British Journal of Hdematology, 1979, 42, 39p-402.
Fatty Acid Composition of Erythrocytes in Hereditary Spher o cy to sis S. S. ZAILAND A. PICKERING
Department of Haematology, School of Pathology, South African Institute for Medical Research, and University of Wifwatersrand,Johannesburg (Received 6 July 1978; accepted for publication 23 October 1978)
SUMMARY, The fatty acid distribution of the four major phospholipids has been determined in the erythrocytes of five patients with hereditary spherocytosis and in five control subjects. In contrast to a report by other workers, we are unable to confirm their findings of a defect in fatty acid chain lengthening activity leading to disappearance of long chain fatty acid conjugates of more than 20 carbon atoms in some of the phospholipid fractions. Attempts to define the molecular abnormality underlying the spherocytic shape transformation in erythrocytes from patients with hereditary spherocytosis (HS) have been concentrated in several areas of red cell membrane chemistry. These include the role of membrane-bound protein kinase and possible abnormalities in the structural proteins and lipids of these membranes (for reviews see Valentine, 1977, and Zail, 1977). Earlier studies by several investigators had shown no abnormalities in the cholesterol content and phospholipid distribution in HS erythrocyte membranes (Van Deenen & de Gier, 1974). However, a finding of some interest was the demonstration by Kuiper & Livne (1972) that HS erythrocyte membranes have decreased fatty acid chain lengthening activity, resulting in the disappearance of long chain fatty acid conjugates of more than 20 carbon atoms in the sphingomyelin, lecithin and phosphatidylserine fractions. It was postulated that this could lead to a change in the interaction of lipid and membrane protein and that this might be responsible for some of the biochemical abnormalities of HS erythrocytes. Further evidence that a membrane lipid abnormality may be present in HS was provided by Aloni et al (197s) who found that the microviscosity of the membrane lipid core as measured by a fluorescence polarization technique in intact cells as well as ghosts was increased, indicating a tighter packing of these lipids. We have repeated the experiments of Kuiper & Livne (1972) as we.feel their findings may be of some importance in understanding the pathogenesis of the membrane abnormality in HS. In the event, we are unable to confirm their findings of decreased fatty acid chain lengthening activity in HS.
Correspondence: Dr S. S. Zail, Department of Haematology, School of Pathology, South African Institute for Medical Research,Johannesburg, South Africa. 0007-1048/79/0700-0399$02.00
01979 Blackwell Scientific Publications 3 99
400
S . S . Zail and A . Pickering
MATERIAL AND METHODS Venous blood anticoagulated with EDTA was obtained from five healthy control subjects and five patients with HS. All the patients had been splenectomized and had normal haemoglobin values and reticulocyte counts. All showed the classic erythrocyte morphology, osmotic fragility, autohaemolysis and autosomal dominant mode of inheritance characteristic of this disease. The blood was cooled a t 4OC in an ice bath, washed three times in 0.9% NaCl, care being taken to remove the buffy coat as completely as possible after each wash. One volume of packed cells was frozen and thawed three times in a dry-ice acetone bath and the lipids extracted with methanol and chloroform as described by Dodge & Phillips (1967). The antioxidant 2,6-di-tert-butyl-p-cresol (BHT) was added to the methanol a t a concentration of 5 mg/dl to prevent autoxidation of the paraffin chains. After drying in vacuo a t 40°C in a rotary vacuum evaporator, the lipid extract was applied on a 0.25 mm layer of silica gel (Merck, Darmstadt) as a IOO mm long streak and chromatographed for approximately 3 0 min in chloroform-methanol-glacial acetic acid-water 60:30:8.4:3.6 to which was added BHT a t a concentration of 50 mg/dl to prevent autoxidation during chromatography. The individual phospholipids were identified by staining a small section of the silica gel plate with iodine vapour, and the appropriate areas of silica gel were rapidly scraped into ampoules. 