ABSTRACT CARBOHYDRATE METABOLISM I. ACCESSING HETEROZYGOSITY FOR GALACTOSEMIA BY ENZYMATIC MEANS II. METABOLISM OF MANNOSE BY Hansenula holstii by Roger Keith Bretthauer Part I Galactosemia is a metabolic disorder characterized by a deficiency of the enzyme galactose-l-phosphate uridyl transferase. A quantitative assay for this enzyme in erythro- cytes has been developed and applied to 14 families of known galactosemic children and to a group of 106 normal individuals. Three distinct groups of enzyme activities are apparent: affected children having little or none; a group including all parents of affected children having an intermediate level; and normal controls having a full level. These results confirm that this disorder is hereditary and imply that it is transmitted as a simple, autosomal recessive trait. Preliminary studies on comparison of the incidence of galactosemic heterozygotes in a normal and mentally retarded population have resulted in the detection of at least two heterozygotes from the 138 normal individuals examined. Part II The metabolism of mannose by Hansenula holstii NRRL Y—2448 has been investigated in relation to the ABSTRACT (cont) extracellular phosphorylated mannan produced by this organism. Mannose is phosphorylated and isomerized to fructose-B-phosphate, as is glucose. The presence of mutases was demonstrated by the conversion of glucose-l- phosphate and mannose-l—phosphate to the corresponding hexose-B-phosphates. Oxidation of glucose-B-phosphate and 6-phosphogluconic acid occurred in the presence of triphosphopyridine nucleotide. The presence of pyro— phosphorylases capable of catalyzing the formation of guanosine diphosphate mannose, guanosine diphosphate glucose, and uridine diphosphate glucose from the respective nucleoside triphosphates and hexose-l—phosphates was established. No interconversion of guanosine diphosphate hexoses or oxidation of uridine diphosphate glucose was detected. Acid soluble nucleotides, including guanosine diphos- phate hexoses and uridine diphosphate hexoses, were isolated and identified. The guanosine diphosphate hexoses contained mannose and minor amounts of glucose and fructose. The synthesis of phosphomannan by washed, resting cells from glucose and inorganic phosphate and by whole C14 cultures from -mannose is described. No incorporation of radioactivity from C14-guanosine diphosphate mannose C14 or -mannose-1-phosphate into phosphomannan was achieved. CARBOHYDRATE METABOLISM I. \ACOESSJLNG‘ HETEROZYGOSITY FOR GALACTOSEMIA BY ENZYMATIC MEANS II. METABOLISM OF MANNOSE BY Hansenula holstii by Roger Keith Bretthauer A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY 1961 ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr. R. G. Hansen for his guidance and encouragement throughout the course of this work. The technical assistance of Mrs. JoAnne Wilken and Miss Diane DeNio in performing the galactosemia studies is also greatly appreciated. To the author's wife, Donna, go many words of thanks for help in the preparation and typing of this thesis. ii To Donna iii VITA Roger K. Bretthauer was born on January 4, 1935, at Morris, Illinois. He graduated from the Yorkville High School in 1952 and attended Blackburn College for two years. After transferring to the University of Illinois, he received the B.S. degree in Agriculture in 1956 and the M.S. degree from the Department of Dairy Science in 1958. Remaining formal academic training and research requirements for the Ph.D. degree were completed in the Department of Biochemistry at Michigan State University. A postdoctoral position at the Department of Bacteriology, University of Wisconsin, has been accepted. iv TABLE OF CONTENTS INTRODUCTION . . . . LITERATURE . . . . . MATERIALS AND METHODS Part I EXPERIMENTAL PROCEDURES AND RESULTS . . . . . . Development of quantitative assay . . . . Studies on families children . Studies on a normal population DISCUSSION . . . . . SUMMARY . . . . . . LITERATURE CITED . . INTRODUCTION . . . . LITERATURE . . . . . Glycogen Starch . . . Cellulose . . Chitin . . Hyaluronic acid Type III pneumococcal B-1,5-Glucan xylo-oligosaccharides . . . . . . . . . Phosphomannan MATERIALS AND METHODS with galactosemic and mentally retarded Quantitative measurements . . . . . . . . Qualitative measurements . . . . . . . . Chemicals . . Page \OKOCEUJ 40 42 42 1+4 ’45 1+8 49 51 2 52+ 56 60 6O 61 62 TABLE OF CONTENTS (cont.) EXPERIMENTAL PROCEDURES AND RESULTS . . . . . . . Nucleotide Components . . . . . . . . . . . . . Enzymes Involved in Glucose and Mannose Metabolism . . . . . . . . . Preparation of cell extracts Hexokinase . . . . Dehydrogenases . . Isomerases . . . - Mutases . . . . . . Pyrophosphorylases . . . Experiments Concerned with Pre ursors Phosphomannan . . . . . . . . . . . . . . . Phosphomannan synthesis by washed, resting cells . . . . . . . . . . . . . . Synthesis of phosphomannan by whole cultures . . . . . . . . . . . . . . . Studies on synthesis of phosphomannan by cell- free extracts . . . . . . . . . . . 00000000 0 O C DISCUSSION 0 O O O O O O O O O O O O O O O O O O I SUMMARY 0 O O O O O O O I O O O O O O O O O O O 0 LITERATURE CITED . . . . . . . . . . . . . . . . . vi Table II. III. IV. VI. VII. VIII. IX. XI. TABLES Gal-l-P Uridyl Transferase Levels in Erythrocytes . . . . . . . . . . . . . Activity of Gal-l—P Uridyl Transferase in Erythrocytes from Normal Controls . Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Affected Families Activity of Gal-l-P Uridyl Transferase from Relatives of Affected Families . Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Normal Controls . Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Suspected Hetero- zygous Families . . . . . . . . . . . Analysis of Guanosine Diphosphate Hexose and Uridine Diphosphate Hexose Fractions Nonspecific Hexokinase Activity . . . Ketohexose Formation from Hexose-6- Phospha te O O O O O O O O O O O O O 0 Synthesis of Phosphomannan by Washed, Resting Cells . . . . . . . . . . . . Synthesis of Phosphomannan in Whole Cultures . . . . . . . . . . . . . . . vii Page 14 19 20 23 26 28 71 77 104 107 Figure \N 10 11 12 14 FIGURES Page Gal-l-P uridyl transferase activity as a function of hemolyzate concentration, time of incubation, and substrate concentration . . . . . . . . . . . . . . . ll Scattergram of values for Gal-l-P uridyl transferase activity from normals and fourteen families with galactosemic children . . . . . . . . . . . . . . . . . . l7 Scattergram of values for Gal-l-P uridyl transferase activity from normal prison population and parents of assumed heterozygotes . . . . . . . . . . . . . . . 25 Pedigrees of Families 12 and 14 of children with galactosemia . . . . . . . . . . . . . 34 Structure of phosphomannan . . . . . . . . . 58 Elution pattern of acid soluble nucleotides from Dowex-l (formate) column . . . . . . . 67 Ultraviolet absorption spectra for isolated sugar nucleotides . . . . . . . . . . . . . 69 Hexokinase activity . . . . . . . . . . . . 76 Glucose-6-phosphate and 6-phosphog1uconic acid dehydrogenase activity . . . . . . . . 79 Six-phosphogluconic acid dehydrogenase aCtivj-ty O O O O O O O O O C O O O O O 0 C C 82 Phosphohexoisomerase activity . . . . . . . 85 Phosphomannomutase activity . . . . . . . . 87 Phosphoglucomutase activity . . . . . . . . 9O Uridine diphosphate glucose pyrophos- phorylase activity . . . . . . . . . . . . . 92 viii Figure 15 l6 l7 l8 FIGURES (cont.) Page Guanosine diphosphate mannose pyrophos- phorylase activity . . . . . . . . . . . . . 96 Sugars from hydrolyzed guanosine diphosphate hexoses . . . . . . . . . . . . . . . . . . 98 Retention of label in mannose in guanosine diphosphate hexoses . . . . . . . . . . . . 102 Summary of enzymic reactions demonstrated in Hansenula holstii . . . . . . . . . . . . 112 ix CARBOHYDRATE METABOLISM I. ACCESSING HETEROZYGOSITY FOR GALACTOSEMIA BY ENZYMATIC MEANS INTRODUCTION Recent advances in biochemical studies of hereditary metabolic disorders have revealed in many cases that the disorder is characterized by an enzymic deficiency. Hetero- zygous carriers of these disorders have been detected by various techniques, in some instances by direct measurement of the deficient enzyme. Many of these disorders are thought to be transmitted by autosomal recessive genes, as both parents of an affected individual appear to be heterozygous with respect to the trait. One such metabolic disorder which has been shown to be characterized by a single enzymic deficiency is galactosemia. The lack of Gal-l-P uridyl transferasel results in an inabil- ity to convert galactose, a normal constituent of milk, to the common metabolite glucose. The affect is clearly illus- trated by the in vitro accumulation of Gal-l-P in erythro- cytes when incubated with galactose. If death does not occur, the secondary abnormalities which result, if all dietary galactose is not removed from the affected person, are cataracts, mental retardation and enlargement of the liver. 1The following abbreviations are used: Gal-l-P, galactose-l-phosphate; G-l-P, glucose-l-phosphate; ATP and ADP, adenosine tri- and diphosphate; UTP and UDP, uridine tri- and diphosphate; UDPG, uridine diphosphate glucose; UDPGal, uridine diphosphate galactose; PP, inorganic pyro- phosphate: DPN, oxidized diphosphopyridine nucleotide. The discovery of the biochemical defect of galactosemia presents an unique opportunity to study the enzymic expression of a heterozygous state and to elucidate the mode of inheri— tance of this disorder. Previous methods have not been sufficiently sensitive to rigorously define this problem. The purpose of this thesis is to present a method which defines the heterozygote, and which, when applied -to family studies, has revealed an autosomal recessive mode ()f inheritance. Preliminary studies on the incidence of Inerterozygosity in a normal and mentally retarded population a re also presented . LITERATURE The first report of galactosemia was in 1908 by Von Reuss (1). Since that time, many other cases with similar clinical features have appeared in the literature. In general the symptoms of the fully manifested disease are jaundice, hepatomegaly, cirrhosis, cataracts, galactosuria, amino- aciduria, and failure to grow. Normally these symptoms occur soon after birth and if not detected, may result in severe cataracts, mental retardation, and often death. Early exclusion of all dietary galactose was found to re- lieve this condition, and thus the eventual finding of an enzymic deficiency in galactose metabolism as the bio- chemical defect (2,3). Schwarz BE 11. (4) noted an accumulation of Gal-l-P in erythrocytes of galactosemic persons when incubated in the presence of galactose. The significance of this dis- covery was realized when Kalckar and associates (5-7) demonstrated a deficiency of Gal-l-P uridyl transferase (Insection 2 below) in both erythrocytes and liver of affected 1i11CLividuals. Other enzymes involved in galactose metabolism (reactions 1, 3, and 4 below) were present in normal amounts. The discovery of such an enzymatic deficiency was I>C>=3ssib1e only after much detailed experimentation with v’a-ZIZ‘ious biological sources revealed the following pathway of galactose metabolism: 4 1) Galactose + ATP GalaCtOklnaS§> Gal—l-P + ADP Gal-l-P uridyl 2) Gal—l-P + UDPG ..tran3ferases G-l—P + UDPGal