- Email: [email protected]
Uridine-diphosphoglucose glucosyltransferase in human erythrocytes
CLINICA CHIMICA ACTA
27
URIDINE-DIPHOSPHOGLUCOSE
GLUCOSYLTRANSFERASE
IN
HUMAN
ERYTHROCYTES* M. CORNBLATH,
D. F. STEINER,
P. BRYAN
Departments of Pediatrics, University of Illinois School of Medicine, Chicago, Ill. (U.S.,4 .)
(Received
AND J. KING College
of Medicine,
and
University
of Chicago
August zest, 1964)
SUMMARY
The presence of glucosyltransferase activity in mature human erythrocytes and leukocytes has been established both by demonstration of UDP formation from UDPG, and by incorporation of [‘“Cl glucose from UDP-[WI glucose into glycogen. The erythrocyte and leukocyte enzymes have similar properties and both require glucose 6-P for activity. The enzyme activity per white cell is about 400 times that per red cell. The erythrocyte glucosyltransferase activities did not differ from that in the normal in two patients with hepatic glucosyltransferase deficiency and in other patients with a variety of carbohydrate abnormalities.
INTRODUCTION
The uridine-diphosphoglucose (UDPG) pathway for glycogen synthesis, described by Leloir l, has been found in all tissues examined 2, except rabbit uterus 3. The presence of glycogen in the mature human erythrocyte prompted an investigation of UDPG-glycogen glycosyltransferase (EC 2.4.1.11) in the red cell. This was particularly pertinent because phosphorylase activity has been demonstrated in these cells 4. Glucosyltransferase has recently been demonstrated both chemically and histochemically in leukocytes 5po.The purpose of this paper is to report the activity and characteristics of glucosyltransferase in human erythrocytes and leukocytes. Levels of enzyme activity in red blood cells from normal infants, children and adults, patients with hepatic glucosyltransferase deficiency and a variety of other carbohydrate disorders are presented.
* This National
investigation was supported in part by Research Grants Institute of Arthritis and Metabolic
6015-04 and 4931-04 from Disease, United States Public Health Service.
Clin. Chim.
Acta,
I2 (1965)
the
27-32
MATERI:\LS ANI) METHOI)S
A. I’vepavation
ofevythvocytes *
Blood was drawn into EDTA . tia, (1.2 mgjml), centrifuged at 4 at L -poo rev./min for IO min. The plasma and buffy coat were removed. The red ~~11s were washed three times in cold 0.15 df NaF containing 0.001 M EDTA . Na,, and then suspended in an equal volume of 0.25 32 sucrose containing 0.02 dl NaF and 0.001 32 EDTA. The hemolysate was prepared by quick-freezing in dry ice and alcohol twice. This preparation stored at --IO’ is stable for at least 2 months. Hemoglobin dettarminations were made prior to assay by standard methods. B. Preparatiofz of leukocytes
Blood was drawn into EDTA . Na, (1.2 mg/ml). The leukocytes were then concentrated by differential sedimentation with polybrene according to Lalezari7. The leukocyte-rich plasma was centrifuged at 4’ at 10,000 rev./min for 30 min. The packed cells were washed once with 0.15 M NaF containing 0.001 M EDTA and resuspended in 0.25 M sucrose containing 0.02 M NaF and 0.001 J4 EDTA. Samples were taken for white blood cell counts and the remainder frozen in dry ice and alcohol and thawed three times, and stored in a freezer at -IO’. Enzyme
assays
The activity of glucosyltransferase in erythrocyte and leukocyte preparations was assayed by two methods: (I) chemically by measuring UDP formation by a modification of the method of Leloir and Goldemberg*, and (2) by the incorporation of [Cl41 glucose into glycogen from [Cl41 UDPG as described by Steiner and Kin@. In the former method, UDP formation was measured at o, 45, 60, and 75 min. The reaction mixtures contained 50 pmoles Tris maleate, 5 pmoles UDPG, IO pmoles G 6-P, 5 pmoles EDTA . Na,, 8 mg glycogen and 0.2 ml enzyme in a total volume of 1.0 ml. The red cell enzyme was added in concentrations equivalent to 20 to 30 mg of hemoglobin and the white cell enzyme in amounts equivalent to approximately z . 106 leukocytes. The reaction mixtures were incubated at 37’ and the reaction stopped by boiling for 5 min. The tubes were cooled, centrifuged at 1500 rev./min for 5 min. and the supernates filtered through glass wool into IO x 75-mm tubes. The filtered supernates were kept iced until assayed. UDP concentration was analyzed as described by Leloir and Goldemberg*. Enzyme activity was expressed as pmoles of UDP formed per g of Hb or per 10~ WBC per IO min, or of [Cl41 glucose incorporated into glycogen. RESULTS
