichiometry (see Table III) , it may be suggested that the irregularly appearing substituent on the mannosyl residue in the backbone is T-Man; that the 2,3,4—Glc in the polysaccharide has, as regular substituents, T-Glc and (linked to C-2 or C-4) Glc-(1+4)-Glc; and that the irregularly appearing substituents on the other two backbone glucosyl residues are, to a large extent, galactose at the 6-position and xylose at the 4 position. Because the sequence of substitutions of backbone glucosyl residues is unknown, an arbitrarily chosen sequence is illustrated. Unlinked carbons-6 are omitted from the drawing for the sake of simplification. --o Glc 102 0 Ole 0 0 4-Glc 0 T-S T-S Figure 14 103 growth medium (28), do not produce evident sheath material. Sheath polysaccharides, when formed, are much richer in xylose (21%) and poorer in glucose (50%) than are the envelope polysaccharides of the differentiated cells, and also contain a trace of fucose (28). In addition, the wall of a vegetative cell contains only between 1.6 and 3.2% as much glucose as does the envelope of the heterocyst (28,29). The heterocyst envelope is deposited in approximately one-fifth of a vegetative cell doubling time (29). Thus, if a vegetative cell forms a polysaccharide comparable‘to that in the envelope of hetero- cysts, it does so at a rate less than 1% as rapid as the rate of synthesis of the polysaccharide in a heterocyst. It follows that if the polysaccharides in the envelopes of heterocysts and spores are identical, it is probable that the same set of polysaccharide synthesizing enzymes is produced or activated during the alternative differentiation processes. Such a finding would be consistent with the idea (38,39) of evolutionary and ontogenetic relationship between the two processes. The amino compounds in the envelopes of heterocysts and spores are amino acids. Amino sugars were not present. It therefore appears that a polypeptide or protein is a component of the envelopes of heterocysts and spores. The protein-like component may be very similar, but appears not to be identical, in the two envelopes. As shown in Figure 13, methionine is present, although in very small amounts (1.5%), only in the heterocyst envelope. Ornithine comprises 9.8% of the peptidic material of the spore envelope, but is absent from the heterocyst envelope, whereas arginine is more abundant in the envelope of the 104 heterocyst than in the envelope of the spore (20% vs. 4.2%). However, there remains the possibility that ornithine either originates from arginine during the processing of spore envelopes or, because of its chemical similarity to arginine, replaces arginine as component of an otherwise very similar polymer, in the spore envelope. Although envelopes have been washed with a 61.8% solution of CsCl (density 1.4 g/cm3), the high levels of arginine and aspartic acid (and perhaps ornithine) may be due to contamination of the envelope fractions with cyanophycin granules.* The distribution of other amino acids (with the exception of methionine) are very similar for the envelopes of the two types of cells and may correspond to the occurrence of a similar protein in these envelopes. * Cyanophycin granules, the principal nitrogenous reserve of blue-green algae, is a copolymer of arginine and aspartic acid (130). These granules are abundant in spores and are thought to be present also in heterocysts (162). They are dissolved by a 61.8% solution of CsCl (88). REFERENCES 10. REFERENCES Albersheim, P., D. J. Nevis, P. D. English, and A. Karr. 1967. A method for the analysis of sugars in plant-cell walls polysaccharides by gas-liquid chromatography. Carbohyd. Res. 5:340-345. Allen, M. M. 1968. Ultrastructure of the cell wall and cell division of unicellular blue-green algae. J. Bact. 96: 842-852. Anderson, C. G., W. N. Harworth, N. Rastrick, and M. Stacey. 