2 ml boron trifluoride-methanol reagent (British Drug Houses, Dorset) was added and the mixture refluxed a t 90°C for 10 min in the case of the phosphatidylcholine, phosphatidylserine and phosphatidyl-ethanolamine fractions and for 90 min for the sphingomyelin fraction. The fatty acid methyl esters were extracted into hexane and stored a t -20°C until analysis. Gas-liquid chromatography was performed with a Uecker model 417 instrument equipped with a flame ionization detector. The column was composed of 10% diethylene glycol succinate as the stationary phase on 80/1oo Supelcoport (Supelco Inc., Pennsylvania) as a support. Isothermal separations were performed at ~ 0 0 ° CIndividual . peaks were indentified by comparison to retention times of mixtures of known fatty acid methyl esters supplied by Alltech Associates, Illinois, which included the methyl esters of 16:o, 16:I , 18:0, I 8: I , 1 8 : 2 , 1 8 : 3 , zo:o, 20:1,20:2,20:3, 20:4, 22:0, 22:2,22:3, 22:4,22:6, 24:0,24:1,as well as by plots of the log retention time versus carbon number. In this abbreviated notation of the fatty acids, the first two digits represent the numbers of carbon atoms and the third digit the number of double bonds in the acyl chain. Because of interference by a BHT artefact, 16:1 was not estimated (Dodge & Phillips, 1967).In addition, in one control and one HS extract, identification of the individual peaks was confirmed by comparison to the known Alltech fatty acid methyl esters which were separated on a less polar column of 10%silicone OV-IOI as the stationary phase on 80/1oo mesh Chromosorb W-AW (Chemlab, Pinegowrie). Quantitation of peak areas was performed by an Autolab ‘Minigrator’ (Spectrophysics) linked to the gas chromatograph, and were considered to be directly proportional to the weight of the component fatty acids. RESULTS AND DISCUSSION The fatty acid composition of the four major phospholipids in five control and five HS erythrocyte preparations is shown in Table I. There were no significant differences in the distribution of any of the fatty acids as determined by two-tailed unpaired t-tests ( P > 0 . I ) . The
Fatty Acid Composition of Erythrocytes
401
TABLE I. Fatty acid composition of phospholipids in normal and hereditary spherocytosis (HS) erythrocytes Fatty acid distribution ( g / i o ogfatty acid) Sph ingomyelin HS
Normal Fatty arid* 16:o 18:o 18:1
18:~ 20:o 20:1+
18:3
20:2
Mean
Range
Mean
20.9-3 1.0 26.8 5.5-11.6 8.9 2.0-5.3 3.3 0.9-4.6 2 .o 0.7-1.6 1 .4 0.6-3.9 0.5 -
27.3 9.4 4.4 2.4 1 .4 1.9 -
z0:3 +22:0 8.8 20:4 + 2 2 : I 0.8 24:o 22.7 24:1+22:3 +22:4 20.3 22:6 0.6
Phosphatidy lserine
5.1-10.1 9.4 0.3-1.5 1.4 1 6 . ~ ~ 2 6 . 22.8 2 16.1-2 I .8 22.3 0.2-1 .o 1.2
Range
Normal Mean
22.1-28. I 4.9 4.7- 12.5 47.3 8.1 2.3-3.7 I .2-2.5 4-2 I .o-I.6 0.4 1.1 0.2-1 .o 0.6 6 . ~ 1 0 . 8 1.8 19. I 1.-1.7 1 7 . ~ 2 5 . 8 0.8 I 8.7-23.5 4.9 0.6-1.8 6.7
Fattyacid
0.1-0.5
0.5-2.3 0.2-2.0 1.2-2.9 14.8-22.8 0-2.0 3.5-5.4 3.7-9.8
Range
Mean
18:1
37.5 12.9 16.8
18:2
22.5
40.2 13.4 17.7 19.6
20:o 20: I
0.2
27.2-40.0 7.4-15.1 7.0-19.1 14.0-26.6 0-0.8 0.6-1.6 0-1.6 0-1.8 2.9-4.8 0.1-0.4 0.3-1.6 1.24.4
16:o 18:o
+ 18:3
0.5
20:2
0.5
20:3 +22:0 20:4+ I I : I 24:o 24:1+22:3+22:4 22:6
1.3 4.0 0.2 0.8 2.7