UDPGal-4-epimerase 3) UDPGal‘l, T’UDPG UDPG pyrophosphorylase 4) G-l-P + UTP \ s UDPG + PP UDPGal4pyrophosphorylase‘ 5) Gal-l-P + UTP UDPGal + PP The irreversible phosphorylation of galactose resulting in Gal-l-P is followed by its conversion to G-l-P (reaction 2) which involves UDPG. UDPGal is formed which supplies the galactose essential for galactolipids, mucopolysaccharides and other galactose containing compounds. In the absence of this reversible reaction, Gal-l-P accumulates and thus the utilization of galactose is affected. In reaction 5 UDPGal is converted to UDPG by the enzyme UDPGa1-4-epimerase. As this reaction is reversible, UDPGal can therefore be formed from UDPG, which in turn can be synthesized from <}-l-P via reaction 4. These reactions offer an explanation Of why the body can satisfy its needs for galactose even on a galactose free diet. Reaction 5, catalyzed by UDPGa1- Pyr0phosphorylase, offers an additional alternative pathway 135? Vvhich Gal-l-P can be incorporated into the nucleotide UDPGal. Whether Gal-l-P uridyl transferase is completely absent :iI)- syalactosemic persons is not clear. Simon and Topper (2) fouhd that erythrocytes of galactosemic persons oxidized galactose at about 0.5% of normal. Eisenberg et_g;, (8) administered C14-ga1actose and menthol to a galactosemic adult and found about 3% of the radioactivity in the urine in 24 hours as the glucuronide of menthol. Thus it appears that the block at reaction 2 may not be complete, or that galactose is being metabolized via alternate routes. One such possible by-pass, UDPGal-pyrophosphorylase (reaction 5), has been demonstrated in mammalian tissues (9). Earlier reports of this disorder mentioned that siblings of both sexes were affected and that parents were pheno- typically normal. Such reports suggest an autosomal, re- cessive mode of inheritance. Holzel and Komrower (10), employing an oral galactose loading test, demonstrated ab- normal galactose tolerance in one parent in each of four families. In one instance of consanguinity, both parents exhibited abnormal tolerance. While suggesting a hereditary etiology, the data did not prove a recessive mode of inheritance, in which.both parents should be heterozygous. In a more esctensive application of this method, Donnell §t_§1, (11) found that a group of mothers and fathers of affected children <=<>tnld be distinguished from a group of normals on a statistical 133813. However, the overlap of individual values was so c<>Ilsiderab1e that the technique has to be regarded as un- suitable for defining the heterozygous condition. Similar lfiidmnuitations apply to the UDPG consumption test, originally de Signed by Kalckar and coworkers (5-7) to detect the homozygous condition. The kinetic limitations of this test were pointed out by Anderson _t__l, (12), who were unable to detect decreased activity of Gal-l-P uridyl transferase in parents of affected children. Hsia et a1. (13) have employed this previous method in attempting to identify heterozygotes, but examination of their data reveals ex- tensive overlap between normals and assumed heterozygotes. Concurrent with the studies to be reported in this thesis, Kirkman and Bynum (14) developed a manometric assay based on the oxidation of Gal-l-P by hemolyzates under conditions of limiting transferase levels. Application of this method to 18 parents of galactosemic children re- sulted in approximately 64% of the activity as in normal controls. However, significant overlapping of groups was still apparent. This technique is also quite complex and costly, thus limiting its practical usefulness. In an attempt to preserve the relative simplicity cxf the original UDPG consumption test as described by Kalckar arnd coworkers (5-7), and unequivocally define the mode of ilulneritance of this disorder, the assay was quantitated kfi?’ altering substrate concentrations, hemolyzate concen- ‘tJrEitzion, and time of incubation. The description of this Eissfiséay and results of its application to 14 families each ha-‘7.:i_ng at least one member affected by galactosemia are Pre sented in this report. The numerous case reports of galactosemic persons have revealed that frequently mental development is impaired. It was, therefore, of interest to apply the quantitative transferase assay to a mentally retarded population and see if a significant number of these individuals have ab- normal galactose metabolism. Concurrent with these studies, examination of a normal population as controls might reveal the incidence of heterozygosity and homozygous recessives. Preliminary results of these studies are also presented in this report. MATERIALS AND METHODS UDPG and DPN were purchased from Sigma Chemical Company. The a-D-Gal-l-P was synthesized according to the procedure described by Hansen §;_gl, (l5). UDPG dehydrogenase was isolated from calves liver and purified through Step 5 of the procedure of Strominger §t_a1. (16). This prepara- tion contained 37,500 units/m1. UDPGa1-4-epimerase was purified from Saccharomyces fragilis and kindly furnished by Dr. R. J. Forrest (unpublished procedure). This enzyme was also purified by the method of Maxwell and Szulmajster (17) and used successfully. Both preparations contained about 50,000 units/m1. Spectrophotometric determinations were carried out in a Beckman DU spectrophotometer with a photomultiplier attachment and pinhole filter. EXPERIMENTAL PROCEDURES AND RESULTS Development of quantitative assay Blood samples were drawn by venipuncture into balanced oxalate tubes. They were iced and centrifuged in the cold within 30 to 60 minutes. The erythrocytes were washed three times with cold 0.9% NaCl. The last centrifugation was done under standard conditions to provide uniform packing. The erythrocytes were stored frozen until use. Prior to measurement of the enzyme, a hemolyzate was prepared by thawing the frozen erythrocytes and adding an equal volume of cold distilled water. The principle of the assay is similar to the most recent enzymic assays published by Anderson et a1. (12) and Maxwell gt a1. (18) with the following major exceptions: the present procedure employs a higher concentration of UDPG and a higher ratio of Gal-l-P to UDPG, less hemolyzate, and a shorter incubation period. A four-fold excess of (ial-l-P over that of UDPG was necessary for optimum trans— fExrase activity. ‘While using these assay conditions, namely 0. 20 umole UDPG and 0.80 Limole Gal-l-P, the amount of hemoly- zate was varied in a 15 minute incubation period to determine 't1163 optimum.enzyme concentration. In similar experiments, ‘tllfiar substrate concentration and the incubation time were V13J2?jied.while the amount of hemolyzate was held constant ‘3‘: (3.20 ml (Fig. 1). Reasonably good linearity was achieved 10 Figure 1. Gal-l-P uridyl transferase activity as a function of hemolyzate concentration, time of incubationII and substrate concentration. Unless otherwise indicated, the reaction mixture contained 100.0 umoles of glycine, pH 8.7, 0.2 umole UDPG and 0.8 umole Gal-l-P (always added in a 1:4 ratio), and 0.2 m1 hemolyzate in a total volume of 0.45 ml except for those incubations with varying hemolyzate concentration where the final volume was 0.7 m1. Incubation time was 15 minutes at 37° C. Remaining details of assay are described in the text. ‘A, variable hemolyzate concentration; B, variable time cxf incubation; C, variable substrate concentration. 11 ll. amoo< 30.3: «0.06: owe 9.0 3.0 o 1 5:43:02. mmhnzi 1 8.0 O 1 O. .0 1 9.0 m I :mENfionI J: - ON 0 GINO owd medon SNOW" GBWflSNOO 12 until the UDPG became limiting at approximately 0.50 ml hemolyzate (curve A) or at 20 minutes incubation time (curve B). In the presence of excess substrate (more than 0.12 umole UDPG), incubation time and/or hemolyzate became limiting (curve C). The difference in slopes of these three curves resulted from using different blood samples for each of the three experiments. For routine determinations the following conditions were selected which afforded a wide range of measurable values with the same assay conditions: 0.20 ml hemolyzate, 0.20 umole UDPG, 0.80 umole Gal-l-P, 100.0 umoles glycine, pH 8.7, in a total volume of 0.45 ml incubated 10 minutes at 37° C. A control incubation lacking the Gal-l-P was run with every sample under identical condi- tions. All incubations were run in cellulose nitrate tubes. The reagents were all added to the cellulose nitrate tubes in an ice bucket, the hemolyzate being added last. At the moment of placing the tubes in a 37° C water bath, timing was started. At 10 minutes the tubes were placed in a boiling water bath for one minute with shaking, cooled, and the pre— cipitate centrifuged off. The clear supernatant was used for the Spectrophotometric UDPG determination. The microcuvettes contained 0.10 umole DPN, 45.0 umoles glycine, pH 8.7, and 0.01 ml of incubation supernatant in a total volume of 0.48 ml. Having the control cuvette (lacking incubation supernatant) set at zero absorption, the absorbancy at 340 mu of the two test cuvettes was recorded. 13 UDPG dehydrogenase (0.01 ml) was then added to the control cell and reset to zero absorption before a similar addition to each test cell. After the increase in optical density had ceased, 0.01 ml UDPGa1-4-epimerase was added to each cell in the same manner as described for the dehydrogenase and further reduction of DPN recorded. The difference in optical density increases of the two test cells before addition of the epimerase was calculated as umoles UDPG consumed in the transferase catalyzed reaction. The differ- ence in increases of optical density of the two test cells after addition of epimerase was calculated as umoles UDPGal formed. The UDPGal recovered was 85 to 105% of the UDPG consumed. In the application of the method, three families were available for examination to determine if the modified assay would clearly distinguish the heterozygotes from the clinical patient and normal controls. The results of such studies are presented in Table I. Enzyme levels are expressed as umoles UDPG consumed per hour of incubation per ml of erythrocytes and per gram of hemoglobin (determined according to Anderson et a1. (12)). Normal controls ranged from 4.8 to 8.1 umoles UDPG/hr/ml erythrocytes. Of special significance is that in every family, both parents exhibited a lower value of transferase, ranging from 2.9 to 3.7 umoles UDPG/hr/ml erythrocytes. AS expected, no activity was exhibited by the galactosemics. 11+ Table I. Gal-l—P Uridyl Transferase Levels in Erythrocytes umole UDPG/hr Individual sex) per m1 erythrocytes per gm Hb Family 1 Affected child 0 0 Mother 2.9 8.7 Father 3.5 10.3 Sibling (F) 3.0 9.0 Sibling (M) 7.0 21.2 Sibling (F) 6.0 '- 2 Affected child 0 0 Mother 3.7 10.8 Father 3.6 10.7 3 Affected child 0 0 Mother 3.0 9.4 Father 3.6 10.7 Controls 9 Child male 7.6 23.0 10 Child male 6.9 21.5 11 Adult male 8.1 25.4 12 Adult female 5.9 18. 