British Journal of Haernatology. 1990, 76, 282-287
Effects of deficiencies of glycophorins C and D on the physical properties of the red cell G. B. NASH,J . P A R M A RA N D M. E. REID*Departnzent of Haematology, St George's Hospital Medical School, London, and *International Blood Group Reference Laboratory. South West Regional Transfusion Centre, Bristol
Received 16 March 1990; accepted for publication 24 May 1990
Summary. Red cells of the rare Leach phenotype lack the membrane glycophorins C and D, and a proportion of the red cells are elliptocytes.Judging from tests on suspensions of red cell ghosts sheared rotationally in an ektacytometer, it has previously been suggested that these membranes are relatively fragile and poorly deformable. We have carried out analyses of individual red cells to investigate possible factors which underlie the physical changes in these glycophorindeficient cells. Micropipette analysis of the red cell membrane showed that the rigidity and viscosity were normal, both for elliptocytes and discocytes, for three donors deficient in glycophorins C and D. Red cell transit times through 5 pm pores, measured electronically for 2000 individual cells,
showed no differences from controls. It was confirmed that the index of deformation obtained using an ektacytometer was reduced, but our results suggest that this arises from shape rather than membrane changes. The elliptocytes were found to have a lower volume and surface area than discocytes from the same donor (measured by micropipette aspiration of single red cells), and were rarely found in less dense red cell fractions. No reticulocytes were found to be elliptical.These data suggest that the elliptocytes are older red cells, and are formed from red cells which are initially released into the circulation with normal shape. Their elongated shape might arise from permanent distortion of the unstable membrane by shear forces in the circulation.
Hereditary absence of the red cell membrane proteins glycophorins C and D (sialogylcoproteins and y), is associated with a variable degree of elliptocytosis (Daniels rt al, 1986). Recently, red cell membranes with this disorder were found to have reduced stability when subjected to shear stress, and also a relatively poor 'membrane deformability' (Reid rt al, 1987). This is particularly interesting because it suggests functional roles for these proteins in maintaining membrane mechanical properties which have previously largely been associated with the membrane skeleton. We have examined red cells from the same donors, using additional techniques capable of testing the properties of individual red cells. More detailed information has been obtained on the membrane and cellular mechanical properties. The data call into question the previous findings regarding reduced membrane deformability and illustrate how care should be taken in interpreting information from different rheological procedures. The results also suggest a mechanism for the transformation from discocyte to elliptocyte in the circulation of these donors.
MATERIALS AND METHODS After obtaining informed consent, blood was drawn into acid citrate dextrose and stored at 4°C for up to 10 d. Storage was required to transport the samples from their country of origin, and to allow measurements over several days. Blood was obtained at the same time from three homozygous glycophorin C- and D-deficient donors and from three controls (two obligate heterozygotes, D1.C and Dd.C., and a normal adult donor, P.C.). Prior to measurements, samples of the blood were diluted 10-fold in culture medium (RPMI 1640, Flow Laboratories, Irvine, U.K.) and incubated for 2 h at 37°C to restore cells to biconcave discoidal shape. On a separate occasion, a single sample of glycophorin A-deficient red cells (donor E.P.) was compared to a normal control which had been stored with it: in this experiment membrane viscoelasticity alone was tested. Micropipette analysis. The incubated blood was diluted 1: 100 in HEPES (20 mmol/l) buffered saline (pH 7.4) plus 2% AB serum and a portion was placed in an open-sided microscope chamber. Under microscope/video observation, a micropipette attached to a hydrostatic pressure system was used to manipulate the red cells. All measurements were performed at room temperature. Membrane viscoelasticity was tested by measuring the
Correspondence: Dr G. B. Nash, Department of Haematology, The Medical School, University of Birmingham. Birmingham B15 2TH.
282
Glycophorin Deficient Red Blood CeZls
viscosity of 2 0 mPa.s at 20°C. They were exposed to a gradually increasing shear stress, up to 16 Pa, and a deformation index was continuously derived from the ellipticity of the diffraction pattern of a laser beam passed through the suspension (Reid et al, 1987). The ektacytometer was also used in its osmoscan mode (Clark et a]. 1 9 8 3 ) , where deformation is measured at a constant stress (16 Pa) but with continuously decreasing osmolarity, ranging from 400 to 100 mOsm/kg. The deformability of individual red cells was quantitated using a cell transit analyser (Zhu et al, 1989). Pore transit times were electronically recorded for 2000 separate red cells as they flowed through a filter containing 30 pores of 5 pm diameter. Automated computer analysis yielded a frequency distribution of transit times, from which measures of central tendency and percentiles were derived. Ofliev teclrrtiqrtes. Reticulocyte counts were made using blood films stained with new methyl blue. MCV and MCHC were determined by Coulter Counter model S plus IV. The morphology of red cells was assessed by light microscopy after mixing incubated, diluted blood 1: 1 with 2% glutaraldehyde in phosphate-buffered saline. These unfractionated red cells were compared to two density fractions. The fractions were prepared as described by Pfafferott et RI(1982) by loading packed red cells, which had been centrifuged at 800 g for 5 min, into microhaematocrit tubes. The tubes were spun in a microhaematocrit centrifuge at 1 3 000 rpm for 15 min. The packed cell column was divided by cutting the tubes with a diamond scribe. The top, least dense 20% and bottom, most dense 20% were harvested, diluted and glutaraldehyde fixed. The red cells were placed in three
membrane shear elastic modulus ( p , resistance to shear deformation at constant area) and the time constant for shape recovery after elongation of the red cell ( t J . The product p . f crepresents the membrane surface shear viscosity (Hochmuth et al, 1979). The time constant was measured for red cells which had settled and attached to the bottom of the chamber. The red cells were elongated by pulling them from one side using a micropipette of approximately 1 . 5 pm diameter. and then releasing them. The rate of their shape recovery was measured from a video recording and analysed as previously described (Nash & Meiselman, 1983). The elastic modulus was determined for the same red cells after detaching them from the chamber surface. A membrane tongue was aspirated from a flattened region of the red cell into the pipette. The length of the tongue ( L ) was typically measured at four pressures ( P ) and the elastic modulus calculated from dZ/dP (Evans & La Celle, 1975). The same pipette was used for each comparison of glycophorindeficient red cells and controls. The surface area (SA) and volume ( V ) of individual cells were measured by aspirating them into 1.6-2.0 p m diameter pipettes using a pressure of 2 0 0 Pa. The red cells were thus constrained to assume a cylindrically symmetrical form with a spherical portion outside the pipette and cylindrical portion inside. By measuring the dimensions of these portions, SA and V were calculated (Nash & Wyard. 1981). Red cell deformability An ektacytometer was used to measure the average deformation of red cells in suspension under the action of a uniform fluid shear stress. Red cells were suspended in a medium made u p of polyvinylpyrollidone (3'5% w/v) dissolved in phosphate buffered saline, with a
Table I. Red cell shape and haematological indices Cell shapes (%)t
Retics Donor
Sample*
Discs
Ovals
Clycophorins C- and D-deficient J.C. Unfract. 42 38
S.S.