1939. Polysaccharides synthesized by microorganisms. IV. The molecular constitution of luteose. Biochem. J. 33: 272-279. Annan, W. D., E. L. Hirst, and D. J. Manners. 1965. The con- stitution of laminarin. V. The location of 1,6—glucosidic linkages. J. Chem. Soc. 885-891. Ashmed, M. V., and F. 8. Cook. 1964. Separation of an a- galactosidase from watermelon seeds. Can. J. Biochem. Physiol. 42:605—612. Balls, A. K., M. K. Walden, and R. R. Thompson. 1948. A crystalline B—amylase from sweet potatoes. J. Biol. Chem. 173:9-19. Barker, S. A., P. J. Sommers, and M. Stacey. 1967. Sequence studies on Diplococcus pneumoniae Type II polysaccharide. Carbohyd. Res. 3:261-270. Battista, O. A. 1950. Hydrolysis and crystallization of cellulose. Ind. Eng. Chem. 42:502-507. Bauer, W. D., K. W. Talmadge, K. Keegstra, and P. Albersheim. 1973. The structure of plant cell walls. II. The hemi- cellulose of the walls of suspension-cultured sycamore cells. Plant Physiol. 51:174-187. BeMiller, J. N. 1965. Acid hydrolysis and other lytic reac— tions of starch. In Starch: Chemistry and Technology, Vol. I, p. 495-527. R. L. Whistler and E. F. Paschall, eds. Academic Press, New York. 105 ll. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 106 BeMiller, J. N. 1967. Acid-catalyzed hydrolysis of glycosides. Advan. Carbohydrate Chemistry 22:25—108. Bernfeld, P. 1955. Amylase, a and B. In Methods in Enzymology, Vol. I, p. 149-158. S. P. Colowick and N. 0. Kaplan, eds. Academic Press, New York. Bishop, C. T., G. A. Adams, and E. 0. Hughes. 1954. A poly- saccharide from the blue-green alga, Anabaena cylindrica. Can. J. Chem. 32:999-1004. Bj6rndal, H., B. Lindberg, and S. Svensson. 1967. Gas-liquid chromatography of partially methylated alditols as their acetates. Acta Chem. Scand. 21:1801-1804. Bj6rndal, H., B. Lindberg, and S. Svensson. 1967. Mass spec— trometry of partially methylated alditol acetates. Carbohyd. Res. 5:433-440. BjOrndal, H., C. G. Hellerqvist, B. Lindberg, and S. Svensson. 1970. Gas-liquid chromatography and mass-spectrometry in methylation analysis of polysaccharides. Angew. Chem. Int. Ed. 9:610—619. Bj6rnda1, H., B. Lindberg, J. LOnngren, M. Mérzéros, and J. L. Thompson. 1973. Structural studies of the capsular poly- saccharide of Klebsiella type 52. Carbohyd. Res. 31:93-100. Bryce, T. A., D. Welti, A. E. Walsby and B. W. Nichols. 1972. Monohexoside derivatives of long-chain polyhydroxy alcohols a novel class of lipid specific to heterocystous algae. Phytochem. 11:295-302. ‘0 Buchala, A. J., and H. Meier. 1973. A galactoglucomannan from the leaf and stem tissue of red clover (Trifolium pratense). Carbohyd. Res. 31:87-92. Cadotte, J. E., G. G. S. Dutton, I. J. Goldstein, B. A. Lewis, F. Smith, and J. W. Van Cleve. 1957. Reduction of the products of periodate oxidation of carbohydrates. IV. Hydrogenation with palladium-charcoal of the dialdehydes from methyl glycosides. J. Am. Chem. Soc. 79:691-695. Clark, J. M. 1964. Experimental Biochemistry, p. 12-13. W. H. Freeman and Co., San Francisco. Clark, R. L., and T. E. Jensen. 1969. Ultrastructure of akinete development in a blue-green alga, Cylindrospermum sp. Cytologia 34:439-448. Conrad, H. E., J. R. Bamburg, J. D. Epley, and T. J. Kindt. 1966. The structure of Aerobacter aerogenes A3(Sl) poly- saccharide. II. Sequence analysis and hydrolysis studies. Biochemistry 5:2808-2817. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 107 Courtois, J. E., F. Petek, and To Dong. 1963. Séparation de deux o-galactosidases a propriétés distinctes de grains de Plantago ovata. Bull. Soc. Chim. Biol. 45:95-100. Danishefsky, I., R. L. Whistler, and F. A. Bettelheim. 