25 Adult male 6.0 17.7 51 Adult female 6.0 18.3 52 Adult female 5.0 14.9 53 Child female 6.5 21.7 58 Child male 5.0 15.8 60 Adult male 4.8 14.7 61 Adult female 5.0 14.7 man Galactosemics 0 0 Carriers 3.33 9.911L Normals 6.14 19.0 Barge Galactosemics 0 0 Carriers 2.9-3.7 8.7-10.8 Normals 4.8-8.1 14.7-25.4 15 Other members of the family resulted in either normal or intermediate levels of enzyme. This data clearly demon- strates that the assay is quantitative and capable of de- tecting the heterozygous condition with no overlap of values. As both parents exhibited an intermediate level of transferase, a recessive mode of inheritance is implicated. Studies on families with galactosemic children A more detailed genetic study of several more known galactosemic families was undertaken to more clearly define the expected range of transferase levels for galactosemics, heterozygotes, and normals. Fourteen families, including the three previous families, were available in which a known case of galactosemia had occurred. Blood samples were obtained from affected children, their siblings, their mothers, and from all but three of the fathers. Also, blood samples were obtained from a great number of other relatives. Normal controls were chosen from resident physicians, nurses, and other adults and children in which no family history of galactosemia was evident or in which no history of mental retardation, liver dysfunction, or metabolic disease was evident. All of the previous samples were taken at the Childrens Hospital of Los Angeles, packed in dry ice, and shipped by air to East Lansing. A total of 278 individuals were examined. Fig. 2 presents a scattergram of all the data obtained divided 16 Figure 2. Scattergram of values for Gal-l-P uridyl transferase activity from normals and fourteen families with galactosemic children. The lines in each column represent the mean value. 0405.200 Somhzoo 82.5.5.5 mu>F<4wm dmhzoo w.._<2 m4<2mu 01:10 .5304 04.10 moz_._m.m .530< mmehs. .5304 91.5.2.0! l7 . ...m u I.) .H ...m In: I“: .1 .. m ... .m . 0 cm on o. c on O M O! O llml ..I...l. .m . ..u :I .H ..mm In») ll." .lall .l.... on 000m an o u com com o H o u .u u . u m o on H o 0.. 0.N 0.n 0.? 0.0 0.5 0.0 0.0 m /so:£aomuuo [w Amman» msdan clown 18 into the various groups considered. The distribution into normal and heterozygous categories is evident, including the expected division among relatives and siblings of affected children. Table II summarizes the mean values for Gal-l-P uridyl transferase from erythrocytes of 106 normal controls, sub- divided by age and sex. The mean values of the groups and their standard deviation approximate each other. The mean value for normal controls has been taken as 5.9 i_1.0 umole UDPG consumed/ml erythrocytes/hr. Data from members of affected families are presented in Table III. Some transferase activity was found in erythro— cytes from 3 of the 14 affected children. Two of these had previously received blood transfusions. Values of transferase in the erythrocytes of the parents studied averaged 2.9 i 0.83, and all values were less than 3.9 umoles UDPG consumed/ml erythrocytes/hr. All controls had values above 4.0, with one exception at 3.8. A value of 4.0 was arbitrarily selected as the value which differ- entiated heterozygotes from normals. Values of transferase in erythrocytes from unaffected siblings of affected children ranged from 1.1 to 7.0. Application of the arbitrary di- viding value of 4.0 yielded two groups with values for means and standard deviations of 3.0 i_0.87 and 6.0 i 0.78, respectively. The similarity of these means and standard deviations to those obtained with known heterozygotes 19 Table II. Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Normal Controls Gal-l-P Uridyl Transferase (umole UDPG/m1 erythrocytes/hr) Mean ...... 3:322:12. Adult males - 28 5.9 4.0-8.3 1.06 Adult females 39 5,8 3,8-7,8 0.95 Child males 21 6.0 4.3-7.4 0.98 Child females 18 6.0 4.1-8.0 1.14 All adults 67 6.0 3.8-8.3 0.99 All children 39 6.0 4.1-8.0 1.05 All males 49 5.9 4.0-8.3 1.02 All females 57 5.9 3.8-8.0 1.01 Total control group 106 5.9 3.8-8.3 1.01 k 20 Table III. Activity of Gal—l—P Uridyl Transferase in Erythrocytes from Affected Families Gal-l-P Uridyl Transferase Faggly Inggigiual (“mole apps/m1 ‘ erythrocytes/hr) 1 .Affected child 0 Mother 3.7 Sibling (F) 4.9 Sibling (M) 3.4 2 Affected child 0 Father 3.4 Mother 2.8 Sibling (F) 2.9 Sibling (M) 6.9 Sibling (F) 5.9 3 Affected child 0 Father 3.5 Mother 3.7 4 Affected child 0 Father 3.5 Mother 2.9 5 Affected child 0.6 Father 3.8 Mother 1.8 Sibling (F) 3.4 6 Affected child 0 Father 3.2 Mother 3.3 Sibling (M) 5.8 Sibling (F) 2.1 Sibling (F) 2.3 7 .Affected child 0 Father 3.0 Mother 0.9 Sibling (M) 1.1 8 Affected child 0.3 Mother 3.0 21 Table III. (cont.) . . . Gal-l-P Uridyl Transferase Fagily Indégidual (umole UDPG/m1 ‘ erythrocytes/hr) 9 Mother 309 Sibling (M) 3 , 9 Sibling (M) 3.9 Sibling (M) 3.1 Sibling (M) 5.7 Sibling (F) 7.0 10 Affected child (M) 0 Affected child (F) 0 Father 3.3 Mother 2.9 11 Father 1.6 Mother 2.0 Sibling (M) 5.6 12 Affected child 0 Father 3-8 Mother 2.4 Sibling (M) 3.4 13 Affected child 0 Father 2.8 Mother 1.5 l4 Affected child 0.64 Father 1.9 Mother 3-0 Summary Mean Range S.D. Affected children (13) 0.1 Fathers and mothers (25) 2.9 0.9-3.9 0.83 Fathers (11) 3.1 1.6-3.8 0.72 Mothers (14) 2.7 0.9-3.9 0.89 Siblings (17) 4.1 1.1-7.0 1.7 Assumed heterozygote (11) 3.0 1.1-3.9 0.87 .Assumed normal (6) 6.0 4.9-7.0 0.78 g 22 (2.9 i 0.83) and normal individuals (5.9 i_1.0) is apparent. On a theoretical basis, two-thirds of the unaffected siblings would be expected to be heterozygotes. In the division into groups as made, 11 siblings fall into the category of heterozygotes out of a total of 17. The results from the 117 relatives studied are given in Table IV. When the data are divided into the same two groups, close correspondence with mean values for hetero- zygotes and normals is again evident. Studies on a normal and mentally retarded population Samples of blood were obtained from inmates at the State Prison of Southern Michigan for a control group of physically and mentally normal individuals. As a random population was desired, volunteer inmates were sampled with no attempt at selection. Samples were drawn, washed, and assayed for transferase activity under conditions identical to those previously described. A total of 132 individuals were studied. The transferase activities of these subjects are presented in Fig. 3 as a scattergram. Pertinent data of these studies are presented in Table V. As the previous studies revealed an arbitrary dividing line of 4.0 umoles UDPG consumed/ml erythrocytes/hr for the normal and heterozygous groups, this group was therefore divided into a subgroup of values 4.0 and above and a subgroup of 3.9 and below. In addition the group was analyzed 23 Table IV. Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Relatives of Affected Families Gal-l-P Uridyl Transferase (umole UDPG/ml erythrocytes/hr Mean Renae 22222222. Individuals assumed as heterozygotes* Adult males 10 2.7 2.4-3.7 0.71 Adult females 19 2.9 1.8-3.9 0.66 Child males 6 2.8 1.6-3.8 0.81 Child females 4 2.6 1.4-3.3 0.86 Total group 39 2.8 1.4-3.9 0.70 Individuals assumed as normals* Adult males 12 5.6 4.1-7.8 1.14 Adult females 38 5.7 4.1-7.2 0.91 Child males 12 6.1 4.3-8.5 1.17 Child females 16 5.6 4.3-7.4 0.90 Total group 78 5.7 4.1-8.5 0.99 * Arbitrary division at 4.0 on basis of results from parents of known affected children. 24 Figure 3. Scattergram of yalues for Gal-l-P uridyl transferase activity from normal prison population and parents of assumed heterozygotes. 25 A G O A m . o «moan u— cm a . «rm gwfidmmmmummmrr. r" . A. O O. D O. O. O. O. O. O. D O E \ 3382.»... .E \toeaucoo some: 3.2:: FATHERS MOTHERS PRISONE RS 26 Table V. Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Normal Controls umoles UDPG/ml erythrocytes/hr Group No. Mean Range Standard ev1ation 4.0 and above 126 6.0 4.0-7.9 0.60 3.9 and below 6 3.1 3.3-3.9 0.58 Complete 132 5.9 2.2-7.9 0.60 27 as a whole. The group above 4.0 has a mean value of 6.0, which corresponds very well to the normal mean value found previously of 5.9. Values below 3.9 have a mean of 3.1, corresponding closely to 2.9 which was found for fathers and mothers of galactosemic persons. To ascertain that the values below 4.0 were real and reproducible, the three lowest values were twice redetermined, each time on a fresh sample of blood. Similar values were obtained. Parents of two of these suspected heterozygotes were sampled, along with the wife and two children of one. The data obtained are presented in Table VI. In Family 1, the father exhibited a transferase value characteristic of the heterozygous range, while the mother was normal. In Family 2, both parents exhibited heterozygous levels of transferase while the wife and children both showed normal values. The heterozygous condition of the two inmates are thus confirmed in that at least one of the parents in each case carries the recessive trait. Preliminary studies on a mentally retarded population have revealed that the overall distribution of transferase activities is lower, averaging around 5.0. Many values below 4.0 were observed with apparently no sharp dividing Line. Whether this lower distribution is significant must await further study with a larger population. Also, repeat determinations of the low values and family studies are 28 Table VI. Activity of Gal-l-P Uridyl Transferase in Erythrocytes from Suspected Heterozygous Families Family . . umoles UDPG/ml No. Ind1v1dual erythrocytes/hr 1 Suspected heterozygote 2.2 Mother 6.0 Father 3-7 2 Suspected heterozygote 2.7 Mother 3.7 Father 3-7 Wife 6.2 Daughter 4.6 Son 5.1 29 required to ascertain whether or not these are truly heterozygotes. It is possible that in such a population, the normal range of transferase activities is signficantly lowered so that the heterozygous range would have to be reevaluated. Further studies on this problem are in progress. DISCUSSION The present procedure for the assay of Gal-l—P uridyl transferase has proven practical and sufficiently sensitive to permit identification of the heterozygote. Application of the method in studies of families was successful in making this identification before knowledge of the family relationships. It is apparent from the results presented ,that three different groups are defined. The first con- sists of the homozygotes with essentially no transferase activity. The second consists of the normal controls and some family members with normal values for transferase activity. A third group, heterozygotes, with intermediate levels of transferase activity was derived entirely from members of affected families. No significant statistical difference exists between the means of the various groups into which the 106 normal controls were divided (t = .99, P 2*.05). Therefore, the overall mean value and standard deviation of 5.9 i_1.0 have been taken as the mean and standard deviation of all normal individuals (Table II). ‘When fathers and mothers of affected children are considered together as one group, the mean value for transferase and the standard deviation are 2.9 1 0.83 (Table III). No statistically significant difference could be shown to exist between fathers and mothers (t = 1.3, 30 31 p < .10) A highly significant difference was found when mean values from fathers and mothers of known homozygotes are compared with the mean of the entire group of 106 normal controls (t = 13.9, p < 10‘6). In Table III it is noted that transferase values of some parents are below 2.0 (Families 5, 7, ll, 13, 14). The possibility exists that these low values are in the upper end of the range for homozygotes, in fact