(%)
MCVf (fl)
MCHCf (g/dl)
20 4
1.0
87
32.0
23 41
--
41 63 35
40 31 43
19 6 22
2.3
96
32.6
46 28 48
5 2 1
1.0
96
32.7
Bottom
49 70 51
Unfract. Unfract. Unfract.
95 83 81
5
0 1 1
-
16 18
91 87 86
31.1 31.2 31.1
Bottom Unfract. Top
T.L.
Ellipses
73 37
Top
Bottom Unfract. TOP
77
Controls P.C. D1.C.
Dd.C.
28 3
-
* Unfract.= unfractionated red cells: Top= least dense 20% cells: Bottom = most dense 20% of cells. t 400-600 cells were classified according to shape. f MCV =mean cell volume: MCHC = mean cell haemoglobin concentration:both measured by Coulter S plus IV.
284
G. B. Nash, J. Parmar and M . E. Reid
morphological categories: (1)normal discocytes with circular outline. (2)oval cells with clear elongation of one axis, but distinct from, ( 3 ) ellipses with marked elongation and axial ratio approximately 2 or greater. RESULTS Blood samples were tested over a period of 7-10 d after withdrawal. This applied equally to control and glycophorindeficient cells which originated from the same laboratories. By this time, the red cells in the whole blood had become slightly echinocytic. However, incubation for 2 h at 37°C in culture medium restored the non-elliptical red cells to normal bi-concave shape. The elliptical cells (see below) had smooth outlines. Values for mean cell volume (MCV) and mean cell haemoglobin concentration (MCHC) were then essentially equal to those measured in the country of origin: MCV= 93 f 5 fl: MCHC= 325 f 5 g/I (meanf SD in the U.K. for the three glycophorin-deficientsamples): MCV = 9 5 f4 fl MCHC= 338 f11 (meanf SD in the country of origin). This indicates that no detectable change in cellar hydration had occurred. The control samples gave values for cellular mechanical parameters within the range for normal samples recently measured in the test laboratories: shear elastic modules 5.0-75 x dyn/cm; Cell Transit Analyser, mean transit time= 1.15-1 5 5 ms: ektacytometer deformation index=0.51-0.62. We have no established normal values for the other parameters tested, although shape recovery times were consistent with other recent studies (Nash et 01. 1986: Waugh & Agre, 1988). Data describing red cell shape are shown in Table I. Blood from the three homozygous glycophorins C- and D-deficient donors contained a proportion of markedly elliptical red cells. and numerous less easily defined oval red cells. The elliptical and oval red cells were reduced in concentration in the least
dense (20%) fraction of red cells, where discocytes made up the majority. Reticulocytes were counted (Table I) and were not elliptical in shape. Sufficient reticulocytes were counted so that the probability of zero being elliptical was less than 0.05 for each glycophorin-deficient sample. Red cell geometry was further investigated by micropipette measurement of the surface area and volume of individual red cells (Table 11). Unfortunately the pipette initially in use was broken after three donors and another, smaller one was substituted. Results for volume and surface area were generally higher for the second pipette. The second group also had higher mean cell volumes as judged by Coulter Counter (Table I), but the increase was not so large as that indicated by the pipette measurements (Table 11). Therefore, direct comparison between the first and second groups is not justified, but comparisons between red cells in a single sample, and between samples measured with the same pipette are valid. Comparisons of discs and eliptocytes in the samples from the glycophorin-deficient donors showed that the elliptocytes were smaller in volume (by 15% on average) and in surface area (by 9% on average). The elliptocytes thus had no deficit in surface/volume, and their sphericity index was unaltered compared to discocytes in the same sample (mean sphericity for ellipses was 0.77 versus 0.78 for discs). On average, these geometric indices were similar for the glycophorin-deficient donors and controls, when measured using the same pipette. The specificviscoelastic properties of the red cell membrane were tested by micropipette aspiration. There was no difference in the shear elastic modulus between elliptocytes and discocytes from glycophorin-deficient donors and nor were the values different from controls (Table 111). It was observed, however, that when tongue length was plotted against aspiration pressure (Fig 1).the elliptical red cells in glycophorin-deficient samples tended to have consistently shorter
Table 11. Red cell volume and surface area measured by micropipette aspiration
Discs Volume bm3)
Donor Group 1: pipette diameter=2.2 pm Glycophorins C- and D-deficient J.C. Controls D1.C. Dd.C. Group 2: pipette diameter= 1 . 6 pm Glycophorins C- and D-deficient S.S. T.L. Control P.C. Data are mean
Ellipses
(d)
Volume (pm3)
Surface area (pm?
84+11
130+10
73f10
116110
80f12 86113
125f12 1323~11
-
-
-
-
1034~11 112fl3
132+10 13818
89f12 90+12
124fll 1231.11
100113
134111
-
-
+ SD for 2 0 or more red cells.
