- Email: [email protected]
Rosette formation between immobilised human lymphocytes and erythrocytes sensitised with monoclonal anti-D
Immunology Letters, 23 (1989/1990) 109-112
Elsevier IMLET 01317
Rosette formation between immobilised human lymphocytes and erythrocytes sensitised with monoclonal anti-D K. A. Leader, B. M. K u m p e l a n d B. A. Bradley United Kingdom Transplant Service, Southmead, Bristol, U.K.
(Received 14 August 1989; revision received 7 September 1989; accepted 8 September 1989)
1. Summary A rapid, reproducible and sensitive assay was developed to investigate the ability of human lymphocytes to form rosettes with erythrocytes sensitised with human monoclonal anti-D. Erythrocytes sensitised with a known number of anti-D molecules per cell were incubated with lymphocytes immobilised on plastic by poly(L-lysine), the resulting rosettes fixed, unbound erythrocytes removed by washing and the cell preparation stained. IgG1 and IgG3 antiD-coated erythrocytes gave similar rosette formation at sensitisation levels in the range of 5000-15000 molecules per cell, although at lower sensitisation levels IgG3 gave greater rosette formation than IgG1. A minimum of 500 IgG3 and 1000 IgG1 anti-D molecules per erythrocyte were required for rosetting.
2. Introduction The interaction of Fc receptor (FcR)-bearing human lymphocytes with cell-bound immunoglobulin (Ig) may be assessed by their ability to form rosettes with Ig-sensitised erythrocytes [1-7] or by measuring the extent of red cell lysis which may occur subsequently [6-10]. Rosette assays reflect the ability of cell-bound antibodies to bind to lymphocyte FcR, while antibody-dependent cell-mediated cytotoxici-
ty (ADCC) assays determine the ability of these antibodies to activate lytic responses. These cellular interactions may be of clinical significance. Lymphocyte-mediated ADCC has been shown to correlate with the clinical severity of haemolytic disease of the newborn (HDN) [11], and FcR-bearing lymphocytes may participate in antibody-mediated immune suppression (AMIS), which prevents maternal allo-immunisation to D-positive foetal erythrocytes following prophylactic administration of antiD [12]. It has been shown previously that some polyclonal and monoclonal IgGl anti-D may promote the lysis of D-positive erythrocytes by human lymphocytes [10, 11]. However, these findings are inconsistent with the failure of several workers to demonstrate rosette formation between human lymphocytes and erythrocytes sensitised with anti-D monoclonal antibodies (Mabs) of the IgG1 subclass [5, 6]. These assays were performed using lymphocytes and sensitised erythrocytes in fluid phase. We report here a lymphocyte rosette assay with erythrocytes sensitised with known levels of IgG1 and IgG3 monoclonal anti-D and lymphocytes attached to a solid phase, permitting a greater red cell-tolymphocyte ratio, resulting in increased sensitivity, and allowing preservation of the results to facilitate reading.
3. Materials and Methods Key words." Lymphocyte;Rosetting; Poly-(L-lysine);Monoclonal
anti-D Correspondence to: K. A. Leader, United Kingdom Transplant Service, Southmead Road, Bristol, BS105ND, U.K.
3.1. Preparation o f lymphocytes
Peripheral blood mononuclear cells (PBMC) were prepared by separation of whole blood over Lym-
0165-2478 / 89 / $ 3.50 © 1989 ElsevierScience Publishers B.V.(BiomedicalDivision)
109
phoprep (Nyegaard, Sweden) [13]. The PBMC were washed three times by repeated centrifugation (400 xg) and resuspension in RPMI-1640 (Imperial Laboratories). The cells were resuspended at 5 x 106 cells per ml in RPMI supplemented with 10°70 foetal calf serum (FCS) (RPMI/FCS). Monocytes were removed by adherence to plastic in tissue culture flasks (Nunclon) for 1 h at 37°C in RPMI/FCS. N o n adherent cells were removed, washed and resuspended at 1 x 105 cells per ml in PBS. Monocyte contamination was assessed by immunofluorescence using a CD14 antibody (Serotec) and was found to be less than 5°70.
3.2. Immobilisation of lymphocytes Ninety-six-well ELISA plates (Nunclon) were coated with 50/A per well of poly-(L-lysine) (0.1 mg/ml in PBS) for 1 h at room temperature. The plates were then washed three times in PBS and 100 #1 (1 x 104 cells) of the lymphocyte suspension added to each well. The plates were centrifuged at 8 5 0 x g for 10 min, the PBS was removed and 100 tzl of R P M I / F C S added to each well for 1 h to block the poly-(L-lysine) to which no cells had bound [14].