that the parents are homozygotes with a very mild form of the disease. However, every well-documented galactosemic studied has essentially no transferase activity in erythrocytes. Further, patients with a mild form of galactosemia should show milk intolerance or an abnormal response to galactose loading, and none of the parents in question showed any such indications (11). All of the parents in this study were without clinical symptoms and physical characteristics of galactosemia. The most convincing evidence for the heterozygosity of the parents lies in the findings on grandparents which were available for study in Families 5, 6, 8, 11, 12, and 14. In Families 5, 6, 11, and 14, one parent of each heterozygous parent of the affected child was found to have an intermediate transferase level. In Families 8 and 12, grandparents on both sides of the affected childs family were not available, but those which were examined again revealed one heterozygous grandparent. In Families 12 and 14 (Fig. 4), the heterozygous condition was found in one great-grandparent. Consequently, the data on the grand— parents indicate heterozygosity in the parents with low values. Only one normal case overlaps values assigned as heterozygous (Table II) but some values in both normal and heterozygous groups are close to the arbitrary dividing line. Normal distribution curves constructed from the data revealed a theoretical overlap of only 5% of the total area of both curves combined. This is insignificant as a population with an equal number of heterozygotes and normal individuals is entirely hypothetical. In a random population, the incidence of galactosemia must be so low that the possibility of heterozygosity is minimal in an indi- vidual whose erythrocytes yield a transferase value near 4.0. In families of children with galactosemia, it is probable that an individual near the 4.0 region is hetero- zygous. In both cases, family studies are required to clarify the point. The studies on prisoners, which were assumed to be a normal population, have revealed that the heterozygote appears with an incidence of at least 2 per 132 persons. This is a minimum figure, as only 2 of the 4 persons reveal- ing a low transferase level in Fig. 3 were further clarified by showing at least one parent in each case to also have \N \N Figure 4. Pedigrees of Families ;2 and 14 of chiLdren with galactosemia . ., galactosemia; O, heterozygous range; CD, normal range; 0, not tested; d, deceased. a transferase level characteristic of the heterozygous range. Assuming two heterozygotes (aA) from a total of 132 individuals, the frequency (q) of the recessive gene (a) would be one in 132. Therefore, the frequency (q2) of the homozygous recessive (aa) would be approximately one in 17,400. For comparison, the homozygous recessive dis- order phenylketonuria has been reported (19) to occur with an incidence of one in 25,000 (U.S.A.). Comparison of the incidence of the heterozygous state Of galactosemia in a mentally retarded population might reveal a significant difference from normal. Indeed, Lippman (19) has proposed that mental deficiency may be correlated with heterozygosity for many different metabolic errors occurring simultaneously. Preliminary studies have revealed a large number of mentally retarded persons having transferase levels in the heterozygous range (below 4.0). However, these must be clarified by family studies and evaluated with respect to the overall distribution obtained from such a population. As opposed to the possible complications of the galactose tolerance test, the direct measurement of the defective enzyme in erythrocytes is a practical approach to detection of the heterozygote for galactosemia. A Simplified colorimetric method, based on the quantitative procedure described in this thesis, has recently been developed by Nordin gp_§;, (20). Thionicotinamide 36 diphosphopyridine nucleotide is used in place of DPN in the UDPG dehydrogenase catalyzed oxidation of UDPG, and as the reduced form of this DPN analog has an absorption maximum at 400 mu, a standard visual spectrophotometer will serve for the assay. This method has been demonstrated to be as equally reliable in differentiating the hetero- zygote as the method described in this thesis, both being useful tools in confirming a previous diagnosis, in clinical study, and in genetic counseling. S UMMARY A quantitative assay for Gal-l-P uridyl transferase in erythrocytes has been developed and applied to 14 families in which a known case of galactosemia has occurred. The method is sensitive enough to categorize the individuals examined into three groups: the galactosemic children having little or no enzyme (homozygous recessives); a group, including all parents, having an intermediate level of enzyme (heterozygotes); and a group with a full level of enzyme. These results substantiate the conclusion that galactosemia is a hereditary metabolic disorder and imply that it is transmitted as a simple autosomal recessive trait. Preliminary data on the incidence of heterozygosity in a normal and mentally retarded population are presented. ll. 12. 15. 16. LITERATURE CITED Von Reuss, A., Wien. Mad. Wghpschr., 58, 799 (1908). Kirkman, H. N., Metabolism,_9, 316 (1960). Holzel, A., Brit Med Bull., 11, 213 (1961). Schwarz, V., Goldberg, L., Komrower, G. M., and Holzel, A., Biochem. J., 62, 34 (1956). Kalckar, H. M., Anderson, E. P., and Isselbacher, K. J., PpocI NatlI AcadI Sci., 42, 49 (1956). Isselbacher, K. J., Anderson, E. P., Kurahashi, K., and Kalckar, H. M., Science, 123, 635 (1956). Kalckar, H. M., Anderson, E. P., and Isselbacher, K. J., BiochimI et BiophysI Acta, 20, 267 (1956). Eisenberg, F., Isselbacher, K. J., and Kalckar, H. M., Science,;25, 116 (1957). Isselbacher, K. J., J Biol Chem., 232, 429 (1958). Holzel, A., and Komrower, G. M., Arch Dis Childhood, .10, 155 (1955). Donnell, G. N., Bergren, W. R., and Roldan, M., Pediatrics, 4, 418 (1959). Anderson, E. P., Kalckar, H. M., Kurahashi, K., and Isselbacher, K. J., J Lab and Clin Med., 50, 469 (1957). Hsia, D. Y. Y., Huang, 1., and Driscoll, S. G., Nature, 102, 1389 (1958). Kirkman, H. N., and Bynum, E., Ann Human Genet., 23, 117 (1959). Hansen, R. G., Rutter, W'. J., and Krichevsky, P., in W V 1 4, Westerfeld, ww., ed., New York, John‘Wiley and Sons, 1955, pg 1. Strominger, J. L., Maxwell, E. S., Axelrod, J., and Kalckar, H. M., J Biol Chem., 224, 79 (1957). 38 17. l8. 19. 20. 39 Maxwell, E. S., and Szulmajster, H. de R., JI Biol. Shem” 23.5., 308 (1959). Maxwell, E. S., Kalckar, H. M., and Bynum, E., JI Lab. gnd Clin. Med., 1m, 478 (1957). Lippman, R. W., Am J of Mental Def., 63, 320 (1958). Nordin, J. H., Bretthauer, R. K., and Hansen, R. G., c;in. Chim. Acta, 6, 578 (1961). CARBOHYDRATE METABOLISM II. METABOLISM OF MANNOSE BY Hansenula holstii INTRODUCTION Recent investigations have revealed that sugar nucleotides serve as glycosyl donors in the synthesis of polysaccharides. Primer molecules are usually required which serve as the glycosyl acceptors, resulting in the release of nucleoside diphosphate. From thermodynamic considerations, synthesis from sugar nucleotides is a much more favorable route than the classical phosphorylase type of reaction, suggesting the possibility that the former reaction is the physiological synthetic route and the latter the route for breakdown. Demonstrated thus far has been the biological synthesis of glycogen, starch, cellulose, and a B-l,3-—glucan from UDPGl hyaluronic acid from UDPGA and UDPAG, Type III pneumoccal polysaccharide from UDPGA and UDPG, and chitin from UDPAG. Evidence for the participation of sugar 1Abbreviations used are: UMP, UDP, and UTP, uridine mono-, di-, and triphosphate; AMP, ADP, and ATP, adenosine mono-, di-, and triphosphate; GMP, GDP, and GTP, guanosine mono-, di-, and triphosphate; DPN and DPNH, oxidized and reduced diphosphopyridine nucleotide; TPN and TPNH, oxidized and reduced triphosphopyridine nucleotide; UDPG, UDPGA, UDPGalA, and UDPAG, uridine diphosphate glucose, glucuronic acid, galacturonic acid, and N-acetylglucosamine; GDPM and GDPG, guanosine diphosphate mannose and glucose; M—l-P, mannose-l-phosphate; M-6-P, mannose-6-phosphate; G-l-P, glucose-l-phosphate; G-6-P, glucose-6-phosphate;, G-l,6-diP, glucose-1,6-diphosphate; 6-PGA, 6-phosphogluconic acid; F-6-P, fructose-6-phosphate; Pi, inorganic phosphate; PP, inorganic pyrophosphate; EDTA, ethylenediaminetetra- acetic acid; Tris, tris(hydroxymethyl) aminomethane. 40 41 nucleotides in polysaccharide synthesis received further support upon demonstration of the presence of acid soluble nucleotides containing the glycosyl monomer of the polysacchar- ide and enzymes required for the synthesis of these sugar nucleotides from free hexoses. The monomeric units of an exocellular phosphorylated mannan produced by various strains of Hansenula was shown by D. R. Wilken (unpublished data) to result from incorporation of glucose units intact. From known enzymic reactions, incorporation of glucose into this polysaccharide might occur by several pathways such as: l) phosphorylation of glucose and subsequent isomerization and mutation to M-l-P which might serve as a glycosyl donor in a classical phosphorylase type reaction; 2) pyrophosphorolysis of the M-l-P with GTP to form GDPM which could serve as a glycosyl donor; 3) pyrophosphorolysis of G-l-P with GTP to form GDPG followed by epimerization to GDPM (not a known enzymic reaction). The present report is concerned with the bio— synthesis of this phosphomannan, including identification of nucleotides and enzymes which might be assumed to function in the conversion of hexose to polymer, and attempts to synthesize polymer directly from GDPM. LITERATURE Gl c en The pioneering work in glycosyl transfer from sugar nucleotides to polysaccharides, reported by Leloir and Cardini (l) in 1957, was the discovery of the catalysis by a rat liver enzyme preparation of glucose transfer from UDPG to primer glycogen. More detailed studies of this reaction carried out by Leloir e; i1. (2) with a rat muscle enzyme revealed the requirement of G-6-P for maximal activity. Among many other compounds tested, glucosamine-6-P, F-6-P, and galactose-6-P were found to be active. The function of the activator remains obscure, however, as radioactive G-6-P was not incorporated into the synthesized glycogen in the presence of unlabeled UDPG. Primer polysaccharide was needed to attain maximal activity. Native glycogen was a better primer than partially degraded products while soluble starch was approximately 20% as active as glycogen. The polysaccharide formed in this system was carefully characterized. In addition to being stable in hot alkali and insoluble in 50% ethanol as anticipated, the product was degraded to its expected components. Glycogen formed from C14-UDPG yielded radioactive maltose and a trace of glucose upon hydrolysis with B-amylase. Hydrolysis of this maltose with maltase produced only radioactive glucose. Also, degradation of radioactive glycogen with 42 a; phosphorylase followed by acid hydrolysis of the phosphate ester formed resulted in only radioactive glucose. Robbins pp a1. (3) found a similar enzyme in pigeon breast muscle which was concentrated in microsomal particles. Since this enzyme contained a high concentration of glycogen, it was suggested that the