Surface area
Glgcophorin Deficient Red Blood Cells
285
...'
,..." ._.'
I
I
I
I
I
Fig 1. Change in membrane tongue length with pressure during measurement of membrane shear elastic modulus. Data points are meanf SD of sample means for three samples of control cells ( O ) , three samples of glycophorin-deficient discocytes (B) and three samples of glycophorindeficient ellipsoidal cells (A).Lines were drawn by linear regression and show different intercepts on the length axis for zero pressure.
tongues at each pressure. The slope (dL/dP) from which the elastic modulus is calculated was the same for all types of red cell. Two possible explanations are that the membrane of the distorted elliptical cells is already under stress, or that the membrane has a higher resistance to bending. Both could cause greater resistance to the initial entry of the membrane into the pipette, but not necessarily affect the subsequent shear deformation with increasing pressure. The time constant for shape recovery after extension was slightly faster for glycophorin-deficient discocytes than controls (Table 111). Elliptocytes were not tested because their shape was already markedly distorted. The data suggest a slight reduction in membrane viscosity associated with Table 111. Viscoelasticity of membranes of glycophorin-deficient red cells and controls
Donor
Time constant for extensional recovery (s)
Shear elastic modulus ( 10- dyn/cm)
Glycophorins C- and D-deficient Discs 6 . 5 f 1.0 J.C. S.S. 7.531.1 T.L. 7.1 1 0 . 8 Pooled data
glycophorin deficiency, but are not statistically significant. Any such change should not have a deletorious effect on deformability. Red cells from a single glycophorin A-deficient donor were also tested. These had normal morphology and normal membrane viscoelasticity (data not shown). This agrees with a previous report of normal ektacytometric deformability (Reid et al, 1987). The deformability of intact, whole red cells was tested by ektacytometer and Cell Transit Analyser. As expected from previous reports (Reid et al, 1987),the ektacytometer showed reduced deformation indices for the glycophorin-deficientred cells. Table IV shows values for the maximum deformation index obtained during the osmoscan. The same comparative loss of deformability was evident for the glycophorin C- and D-deficient cells when deformation was measured at varying shear stress but constant osmolarity. The loss was least for the sample (TL) with the lowest number of elliptocytes. Cell transit analysis showed that red cells from deficient donors
__ Table IV. Deformabilityof glycophorin-deficientred cells and controls
Ellipses 6.9f0.8 7.4f2.0 6.43~1.1
7.Of-1.2
Discs only 0.082 f0.021 0.090+0.020 0.077f0.017 0.083k0.027
Cell transit analyser: transit time (ms) Donor
Ektacytometer deformation index
Mean
Mode
90th percentile
Glycophorins C- and D-deficient 0.40 1.17 J.C. 0.39 1.32 S.S. T.L. 0.44 1.33
1.05 1.25 1.25
1.42 1.64 1.60
Contro1s P.C. D1.C. Dd.C.
1.25 1.15 1.15
1.65 1.52 1.55
~~
Controls P.C. D1.C. Dd.C.
7.1 f l . 1 6.4f0.8 6.4f0.9
Pooled data
6.7f 1.0
-
-
0.123 f 0 4 3 3 0.108 f0.019 0.091 f0.017 0.108 &0.027
Data are mean fSD for eight cells for controls, and six discs and six ellipses for glycophorin-deficientcells. Pooled data are mean +SD of values for all deficient or all control cells.
0.57 0.54 0.54
~
1.41 1.29 1.34
286
G. B. Nash, 1, Parmar and
M.E. Reid
did not have slower passage through 5 pm pores (Table IV), and nor was there any evidence of the existence of a slow flowing sub-population. DISCUSSION The measurements carried out in this study show that deficiency of glycophorins C and D has little effect on the shear elastic modulus or shear viscosity of the red cell membrane. These properties were tested using micropipette methods specific to the membrane and unaffected by haemoglobin concentration, except under exceptional circumstances (Nash & Meiselman. 1983; Evans & Mohandas, 1987). In addition, we found that cellular deformability, as measured by transit time through 5 pm pores, was normal. Previously, membrane ghosts from such red cells were found to give a lower index of deformation when subjected to shear flow in an ektacytometer (Reid et al, 1987). This was interpreted as a defect in ‘membrane deformability’. Preparation of membrane ghosts itself reduces membrane viscosity but not the elastic modulus (Nash et al, 1986). Our interpretation of the discrepancy in findings is that the ektacytometer result is influenced by red cell shape rather than membrane properties in this case. In this device, normal red cells or ghosts become aligned parallel with the flow, their membrane rotates (tank-treads) and they elongate progressively into prolate ellipsoids as the shear stress increases. The deformation index is determined from the ellipticity of a laser beam diffracted as it passes through the red cell suspension. and is averaged over the whole red cell population. Elliptocytic red cells may align perpendicular to the flow and rotate round their long axis, rather than align parallel to the flow. Their deformation index would be negative. As the shear stress increases they will gradually deform parallel to the flow, but less than discocytes (Bull & Bessis, 1983).A mixture of ellipses, oblongs and discs and may give an average signal dependent on their relative proportions and alignment. One caveat is that the samples used in the present study were stored in the cold for 7-10 d before measurement. As outlined above, based on observations of cell shape, volume, membrane rigidity and cell deformability, this had no evident effect on the properties of interest. The stored glycophorindeficient samples had unaltered cellular hydration, and the control samples gave values for rheological parameters within normal laboratory ranges, where these were available. The remaining assumption is that glycophorin-deficient red cells or elliptocytes underwent no preferential deterioration during storage, compared to the controls or nonelliptical glycophorin-deficient red cells stored with them. According to the previous study (Reid et al, 1987). the glycophorin-deficient membrane is also mechanically unstable. We have no reason to dispute this. Since all red cells in homozygotes are deficient in these glycophorins, one wonders why some are discocytic and others elliptocytic.Our data show that the ellipsoidal red cells are denser and smaller and have less surface area than average discocytes from the same donor. These trends in geometric properties are the same as those seen for normal density fractionated red cells (Nash et al, 1988). Also, reticulocytes were not found to be