The glucosyltransferase activity in the erythrocyte was estimated by measuring UDP formed. Initially, the specificity of the reaction to assay UDP was determined. At o time, there was approximately 0.11 to 0.12 ,umoles of high-energy * Abbreviations used: Ethylenediaminetetraacetate disodium: EDTX Na,; glucose 6-phosphate : G 6-P; uridine diphosphate: UDP; uridine diphosphoglucose : UDPG; uridine monophosphate : UMP: tris (hydroxymethyl) aminomethane maleate : Tris maleate; erythrocyte: rbc ; leukocyte: wbc. Clin. China. Acta,
12 (1965)
27--32
GLUCOSYLTRANSFERASE
IN ERYTHROCYTES
29
phosphate acceptors present per ml of red cell hymolysate and this did not increase with incubation in the absence of G 6-P of substrate (Fig. I). Therefore, increments over the zero time value were due to UDP formation from UDPG and glycogen synthesis. The reaction rate was linear for 60 to 75 min at optimal substrate (5pmoles/ml) and activator concentrations as measured by UDP formation or by incorporation of [Cl&j glucose into glycogen (Fig. I). With the latter assay, on paper chromatography of the reaction products, no UMP or free labelled glucose was detected, and the UDP formed was equivalent to the amount of Cl* incorporated into glycogen. Activity of the enzyme was influenced by UDPG concentration and no activity due to nonspecific phosphate acceptors occurred in the absence of UDPG (Fig. I). UDPG, G 6-P 14-
UDPG,G
&P
Fig. I. The effect of the concentration of UDPG and G 6-P on the assay of glucosyltransferase activity in erythrocytes as measured by UDP formation and the incorporation of [“Cl glucose into giycogen.
The enzyme was completely G 6-P dependent. No activity could be measured in red or white cell preparations in the absence of G 6-P. The optimal concentration of G 6-P was approximately IO pmoles/ml. In the absence of glycogen primer, only 40 to 50% of the original activity was present. The small quantities of glycogen in the red or white cells may be sufficient to prime the reaction. In order to exclude the possibility that the glucosyltransferase activity in the red cell hymolysates was due to contaminating leukocytes, the relative activitiesof the red and white cells were estimated. This was done by counting residual leukocytes in 7 red cell preparations and residual red cells in II white cell preparations, The activity per white cell was approximately 400 times that in the red cell. Since there was approximately one white cell per r600 rbc in the red cell preparations, the activity in the leukocyte could at most account for only 25% of the total activity (Table I). It can be concluded that the major fraction of glucosyltransferase activity in these preparations is derived from the erythrocytes. Attempts were made to determine whether the leukocyte and erythrocyte enzymes differed in any characteristic properties. The PH optimum in Tris maleate buffer (0.05 M) was approximately 7.6 for both enzyme preparations (Fig. 2). Both C&T&. Chim. Acta,
I2
(~965)
27-32
bJ. COJINJIJ_\‘l’H~‘f n/.
20-
68
7.2
Z6
8.0
a4
8.a
9.2
PH
Fig. 2. The PH optimum of glucosyltransferasc in erythrocytes and leukocytes in 0.5 M Tris malcate buffer. Each point represents the mean of 3 separate assays. The results are expressed as percentage of maximal activity.
red and white cell enzymes required G 6-P and the same concentration of substrate for optimal activity. Preincubation of the cell preparations at 37’ for 20 to 40 min resulted in some loss of activity from the white cell but not from the red cell preparations. However, the significance of this observation is uncertain and further purification of the enzyme from both sources will be necessary the same or different
to establish
whether these are
forms of glucolsyltransferase.