enzyme is actually adsorbed to the polysaccharide. Leloir and Goldenberg (4) extended the studies with the liver enzyme and found it also to be associated with polysaccharide in the particulate fraction. Washing this fraction with soluble starch resulted in a 300-fold puri- fication of the enzyme, which still retained requirements for G-6-P and primer. Reversal of the reaction was attempted but not achieved. Similar glycogen-synthesizing enzymes have been described from human skeletal and heart muscle (5), and from brain tissue (6). Some insight into the possibility of oligosaccharides being an intermediate glycosyl acceptor of glucose from the sugar nucleotide for glycogen synthesis was furnished by Olavarria (7). Having detected radioactive oligosaccharides when radioactive UDPG and the rat muscle enzyme were incubated, the problem was further investigated with the rat liver particulate enzyme. Incubation of radioactive oligosaccharides and unlabeled UDPG with the enzyme did not result in radioactive glycogen. Alternatively, incu- bation of the enzyme with radioactive glycogen and unlabeled 44 UDPG did result in the formation of radioactive oligo- saccharides. Furthermore, in short time incubations with labeled UDPG, glycogen formed had a higher specific activity than did the oligosaccharides. The differences were still greater if only the exterior chains of the glycogen were measured for radioactivity. These data indicate that the oligosaccharides are not intermediates in the synthesis of glycogen, but rather are degradation products. Starch The biosynthesis of starch and oligosaccharides from UDPG was reported by Leloir pp 2;. (8). The active enzyme, in association with starch granules, was isolated from the embryos and cotyledons of Phaseolus vulgaris. All attempts to separate the enzyme from the starch granules were negative. Starch biosynthesized from C14-UDPG was separated into its components, amylose and amylopectin. As both fractions C14-glucose, there was no indication of one contained fraction being a precursor of the other. That the label was indeed incorporated into starch was shown by B-amylase degradation to labeled maltose. This enzyme preparation catalyzed the transfer of glucose from UDPG to malto-oligosaccharides. This was clearly demonstrated by the synthesis of higher oligo- saccharides from UDPG and lower oligosaccharides. Maltose but not glucose was active in this system. The oligo- saccharides were not intermediates in the transfer 45 of glucose from UDPG to starch as shown by the lack of incorporation of label from labeled maltotriose into starch when incubated with or without UDPG and the enzyme. Alterna- tively, the oligosaccharides were not formed from starch as there was no incorporation of label into oligosaccharides when radioactive starch was incubated with the enzyme. It appears then that the malto-oligosaccharides compete ‘with starch for accepting glucose from UDPG. The addition of the glucosyl group to the nonreducing end of the acceptor was demonstrated by the formation of labeled gluconic acid and unlabeled glucose after bromine oxidation and hydrolysis of maltotetraose enzymically synthesized from unlabeled UDPG and maltotriose labeled in the reducing end group. Similarly, borohydride re- duction and hydrolysis of the same maltotetraose resulted in the formation of labeled sorbitol and unlabeled glucose. Alternatively, borohydride reduction of maltotetraose synthesized from C14-UDPG and unlabeled maltotriose re- sulted in the formation of labeled glucose and unlabeled sorbitol. Cgllulose The biosynthesis of cellulose has been studied by numerous investigators using different approaches to the problem. Minor, Greathouse, and coworkers in a series of papers (9-12) established that Apgpppagpgp x linum, when grown on specifically labeled glucose, produced 46 cellulose of which the monomer units retained in general the position of the label. Greathouse (13) and Colvin (14) reported the first cell-free synthesis of cellulose. As with growing cells, cell-freeextracts incorporated labeled glucose intact into the cellulose but ATP was required to obtain good yields. The role of hexose-phosphates in cellulose synthesis in A" xylinpm was investigated by Schramm.§p.11. (15). Both extracts and cell suspensions were used to demonstrate individual enzymes which catalyzed phosphorylation of hexoses and their subsequent conversion to C02, acetate, ethanol, and cellulose. Distribution of label in glucose units of cellulose produced from specifically labeled substrates suggested that hexose-phosphates are involved in the overall process of cellulose synthesis, probably without cleavage of the hexose residue. However, resting cells were not capable of metabolizing d- or B-G-l-P, G-6-P, or UDPG. As cellulose is found extracellularly, it was suggested that the role of the hexose-phosphates is to furnish a diffusible intermediate which contains the carbon skeleton but in which the phosphate group is modified or replaced. The transfer of glucose residues from UDPG to cellulose in the presence of a particulate enzyme system from A. gylippm_and soluble cellodextrins was demonstrated by 47 Glaser (16). Although incorporation of label from UDPG was low, partial acid hydrolysis of the synthesized poly- saccharide revealed radioactive glucose, cellobiose, cellotriose, cellotetraose, and larger oligosaccharides. Further digestion of the partial acid digest with Agpgrgillps pipe; cellulase produced only radioactive glucose. Digestion with Myrothecium verrucaria cellulase produced radioactive glucose and cellobiose, as this enzyme splits cellobiose at a very slow rate. These characteristics imply that the structure of the polysaccharide produced is a B-l,4—linked glucose polymer (cellulose). Primer dependence was established by pretreatment of the particulate enzyme with M, verrucaria cellulase. ‘Without added primer, this enzyme had essentially no activity. Cellodextrins of higher molecular weight eXhibited more activity as primer than did those of low molecular weight. Net synthesis and not simply exchange of glucose residues was demonstrated by the lack of isotope dilution in the cellulose upon incu- bation of enzyme labeled cellulose, and unlabeled UDPG. Some doubt as to whether UDPG is the immediate pre- cursor of cellulose has been cast by recent work of Colvin (17). Ultra-filtered 80% ethanol extracts of active A. xylinnm,cells were found to contain a compound which was rapidly converted to cellulose microfibrils when slowly evaporated to dryness at 400 C. This conversion was accelerated by a heat-labile, extracellular material in 48 the medium of the active cells, presumably an enzyme. When Cl4-glucose was utilized by the active cells, no labeled cellobiose, cellotriose, cellotetraose, or UDPG could be recovered. Further investigations on the nature of this compound revealed that after adsorption of the ethanol extract on a trisilicate-celite column, the active compound could be eluted off (18). The compound, having chemical characteristics of a glyco-lipid, was radioactive when extracted from cells grown on C14-glucose. The radioactivity was transferred to cellulose microfibrils when the compound was incubated with ultra-filtered medium. As no lipid was found to accumulate in the medium, a cyclic transport system was suggested whereby the lipid alternately leaves and enters the cell wall, transporting an activated glucose molecule to the extracellular acceptor. Chitin Glaser and Brown (19) demonstrated the synthesis of Cl4—UDPAG by a particulate fraction of Neupospora chitin from gpassa. The product of the enzymic reaction was character- ized by chitinase degradatiOn of partial acid digests to C14-N-acetylglucosamine. Carbon14-N'-N—diacetylchitobiose was also identified in the partial acid digest. For maximal activity, the particulate enzyme required soluble chito- dextrins as primer and N-acetylglucosamine. Labeled soluble primer was incorporated into insoluble chitin only in the presence of UDPAG. Net synthesis of chitin was demonstrated 49 by a lack of isotope dilution when C14—chitin was incubated with unlabeled UDPAG and also by direct chemical measurement of the hexosamine content. Reversibility of the reaction was attained by incorporation of label into UDPAG upon incubation of Cl4 -chitin and UDP with the enzyme. A soluble enzyme was obtained by butanol treatment which had an absolute primer requirement, but was not activated by N-acetylglucosamine. Hyaluronic Acid Investigations concerned with the biosynthesis of hyaluronic acid prior to 1955 are thoroughly reviewed by Dorfman (20). Briefly, it has been shown that in Group A streptococci the glucosamine and glucuronic acid residues of hyaluronic acid are derived from glucose without cleavage of the glucose molecule. Cifonelli and Dorfman (21) in further investigations on the biosynthesis of this polymer examined the nucleotides present in streptococcus capable of synthesizing hyaluronic acid. Boiling water extracts of waShed cells were chromatographed on Dowex-l (C1) columns. Sugar nucleotides found were UDPG, UDPAG, and large amounts of UDPGA. The authors suggested that these compounds may be intermediates in hyaluronic acid synthesis. The first cell-free synthesis of hyaluronate chains was reported in 1955 by Glaser and Brown (22). Cell-free homogenates of Rous chicken sarcoma were capable of catalyzing the incorporation of label from 50 Cl4-N-acetylglucosamine-6-P and UTP or from Cl4-UDPAG, along with UDPGA into hyaluronic acid chains. However, much of the label was lost on dialysis or reprecipitation of the carrier hyaluronic acid, indicating that the actual products formed were oligosaccharide chains. Markowitz §£.