0.1
0.7 0.3 I .9 21.4 0.1
4. 1 7.2
Range 3.3-7.8 39.2-51.2 7.4-1 I . 1 2.54.1 0-0.2 0.3-1.6 0.2~,.7 1.4-2.2 I 7.8-23.2 00.4 3.7-4.4 4.7-8.0
Phosphatidylethanolamine
HS
Mean
Mean
5.0 3.0-6.3 40.9-53.7 45.9 6.5-8.8 9.4 2.3- 5.9 3.9
Phosphatidylcholine Normal
Range
HS
Range
35.F47.5 12.3-14.0 15.8-19.9 14.6-25.7 0.1 0-0.2 6.4 , 00.6 0.3 ' 00.5 0.4-1.7 1.4 3.24.7 5.0 0.1 o.o5*.15 0.5 0.4-0.6 1.3 0.9-1.7
HS
Normal Mean
Range
Mean
Range
17.9 12.4 15.9 8.7 0.6 1.4
13.5-20.4 8.1-13.7 12.8-19.8 4.8-12.5 0.1-1.5 0.4-2.5 00.3
17.8 12.3 15.5 6.4 0.2 0.7 0.4
11.6-20.3 9.5-14.0 11.8-17.7 4.8-8.2
0.2 1.0
23.6 0.3 9.7 8.3
0 . ~ 1 . 2 1.1
18.9-24.6 25.9 0.3 m.7 7.0-11.6 10.8 6.8-9.6 8.5
-1.0
0.4Q.9 0.2-0.6 0.8-1.4 22.3-27.0 0-0.6 8.0-12.9 7.4-9.1
* The first two digits state the number of carbon atomsand the third digit states the number of double bonds of the individual fatty acids. distribution of the fatty acids in all the phospholipid fractions is very similar to that found by other workers (Ways & Hanahan, 1964; Dodge & Phillips, 1967). O f particular importance was the similar percentage composition of the 24:o and 24: I fractions of sphingomyelin in the control and HS erythrocytes, as these fatty acids in sphingomyelin comprise the highest proportion of long chain fatty acids above 20 carbon atoms found in any of the phospholipid fractions. Only trace quantities of these two fatty acids were found in HS erythrocytes by
402
S. S. Zail and A . Pickering
Kuiper & Livne (1972). In our gas chromatographic separation the methyl esters of 22:3 and 22:4 have the same retention times as 24:r. However, 22:3 and 22:4 are present only in trace amounts particularly in the sphingomylin fraction (Dodge & Phillips, 1967), and do not contribute significantly to the values for 24:i found in this study. This was confirmed in the separations on the less polar silicone OV-ror column, where 22:3 and 22:4 were clearly separated from 24: I . At present we cannot offer a reasonable explanation for the difference between our findings and those of Kuiper & Livne (1972). One possibility is that hereditary spherocytosis may be a heterogenous disease and that the cases described by the latter workers may come from a single ethnic group showing the described defect. Whatever the reason, we feel that it would be of interest to determine the fatty acid composition of HS membranes in other laboratories to try and resolve this problem. ACKNOWLEDGMENTS
These studies were supported by grants from the South African Medical Research Council and the Atomic Energy Board. W e thank Dr 0. Abrahams for expert advice and assistance. REFERENCES
ALONI, B., SHINITZKY, M., MOSES,S. Ik LIVNE,A. (1975) Elevated microviscosity in membranes of erythrocytes affected by hereditary spherocytosis. BritidtJoirrnal ofHaenzarolofy, 31, I 17-123. G.B. (1967) Composition of I ~ O D GJ.T. E , 1(( PHILLIPS, phospholipids and of phospholipid fatty acids and aldehydes in human red cells. Journal O J Lipid Research, 8,667-475. KUIPER, J.C.P. & LIVNE,A. (1972) Differences in fatty acid composition between normal erythrocytes and hereditary spherocytosis affected cells. Biochimira ef Biophysica Acta, 260, 755-758.
VALENTINE, W.N. (1977) The molecular lesion ofhereditary spherocytosis (HS): a continuing enigma. Blood, 49. 241-245. VANDEENEN, L.M. & DE GIER,J. (1974) Lipids of the red cell membrane. T h e Red Blood Cell, 2nd edn (ed. by 11. MacN. Surgenor), p. 147. Academic Press, New York. WAYS,P. & HANAHAN, D.J. (1964) Characterization and quantification of red cell lipids in normal man. Journal ofLipid Research, 5 , 3 18-328. ZAIL.S.S. (1977) The erythrocyte abnormality of hereditary spherocytosis. (Annotation). BririshJournal of Haemafology, 37, 305-3 10.