elliptical. These results suggests that the elliptocytes are older red cells, or at least not very young cells, and are formed from the discocytes. The inherent instability of the membrane could thus be the cause of progressive distortion of shape as the red cells are exposed to shear stress and elongation in the circulation. Our results suggest that the ability of the glycophorin-deficient red cells to circulate is not diminished, in the sense that their ability to pass through small vessesls should be normal. However, their instability could lead to an increased rate of destruction. This is not, however, great enough to cause overt haemolytic anaemia. It has previously been pointed out that the structural elements of the red cell membrane skeleton which regulate stability and deformability are probably separate (Chasis & Mohandas, 1986). Glycophorins C and D span the lipid bilayer and interact with band 4.1, which in turn binds to spectrin and actin in the skeleton (Bennett, 1985). The quantity of spectrin has been suggested to be the major determinant of membrane viscoelasticity (Waugh & Agre, 1988), while abnormal interaction between band 4.1 and spectrin does not alter this property, but does cause membrane instability (Chasis & Mohandas. 1986). Thus, lack of glycophorins C and D might be expected to affect protein interactions regulating stability rather than viscoelasticity, and our results support this case. Another interesting point is that deficiency in glycophorin C inhibits invasion by Plasrnodium fakiparum (Pasvol et al, 1984). We conclude that this cannot be attributed to impaired membrane deformability (Reid et al, 1987), but may reflect some other change in protein organization which affects the ability of the merozoite to pass through the membrane (Rangachari et al, 1989). ACKNOWLEDGMENTS We gratefully acknowledge help given by P. Colpitts in obtaining the blood samples used in this study, and by Philip Stone in carrying out measurements with the ektacytometer, and cell transit analyser. This investigation received the financial support of the UNDP/World Bank/WHO Special Programme for Research Training in Tropical Diseases. REFERENCES Bennett, V. (1985) The membrane skeleton of human erythrocytes and its implications for more complex cells. Annual Review of Biochemistrg. 54, 273-304. Bull, B., Feo, C. & Bessis, M. (1983)Behaviour of elliptocytes under shear stress in the rheoscope and ektacytometer. Cgtornetrg, 3 , 300-304. Chasis.J.A.& Mohandas. N. (1986)Erythrocyte membrane deformability and stability: two distinct membrane properties that are independently regulated by skeletal protein associations. Journal of CeII Biology. 103, 343-350. Clark, M.R., Mohandas. N. & Shohet, S.B. (1983) Osmotic gradient ektacytometry: comprehensive characterization of red cell volume and surface maintenance. Blood. 61, 899-910. Daniels, G.L.. Shaw, M.-A., Judson, P.A.. Reid, M.E.. Anstee, D.J.. Colpitts. P., Cornwall, S . . Moore, B.P.L. & Lee. S. (1986) A family demonstrating inheritance of the Leach phenotype: a Gerbichnegative phenotype associated with elliptocytosis. Vox Sanguinis. 50, 117-121.
Glycophorin Deficient Red Blood Cells Evans, E.A. & LaCelle. P.L. ( 1 975) Intrinsic material properties of the erythrocytes membrane indicated by mechanical analysis of deformation. Blood. 45, 29-43. Evans, E.A. & Mohandas. N. (1 987) Membrane-associated sickle hemoglobin: a major determinant of sickle erythrocyte rigidity. Blood, 70, 1443-1449. Hochmuth. R.M., Worthy, P.R. & Evans, E.A. (1979) Red cell extensional recovery and the determination of membrane viscosity. Biophysical journal, 26, 101-1 14. Nash, C.B., Linderkamp. 0.. Pfafferott, C. & Meiselman, H.J. (1988) Changes in red cell mechanics during in vivo ageing: possible influence on removal of senescent cells. Blood Cells, Rlieology and Aging (ed. by D. Platt). p. 99. Springer, Heidelberg. Nash. G.B. & Meiselman. H.J. (1983) Red cell and ghost viscoelasticity: effects of hemoglobin concentration and in vivo aging. Biophysical journal, 43, 63-73. Nash, C.B., Tran-Son-Tay. R. & Meiselman. H.J. (1986) Influence of preparative procedures on the membrane viscoelasticityof red cell ghosts. Biochimica et Biophysica Acta, 8 5 5 , 105-1 14. Nash. G.B. & Wyard. S.J.(1981) Changes in surface area andvolume
28 7
measured by micropipette aspiration for erythrocytes ageing in vivo. Biorheology. 17, 479-484. Pasvol. G.. Anstee, D. &Tanner.M.J.A. (1984)Glycophorin C and the invasion of red cells by Plasmodiumfalciparum. Lancet. i, 907-908. Pfafferott. C., Wenby, R.B. & Meiselman, H.J. (1982) Morphologic and internal viscosity aspects of RBC rheologic behaviour. Blood Cells. 8, 65-78. Rangachari, K., Beavan. G.W., Nash. G.B., Clough, B.. Dluzewski, A.R.. Myint-Oo, U.. Wilson, R.J.M. & Grazer, W.B. (1989)Astudy of red cell membrane properties in relation to malarial invasion. Molecular arid Biochemical Parasitology, 34, 6 3-74. Reid. M.E., Chasis. J.A. & Mohandas. N. (1987) Identification of a functional role for human erythrocyte sialoglycoprotein fl and y. Blood, 69, 1068-1072. Waugh, R.E. & Agre, P. (1988)Reductions of erythrocyte membrane viscoelastic coefficients reflect spectrin deficiencies in hereditary spherocytosis. lournal of Clinical Investigation. 81, 133-141. Zhu, J-C., Stone. P.C.W. & Stuart, J. (1989) Measurement of erythrocyte deformability by cell transit analyser. Clinical Hemorheology, 9, 897-908.