§1, (23) were able to confirm the work of Glaser and Brown (22) and conclusively show the net synthesis of hyaluronic acid from UDPGA and UDPAG. A particulate fraction from Group A streptococcus catalyzed the incorporation of radioactivity of N—acetylglucosamine from tritiated UDPAG into the acetylglucosamine moiety of hyaluronic acid in the presence of nonlabeled UDPGA. The incorporation of the radioactivity of glucuronic acid from UDPGA into the glucurOnic acid moiety of hyaluronic acid in the presence of nonlabeled UDPAG was also shown. Thelabeled product was characterized first by adsorption and elution from Dowex-l (C1) resin without change in radioactivity and secondly, hydrolysis of the polymer yielded labeled glucosamine only from polymer synthesized from labeled UDPAG, and labeled glucuronic acid only from polymer synthesized from labeled UDPGA. Finally, degradation of the synthesized hyaluronate by Clostridium perfringens hyaluronidase to the unsaturated disaccharide 3-0-(B-D—A, 4,5-glucoseenpyranosyluronic acid) 2-deoxy-2-acetamido-D- glucose resulted in recovery of 80% of the label in this disaccharide. This enzyme system had an absolute MgClz 51 requirement and was slightly activated by ATP and N-acetyl— glucosamine-l-P. The presence of UDPAG-pyrophosphorylase in the enzyme preparation was shown to be the reason for the N-acetylglucosamine-l-P activation, as incorporation of this phosphate ester proceeded only if UTP were present. No absolute primer requirement was demonstrated, although the washed particulate enzyme contained traces of nondialyzable uronic acid. Synthesis of hyaluronic acid from the respective sugar nucleotides has more recently been reported by Schiller at B1- (24) with enzymes from rat fetus skin. Type III Ppegmpgoccal Polysaccharide Bernheimer (25) in 1953 showed that resting suspensions of Type III pneumococci could synthesize capsular poly- saccharide from glucose, but little or none was formed from glucuronic acid or cellobiuronic acid. The importance of these observations was realized after later investigations of the nucleotides and enzymes present in various types of pneumococci. Smith ep_§1. (26) demonstrated in a non- capsulated strain derived from a Type II organism and a capsulated Type III organism the presence of UDPAG, UDPGA, and UDPG. UDPG pyrophosphorylase was present, along with enzymes necessary to transform glucose to G-l-P. Further investigations revealed the presence of UDPG dehydrogenase in.a noncapsulated strain (27). UDPGa1A.was isolated from a strain of pneumococcus Type I (28), and later, 52 the epimerase which interconverted UDPGA and UDPGalA with the mediation of DPN (29). As Type I capsular polysaccharide contains galacturonic acid, the finding of UDPGalA and enzymes necessary for its formation in these bacteria strongly suggested their participation in capsule formation. This ‘was further substantiated by Austrian pp al. (30) who showed by transformation and genetic studies that synthesis of capsular polysaccharide by certain strains of pneumococci is dependent on an intact pathway of uridine diphosphate glycosyl metabolism. The lack of only one enzyme resulted in a block of polysaccharide synthesis. Smith pp pi, (31,32) reported the incorporation of radioactivity from Cl4—UDPG and C14-UDPGA into glucose and glucuronic acid residues, respectively, of Type III pneumococcal capsular polysaccharide (B-l,4-glucuronosido- glucose), resulting in net synthesis of polymer. A partially fractionated, cell-free enzyme preparation was used which subsequently was shown to contain all activity in a sedimentable fraction. The labeled polysaccharide\ was characterized in several ways. It was precipitated with the specific rabbit anti-Type III capsular serum. €14 Upon acid hydrolysis of the polysaccharide, -glucose C14-glucuronic acid were obtained, depending on and/or which labeled sugar nucleotide was used as precursor. Also, the labeled polysaccharide was depolymerized by the specific Type III capsular depolymerase, rendering 90% 53 of the label dialyzable. Primer dependence was not rigorously established, as addition of primer polysaccharide did not stimulate synthesis. However, various enzyme preparations were found to contain different amounts of polysaccharide, and the rate of synthesis was proportional to this polysaccharide concentration. It appears, therefore, that in this case primer molecules are associated with the enzyme. fi-l.§-G1ucan Feingold pp pl, (33) reported the synthesis of a B-l,3-linked glucan from UDPG, the reaction being catalyzed by a particulate enzyme found in various plants. The polysaccharide formed was immobile in chromatographic solvents in which oligosaccharides migrated. Upon partial acid hydrolysis and subsequent chromatography of the polymer synthesized from C14-UDPG, seven radioactive oligosaccharides were observed. These migrated identically tociigosaccharides prepared from laminaran, a B-l,3—linked glucan. The 1,3 linkage was further established by lead tetraacetate oxi- dation of the disaccharide obtained from a partial acid hydrolyzate. As previously shown (34), such oxidation of a reducing sugar having a 1,3 linkage with glucose as the reducing moiety produces arabinose. Arabinose was obtained, indicating a 1,3 linkage. In addition, the arabinose was radioactive, proving the incorporation of label from C14-UDPG into the polymer. The B-type linkage 54 of the radioactive oligosaccharides obtained by partial acid hydrolysis of the polymer was proven by degradation to glucose with B-glucosidase. The particulate enzyme was capable of catalyzing polymer formation from UDPG without added primer. The digitonin solubilized enzyme exhibited an absolute primer or activator requirement, glucose or oligosaccharides of cellulose and laminarin being active. Primer glucose was not, however, incorporated into the polymer. Xylo-O;igosaccharides Feingold, §p_a1, (35) reported the ability of an asparagus extract, prepared similarly to the B-l,3-glucan synthetase from mung bean seedlings (33), to catalyze the transfer of xylose from UDnylose to B-l,4-linked xylo- oligosaccharide acceptors. It was shown by cochromatography and coelectrophoresis that radioactive xylo-oligosaccharides were formed which contained one more pentose unit (3 to 6 units) than the initial xylo-oligosaccharide acceptor (2 to 5 units). Also, total acid hydrolysis yielded xylose as the sole radioactive product. A 6-1,4-linked xylo—oligosaccharide primer was required in very high concentration for maximal transfer to occur ____£L_JL_ acce to > 100) . UDnylose Similar concentrations (molar ratio of di-, tri-, tetra-, and penta-xylo-oligosaccharides were necessary as primer. This apparent identical affinity of the enzyme for each of the acceptors accounts for the 55 observed one-step reaction, without formation of xylan. The possibility of xylose transfer by this enzyme to a higher molecular weight primer was not investigated. Synthesis of polysaccharides by glycosyl transfer from sugar nucleotides have in most cases required primer. Lack of primer dependence in hyaluronic acid and Type III pneumococcal polysaccharide synthesis has not been rigorously established, as traces of polysaccharide con- stituents were found associated with the enzyme. These findings lead to the very interesting question of how does a primer molecule arise. A possible explanation was furnished by Illingworth pp pl, (36) who demonstrated the dp_ppyp synthesis of amylose chains from G-l-P by phosphorylase in the absence of primer. Crystalline muscle phosphorylase, free of amylo-1,4-+ 1,6-transglucosidase (branching enzyme), d-amylase, amylo-l,6-glucosidase, and essentially free of any carbohydrate (0.3 glucose monomer per monomeric unit of phosphorylase), when incubated in large quantities with G-l-P after an initial lag period catalyzed the release of Pi and the formation of amylose chains. This lag period could be abolished by addition of minute quantities of glycogen. When catalytic quantities of phosphorylase were incubated with G-l-P, no release of Pi occurred. However, addition of catalytic quantities of branching enzyme resulted in release of Pi, the lag period and running rate of the reaction being dependent 56 on the branching enzyme concentration. The stimulation of synthesis by branching enzyme and glycogen are both apparently due to the creation of multiple end groups. It is suggested that such a catalytic system (phosphorylase and branching enzyme) could conceivably account for the initial synthesis of glycogen in embryonic tissues. Phosphomannan The production of extracellular phosphorylated mannans by various strains of Hansenula has recently been reported (37,38). Laboratory production (39), characterization, and properties (40) of the phosphomannan produced by Hgnsgnula holstii NRRL Y-2448 have been described in more detail. This particular polymer consists of mannose and phosphate in a ratio of 5:1. Investigations on the structure of the polymer (private communication) produced by Hansenpla gapsplata NRRL Y-1842 (mannose to phosphate ratio of 2.5:1) are consistent with that shown in Fig. 5. The biosynthesis of Hansenula holstii NRRL Y-2448 polymer was investigated by Wilken (unpublished data). C14 produced polymer which retained Cells grown on glucose-l- essentially all the label (84.3%) in carbon 1 of the mannose residues. The remaining label in carbon 6 was ascribed to a slight equilibration of the hexose-phosphates with triose phosphates and subsequent incorporation of hexose into phosphomannan. The suggestion by Wilken that GDPM may be the glycosyl donor was enhanced by a preliminary 57 Figure 5. Structure of phosphomannan. 58 I N:oluol d lol : 0 59 report (41) that sugar nucleotides are present in these organisms. MATERIALS AND METHODS Quantitative Measurements Radioactivity was measured in a Packard Instrument Company Tri-Carb liquid scintillation counter. The counting solution consisted of 4.0 g 2,5-diphenyloxazole and 50 mg 1,4-bis-2-(5-phenyloxazole)—benzene made up to 1 liter with toluene (42). Usually 0.1 ml of aqueous solution to be counted was added to 5 ml of absolute ethanol and 10 ml of counting solution. Polysaccharide samples were counted by adding 1 to 5 mg of dry sample to 10 m1 of a thixotropic gel suspension (2.5% thixotropic gel in the previously described solution). The counting vial was shaken vigorously to evenly disperse the sample in the gel. This technique resulted in proportionality of counts to sample added over a range of l to 10 mg with an efficiency of 56%. Radioactive spots were detected and counted on paper chromatograms with a gas flow counter equipped with a thin Mylar window Geiger-Mueller tube. Pyridine nucleotide reduction and oxidation were measured on a Beckman model DU spectrophotometer equipped with a Gilford cuvette changer and recording attachment (43). Phosphorus was determined by the Fiske—SubbaRow procedure as described by Leloir and Cardini (44). Ketoses were detected by the resorcinol method of Roe as described by Ashwell (45) and reducing sugars were 60 61 measured by the procedure of Park and Johnson (46). Phosphomannan was estimated by the phenol-sulfuric acid method of Dubois pp pi. (47). Phosphomannan gave identical color development as equimolar amounts of mannose (assuming that the polymer contained 14% moisture (40)). Protein was measured Spectrophotometrically (48). Qualitative Measurements Nucleotides were separated on ion exchange resin columns by the method of Hurlbert pp al. (49) as described by Forrest and Hansen (50). Norite A (Pfanstiehl Labora- tories, Incorporated) was prepared by washing with 2 N HCl, 0.05 EDTA, 95% ethanol, and water, followed by drying. Nucleotides were chromatographed on.Whatman No. 1 and/or No. 3 filter paper in a descending direction with the following solvents (51): isobutyric acid-ammonium hydroxide-water, pH 3.7 (66:1:33 v/v), 95% ethanol-1.0 M ammonium acetate, pH 7.5 (7:3 v/v), or isobutyric acid- ammonium hydroxide-water, pH 4.3 (57:4:39 v/v). Nucleotides were detected on paper with an ultraviolet lamp (Mineralight SL-2537). Photographs of ultraviolet absorbing spots on paper chromatograms were made by contact prints prepared by a slight modification of the procedure of Markham and Smith (52) as described by Wilken and Hansen (53). Carbo- hydrates were chromatographed on Whatman No. 1 filter paper and developed with water-saturated phenol (54) or butanol-pyridine-water (6:4:3 v/v) (55). Sugars were detected 62 on paper with alkaline-silver nitrate (56). Radioactive spots were detected by exposing paper chromatograms to No-Screen X-ray film. Chemicals The tricyclohexylammonium salt of 6-PGA was obtained from the California Corporation for Biochemical Research. K2-G-l-P was purchased from Sigma Chemical Company. Ba-F-6-P and Ba-M-6-P were purchased from Nutritional Biochemicals Corporation. Ba-G-6-P was made by the procedure of Wood and Horecker (57). Ba-M—l-P was synthesized according to Colowick (58) with the following modifications: penta— acetyl mannose was prepared by the method of Hudson and Dale (59) from D (+) mannose (Pfanstiehl Laboratories, Incorporated) and acetobromomannose by the method of Levene and Sobotka (60), neither product being crystallized. All barium salts were converted to the sodium salts by precipitation of the barium as BaSO4 by addition of Na2804. Tricyclohexylammonium salts were treated with Dowex-50 (H) and neutralized with NaOH. GDPM was synthesized by the method of Roseman ep_a1, (61) and used as the lithium salt. All other nucleotides were purchased from Sigma Chemical Company and Pabst Brewing Company. The barium salt of Cl4-M-l-P and the lithium salt of C14-GDPM were synthesized as previously described from C14 uniformly labeled -D-mannose (Schwarz Bioresearch, Incorporated) and had specific activities of 8 x 104 63 dpm/umole. The lithium salt of C14—M—1—P was recovered from the Dowex-l (Cl) resin column used in purification of the C14-GDPM by precipitation from methanol with acetone in the same manner as for the sugar nucleotide (61). C14-G-1-P was prepared as the crystalline dipotassium salt from uniformly labeled leaf starch (62) by phosphorolysis with crystalline muscle phosphorylase (63) and had a specific activity of 5 x 105 dpm/umole. EXPERIMENTAL PROCEDURES AND RESULTS Nucleotide Components Hapsenpla ho;stii NRRL Y-2448 was grown on 6% glucose for 36 hours and harvested by centrifugation at 12,000 x g for 15 minutes. Approximately 215 gm of wet packed cells were obtained from 3 liters of growth medium. The wet cells were extracted in an Omni-mixer at 50,000 rpm for 15 minutes with 2 volumes of cold 0.6 N perchloric acid. After centrifugation, the supernatant was immediately neutralized to pH 6.8 with 5 N KOH. The pellet was re- extracted as outlined above. The neutralized extracts were allowed to Stand at 3° C for 12 hours before filtering through Whatman No. 50 filter paper to remove potassium perchlorate. Based on an approximate molar extinction coefficient at 260 mu of 10,000, 570 umoles of nucleotides were obtained. The following procedures including elution of the column were carried out at 3° C. The total extract was placed on a 1.8 x 40 cm Dowex-l-lO x 200-400 mesh formate resin column which had been filled in 2 cm layers with no external pressure applied. After washing with 500 ml water, about 340 umoles were calculated to have been retained by the resin. The column was developed as described by Forrest and Hansen (50) using a 500 m1 mixing 64 65 flask and collecting 12.5 ml fractions at about 1 ml per minute. The various peak tubes (Fig. 6) were combined, lyophilized, and taken up in 20 ml water. (For quantitative purposes, all tubes from 121 through 199 were collected in fractions as: K, 121-139; L, 140-155; M, 156-167; N, 168-181; and N1, 182-199). Norite (0.5)gm was added, collected on a Buchner funnel, and washed with 10 ml water. The nucleotides were eluted by stirring the Norite with 50 ml of 15% aqueous pyridine. Following filtration, pyridine was removed by extracting the filtratetwice with 50 ml of CHC13 and the aqueous layer was lyophilized. Further purification of peaks K, L, M, N, and N; was achieved by applying the freeze dried material to Whatman No. 3 filter paper and chromatographing in ethanol-ammonium acetate, pH 7.5. Bands migrating with authentic GDPM were found in peaks K, L, M, N, and N1 and UDPG in peaks N and N1. These were eluted with water, lyophilized, and rechromato- graphed on Whatman No. 1 filter paper (previously washed with l N acetic acid, 95% ethanol, and water) in isobutyric acid-ammonium hydroxide-water, pH 4.3. The bands corresponding to GDPM and UDPG were eluted and analyzed. Spectra were identical for the GDPM and UDPG isolated from the various peaks of the column with authentic nucleo- tides. Typical absorption spectra of GDPM and UDPG from peak N; are shown in Fig. 7. Quantitative analysis by ultraviolet absorption and reducing values are presented 66 Figure 6. Elution pattern of acid soluble nucleotides from Dowex-l (formate) column. The column was developed with: l) 2 N formic acid; 2) 4 N formic acid: 3) 4 N formic acid and 0.2 N ammonium formate; 4) 4 N formic acid and 0.4 N ammonium formate; 5) 4 N formic acid and 0.8 N ammonium formate. Principal components of each peak are: c and D, DPN and AMP; E, TPN; F, GMP; J, ADP; K and L, GDPHexose; M, UDP-N-acetylhexosamine; N, UDPHexose; N1, mixutre of K, L, M, and N; P, ATP; P2, UDP; R, UTP and GTP. 67 mmmsSz 20_._.0aonowg .oonumfi comc£0hlxumm was an mflmhamc¢++ .o.b mm um ooo.oH mo mannans How ocmaowmmmoo COHOUCHuxO HMHOE o co ompmaooamo+ .o.h mm um oom.ma mo mnemocmsm mom ucmfloflwwmoo coauocfluxm HOHOE m :0 UTOMHDUHMO * ms.m mmoxmmmoo 25m om.o mm.o mo.o Hm.o = Hz oe.a ew.a Sa.o +sm.a mmoxmmmoo z mm.a mmoxmmmoo saw ma.o mm.o ma.o sm.o = Hz -- -- -- dA.o . z HH.o mH.o no.0 ea.o . z Hm.o mm.o :H.o mm.o = A wa.o Hm.o ma.o *mm.o mmoxmmmoo M mmaofid mmaofii mocmuwmmwn mflmhaonomm mammaono>m Hound muommm mmmm mofluowaosz coauomum m m ++Hm om cflosomm mcoauomum mmoxmm Opmnmmonmfin Teflon“: pom mmoxmm mpmnmmonmwn OCHmocmsO mo mflmhamcfl .HH> magma 72 GMP and GDP. Figure 6 reveals that GDPM was not eluted off in a discreet area as were other nucleotides. This may be due in part to the overloading of this column which was necessary to obtain sufficient quantities of nucleotides for identification. However, calculated on umoles per ml eluate basis, GDPM is most concentrated in peak L. It should be mentioned that the preparative chromatograms of these peaks (K through N1) developed in ethanol-ammonium- acetate, pH 7.5, contained several ultraviolet absorbing bands in addition to GDPM and UDPG. One of these was also ninhydrin positive. The main components of other peaks eluted from the column as revealed by spectral analysis and chromatography in ethanol-ammonium acetate, pH 7.5, isobutyric acid- ammonium.hydroxide-water, pH 4.3, and isobutyric acid- ammonium hydroxide-water, pH 3.7, are given in Fig. 6. Further rigorous characterization of these compounds was not undertaken. Enzymes Involved in Glucose and Mannose Metabolism Preperetion of Cell Extracts Hapsenule holstii NRRL Y-2448 was grown on a medium consisting of 6.0% glucose, 0.1% corn steep liquor, 0.1% bactotryptone, 0.5% KH2P04, and 0.5% salt solution as described by Anderson _p._1. (39), with 200 ml of medium in each Fernbach flask inoculated with 10 m1 of a 24 hour culture. The cultures were incubated at 280 C with shaking for various lengths of time, depending on the nature of the experiment for which the yeast was to be used. Cells were harvested in either a refrigerated Lourdes or Servall SS-l centrifuge at 20 C, washed twice with cold water, and frozen until used. Various methods of attaining soluble extracts were tried. Usually the cells were acetone dried by adding 20 volumes of acetone at -100 C to the wet cell paste and blending in an Omni-mixer for 2 minutes. After filtra- tion, the powder was reextracted with cold acetone and filtered again. The dry powder was then worked on a filter paper until no acetone vapors were detected. The dried cells were extracted with 10 volumes of buffer (0.1 M Tris-0.01 M MgC12-0.001 M EDTA-0.001 mercapto- ethanol, pH 7.6) at 00 C for 2 hours. Ammonium sulfate was added to the supernatant to 70% of saturation and the precipitate dissolved in buffer. An alternative procedure consisted of blending the wet cell paste with 4 times the weight of glass beads (Superbrite size No. 110, Minnesota Mining and Manu- facturing Company) and 1 volume of buffer at 00 C in the Omni-mixer at 5,000 rpm for one hour. Remaining whole cells and cell debris were removed by centrifugation at 74 15,000 g for 10 minutes. The supernatant was used directly or fractionated with ammonium sulfate as previously des- cribed. In some cases, the extract was used without centrifugation to remove cell debris. Hexokinase Spectrophotometric evidence for the presence of hexokinase was obtained by the reduction of TPN in the presence of glucose, ATP, G-6-P dehydrogenase and an extract from acetone dried cells. As seen in Fig. 8, reduction of TPN is dependent on the presence of ATP and glucose. To test the specificity of the kinase, extracts obtained from cells ruptured with glass beads were treated with ammonium sulfate and the precipitated protein assayed by measuring DPNH oxidation in the presence of hexose, ATP, phosphoenolpyruvate, and crystalline lactic dehydro- genase (containing pyruvate kinase). Table VIII shows that mannose is phosphorylated 71% as rapidly as is glucose. However, when both hexoses were present in the incubation mixture, no increase or additive rate was observed above that of ghcose alone, indicating that the same enzyme is phosphorylating both hexoses. That mannose will serve as substrate for phosphomannan production as well as glucose will be shown later. W Glucose-6-phosphate may be oxidized to 6-PGA in the presence of TPN and acetone powder extract. Fig. 9 75 Figure 8. Hexokinase activity. Complete system in 0.5 ml total volume contained: 1.0 umole glucose, 1.0 umole ATP, 0.5 umole TPN, 0.5 umole MgClg, 0.01 ml G-6-P dehydrogenase, 41.0 umoles glycylglycine, pH 8.0, and enzyme (approximately 0.02 mg protein). 76 IN MINUTES TIME 77 Table VIII. Nonspecific Hexokinase Activity Hexose A OD34o/min -_ 0.09 Mannose 0.15 Glucose 0.21 Mannose + Glucose 0.21 The complete system contained in a 0.5 m1 total volume: 0.5 umole each hexose, 0.5 umole phosphoenolpyruvate, 0.5 umole DPNH, 0.01 ml crystalline lactic acid dehydrogenase diluted 1:10, 0.5 umole MgClg, 2.5 umoles KF, 0.5 umole ATP, 42.5 umoles glycylglycine, pH 8.0, and enzyme (approximately 0.05 mg protein). 78 Figure 9. Glucose-fi-phosphate apd Q-phosphogluconic acid dehydrogenase ectivity. Complete system in 0.5 ml total volume contained: 0.1 umole G-e-P, 0.5 umole TPN, 0.5 umole MgClg, 46.0 umoles Tris, pH 7.6, and enzyme (approximately 0.1 mg protein). The dash lines indicate the theoretical end point for 1.0 mole of TPN reduced (bottom line) and 2.0 moles of TPN reduced (top line) per mole of G—6-P added. 