Glucosyltransferase activity in erythrocytes from controls and $atients Glucosyltransferase activity was measured in the red cells of normal
infants,
children and adults. The mean activity (I_t S.D.) was 1.57 * 0.14 pmoles UDP/g Hb/Io min in 6 premature and 1.45 & 0.18 pmoles UDP/g Hb/ro min in 8 full-term newborn infants. Values of 0.76 f 0.16 and 1.06 f 0.39 pmoles UDP/g Hb/ro min were found in erythrocytes from 6 children aged 3 to IO years and in 15 adults respectively. Normal enzyme activities of 1.9 and 2.0 pmoles UDP/g Hb/Io min were found in the red blood cells of z patients who had fasting hypoglycemia due to a deficiency of glucosyltransferase in the liver as reported by Lewis et a1.1”. The siblings and parents of these twins also had normal levels of glucosyltransferase activity in the rbc, ranging from 1.7 to 2.2 ,umoles UDP/g Hb/ro min*. * The autors wish to thank Drs. G. M. Lewis and J. Spencer-Peet for providing the prepared hemolysates from the hepatic glucosyltransferase-deficient twins and their family. Thanks are also due to Drs. Robert A. Ulstrom and William A. Cochrane for sending red cell preparations from patients with hypoglycemia and glycogen storage disease. C&n. Chim. Acta,
12
(1965)
27-32
GLUCOSYLTRANSFERASE
IN ERYTHROCYTES
31
Normal enzyme activities were found in the erythrocytes of a number of patients with abnormalities of carbohydrate metabolism. The red cell preparations from 7,patients with Type I glycogen storage disease (Von Gierke), 3 with juvenile diabetes mellitus, I with hereditary fructose intolerance, 3 with idiopathic hypoglycemia and I with leucine-sensitive hypoglycemia were within the normal range of activity (0.65 to in the red blood 3.2 pmoles UDP/g Hb/xo min) *. The activity of glucosyltransferase cell could not be correlated with a variety of aberrations of carbohydrate metabolism. The activity of glucosyltransferase in the leukocytes from II normal adults ranged from 1.0 to 4.2 ,umoles UDP/ro* wbc/ro min. Leukocyte preparations from 3 children with Down’s syndrome (Trisomy 21) had similar enzyme activities (2.4, 2.5 and 32 pmoles UDP/108 wbc/ro mm). DISCUSSION
The presence of glycogen in the leukocyte’” and erythrocyte*a prompted the investigation of glucosyltransferase activity in these cells. In addition, since the enzymatic defect in galactosemia can be demonstrated in the red ce1113, glucosyltransferase assays were done in red cells from twins who had clinical hypoglycemia and proven glucosyltransferase deficiency of the liver in onelo. The specificity of the assay methods for glucosyltransferase was established by anumber of parameters. The reaction was substrate-, primer-, and activator-dependent. Furthermore, comparable activities were found by measuring UDP formation or 1% incorporation into glycogen. By counting residual white cells in the erythrocyte preparations and residual red cells in leukocyte preparations, it was possible to determine that the major portion of the glucosyltransferase activity in red cell hemolysates was from the erythrocytes. No glucosyltransf&ase activity could be measured in either rbc or wbc preparations in the absence of G 6-P. The pH optimum of both preparations was identical. Although preincubation at 37” resulted in no loss of activity in the rbc and some in the wbc, the leukocyte preparation could be protected by the addition of amino acid buffers and could not be reactivated with KF and G 6-P’*. Thus, the enzyme in the peripheral blood cells in crude hemolysates differs from hepatic glucosyltransferase which is labile if preincubated at 37”, can be reactivated, retains some activity in the absence of G G-P, and has a PH optimum of PH 8.2 8, It. Muscle glucosyltransferase is also partially active in the absence of G 6-P, but does resist preincubation inactivations. The inability to distinguish levels of glucosyltransferase activities in the red cells of normal infants and children and of a patient with a proven deficiency of glucosyltransferase in the liver was disappointing. One explanation for this observation may be that the glucosyltransferase enzyme of liver cells differs from the correspondingenzyme of red cells, as suggested in the case of the liver and muscle enzymes by Steiner and Kings. The possibility of an inhibitor of this enzyme in the liver of the patient reported was ruled out by Lewis et aE.l”. No significant differences in red cell enzyme activities were observed in a variety of clinical diseases associated with low (juvenile diabetic in ketosis) or high stores (glycogen storage disease) of glycogen in the liver or hypoglycemia. C&n. Chim. Acta, IZ (1965)
27-32
hl.(‘ORNliL~l‘H Pf id.