Fatty Acid Composition of Erythrocytes in Hereditary Spher o cy to sis S. S. ZAILAND A. PICKERING
Department of Haematology, School of Pathology, South African Institute for Medical Research, and University of Wifwatersrand,Johannesburg (Received 6 July 1978; accepted for publication 23 October 1978)
SUMMARY, The fatty acid distribution of the four major phospholipids has been determined in the erythrocytes of five patients with hereditary spherocytosis and in five control subjects. In contrast to a report by other workers, we are unable to confirm their findings of a defect in fatty acid chain lengthening activity leading to disappearance of long chain fatty acid conjugates of more than 20 carbon atoms in some of the phospholipid fractions. Attempts to define the molecular abnormality underlying the spherocytic shape transformation in erythrocytes from patients with hereditary spherocytosis (HS) have been concentrated in several areas of red cell membrane chemistry. These include the role of membrane-bound protein kinase and possible abnormalities in the structural proteins and lipids of these membranes (for reviews see Valentine, 1977, and Zail, 1977). Earlier studies by several investigators had shown no abnormalities in the cholesterol content and phospholipid distribution in HS erythrocyte membranes (Van Deenen & de Gier, 1974). However, a finding of some interest was the demonstration by Kuiper & Livne (1972) that HS erythrocyte membranes have decreased fatty acid chain lengthening activity, resulting in the disappearance of long chain fatty acid conjugates of more than 20 carbon atoms in the sphingomyelin, lecithin and phosphatidylserine fractions. It was postulated that this could lead to a change in the interaction of lipid and membrane protein and that this might be responsible for some of the biochemical abnormalities of HS erythrocytes. Further evidence that a membrane lipid abnormality may be present in HS was provided by Aloni et al (197s) who found that the microviscosity of the membrane lipid core as measured by a fluorescence polarization technique in intact cells as well as ghosts was increased, indicating a tighter packing of these lipids. We have repeated the experiments of Kuiper & Livne (1972) as we.feel their findings may be of some importance in understanding the pathogenesis of the membrane abnormality in HS. In the event, we are unable to confirm their findings of decreased fatty acid chain lengthening activity in HS.
Correspondence: Dr S. S. Zail, Department of Haematology, School of Pathology, South African Institute for Medical Research,Johannesburg, South Africa. 0007-1048/79/0700-0399$02.00
01979 Blackwell Scientific Publications 3 99
400
S . S . Zail and A . Pickering
MATERIAL AND METHODS Venous blood anticoagulated with EDTA was obtained from five healthy control subjects and five patients with HS. All the patients had been splenectomized and had normal haemoglobin values and reticulocyte counts. All showed the classic erythrocyte morphology, osmotic fragility, autohaemolysis and autosomal dominant mode of inheritance characteristic of this disease. The blood was cooled a t 4OC in an ice bath, washed three times in 0.9% NaCl, care being taken to remove the buffy coat as completely as possible after each wash. One volume of packed cells was frozen and thawed three times in a dry-ice acetone bath and the lipids extracted with methanol and chloroform as described by Dodge & Phillips (1967). The antioxidant 2,6-di-tert-butyl-p-cresol (BHT) was added to the methanol a t a concentration of 5 mg/dl to prevent autoxidation of the paraffin chains. After drying in vacuo a t 40°C in a rotary vacuum evaporator, the lipid extract was applied on a 0.25 mm layer of silica gel (Merck, Darmstadt) as a IOO mm long streak and chromatographed for approximately 3 0 min in chloroform-methanol-glacial acetic acid-water 60:30:8.4:3.6 to which was added BHT a t a concentration of 50 mg/dl to prevent autoxidation during chromatography. The individual phospholipids were identified by staining a small section of the silica gel plate with iodine vapour, and the appropriate areas of silica gel were rapidly scraped into ampoules. 