79 mw.52=2 2. ms.:. '0. O Q IQ 0. (\l ““1 OPE '0 'O 80 demonstrates that per mole of G—6-P, almost 2 moles of TPN are reduced, indicating that 6-PGA dehydrogenase is also present. The latter enzyme was confirmed by showing that 6-PGA is oxidized by the same extract, producing 1 mole of TPNH per mole of 6-PGA (Fig. 10). Ispmerases The possibility that G-6-P may be isomerized to F-6—P was tested by measuring ketohexose formation from G-6-P by acetone powder extract (Table IX). Isomerization of the F-6-P formed to Me6-P by phosphomannoisomerase is minimized by running the reaction at pH 9.0 (64). The presence of the latter enzyme was shown by incubating M—6-P with acetone powder extract and measuring ketohexose formation. As expected, more F-6-P is formed at pH 5.5 than at pH 7.6; however, phosphomannoisomerase is present in much lower amounts than phosphoglucoisomerase. Compar- able activities of the two isomerases at pH 8.0 may be seen in Fig. 11. Mpteses Evidence for the ability of Mel-P to be metabolized via MPG-P, F-6-P, G-6-P and 6-PGA by acetone powder extract was obtained (Fig. 12). As the extract had previously been shown to contain phosphomannoisomerase, phospho- glucoisomerase, and G-6-P dehydrogenase, it was not necessary to add further quantities of these enzymes. The initial step in this sequence of reactions (conversion 81 Figure 10. Six-phosphogluconic acid dehydrogenase Complete system in 0.5 m1 total volume contained: 0.5 umole TPN, 0.1 umole 6-PGA, 0.5 umole MgClz, 45.0 umoles Tris, pH 7.6, and enzyme (approximately 0.1 mg protein). The dash line indicates the theoretical end point for 1.0 mole of TPN reduced per mole of 6-PGA oxidized. O.D. , 340 mu 3'6 O .b O Q m T 82 I —- NO G-PGA? l O 1 I 1 IO 20 30 .49 TIME IN MINUTES Table IX. Ketohexose Formation from Hexose-6-Phosphate Substrate Tlme F‘5‘P formed (umoles) (min) PH 5.5 pH 7.6 pH 9.0 G-6-P Q 0 10 0.22 20 0.39 M-G-P O O 0 30 0.13 0.07 Separate incubations were run for each time and each pH. The mixture contained in a 0.5 ml total volume: 1.0 umole hexose-6-phosphate, 21.6 umoles acetate, pH 5.5, or 36.0 umoles Tris, pH 7.6 or 44.0 umoles Tris, pH 9.0, and enzyme (approximately 0.04 mg protein for the G-6-P incubations and 0.2 mg protein for the Mb6-P incubations). At the time intervals shown, ketohexose was determined using F-6-P as standard. 84 Figure 11. Phosphohexoisomerase activity. Complete system in 0.5 ml total volume contained: 0.5 “mole MgClg, 0.5 Hmole TPN, 43.0 umoles glycyl- glycine, pH 8.0, 0.5 umole F—6-P or 0.5 Hmole M—6-P, and enzyme (approximately 0.04 mg protein). 0.0. 340 mu .0 m .0 4:. .0 N O ‘85 F" 6" P} M-G-P NO Substrate} * 1 - l 1 5 IO I5 TIME IN MINUTES 86 Figure 12. Phosphomannomutase activity. Complete system in 0.5 ml total volume contained: 1.0 umole M-l-P, 0.5 Hmole TPN, 0.5 umole MgClg, 42.0 umoles Tris, pH 7.6, and enzyme (approximately 0.4 mg protein). Optical density was recorded at various time intervals. At the arrow (135 minutes) 1.0 umole potassium pyruvate and 0.1 mg crystalline lactic acid dehydrogenase was added to each cuvette. 87 mntbzi z. NEE. 0m. 00 0 q _ ”2.: oz 88 of M—l—P to M—6-P) was shown to be limiting by a greatly increased rate of TPN reduction upon addition of Me6-P. The initial lag period of approximately 10 minutes is probably the time required to build up intermediates to make the chain of enzymic reactions proceed through G-6-P dehydrogenase. That the slow reaction was actually resulting in TPN reduction is substantiated by the rapid oxidation of TPNH upon addition of potassium pyruvate and crystalline lactic dehydrogenase. In one experiment, acid labile phosphorus was determined in the incubation mixture at the end of the Spectrophotometric assay. The decrease in the amount of acid labile phosphorus (M—l-P) corresponded very well to the amount of TPNH formed, thus confirming the conversion of M-l-P to acid stable hexose-6-phosphates. No apparent effect of G-l,6-diP on the reaction rate was observed. A very active phosphoglucomutase was demonstrated in acetone powder extracts by adding G-l-P, TPN, and additional G-6-P dehydrogenase (Fig. 13). A slight stimulation by G-l,6-diP was observed. P o hos o lases The phosphoglucomutase reaction provides a source of G-l-P for the synthesis of UDPG. The presence of UDPG pyrophosphorylase was detected by measuring DPN reduction in the presence of UTP, G-l-P, and excess UDPG dehydro- genase. As seen in Fig. 14, this enzyme is very active in acetone powder extracts and was also active in one 89 Figure 13. Phosphpglucomutase activity. Complete system in 0.5 ml total volume contained: 0.5 umole G-l-P, 0.5 umole MgClz, 0.01 ml G—6-P dehydro- genase, 1.0 umole TPN, 5.0 umoles L-cysteine, pH 7.5, 0.2 mumole G-l,6-diP, 40.0 umoles Tris, pH 7.5, and enzyme (approximately 0.02 mg protein). .0 :— (I) O 0. D. 340 mu ,0 o A m .0 N O o . G-6-P and G-l,6-diP~. ‘G-e-P No G-l,6-diP No G-6-P3 __.__, l l 5 IO TIME IN MINUTES l5 91 Figure 14. Uridine diphosphate glucose pyrophgs- phorylase activity. Complete system in 0.5 ml total volume contained: 0.4 umole UTP, 1.0 umole G-l-P, 0.5 umole MgClg, 0.25 umole DPN, 5.0 umoles NaF, 39.0 umoles Tris, pH 7.6, and enzyme (approximately 0.2 mg protein). 92 gm 0N mmzbzi z. m_>__._. N. w v _ 0. i q _ - £5 oz) rifle oz 0 “I 0 VI? 0 LO 0 00. o "w 0172 '0 'O 93 preparation of air dried cells extracted with 0.1 M Tris buffer, pH 7.6. Attempts to demonstrate UDPG dehydrogenase spectro- photometrically were consistently negative, both in acetone powder extracts and cells ruptured with glass beads. This is consistent with the absence of UDPGA in the acid soluble nucleotides from this yeast. The formation of GDPM from GTP and M-l-P by acetone powder extracts was detected chromatographically. After incubation of the extract with GTP and M-l-P, the reaction mixture was heated in a boiling water bath for one minute to denature protein. The supernatant was lyophilized and chromatographed on Whatman No. 1 filter paper in ethanol- ammonium acetate, pH 7.5. For better separation of GDPM and GMP, the GDPM was eluted with water, lyophilized, and rechromatographed in isobutyric acid-ammonium.hydroxide- 'water, pH 4.3. The GDPM.was then eluted with water and quantitated by ultraviolet absorption, using a molar extinction coefficient of 13,700 at pH 7.0. Resolution of the various nucleotides in ethanol-ammonium acetate, pH 7.5, was enhanced by first adsorbing the nucleo- tides on charcoal after the protein denaturation step, then washing the charcoal with water before eluting the nucleotides with 15% pyridine. Following extraction of the pyridine with chloroform, the aqueous layer was lyo- philized and chromatographed. Results of such an experiment 94 are shown in Fig. 15. No GDPM was detected in the absence of Mel—P. For further characterization of the product of this enzymic reaction, larger scale incubations consisting of 6.0 umoles GTP, 12.0 umoles M—l-P, 60.0 umoles NaF, 108.0 umoles Tris, pH 7.6, 10.8 umoles MgClz, 1.1 umole EDTA, 1.1 umole mercaptoethanol, and 2.4 mg protein in a total volume of 1.8 ml were incubated two hours. The GDPM formed was isolated by chromatography in ethanol-ammonium acetate and isobutyric acid-ammonium.hydroxide-water, pH 4.3 as described previously. In this manner, 0.5 umole of chromatographically pure GDPM was obtained which exhibited a typical guanosine spectrmnat pH 1, 7 and 11. Reducing sugar analysis by the Park and Johnson method resulted in a ratio of reducing sugar to base of 0.7:1. An aliquot containing 0.15 umole was hydrolyzed in 0.1 N HCl for 15 minutes at 1000 C and then deionized with Amberlite IR-l20 (H) and IR-45 (OH). The sugars obtained were chromatographed in phenol-water and the spots detected with alkaline AgNOa. A heavy mannose spot was observed along with lighter spots of glucose and fructose. Hydrolysis of an equal amount of chemically synthesized GDPM resulted in the detection of only mannose (Fig. 16). The occurrence of glucose and fructose in hydrolyzates might arise by interconversion of the intact GDPM molecule or alternatively by formation of the respective GDPHexoses 95 Figure 15. Guanosine diphosphate mennose pypophgs- phopylase activity. Complete system in 0.3 m1 total volume contained: 1.0 umole GTP, 2.0 umoles M—l-P, 10.0 umoles NaF, 16.0 umoles Tris, pH 7.6, 1.6 umoles MgClg, 0.16 umole EDTA, 0.16 umole mercaptoethanol, and enzyme (approximately 0.6 mg protein). The control lacked Mel-P. Incubation was for 1.5 hours at 25° C. The tubes were then heated in a boiling water bath for one minute, cooled, and centrifuged. The nucleotides were adsorbed onto charcoal, eluted off, and chromato- graphed in ethanol-ammonium acetate, pH 7.5. Lane 1, control incubation; lane 2, complete incubation; lane 3, standards. 97 Figure 16. S rs f om h rol ze uanosine di os- phate hexoses. GDPHexose synthesized enzymically from GTP and M-l—P was hydrolyzed, deionized, and the sugars chromatographed in phenol-water. Chemically synthesized GDPM was treated identically. Lane 1, sugars from chemically synthesized GDPM; lane 2, standards; lane 3, sugars from enzymically synthesized GDPHexose. 98 GLUCOSE MANNOSE FRUCKEE 99 from GTP and hexose-l-phosphate. In an attempt to elucidate the mechanism of formation of these GDPHexoses, an experiment was carried out as follows: 4.0 umoles C14-GDPM, 10.0 umoles NaF, 18.0 umoles Tris, pH 7.6, 1.8 umoles MgClg, 0.2 umole EDTA, 0.2 umole mercaptoethanol, and 1.1 mg protein in a total volume of 0.33 ml were incubated three and a half hours at room temperature before heating two minutes in a boiling water bath. An identical mixture was heated without incubation. About 2.5 umoles of GDPHexoses in each case were reisolated by chromatography in the two solvents previously used. An aliquot of each sample was counted in the liquid scintillation spectrophotometer and identical specific activities were observed (51,800 dpm/umole). Apparently a dilution due to endogenous GDPHexose occurred as the specific activity of the C14-GDPM used initially was 80,000 dpm/umole. Another aliquot (0.5 umole) of each sample was hydrolyzed in 0.1 N HCl for 15 minutes at 1000 C andOEionized. Half of each hydrolyzate was chromatographed in phenol-water and developed with alkaline AgNOa. A heavy mannose spot and faint spots of glucose and fructose were observed for each sample. The remaining half of each hydrolyzate was chromatographed in phenol-water under identical conditions as the preceding run. The vertical areas corresponding to authentic glucose, mannose and fructose were cut into 4.0 mm cross sections and each placed in 10 m1 of scintillating fluid. The counting 100 pattern observed (Fig. 17) shows that only mannose is labeled. It is therefore concluded that the intact GDPM was not being enzymically converted to GDPG or GDPFructose. To decide if GDPG pyrophosphorylase was present, the following experiments were performed: 4.0 umoles GTP, 40.0 umoles NaF, 72.0 umoles Tris, pH 7.6, 7.2 umoles MgClg, 0.7 umole EDTA, 0.7 umole mercaptoethanol, and 0.44 mg protein were incubated with l) 2.0 umoles C14-M-l-P, 2) 2.0 umoles C14-G-l-P, and 3) neither hexose-l-phosphate for a control. After three hours, the samples were treated and chromatographed as described for the previous experiment. No ultraviolet absorbing material which corresponded to GDPM was observed in the control sample, but the appropriate areas on the chromatograms were cut out, eluted, and treated the same as for the other two. About 0.17 umole of GDPHexose from GTP and C14-Mrl-P and about 0.06 umole from GTP and Cl4-G-1-P were isolated. All three samples were hydrolyzed, deionized and chromatographed in phenol- water. A radioaudiogram revealed only labeled mannose from the hydrolyzed GDPHexose formed from GTP and C14-Mel-P, and only labeled glucose from the hydrolyzed GDPHexose formed from GTP and C14-G-l-P. The C14-glucose area was eluted and rechromatographed in butanol-pyridine-water. Radioactivity was detected with a gas flow counter only in the glucose area and upon development with alkaline AgNOs, a heavy glucose spot was visible. The remainder 101 Figure 17. Retention of label in mannose ip gpenosipe diphosphate hexoses. Details of the experiments are described in the text. o—o , GDPMannose, 0 time; 0—0 , GDPMannose, 3.5 hours. The difference in peak locations is due to a slight difference in rate of solvent flow. 102 0N 6E3 z_0_m0 20m... moz<._.w_o 5N mm nm _N m. h. 0. III 0 1N f. .o ..a w I‘m . \ X m 10 2 33221134; 8 .9 wom