Effects of deficiencies of glycophorins C and D on the physical properties of the red cell G. B. NASH,J . P A R M A RA N D M. E. REID*Departnzent of Haematology, St George's Hospital Medical School, London, and *International Blood Group Reference Laboratory. South West Regional Transfusion Centre, Bristol
Received 16 March 1990; accepted for publication 24 May 1990
Summary. Red cells of the rare Leach phenotype lack the membrane glycophorins C and D, and a proportion of the red cells are elliptocytes.Judging from tests on suspensions of red cell ghosts sheared rotationally in an ektacytometer, it has previously been suggested that these membranes are relatively fragile and poorly deformable. We have carried out analyses of individual red cells to investigate possible factors which underlie the physical changes in these glycophorindeficient cells. Micropipette analysis of the red cell membrane showed that the rigidity and viscosity were normal, both for elliptocytes and discocytes, for three donors deficient in glycophorins C and D. Red cell transit times through 5 pm pores, measured electronically for 2000 individual cells,
showed no differences from controls. It was confirmed that the index of deformation obtained using an ektacytometer was reduced, but our results suggest that this arises from shape rather than membrane changes. The elliptocytes were found to have a lower volume and surface area than discocytes from the same donor (measured by micropipette aspiration of single red cells), and were rarely found in less dense red cell fractions. No reticulocytes were found to be elliptical.These data suggest that the elliptocytes are older red cells, and are formed from red cells which are initially released into the circulation with normal shape. Their elongated shape might arise from permanent distortion of the unstable membrane by shear forces in the circulation.
Hereditary absence of the red cell membrane proteins glycophorins C and D (sialogylcoproteins and y), is associated with a variable degree of elliptocytosis (Daniels rt al, 1986). Recently, red cell membranes with this disorder were found to have reduced stability when subjected to shear stress, and also a relatively poor 'membrane deformability' (Reid rt al, 1987). This is particularly interesting because it suggests functional roles for these proteins in maintaining membrane mechanical properties which have previously largely been associated with the membrane skeleton. We have examined red cells from the same donors, using additional techniques capable of testing the properties of individual red cells. More detailed information has been obtained on the membrane and cellular mechanical properties. The data call into question the previous findings regarding reduced membrane deformability and illustrate how care should be taken in interpreting information from different rheological procedures. The results also suggest a mechanism for the transformation from discocyte to elliptocyte in the circulation of these donors.
MATERIALS AND METHODS After obtaining informed consent, blood was drawn into acid citrate dextrose and stored at 4°C for up to 10 d. Storage was required to transport the samples from their country of origin, and to allow measurements over several days. Blood was obtained at the same time from three homozygous glycophorin C- and D-deficient donors and from three controls (two obligate heterozygotes, D1.C and Dd.C., and a normal adult donor, P.C.). Prior to measurements, samples of the blood were diluted 10-fold in culture medium (RPMI 1640, Flow Laboratories, Irvine, U.K.) and incubated for 2 h at 37°C to restore cells to biconcave discoidal shape. On a separate occasion, a single sample of glycophorin A-deficient red cells (donor E.P.) was compared to a normal control which had been stored with it: in this experiment membrane viscoelasticity alone was tested. Micropipette analysis. The incubated blood was diluted 1: 100 in HEPES (20 mmol/l) buffered saline (pH 7.4) plus 2% AB serum and a portion was placed in an open-sided microscope chamber. Under microscope/video observation, a micropipette attached to a hydrostatic pressure system was used to manipulate the red cells. All measurements were performed at room temperature. Membrane viscoelasticity was tested by measuring the
Correspondence: Dr G. B. Nash, Department of Haematology, The Medical School, University of Birmingham. Birmingham B15 2TH.
282
Glycophorin Deficient Red Blood CeZls
viscosity of 2 0 mPa.s at 20°C. They were exposed to a gradually increasing shear stress, up to 16 Pa, and a deformation index was continuously derived from the ellipticity of the diffraction pattern of a laser beam passed through the suspension (Reid et al, 1987). The ektacytometer was also used in its osmoscan mode (Clark et a]. 1 9 8 3 ) , where deformation is measured at a constant stress (16 Pa) but with continuously decreasing osmolarity, ranging from 400 to 100 mOsm/kg. The deformability of individual red cells was quantitated using a cell transit analyser (Zhu et al, 1989). Pore transit times were electronically recorded for 2000 separate red cells as they flowed through a filter containing 30 pores of 5 pm diameter. Automated computer analysis yielded a frequency distribution of transit times, from which measures of central tendency and percentiles were derived. Ofliev teclrrtiqrtes. Reticulocyte counts were made using blood films stained with new methyl blue. MCV and MCHC were determined by Coulter Counter model S plus IV. The morphology of red cells was assessed by light microscopy after mixing incubated, diluted blood 1: 1 with 2% glutaraldehyde in phosphate-buffered saline. These unfractionated red cells were compared to two density fractions. The fractions were prepared as described by Pfafferott et RI(1982) by loading packed red cells, which had been centrifuged at 800 g for 5 min, into microhaematocrit tubes. The tubes were spun in a microhaematocrit centrifuge at 1 3 000 rpm for 15 min. The packed cell column was divided by cutting the tubes with a diamond scribe. The top, least dense 20% and bottom, most dense 20% were harvested, diluted and glutaraldehyde fixed. The red cells were placed in three
membrane shear elastic modulus ( p , resistance to shear deformation at constant area) and the time constant for shape recovery after elongation of the red cell ( t J . The product p . f crepresents the membrane surface shear viscosity (Hochmuth et al, 1979). The time constant was measured for red cells which had settled and attached to the bottom of the chamber. The red cells were elongated by pulling them from one side using a micropipette of approximately 1 . 5 pm diameter. and then releasing them. The rate of their shape recovery was measured from a video recording and analysed as previously described (Nash & Meiselman, 1983). The elastic modulus was determined for the same red cells after detaching them from the chamber surface. A membrane tongue was aspirated from a flattened region of the red cell into the pipette. The length of the tongue ( L ) was typically measured at four pressures ( P ) and the elastic modulus calculated from dZ/dP (Evans & La Celle, 1975). The same pipette was used for each comparison of glycophorindeficient red cells and controls. The surface area (SA) and volume ( V ) of individual cells were measured by aspirating them into 1.6-2.0 p m diameter pipettes using a pressure of 2 0 0 Pa. The red cells were thus constrained to assume a cylindrically symmetrical form with a spherical portion outside the pipette and cylindrical portion inside. By measuring the dimensions of these portions, SA and V were calculated (Nash & Wyard. 1981). Red cell deformability An ektacytometer was used to measure the average deformation of red cells in suspension under the action of a uniform fluid shear stress. Red cells were suspended in a medium made u p of polyvinylpyrollidone (3'5% w/v) dissolved in phosphate buffered saline, with a
Table I. Red cell shape and haematological indices Cell shapes (%)t
Retics Donor
Sample*
Discs
Ovals
Clycophorins C- and D-deficient J.C. Unfract. 42 38
S.S.