32
The authors express secretarial assistance.
their
appreciation
to Mrs. De Lore.-; Stratten
for her
27
URIDINE-DIPHOSPHOGLUCOSE
GLUCOSYLTRANSFERASE
IN
HUMAN
ERYTHROCYTES* M. CORNBLATH,
D. F. STEINER,
P. BRYAN
Departments of Pediatrics, University of Illinois School of Medicine, Chicago, Ill. (U.S.,4 .)
(Received
AND J. KING College
of Medicine,
and
University
of Chicago
August zest, 1964)
SUMMARY
The presence of glucosyltransferase activity in mature human erythrocytes and leukocytes has been established both by demonstration of UDP formation from UDPG, and by incorporation of [‘“Cl glucose from UDP-[WI glucose into glycogen. The erythrocyte and leukocyte enzymes have similar properties and both require glucose 6-P for activity. The enzyme activity per white cell is about 400 times that per red cell. The erythrocyte glucosyltransferase activities did not differ from that in the normal in two patients with hepatic glucosyltransferase deficiency and in other patients with a variety of carbohydrate abnormalities.
INTRODUCTION
The uridine-diphosphoglucose (UDPG) pathway for glycogen synthesis, described by Leloir l, has been found in all tissues examined 2, except rabbit uterus 3. The presence of glycogen in the mature human erythrocyte prompted an investigation of UDPG-glycogen glycosyltransferase (EC 2.4.1.11) in the red cell. This was particularly pertinent because phosphorylase activity has been demonstrated in these cells 4. Glucosyltransferase has recently been demonstrated both chemically and histochemically in leukocytes 5po.The purpose of this paper is to report the activity and characteristics of glucosyltransferase in human erythrocytes and leukocytes. Levels of enzyme activity in red blood cells from normal infants, children and adults, patients with hepatic glucosyltransferase deficiency and a variety of other carbohydrate disorders are presented.
* This National
investigation was supported in part by Research Grants Institute of Arthritis and Metabolic
6015-04 and 4931-04 from Disease, United States Public Health Service.
Clin. Chim.
Acta,
I2 (1965)
the
27-32
MATERI:\LS ANI) METHOI)S
A. I’vepavation
ofevythvocytes *
Blood was drawn into EDTA . tia, (1.2 mgjml), centrifuged at 4 at L -poo rev./min for IO min. The plasma and buffy coat were removed. The red ~~11s were washed three times in cold 0.15 df NaF containing 0.001 M EDTA . Na,, and then suspended in an equal volume of 0.25 32 sucrose containing 0.02 dl NaF and 0.001 32 EDTA. The hemolysate was prepared by quick-freezing in dry ice and alcohol twice. This preparation stored at --IO’ is stable for at least 2 months. Hemoglobin dettarminations were made prior to assay by standard methods. B. Preparatiofz of leukocytes
Blood was drawn into EDTA . Na, (1.2 mg/ml). The leukocytes were then concentrated by differential sedimentation with polybrene according to Lalezari7. The leukocyte-rich plasma was centrifuged at 4’ at 10,000 rev./min for 30 min. The packed cells were washed once with 0.15 M NaF containing 0.001 M EDTA and resuspended in 0.25 M sucrose containing 0.02 M NaF and 0.001 J4 EDTA. Samples were taken for white blood cell counts and the remainder frozen in dry ice and alcohol and thawed three times, and stored in a freezer at -IO’. Enzyme
assays
The activity of glucosyltransferase in erythrocyte and leukocyte preparations was assayed by two methods: (I) chemically by measuring UDP formation by a modification of the method of Leloir and Goldemberg*, and (2) by the incorporation of [Cl41 glucose into glycogen from [Cl41 UDPG as described by Steiner and Kin@. In the former method, UDP formation was measured at o, 45, 60, and 75 min. The reaction mixtures contained 50 pmoles Tris maleate, 5 pmoles UDPG, IO pmoles G 6-P, 5 pmoles EDTA . Na,, 8 mg glycogen and 0.2 ml enzyme in a total volume of 1.0 ml. The red cell enzyme was added in concentrations equivalent to 20 to 30 mg of hemoglobin and the white cell enzyme in amounts equivalent to approximately z . 106 leukocytes. The reaction mixtures were incubated at 37’ and the reaction stopped by boiling for 5 min. The tubes were cooled, centrifuged at 1500 rev./min for 5 min. and the supernates filtered through glass wool into IO x 75-mm tubes. The filtered supernates were kept iced until assayed. UDP concentration was analyzed as described by Leloir and Goldemberg*. Enzyme activity was expressed as pmoles of UDP formed per g of Hb or per 10~ WBC per IO min, or of [Cl41 glucose incorporated into glycogen. RESULTS