1970. Introduction to polysaccharide chemistry. In The Carbo- hydrates, Vol. II-A, p. 375-412. W. Pigman and D. Horton, eds. Academic Press, New York. Dey, P. M., and J. B. Pridham. 1968. Multiple forms of a- galactosidase in Vicia faba seeds. Phytochem. 7:1737-1739. Dische, Z. 1962. Color reaction of carbohydrates. In Methods in Carbohydrate Chemistry, Vol. I, p. 477-512. R. L. Whistler and M. L. Wolfrom, eds. Academic Press, New York. Dunn, J. H., and C. P. Wolk. 1970. Composition of the cellular envelopes of Anabaena cylindrica. J. Bacteriol. 103:153-158. Dunn, J. H., R. D. Simon, and C. P. Wolk. 1971. Incorporation of amino sugars into walls during heterocyst differentiation. Dev. Biol. 26:159-164. Echlin, P. 1963. The fine structure of the cell boundary of the blue-green alga Anacystis montana. J. Cell Biol. 17: 212-215. Edward, J. T. 1955. Stability of glycosides to acid hydrolysis. A conformational analysis. Chem. Ind. 1102-1104. Fay, P., and N. J. Lang. 1971. The heterocysts of blue—green algae. I. Ultrastructural integrity after isolation. Proc. Roy. Soc. Ser. B. 178:185-192. Fear, C. M., and R. P. Menzies. 1926. Application of thallium compounds in organic chemistry. Part III. Alkylations. J. Chem. Soc. 937-940. Fleming, M., and D. J. Manners. 1965. A structural difference between the soluble and insoluble forms of laminarin. Biochem. J. 94:17P. Fleming, M., and D. J. Manners. 1966. A comparison of the fine-structure of lichenin and barley glucan. Biochem. J. 100:4P. Frank, H., M. Lefort, and H. H. Martin. 1962. Elektronenoptische und chemische Untersuchungen an Zellwfinden der Blaualgen, Phormidium uncinatum. Z. Naturf. 178:262—268. Frank, H., M. Lefort, and H. H. Martin. 1962. Chemical analysis of a muCOpolymer component in cell walls of the blue-green alga Phormidium uncinatum. Biochem. Biophys. Res. Commun. 7:322—325. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 108 Geitler, L. 1921. Versuch einer Lfisung des Heterocysten- Problems. Sitzungsber. Akad. Wiss. Wien, Math. Naturw. Kl. 130 (Abth. I):223-245. Geitler, L. 1925. Synoptische Darstellung der Cyanophyceen in morphologischer und systematischer Hinsicht. Beih. Botan. Centralb. 41:163-294. Gillman, S. M., K. C. Tsou, and A. M. Seligman. 1962. The colorimetric estimation of a-D-galactosidase activity and some comparisons with B—D-galactosidase. Arch. Biochem. Biophys. 98:229-234. Goldstein, I. J., and F. Smith. 1958. Reaction of periodate- oxidized polysaccharides with alcoholic hydrogen chloride. Chem. Ind. (London) 40-42. Goldstein, I. J., B. A. Lewis, and F. Smith. 1958. The alcohol-binding capacity in mutarotation of the so—called dialdehydes obtained by periodate oxidation of sugar glycosides. Chem. Ind. (London) 595-597. Goldstein, I. J., G. W. Hay, B. A. Lewis, and F. Smith. 1959. A new approach to the determination of the fine structure of polysaccharides. Abstracts, Am. Chem. Soc. 135:3D. Goldstein, I. J., G. W. Hay, B. A. Lewis, and F. Smith. 1965. Controlled degradation of polysaccharides by periodate oxidation, reduction and hydrolysis. In Methods in Carbo~ hydrate Chemistry, Vol. V, p. 361-370. J. N. BeMiller, R. L. Whistler, and M. L. wolfrom, eds. Academic Press, New York. Guthrie, R. D. 1962. Periodate oxidation. Determination of periodate. In Methods in Carbohydrate Chemistry, Vol. I, p. 435-444. R. L. Whistler and M. L. WOlfrom, eds. Academic Press, New York. Guthrie, R. D., and J. F. McCarthy. 1967. Acetolysis. Advan. Carbohydrate Chem. 22:11-23. Hakomori, S. 1964. A rapid permethylation of glycolipid and polysaccharide catalyzed by methylsulfinyl carbanion in dimethyl sulfoxide. J. Biochem. (Tokyo) 55:205-208. Hakomori, S., Y.