2 ml boron trifluoride-methanol reagent (British Drug Houses, Dorset) was added and the mixture refluxed a t 90°C for 10 min in the case of the phosphatidylcholine, phosphatidylserine and phosphatidyl-ethanolamine fractions and for 90 min for the sphingomyelin fraction. The fatty acid methyl esters were extracted into hexane and stored a t -20°C until analysis. Gas-liquid chromatography was performed with a Uecker model 417 instrument equipped with a flame ionization detector. The column was composed of 10% diethylene glycol succinate as the stationary phase on 80/1oo Supelcoport (Supelco Inc., Pennsylvania) as a support. Isothermal separations were performed at ~ 0 0 ° CIndividual . peaks were indentified by comparison to retention times of mixtures of known fatty acid methyl esters supplied by Alltech Associates, Illinois, which included the methyl esters of 16:o, 16:I , 18:0, I 8: I , 1 8 : 2 , 1 8 : 3 , zo:o, 20:1,20:2,20:3, 20:4, 22:0, 22:2,22:3, 22:4,22:6, 24:0,24:1,as well as by plots of the log retention time versus carbon number. In this abbreviated notation of the fatty acids, the first two digits represent the numbers of carbon atoms and the third digit the number of double bonds in the acyl chain. Because of interference by a BHT artefact, 16:1 was not estimated (Dodge & Phillips, 1967).In addition, in one control and one HS extract, identification of the individual peaks was confirmed by comparison to the known Alltech fatty acid methyl esters which were separated on a less polar column of 10%silicone OV-IOI as the stationary phase on 80/1oo mesh Chromosorb W-AW (Chemlab, Pinegowrie). Quantitation of peak areas was performed by an Autolab ‘Minigrator’ (Spectrophysics) linked to the gas chromatograph, and were considered to be directly proportional to the weight of the component fatty acids. RESULTS AND DISCUSSION The fatty acid composition of the four major phospholipids in five control and five HS erythrocyte preparations is shown in Table I. There were no significant differences in the distribution of any of the fatty acids as determined by two-tailed unpaired t-tests ( P > 0 . I ) . The
Fatty Acid Composition of Erythrocytes
401
TABLE I. Fatty acid composition of phospholipids in normal and hereditary spherocytosis (HS) erythrocytes Fatty acid distribution ( g / i o ogfatty acid) Sph ingomyelin HS
Normal Fatty arid* 16:o 18:o 18:1
18:~ 20:o 20:1+
18:3
20:2
Mean
Range
Mean
20.9-3 1.0 26.8 5.5-11.6 8.9 2.0-5.3 3.3 0.9-4.6 2 .o 0.7-1.6 1 .4 0.6-3.9 0.5 -
27.3 9.4 4.4 2.4 1 .4 1.9 -
z0:3 +22:0 8.8 20:4 + 2 2 : I 0.8 24:o 22.7 24:1+22:3 +22:4 20.3 22:6 0.6
Phosphatidy lserine
5.1-10.1 9.4 0.3-1.5 1.4 1 6 . ~ ~ 2 6 . 22.8 2 16.1-2 I .8 22.3 0.2-1 .o 1.2
Range
Normal Mean
22.1-28. I 4.9 4.7- 12.5 47.3 8.1 2.3-3.7 I .2-2.5 4-2 I .o-I.6 0.4 1.1 0.2-1 .o 0.6 6 . ~ 1 0 . 8 1.8 19. I 1.-1.7 1 7 . ~ 2 5 . 8 0.8 I 8.7-23.5 4.9 0.6-1.8 6.7
Fattyacid
0.1-0.5
0.5-2.3 0.2-2.0 1.2-2.9 14.8-22.8 0-2.0 3.5-5.4 3.7-9.8
Range
Mean
18:1
37.5 12.9 16.8
18:2
22.5
40.2 13.4 17.7 19.6
20:o 20: I
0.2
27.2-40.0 7.4-15.1 7.0-19.1 14.0-26.6 0-0.8 0.6-1.6 0-1.6 0-1.8 2.9-4.8 0.1-0.4 0.3-1.6 1.24.4
16:o 18:o
+ 18:3
0.5
20:2
0.5
20:3 +22:0 20:4+ I I : I 24:o 24:1+22:3+22:4 22:6
1.3 4.0 0.2 0.8 2.7
0.1
0.7 0.3 I .9 21.4 0.1