The glucosyltransferase activity in the erythrocyte was estimated by measuring UDP formed. Initially, the specificity of the reaction to assay UDP was determined. At o time, there was approximately 0.11 to 0.12 ,umoles of high-energy * Abbreviations used: Ethylenediaminetetraacetate disodium: EDTX Na,; glucose 6-phosphate : G 6-P; uridine diphosphate: UDP; uridine diphosphoglucose : UDPG; uridine monophosphate : UMP: tris (hydroxymethyl) aminomethane maleate : Tris maleate; erythrocyte: rbc ; leukocyte: wbc. Clin. China. Acta,
12 (1965)
27--32
GLUCOSYLTRANSFERASE
IN ERYTHROCYTES
29
phosphate acceptors present per ml of red cell hymolysate and this did not increase with incubation in the absence of G 6-P of substrate (Fig. I). Therefore, increments over the zero time value were due to UDP formation from UDPG and glycogen synthesis. The reaction rate was linear for 60 to 75 min at optimal substrate (5pmoles/ml) and activator concentrations as measured by UDP formation or by incorporation of [Cl&j glucose into glycogen (Fig. I). With the latter assay, on paper chromatography of the reaction products, no UMP or free labelled glucose was detected, and the UDP formed was equivalent to the amount of Cl* incorporated into glycogen. Activity of the enzyme was influenced by UDPG concentration and no activity due to nonspecific phosphate acceptors occurred in the absence of UDPG (Fig. I). UDPG, G 6-P 14-
UDPG,G
&P
Fig. I. The effect of the concentration of UDPG and G 6-P on the assay of glucosyltransferase activity in erythrocytes as measured by UDP formation and the incorporation of [“Cl glucose into giycogen.
The enzyme was completely G 6-P dependent. No activity could be measured in red or white cell preparations in the absence of G 6-P. The optimal concentration of G 6-P was approximately IO pmoles/ml. In the absence of glycogen primer, only 40 to 50% of the original activity was present. The small quantities of glycogen in the red or white cells may be sufficient to prime the reaction. In order to exclude the possibility that the glucosyltransferase activity in the red cell hymolysates was due to contaminating leukocytes, the relative activitiesof the red and white cells were estimated. This was done by counting residual leukocytes in 7 red cell preparations and residual red cells in II white cell preparations, The activity per white cell was approximately 400 times that in the red cell. Since there was approximately one white cell per r600 rbc in the red cell preparations, the activity in the leukocyte could at most account for only 25% of the total activity (Table I). It can be concluded that the major fraction of glucosyltransferase activity in these preparations is derived from the erythrocytes. Attempts were made to determine whether the leukocyte and erythrocyte enzymes differed in any characteristic properties. The PH optimum in Tris maleate buffer (0.05 M) was approximately 7.6 for both enzyme preparations (Fig. 2). Both C&T&. Chim. Acta,
I2
(~965)
27-32
bJ. COJINJIJ_\‘l’H~‘f n/.
20-
68
7.2
Z6
8.0
a4
8.a
9.2
PH
Fig. 2. The PH optimum of glucosyltransferasc in erythrocytes and leukocytes in 0.5 M Tris malcate buffer. Each point represents the mean of 3 separate assays. The results are expressed as percentage of maximal activity.
red and white cell enzymes required G 6-P and the same concentration of substrate for optimal activity. Preincubation of the cell preparations at 37’ for 20 to 40 min resulted in some loss of activity from the white cell but not from the red cell preparations. However, the significance of this observation is uncertain and further purification of the enzyme from both sources will be necessary the same or different
to establish
whether these are
forms of glucolsyltransferase.