—T. Li, S. C. Li, and C. G. Hellerqvist. 1971. Anomeric structures of globoside and ceremide trihexoside of human erythrocytes and hamster fibroblast. J. Biol. Chem. 246:2271-2277. Halfen, L. N. 1973. Gliding motility in the blue-green alga Oscillatoria. Ultrastructural and chemical characteriza- tion of the fibrillar layer. J. Phycol. 9:248-253. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 109 Halfen, L.N., and R. W. Castenholz. 1971. Gliding motility in the blue-green alga Oscillatoria princeps. J. Phycol. 7: 133-145. Halvorson, H. 1966. a-Glucosidase from yeast. In Methods in Enzymology, Vol. VIII, p. 559-562. E. F. Neufeld and V. Ginsburg, eds. Academic Press, New York. Halvorson, H., and L. Ellias. 1958. The purification and properties of an a-glucosidase of Saccharomyces italicus Y125. Biochim. Biophys. Acta 30:28-40. Haworth, W. N. 1915. A new method of preparing alkylated sugars. J. Chem. Soc. 107:8-16. Haworth, W. N., E. L. Hirst, and J. I. Webb. 1929. Poly- saccharides. V. Glycogen. J. Chem. Soc. 2479-2485. Haworth, W. N., E. L. Hirst, L. N. Owen, S. Peat, and F. J. Averill. 1939. Polysaccharides. XXXIII. The methylation of cellulose in air and in nitrogen. J. Chem. Soc. 1885- 1898. Hebbert, H., and E. G. V. Percival. 1930. Studies on reac- tions relating to carbohydrates and polysaccharides. XXX. The comparative hydrolysis of some disaccharides and poly- saccharides. J. Am. Chem. Soc. 52:3995-4005. Helferich, B., and K. H. Jung. 1958. Zur Spezifitat von Glykosidasen. Verschiedenheit der B-D-glucosidase und der B-D—galactosidase des Sflssmandelemulsins bei Adsorption durch Poly-(hydroxystyrol-glykoside). Hoppe-Seyler's Z. Physiol. Chem. 311:54-62. Helferich, B., and T. Kleinschmidt. 1965. Zur Kenntnis des Sflssmandelemulsins. Hoppe-Seyler's Z. Physiol. Chem. 340: 31-45. Helferich, B., and T. Kleinschmidt. 1966. Kristallisierte B-D-glucopyranosidase des Sussmandelemulsins. Naturwiss. 153:132. Hérissey, H., P. Fleury, and M. Joly. 1934. Comparison of the action of periodic acid upon some hexoses and certain synthetic heterosides derived therefrom. J. Pharm. Chim. 20:149-160. Hers, H. G. 1964. Glycogen storage disease, type II. In Control of Glycogen Metabolism. Ciba Foundation Symp., p. 354-365. W. J. Whelan and M. P. Cameron, eds. Little, Brown, Boston, Mass. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 110 Hestrin, S., D. S. Feingold, and M. Schramm. 1955. Hexoside hydrolases. In Methods in Enzymology, Vol. I, p. 231-257. S. P. Colowick and N. 0. Kaplan, eds. Academic Press, New York. Hodge, J. E., S. A. Karjala, and G. E. Hilbert. 1951. Methyla- tion and ethylation of corn starch, amylose and amylopectin in liquid ammonia. J. Am. Chem. Soc. 73:3312-3316. Hogness, D. S., and E. H. Battley. 1957. The a-galactosidase of Aerobacter aerogenes; its characteristics and induced synthesis. Federation Proc. 16:197. H0116, J., and J. Szejtli. 1959. fiber einige Probleme bei der sauren Starke-hydrolyse. Starke 11:244-246. Hough, L., J. K. N. Jones, and W. H. Wadman. 1952. An investi- gation of the polysaccharide components of certain fresh- water algae. J. Chem. Soc.:3393-3399. Howard, B. H., G. Jones, and M. R. Purdom. 1960. The pento- sanases of some rumen bacteria. Biochem. J. 74:173-180. Huebner, C. F., S. R. Ames, and E. C. Bulb. 1946. Periodate oxidation of certain active methylene groups. J. Amer. Chem. Soc. 68:1621-1628. Hutson, D. H., and D. J. Manners. 1965. Studies on carbo- hydrate-metabolizing enzymes. Biochem. J. 94:783-789. Jackson, E. L., and C. S. Hudson. 1936. The oxidation of a1pha-methyl-d-mannopyranoside. J. Am. Chem. Soc. 58: 378-379. Jackson, E. L., and C. S. Hudson. 