(%)
MCVf (fl)
MCHCf (g/dl)
20 4
1.0
87
32.0
23 41
--
41 63 35
40 31 43
19 6 22
2.3
96
32.6
46 28 48
5 2 1
1.0
96
32.7
Bottom
49 70 51
Unfract. Unfract. Unfract.
95 83 81
5
0 1 1
-
16 18
91 87 86
31.1 31.2 31.1
Bottom Unfract. Top
T.L.
Ellipses
73 37
Top
Bottom Unfract. TOP
77
Controls P.C. D1.C.
Dd.C.
28 3
-
* Unfract.= unfractionated red cells: Top= least dense 20% cells: Bottom = most dense 20% of cells. t 400-600 cells were classified according to shape. f MCV =mean cell volume: MCHC = mean cell haemoglobin concentration:both measured by Coulter S plus IV.
284
G. B. Nash, J. Parmar and M . E. Reid
morphological categories: (1)normal discocytes with circular outline. (2)oval cells with clear elongation of one axis, but distinct from, ( 3 ) ellipses with marked elongation and axial ratio approximately 2 or greater. RESULTS Blood samples were tested over a period of 7-10 d after withdrawal. This applied equally to control and glycophorindeficient cells which originated from the same laboratories. By this time, the red cells in the whole blood had become slightly echinocytic. However, incubation for 2 h at 37°C in culture medium restored the non-elliptical red cells to normal bi-concave shape. The elliptical cells (see below) had smooth outlines. Values for mean cell volume (MCV) and mean cell haemoglobin concentration (MCHC) were then essentially equal to those measured in the country of origin: MCV= 93 f 5 fl: MCHC= 325 f 5 g/I (meanf SD in the U.K. for the three glycophorin-deficientsamples): MCV = 9 5 f4 fl MCHC= 338 f11 (meanf SD in the country of origin). This indicates that no detectable change in cellar hydration had occurred. The control samples gave values for cellular mechanical parameters within the range for normal samples recently measured in the test laboratories: shear elastic modules 5.0-75 x dyn/cm; Cell Transit Analyser, mean transit time= 1.15-1 5 5 ms: ektacytometer deformation index=0.51-0.62. We have no established normal values for the other parameters tested, although shape recovery times were consistent with other recent studies (Nash et 01. 1986: Waugh & Agre, 1988). Data describing red cell shape are shown in Table I. Blood from the three homozygous glycophorins C- and D-deficient donors contained a proportion of markedly elliptical red cells. and numerous less easily defined oval red cells. The elliptical and oval red cells were reduced in concentration in the least
dense (20%) fraction of red cells, where discocytes made up the majority. Reticulocytes were counted (Table I) and were not elliptical in shape. Sufficient reticulocytes were counted so that the probability of zero being elliptical was less than 0.05 for each glycophorin-deficient sample. Red cell geometry was further investigated by micropipette measurement of the surface area and volume of individual red cells (Table 11). Unfortunately the pipette initially in use was broken after three donors and another, smaller one was substituted. Results for volume and surface area were generally higher for the second pipette. The second group also had higher mean cell volumes as judged by Coulter Counter (Table I), but the increase was not so large as that indicated by the pipette measurements (Table 11). Therefore, direct comparison between the first and second groups is not justified, but comparisons between red cells in a single sample, and between samples measured with the same pipette are valid. Comparisons of discs and eliptocytes in the samples from the glycophorin-deficient donors showed that the elliptocytes were smaller in volume (by 15% on average) and in surface area (by 9% on average). The elliptocytes thus had no deficit in surface/volume, and their sphericity index was unaltered compared to discocytes in the same sample (mean sphericity for ellipses was 0.77 versus 0.78 for discs). On average, these geometric indices were similar for the glycophorin-deficient donors and controls, when measured using the same pipette. The specificviscoelastic properties of the red cell membrane were tested by micropipette aspiration. There was no difference in the shear elastic modulus between elliptocytes and discocytes from glycophorin-deficient donors and nor were the values different from controls (Table 111). It was observed, however, that when tongue length was plotted against aspiration pressure (Fig 1).the elliptical red cells in glycophorin-deficient samples tended to have consistently shorter
Table 11. Red cell volume and surface area measured by micropipette aspiration
Discs Volume bm3)
Donor Group 1: pipette diameter=2.2 pm Glycophorins C- and D-deficient J.C. Controls D1.C. Dd.C. Group 2: pipette diameter= 1 . 6 pm Glycophorins C- and D-deficient S.S. T.L. Control P.C. Data are mean
Ellipses
(d)
Volume (pm3)
Surface area (pm?
84+11
130+10
73f10
116110
80f12 86113
125f12 1323~11
-
-
-
-
1034~11 112fl3
132+10 13818
89f12 90+12
124fll 1231.11
100113
134111
-
-
+ SD for 2 0 or more red cells.