4. 1 7.2
Range 3.3-7.8 39.2-51.2 7.4-1 I . 1 2.54.1 0-0.2 0.3-1.6 0.2~,.7 1.4-2.2 I 7.8-23.2 00.4 3.7-4.4 4.7-8.0
Phosphatidylethanolamine
HS
Mean
Mean
5.0 3.0-6.3 40.9-53.7 45.9 6.5-8.8 9.4 2.3- 5.9 3.9
Phosphatidylcholine Normal
Range
HS
Range
35.F47.5 12.3-14.0 15.8-19.9 14.6-25.7 0.1 0-0.2 6.4 , 00.6 0.3 ' 00.5 0.4-1.7 1.4 3.24.7 5.0 0.1 o.o5*.15 0.5 0.4-0.6 1.3 0.9-1.7
HS
Normal Mean
Range
Mean
Range
17.9 12.4 15.9 8.7 0.6 1.4
13.5-20.4 8.1-13.7 12.8-19.8 4.8-12.5 0.1-1.5 0.4-2.5 00.3
17.8 12.3 15.5 6.4 0.2 0.7 0.4
11.6-20.3 9.5-14.0 11.8-17.7 4.8-8.2
0.2 1.0
23.6 0.3 9.7 8.3
0 . ~ 1 . 2 1.1
18.9-24.6 25.9 0.3 m.7 7.0-11.6 10.8 6.8-9.6 8.5
-1.0
0.4Q.9 0.2-0.6 0.8-1.4 22.3-27.0 0-0.6 8.0-12.9 7.4-9.1
* The first two digits state the number of carbon atomsand the third digit states the number of double bonds of the individual fatty acids. distribution of the fatty acids in all the phospholipid fractions is very similar to that found by other workers (Ways & Hanahan, 1964; Dodge & Phillips, 1967). O f particular importance was the similar percentage composition of the 24:o and 24: I fractions of sphingomyelin in the control and HS erythrocytes, as these fatty acids in sphingomyelin comprise the highest proportion of long chain fatty acids above 20 carbon atoms found in any of the phospholipid fractions. Only trace quantities of these two fatty acids were found in HS erythrocytes by
402
S. S. Zail and A . Pickering
Kuiper & Livne (1972). In our gas chromatographic separation the methyl esters of 22:3 and 22:4 have the same retention times as 24:r. However, 22:3 and 22:4 are present only in trace amounts particularly in the sphingomylin fraction (Dodge & Phillips, 1967), and do not contribute significantly to the values for 24:i found in this study. This was confirmed in the separations on the less polar silicone OV-ror column, where 22:3 and 22:4 were clearly separated from 24: I . At present we cannot offer a reasonable explanation for the difference between our findings and those of Kuiper & Livne (1972). One possibility is that hereditary spherocytosis may be a heterogenous disease and that the cases described by the latter workers may come from a single ethnic group showing the described defect. Whatever the reason, we feel that it would be of interest to determine the fatty acid composition of HS membranes in other laboratories to try and resolve this problem. ACKNOWLEDGMENTS
These studies were supported by grants from the South African Medical Research Council and the Atomic Energy Board. W e thank Dr 0. Abrahams for expert advice and assistance. REFERENCES
ALONI, B., SHINITZKY, M., MOSES,S. Ik LIVNE,A. (1975) Elevated microviscosity in membranes of erythrocytes affected by hereditary spherocytosis. BritidtJoirrnal ofHaenzarolofy, 31, I 17-123. G.B. (1967) Composition of I ~ O D GJ.T. E , 1(( PHILLIPS, phospholipids and of phospholipid fatty acids and aldehydes in human red cells. Journal O J Lipid Research, 8,667-475. KUIPER, J.C.P. & LIVNE,A. (1972) Differences in fatty acid composition between normal erythrocytes and hereditary spherocytosis affected cells. Biochimira ef Biophysica Acta, 260, 755-758.
VALENTINE, W.N. (1977) The molecular lesion ofhereditary spherocytosis (HS): a continuing enigma. Blood, 49. 241-245. VANDEENEN, L.M. & DE GIER,J. (1974) Lipids of the red cell membrane. T h e Red Blood Cell, 2nd edn (ed. by 11. MacN. Surgenor), p. 147. Academic Press, New York. WAYS,P. & HANAHAN, D.J. (1964) Characterization and quantification of red cell lipids in normal man. Journal ofLipid Research, 5 , 3 18-328. ZAIL.S.S. (1977) The erythrocyte abnormality of hereditary spherocytosis. (Annotation). BririshJournal of Haemafology, 37, 305-3 10.