1937. Studies on the cleavage of the carbon chain of glycosides by oxidation, a new method for determining ring structures and alpha and beta configuration. J. Am. Chem. Soc. 59:994-1003. Jackson, E. L., and C. S. Hudson. 1939. The periodic acid oxidation of beta-methyl-D-mannopyranoside. J. Am. Chem. Soc. 61:959-960. Jackson, E. L., and C. S. Hudson. 1941. Crystalline B- methyl-D-ribopyranoside. J. Am. Chem. Soc. 63:1229-1231. Jensen, T. E., and L. M. Sicko. 1971. The effect of lysozyme on cell wall morphology in a blue-green alga, Cylindrospermum sp. J. Gen. Microbiol. 68:71-75. Johnson, J., S. Kirkwood, A. Misaki, T. E. Nelson, J. V. Scaletti, and F. Smith. 1963. Structure of a new glucan. Chem. Ind. (London) 820-822. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 111 Jones, J. K. N., and W. H. Nicholson. 1958. The acid catalysed reversion of L-arabinose and of D-mannose. J. Chem. Soc 27-33. J¢rgensen, 0. B. 1963. Barley malt d—glucosidase with iso— maltase activity. Acta Chem. Scand. 17:1765-1770. J¢rgensen, 0. B. 1964. Barley malt a-glucosidase. V. Degra— dation of starch and dextrins. Acta Chem. Scand. 18:1975-1978. Jost, M. 1965. Die Ultrastruktur von Oscillatoria rubescens D.C. Arch. Mikrobiol. 50:211-245. Keegstra, K., K. W. Talmadge, W. D. Bauer, and P. Albersheim. 1973. The structure of plant cell walls. III. A model of the walls of suspension-cultured sycamore cells based on the interconnection of the macromolecular components. Plant Physiol. 51:188-196. Kihara, H. K., A. S. L. Hu, and H. 0. Halvorson. 1961. The identification of a ribosomal-bound B-glucosidase. Proc. Nat. Acad. Sci. U.S. 47:489-496. Kocourek, J., and C. E. Ballou. 1969. Method for finger- printing yeast cell wall mannans. J. Bacteriol. 100: 1175-1181. Krishma Murti, C. R., and B. A. Stone. 1961. Fractionation of the B-glucosidases from Aspergillus niger. Biochem. J. 78:715-723. Kuhn, R., H. Trischmann, and I. L5w. 1955. Zur Permethylierung von Zuckern und Glykosiden. Angew. Chem. 67:32. Lambein, F., and C. P. Wolk. 1973. Structural studies on the glycolipids from the envelope of the heterocyst of Anabaena cylindrica. Biochemistry 12:791-798. Lamont, H. C. 1969. Sacrificial cell death and trichome breakage in an Oscillatoriacean blue-green alga: The role of murein. Arch. Mikrobiol. 69:237-259. Lang, N. J., and P. Fay. 1971. The heterocysts of blue-green algae. II. Details of ultrastructure. Proc. Roy. Soc. Ser. B 178:193-203. Lang, N. J., R. D. Simon, and C. P. Wolk. 1972. Correspondence of cyanophycin granules with structured granules in Anabaena cylindrica. Arch. Mikrobiol. 83:313-320. Lemieux, R. 1954. Some implications in carbohydrate chemistry of theories relating to the mechanisms of replacement reac- tions. Advan. Carbohydrate Chem. 9:1-57. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 112 Li, Y.-T. 1966. Presence of a-D-mannosidic linkage in glyco- proteins. Liberation of D-mannose from various glycoproteins by a-mannosidase isolated from jack bean meal. J. Biol. Chem. 241:1010-1012. Li, Y.-T., S. C. Li, and M. R. Shettlar. 1963. a-Galactosidase from Diplococcus pneumoniae. Arch. Biochem. Biophys. 103: 436-442. Li, Y.-T., and M. R. Shettlar. 1964. Occurrence of a-galactosidase in higher fungi. Isolation of a-galactosidase from Calvatia cyathifbrmis. Arch. Biochem. Biophys. 108:523-530. Li, Y.-T., and S. P. Li. 1972. Alpha-mannosidase, B-N-acetyl- hexosaminidase, and B-galactosidase from jack bean meal. In Methods in Enzymology, Vol. XXVIII, p. 702-704. V. Ginsburg, ed. Academic Press, New York. Lindberg, B. 1972. Methylation analysis of polysaccharides. In Methods in Enzymology, Vol. XXVIII, p. 178-195. V. Ginsburg, ed. Academic Press, New York. Lindberg, B., and J. McPherson. 