Surface area
Glgcophorin Deficient Red Blood Cells
285
...'
,..." ._.'
I
I
I
I
I
Fig 1. Change in membrane tongue length with pressure during measurement of membrane shear elastic modulus. Data points are meanf SD of sample means for three samples of control cells ( O ) , three samples of glycophorin-deficient discocytes (B) and three samples of glycophorindeficient ellipsoidal cells (A).Lines were drawn by linear regression and show different intercepts on the length axis for zero pressure.
tongues at each pressure. The slope (dL/dP) from which the elastic modulus is calculated was the same for all types of red cell. Two possible explanations are that the membrane of the distorted elliptical cells is already under stress, or that the membrane has a higher resistance to bending. Both could cause greater resistance to the initial entry of the membrane into the pipette, but not necessarily affect the subsequent shear deformation with increasing pressure. The time constant for shape recovery after extension was slightly faster for glycophorin-deficient discocytes than controls (Table 111). Elliptocytes were not tested because their shape was already markedly distorted. The data suggest a slight reduction in membrane viscosity associated with Table 111. Viscoelasticity of membranes of glycophorin-deficient red cells and controls
Donor
Time constant for extensional recovery (s)
Shear elastic modulus ( 10- dyn/cm)
Glycophorins C- and D-deficient Discs 6 . 5 f 1.0 J.C. S.S. 7.531.1 T.L. 7.1 1 0 . 8 Pooled data
glycophorin deficiency, but are not statistically significant. Any such change should not have a deletorious effect on deformability. Red cells from a single glycophorin A-deficient donor were also tested. These had normal morphology and normal membrane viscoelasticity (data not shown). This agrees with a previous report of normal ektacytometric deformability (Reid et al, 1987). The deformability of intact, whole red cells was tested by ektacytometer and Cell Transit Analyser. As expected from previous reports (Reid et al, 1987),the ektacytometer showed reduced deformation indices for the glycophorin-deficientred cells. Table IV shows values for the maximum deformation index obtained during the osmoscan. The same comparative loss of deformability was evident for the glycophorin C- and D-deficient cells when deformation was measured at varying shear stress but constant osmolarity. The loss was least for the sample (TL) with the lowest number of elliptocytes. Cell transit analysis showed that red cells from deficient donors
__ Table IV. Deformabilityof glycophorin-deficientred cells and controls
Ellipses 6.9f0.8 7.4f2.0 6.43~1.1
7.Of-1.2
Discs only 0.082 f0.021 0.090+0.020 0.077f0.017 0.083k0.027
Cell transit analyser: transit time (ms) Donor
Ektacytometer deformation index
Mean
Mode
90th percentile
Glycophorins C- and D-deficient 0.40 1.17 J.C. 0.39 1.32 S.S. T.L. 0.44 1.33
1.05 1.25 1.25
1.42 1.64 1.60
Contro1s P.C. D1.C. Dd.C.
1.25 1.15 1.15
1.65 1.52 1.55
~~
Controls P.C. D1.C. Dd.C.
7.1 f l . 1 6.4f0.8 6.4f0.9
Pooled data
6.7f 1.0
-
-
0.123 f 0 4 3 3 0.108 f0.019 0.091 f0.017 0.108 &0.027
Data are mean fSD for eight cells for controls, and six discs and six ellipses for glycophorin-deficientcells. Pooled data are mean +SD of values for all deficient or all control cells.
0.57 0.54 0.54
~
1.41 1.29 1.34
286
G. B. Nash, 1, Parmar and
M.E. Reid
did not have slower passage through 5 pm pores (Table IV), and nor was there any evidence of the existence of a slow flowing sub-population. DISCUSSION The measurements carried out in this study show that deficiency of glycophorins C and D has little effect on the shear elastic modulus or shear viscosity of the red cell membrane. These properties were tested using micropipette methods specific to the membrane and unaffected by haemoglobin concentration, except under exceptional circumstances (Nash & Meiselman. 1983; Evans & Mohandas, 1987). In addition, we found that cellular deformability, as measured by transit time through 5 pm pores, was normal. Previously, membrane ghosts from such red cells were found to give a lower index of deformation when subjected to shear flow in an ektacytometer (Reid et al, 1987). This was interpreted as a defect in ‘membrane deformability’. Preparation of membrane ghosts itself reduces membrane viscosity but not the elastic modulus (Nash et al, 1986). Our interpretation of the discrepancy in findings is that the ektacytometer result is influenced by red cell shape rather than membrane properties in this case. In this device, normal red cells or ghosts become aligned parallel with the flow, their membrane rotates (tank-treads) and they elongate progressively into prolate ellipsoids as the shear stress increases. The deformation index is determined from the ellipticity of a laser beam diffracted as it passes through the red cell suspension. and is averaged over the whole red cell population. Elliptocytic red cells may align perpendicular to the flow and rotate round their long axis, rather than align parallel to the flow. Their deformation index would be negative. As the shear stress increases they will gradually deform parallel to the flow, but less than discocytes (Bull & Bessis, 1983).A mixture of ellipses, oblongs and discs and may give an average signal dependent on their relative proportions and alignment. One caveat is that the samples used in the present study were stored in the cold for 7-10 d before measurement. As outlined above, based on observations of cell shape, volume, membrane rigidity and cell deformability, this had no evident effect on the properties of interest. The stored glycophorindeficient samples had unaltered cellular hydration, and the control samples gave values for rheological parameters within normal laboratory ranges, where these were available. The remaining assumption is that glycophorin-deficient red cells or elliptocytes underwent no preferential deterioration during storage, compared to the controls or nonelliptical glycophorin-deficient red cells stored with them. According to the previous study (Reid et al, 1987). the glycophorin-deficient membrane is also mechanically unstable. We have no reason to dispute this. Since all red cells in homozygotes are deficient in these glycophorins, one wonders why some are discocytic and others elliptocytic.Our data show that the ellipsoidal red cells are denser and smaller and have less surface area than average discocytes from the same donor. These trends in geometric properties are the same as those seen for normal density fractionated red cells (Nash et al, 1988). Also, reticulocytes were not found to be