Glucosyltransferase activity in erythrocytes from controls and $atients Glucosyltransferase activity was measured in the red cells of normal
infants,
children and adults. The mean activity (I_t S.D.) was 1.57 * 0.14 pmoles UDP/g Hb/Io min in 6 premature and 1.45 & 0.18 pmoles UDP/g Hb/ro min in 8 full-term newborn infants. Values of 0.76 f 0.16 and 1.06 f 0.39 pmoles UDP/g Hb/ro min were found in erythrocytes from 6 children aged 3 to IO years and in 15 adults respectively. Normal enzyme activities of 1.9 and 2.0 pmoles UDP/g Hb/Io min were found in the red blood cells of z patients who had fasting hypoglycemia due to a deficiency of glucosyltransferase in the liver as reported by Lewis et a1.1”. The siblings and parents of these twins also had normal levels of glucosyltransferase activity in the rbc, ranging from 1.7 to 2.2 ,umoles UDP/g Hb/ro min*. * The autors wish to thank Drs. G. M. Lewis and J. Spencer-Peet for providing the prepared hemolysates from the hepatic glucosyltransferase-deficient twins and their family. Thanks are also due to Drs. Robert A. Ulstrom and William A. Cochrane for sending red cell preparations from patients with hypoglycemia and glycogen storage disease. C&n. Chim. Acta,
12
(1965)
27-32
GLUCOSYLTRANSFERASE
IN ERYTHROCYTES
31
Normal enzyme activities were found in the erythrocytes of a number of patients with abnormalities of carbohydrate metabolism. The red cell preparations from 7,patients with Type I glycogen storage disease (Von Gierke), 3 with juvenile diabetes mellitus, I with hereditary fructose intolerance, 3 with idiopathic hypoglycemia and I with leucine-sensitive hypoglycemia were within the normal range of activity (0.65 to in the red blood 3.2 pmoles UDP/g Hb/xo min) *. The activity of glucosyltransferase cell could not be correlated with a variety of aberrations of carbohydrate metabolism. The activity of glucosyltransferase in the leukocytes from II normal adults ranged from 1.0 to 4.2 ,umoles UDP/ro* wbc/ro min. Leukocyte preparations from 3 children with Down’s syndrome (Trisomy 21) had similar enzyme activities (2.4, 2.5 and 32 pmoles UDP/108 wbc/ro mm). DISCUSSION
The presence of glycogen in the leukocyte’” and erythrocyte*a prompted the investigation of glucosyltransferase activity in these cells. In addition, since the enzymatic defect in galactosemia can be demonstrated in the red ce1113, glucosyltransferase assays were done in red cells from twins who had clinical hypoglycemia and proven glucosyltransferase deficiency of the liver in onelo. The specificity of the assay methods for glucosyltransferase was established by anumber of parameters. The reaction was substrate-, primer-, and activator-dependent. Furthermore, comparable activities were found by measuring UDP formation or 1% incorporation into glycogen. By counting residual white cells in the erythrocyte preparations and residual red cells in leukocyte preparations, it was possible to determine that the major portion of the glucosyltransferase activity in red cell hemolysates was from the erythrocytes. No glucosyltransf&ase activity could be measured in either rbc or wbc preparations in the absence of G 6-P. The pH optimum of both preparations was identical. Although preincubation at 37” resulted in no loss of activity in the rbc and some in the wbc, the leukocyte preparation could be protected by the addition of amino acid buffers and could not be reactivated with KF and G 6-P’*. Thus, the enzyme in the peripheral blood cells in crude hemolysates differs from hepatic glucosyltransferase which is labile if preincubated at 37”, can be reactivated, retains some activity in the absence of G G-P, and has a PH optimum of PH 8.2 8, It. Muscle glucosyltransferase is also partially active in the absence of G 6-P, but does resist preincubation inactivations. The inability to distinguish levels of glucosyltransferase activities in the red cells of normal infants and children and of a patient with a proven deficiency of glucosyltransferase in the liver was disappointing. One explanation for this observation may be that the glucosyltransferase enzyme of liver cells differs from the correspondingenzyme of red cells, as suggested in the case of the liver and muscle enzymes by Steiner and Kings. The possibility of an inhibitor of this enzyme in the liver of the patient reported was ruled out by Lewis et aE.l”. No significant differences in red cell enzyme activities were observed in a variety of clinical diseases associated with low (juvenile diabetic in ketosis) or high stores (glycogen storage disease) of glycogen in the liver or hypoglycemia. C&n. Chim. Acta, IZ (1965)
27-32
hl.(‘ORNliL~l‘H Pf id.
32
The authors express secretarial assistance.
their
appreciation
to Mrs. De Lore.-; Stratten
for her