1954. Studies on the chemistry of lichens. VI. The structure of pustulan. Acta Chem. Scand. 8:985-988. Lindsey, J. K., B. D. Vance, J. S. Keeter, and V. E. Scholes. 1971. Spheroplast formation and associated ultrastructural changes in a synchronous culture of Anacystis nidulans treated with lysozyme. J. Phycol. 7:65-71. Maclay, W. D., and C. S. Hudson. 1938. The cleavage of the carbon chain of a-methyl-d-1yxopyranoside by oxidation with periodic acid. J. Am. Chem. Soc. 60:2059-2060. Mahadevan, P. R., and B. Eberhart. 1964. The B-glucosidase system of Neurospora crassa. II. Purification and charac- terization of aryl-B-glucosidase. Arch. Biochem. Biophys. 108:22-29. Matsubara, T. 1960. Studies on takamaltase. II. Substrate specificity of taka-maltase I. J. Biochem. 48:138-143. McIntire, F. C., W. H. Peterson, and A. J. Riker. 1942. A polysaccharide produced by the crown-gall organism. J. Biol. Chem. 143:491-496. Mejzler, D., J. Schmorak, and M. Lewin. 1960. A contribution to the study of random degradation of chain polymers. J. Polymer Sci. 46:289-302. Moore, B. G., and R. G. Tischer. 1965. Biosynthesis of extra- cellular polysaccharides by the blue-green alga Anabaena flos-aquae. Can. J. Microbiol. 11:877-885. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 113 Muskat, I. E. 1934. Reactions of carbohydrates in liquid ammonia: Preliminary paper. J. Am. Chem. Soc. 56:693-695. Nichols, B. W., and B. J. B. Wood. 1968. New glycolipid specific to nitrogen-fixing blue-green algae. Nature (London) 217:767-768. Niklowitz, W., and G. Drews. 1956. Beitrage zur Cytologie der Blaualgen. I. Untersuchungen zur Substruktur von Phormidium uncinatum. Gomont. Arch. Mikrobiol. 24:134-146. Nissizawa, K., and Y. Hashimoto. 1970. Glycoside hydrolases and glycosyl transferases. In The Carbohydrates, Vol. II-A, p. 241-300. W. Pigman, D. Horton, and A. Herp, eds. Academic Press, New York. Osborn, M. J. 1969. Structure and biosynthesis of the bacterial cell wall. Annu. Rev. Biochem. 38:501-538. Pacsu, E. 1939. Studies on the ketone sugar series. IX. The validity of Hudson's rules of isorotation in the 1-sorbose series. B-Ethylsorboside and its tetraacetate. J. Am. Chem. Soc. 61:2669-2674. Pankratz, H. S., and C. C. Bowen. 1963. Cytology of blue- green algae. I. The cells of Symploca muscorum. Am. J. Bot. 50:387-399. Payen, J. 1938. Recherches biochimiques sur quelques Cyanophycées. Rev. Algol. 11:1-99. Payen, J. 1953. Lew glucides de Phormidium tenue. Meneg. Gom. C. R. Acad. Sci. 236:1811-1813. Pazur, J. H., T. Budovich, and C. L. Tipton. 1957. The enzymatic synthesis and disproportionations of 3-0-a-D- glucopyranosyl-D—glucose. J. Am. Chem. Soc. 79:625-628. Peat, S., W. J. Whelan, and H. G. Lawley. 1958. The structure of laminarin. I. The main polymeric linkage. J. Chem. Soc. 724-729. Petek, F., and To Dong. 1961. Séparation et etude de deux a-galactosidases des graines du café. Enzymologia 23:133-142. Phillips, A. W. 1959. The purification of yeast maltase. Arch. Biochem. Biophys. 80:346-352. Pigman, W. W. 1944. Specificity, classification and mechanism of action of the glucosidases. Advan. Enzymol. 4:41-74. Purdie, T., and J. C. Irvine. 1903. The alkylation of sugars. J. Chem. Soc.:1021-1037. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 114 Purves, C. B. 1954. Cellulose and cellulose derivatives. In High Polymer, Vol. I, p. 116-140. E. Ott, H. M. Spurlin, and M. W. Graffin, eds. Interscience Publishers, Inc., New York. Putman, E. W., and W. Z. Hassid. 1954. Structure of galacto- sylglycerol from Irideae laminarioides. J. Am. Chem. Soc. 76:2221-2223. Reese, E. T., and M. Mandels. 