elliptical. These results suggests that the elliptocytes are older red cells, or at least not very young cells, and are formed from the discocytes. The inherent instability of the membrane could thus be the cause of progressive distortion of shape as the red cells are exposed to shear stress and elongation in the circulation. Our results suggest that the ability of the glycophorin-deficient red cells to circulate is not diminished, in the sense that their ability to pass through small vessesls should be normal. However, their instability could lead to an increased rate of destruction. This is not, however, great enough to cause overt haemolytic anaemia. It has previously been pointed out that the structural elements of the red cell membrane skeleton which regulate stability and deformability are probably separate (Chasis & Mohandas, 1986). Glycophorins C and D span the lipid bilayer and interact with band 4.1, which in turn binds to spectrin and actin in the skeleton (Bennett, 1985). The quantity of spectrin has been suggested to be the major determinant of membrane viscoelasticity (Waugh & Agre, 1988), while abnormal interaction between band 4.1 and spectrin does not alter this property, but does cause membrane instability (Chasis & Mohandas. 1986). Thus, lack of glycophorins C and D might be expected to affect protein interactions regulating stability rather than viscoelasticity, and our results support this case. Another interesting point is that deficiency in glycophorin C inhibits invasion by Plasrnodium fakiparum (Pasvol et al, 1984). We conclude that this cannot be attributed to impaired membrane deformability (Reid et al, 1987), but may reflect some other change in protein organization which affects the ability of the merozoite to pass through the membrane (Rangachari et al, 1989). ACKNOWLEDGMENTS We gratefully acknowledge help given by P. Colpitts in obtaining the blood samples used in this study, and by Philip Stone in carrying out measurements with the ektacytometer, and cell transit analyser. This investigation received the financial support of the UNDP/World Bank/WHO Special Programme for Research Training in Tropical Diseases. REFERENCES Bennett, V. (1985) The membrane skeleton of human erythrocytes and its implications for more complex cells. Annual Review of Biochemistrg. 54, 273-304. Bull, B., Feo, C. & Bessis, M. (1983)Behaviour of elliptocytes under shear stress in the rheoscope and ektacytometer. Cgtornetrg, 3 , 300-304. Chasis.J.A.& Mohandas. N. (1986)Erythrocyte membrane deformability and stability: two distinct membrane properties that are independently regulated by skeletal protein associations. Journal of CeII Biology. 103, 343-350. Clark, M.R., Mohandas. N. & Shohet, S.B. (1983) Osmotic gradient ektacytometry: comprehensive characterization of red cell volume and surface maintenance. Blood. 61, 899-910. Daniels, G.L.. Shaw, M.-A., Judson, P.A.. Reid, M.E.. Anstee, D.J.. Colpitts. P., Cornwall, S . . Moore, B.P.L. & Lee. S. (1986) A family demonstrating inheritance of the Leach phenotype: a Gerbichnegative phenotype associated with elliptocytosis. Vox Sanguinis. 50, 117-121.
Glycophorin Deficient Red Blood Cells Evans, E.A. & LaCelle. P.L. ( 1 975) Intrinsic material properties of the erythrocytes membrane indicated by mechanical analysis of deformation. Blood. 45, 29-43. Evans, E.A. & Mohandas. N. (1 987) Membrane-associated sickle hemoglobin: a major determinant of sickle erythrocyte rigidity. Blood, 70, 1443-1449. Hochmuth. R.M., Worthy, P.R. & Evans, E.A. (1979) Red cell extensional recovery and the determination of membrane viscosity. Biophysical journal, 26, 101-1 14. Nash, C.B., Linderkamp. 0.. Pfafferott, C. & Meiselman, H.J. (1988) Changes in red cell mechanics during in vivo ageing: possible influence on removal of senescent cells. Blood Cells, Rlieology and Aging (ed. by D. Platt). p. 99. Springer, Heidelberg. Nash. G.B. & Meiselman. H.J. (1983) Red cell and ghost viscoelasticity: effects of hemoglobin concentration and in vivo aging. Biophysical journal, 43, 63-73. Nash, C.B., Tran-Son-Tay. R. & Meiselman. H.J. (1986) Influence of preparative procedures on the membrane viscoelasticityof red cell ghosts. Biochimica et Biophysica Acta, 8 5 5 , 105-1 14. Nash. G.B. & Wyard. S.J.(1981) Changes in surface area andvolume
28 7
measured by micropipette aspiration for erythrocytes ageing in vivo. Biorheology. 17, 479-484. Pasvol. G.. Anstee, D. &Tanner.M.J.A. (1984)Glycophorin C and the invasion of red cells by Plasmodiumfalciparum. Lancet. i, 907-908. Pfafferott. C., Wenby, R.B. & Meiselman, H.J. (1982) Morphologic and internal viscosity aspects of RBC rheologic behaviour. Blood Cells. 8, 65-78. Rangachari, K., Beavan. G.W., Nash. G.B., Clough, B.. Dluzewski, A.R.. Myint-Oo, U.. Wilson, R.J.M. & Grazer, W.B. (1989)Astudy of red cell membrane properties in relation to malarial invasion. Molecular arid Biochemical Parasitology, 34, 6 3-74. Reid. M.E., Chasis. J.A. & Mohandas. N. (1987) Identification of a functional role for human erythrocyte sialoglycoprotein fl and y. Blood, 69, 1068-1072. Waugh, R.E. & Agre, P. (1988)Reductions of erythrocyte membrane viscoelastic coefficients reflect spectrin deficiencies in hereditary spherocytosis. lournal of Clinical Investigation. 81, 133-141. Zhu, J-C., Stone. P.C.W. & Stuart, J. (1989) Measurement of erythrocyte deformability by cell transit analyser. Clinical Hemorheology, 9, 897-908.