1959. B-D-(l+3)—g1ucanases in fungi. Can. J. Microbiol. 5:173-185. Reese, E. T., and M. Mandels. 1966. B-Glucanases other than cellulases. In Methods in Enzymology, Vol. VIII, p. 607- 615. E. F. Neufeld and V. Ginsburg, eds. Academic Press, New York. Reese, E. T., and A. S. Perlin. 1963. Enzymatic preparation of 3-0-B-cellobiosyl-D-glucose. Biochem. Biophys. Res. Commun. 12:194-197. Reese, E. T., and Y. Shibata. 1964. B-Mannanases of fungi. Can. J. Microbiol. 5:173—185. Reese, E. T., E. Smakula, and A. S. Perlin. 1959. Enzymatic production of cellotriose from cellulose. Arch. Biochem. Biophys. 85:171-175. Rosenfeld, L., and C. E. Ballou. 1974. Acetolysis of disac- charides: comparative kinetics and mechanism. Carbohyd. Res. 32:287-288. Ruesink, A. W., and K. V. Thimann. 1965. Protoplast from the Avena coleoptile. Proc. Nat. Acad. Sci. 54:56-64. Sandford, P. A., and H. E. Conrad. 1966. The structure of the Aerobacter aerogenes A3(Sl) polysaccharide. I. A reexamination using improved procedures of methylation analysis. Biochemistry 5:1508-1517. Schmidt, P. E. 1966. Techniques in Amino Acid Analysis. Chertsey: Technicon Instruments Company, Ltd. Seind, H., and K. Weber. 1963. Untersuchungen fiber milchzucker- spaltende Enzyme. XIII. Grosse und Gestalt der B—Galactosi- dase aus E. coli. Biochem. Z. 337:24-34. Simon, R. D. 1971. Cyanophycin granules from the blue-green alga Anabaena cylindrica: a reserve material consisting of copolymers of aspartic acid and arginine. Proc. Nat. Acad. Sci. (Wash.) 68:265-267. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 115 Shibata, Y., and K. Nisizawa. 1965. Microheterogeneity of B-glucosidases in apricot emulsin. Arch. Biochem. Biophys. 109:516-521. Smith, F., and R. Montgomery. 1959. Lichen polysaccharaides. In Chemistry of Plants, Gums and Mucilages, p. 394-401. Reinhold Publishing Corporation, New York. Stanier, R. Y., R. Kunisawa, M. Mandel, and G. Cohen-Bazire. 1971. Purification and properties of unicellular blue- green algae (order Chroococcales). Bacteriol. Rev. 35: 171-205. Steer, E. Jr., G. R. Craven, and C. B. Anfinsen. 1965. Evi- dence for nonidentical chains in the B-galactosidase of Escherichia coli K12. J. Biol. Chem. 240:2478—2484. Stevens, G. 1955. Cellulase preparation from Helix pomatia (snails). In Methods in Enzymology, Vol. 1, p. 173-178. S. P. Colowick and N. 0. Kaplan, eds. Academic Press, New York. Swanson, M. A., and C. F. Cori. 1948. Studies on the struc- ture of polysaccharides. I. Acid hydrolysis of starch- like polysaccharides. J. Biol. Chem. 172:797-814. Sweeley, C. C., B. D. Ray, W. T. Wood, and J. F. Holland. 1970. On-line digital computer system for high-speed single focusing mass spectrometry. Anal. Chem. 42:1505- 1516. Takano, K. 1956. Acceptor relations in the enzymatic transfer of B-glucosyl and B-galactosyl groups. J. Biochem. (Tokyo) 43:205-216. Talmadge, K. W., K. Keegstra, W. D. Bauer, and P. Albersheim. 1973. The structure of plant cell walls. I. The macro- molecular components of the walls of suspension-cultured sycamore cells with a detailed analysis of the pectic polysaccharides. Plant Physiol. 51:158-173. Thompson, A., M. L. Wblfrom, and E. J. Quinn. 1953. Acid reversion in relation to isomaltose as a starch hydrolytic product. J. Am. Chem. Soc. 75:3003-3004. Tsusué,‘Y.,and.Y3 Fujita. 1964. Mono- and oligo-saccharides in the blue green alga, Tolypothrix tenuis. J. Gen. Appl. Microbiol. 10:283-294. Tsusué, Y., and T. Yamakawa. 1965. Chemical structure of oligosaccharides in the blue green alga, Tolypothrix tenuis. J. Biochem. 58:587-594. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 116 Ulmann, M. 1953. Chromatographische Untersuchungen zur saurehydrolyse der Kartoffelstarke. Makromol. Chem. 10: 221-234. Veibel, S. 1950. B-Glucosidase. In The Enzymes, Vol. I, p. 583-620. J. B. Sumner and K. Myrback, eds. Academic Press, New York. Wallenfels, K., M. L. Zarnitz, G. Laule, H. Bender, and M. Keser. 1959. Untersuchungen fiber milchzuckerspaltende Enzyme. III. Reinigung, Kristallisation und Eigenschaften der B-galactosidase von Escherichia coli ML309. Biochem. Z. 331:459-485. Wallenfels, K., G. Bechtler, R. Kuhn, H. Tirschmann, and H. Egge. 1963. Permethylation of oligomeric and polymeric carbohydrates and quantitative analysis of the cleavage products. Angew. Chem. Int. Ed. 2:515-523. Walsby, A. E., and B. W. Nichols. 1969. Lipid composition of heterocysts. Nature (London) 221:673-674. Wan, C. C., M. D. Urso, J. E. Muldrey, S. C. Li, and Y.-T. Li. 1975. Isolation of B-mannosidase from Polyporus sulfureus. Federation Proc. 34:678. Weber, M., and G. Wbber. 1975. The fine structure of the a-D-glucan from the blue green alga Anacystis nidulans: comparison with other bacterial glycogens and phytoglycogen. Carbohyd. Res. 39:295-302. Weckesser, J., A. Katz, G. Drews, H. Mayer, and I. Fromme. 1974. Lipopolysaccharide-containing L-acofriose in the filamentous blue-green alga Anabaena variabilis. J. Bacteriol. 120:672-678. Weise, G., G. Drews, B. Jann, and K. Jann. 1970. Identifica- tion and analysis of a lipopolysaccharide in cell walls of the blue-green alga Anacystis nidulans. Arch. Mikrobiol. 71:89-98. Whistler, R. L., and W. M. Corbett. 1957. The structures of polysaccharides. In The Carbohydrates: Chemistry, Bio- chemistry, Physiology, p. 644-708. W. Pigman, ed. Academic Press, New York. Whistler, R. L., and J. L. Hickson. 1954. Maltotetraose and crystalline pentadecaacetylmaltotetraitol. J. Am. Chem. Soc. 76:1671-1673. Whistler, R. L., and J. Hickson. 1955. Determination of some components in corn syrups Iby quantitative paper chroma- gotraphy. Anal. Chem. 27:1514-1517. 155. 156. 157. 158. 159. 160. 161. 162 O 163. 164. 165. 166. 117 Winkenbach, F., and C. P. Wolk. 1973. Activities of enzymes of the oxidative and the reductive pentose phosphate path- ways in heterocysts of a blue-green alga. Plant Physiol. 52:480-483. Winkenbach, F., C. P. Wolk, and M. Jost. 1972. Lipids of membranes and of the cell envelope in heterocysts of a blue-green alga. Planta 107:69-80. Wolfrom, M. L., E. N. Lassetre, and A. N. O'Neill. 1951. Degradation of glycogen to isomaltose. J. Am. Chem. Soc. 73:595-599. Wolfrom, M. L., J. T. Tyree, T. T. Galkowski, and A. N. O'Neill. 1951. Acid degradation of amylopectin to iso- maltose and maltotriose. J. Am. Chem. Soc. 73:4927-4929. Wolfrom, M. L., A. Thompson, A. N. O'Neill, and T. T. Galkowski. 1952. Isomaltitol. J. Am. Chem. Soc. 74:1062-1064. Wolfrom, M. L., A. Thompson, and C. E. Timberlake. 1963. Comparative hydrolysis rate of the reducing disaccharides of D-glucopyranose. Cereal Chem. 40:82-86. Wolk, C. P. 1968. Movement of carbon from vegetative cells to heterocysts in Anabaena cylindrica. J. Bacteriol. 96: 2138-2143. Wolk, C. P. 1973. Physiology and cytological chemistry of blue-green algae. Bacteriol. Rev. 37:32-101. Wolk, C. P. 1975. Differentiation and pattern formation in filamentous blue-green algae. In Spores VI, p. 85-96. P. Gerhardt, R. N. Costilow, and H. L. Sadoff, eds. American Society for Microbiology, Washington, D.C. Wolk, C. P., and R. D. Simon. 1969. Pigments and lipids of heterocysts. Planta 86:92-97. Zonneveld, J. M. 1972. A new type of enzyme, an exo-splitting a-(l+3)glucanase from non-induced cultures of Aspergillus nidulans. Biochim. Biophys. Acta 258:541-547. Zweig, G., and R. Whitaker. 1967. Paper Chromatography and Electrophoresis, Vol. I, p. 232-279. Academic Press, New York.