5070 (approx 4x105 per ml) in RPMI. 3.5. Lymphocyte rosette assay The R P M I / F C S was removed from the prepared lymphocyte plates and 100/A (4x107 cells) of the sensitised erythrocyte suspension added to each well. The plates were centrifuged at 70 g for 3 min and incubated at 37 °C for 30 min. Without removing the medium the rosettes were fixed by the addition of 200 t~l of 1°70glutaraldehyde in PBS for 10 min. Supernatants were aspirated and unrosetted erythrocytes were gently resuspended in PBS and removed. The lymphocytes were stained with 0.1 °70trypan blue in PBS and the percentage of lymphocytes with 3 or more adherent erythrocytes was assessed. A minimum of 200 lymphocytes were counted per well. 4. Results
All the Mabs brought about rosette formation (Fig. 1), the extent of rosetting depending upon the erythrocyte sensitisation level. Each Mab was tested using lymphocytes from six different donors. Con-
Culture supernatants containing Mabs were diluted in R P M I / F C S and incubated with an equal volume of a 5°70 suspension of group O R1R2 erythrocytes for 2 h at 37 °C, followed by three washes in RPMI. Cell-bound IgG was quantified using a radiometric indirect antiglobulin method [17]. Sensitisation levels of 15 000, 10000, 5000, 1000 and 500 antibody molecules per erythrocyte were obtained by sensitising cells in supernatants appropriately diluted. After sensitisation, the cells were resuspended at 110
R1D7
25-
H27
25 2O
20-
The human Mabs with specificity for the Rh antigen D were derived from EBV-transformed Blymphoblastoid cell lines [10, 15, 16]. Four Mabs were used in this study, all of different isotype or Gm allotype, as follows: R1D7, IgG1 Glm(3); H27, IgG1 Glm(1,2,17); AB5, IgG1 Glm(1,17); 6D10, IgG3 G3m(5). 3.4. Sensitisation and quantitation of cell-bound anti-D
30~
30
3.3. Monoclonal antibodies
z5-~ 10-
15 I' :
10-~
i
i
i o
f5
I
~: : :::! fo
5
5 i
0:5
30
o
~, 1'5
1'0
5
i
015
30 AB5
E
25-
~
20-
20-
15-
15.
1 0 - - -
10--
6D10
25
5-
5 i . . . .
0
I
1'5
fo
I
•
5
t
0'5
0
i/::!iiiiiiiiiiiiiiiii!!i!L 1'5
fo
5
i
.....
0:5
Mab Molecules/Erythrocyte x 103
Fig. 1. Percentage of lymphocyte rosettes formed with erythrocytes sensitised with 15000, 10000, 5000, 1000 and 500 IgG1 and IgG3 Mab molecules per cell. Separate assays were performed with lymphocytes from six donors and the mean values are given here.
siderable heterogeneity was found between the ability of lymphocytes from different donors to form rosettes, but lymphocytes from any particular donor gave consistently high or low values with all the Mabs. For example, the rosetting obtained with erythrocytes sensitised with R1D7 at 10000 molecules per cell ranged between 5 %0and 29%. The mean of the results are expressed here and the standard deviation is not quoted because of the great variability between donors. The interassay standard deviation was determined by assessing the rosette formation between lymphocytes from one donor and erythrocytes sensitised with 10 000 molecules per cell of RID7 in six separate assays, and was found to be 1.8%. A control using unsensitised erythrocytes was included in each assay, and this gave no rosette formation. Each rosette contained too many erythrocytes (20 or more) for them to be counted. No clear relationship was found between IgG isotype or Gm allotype and rosette formation. At the highest sensitisation level tested (15 000 molecules/cell) R1D7 gave a greater mean percentage of rosetted lymphocytes (16%) than H27, AB5 or 6D10 (9%, 10% and 11% respectively). This was consistent with all six lymphocyte donors tested. IgG3 anti-D (Mab 6D10) promoted greater rosette formation than IgG1 at 1000 molecules per cell. A minimum of 500 IgG3 and 1000 IgG1 Mab molecules per erythrocyte were required for rosette formation to be observed.
5. Discussion It was found that both IgG1 and IgG3 monoclonal anti-D-sensitised erythrocytes could form rosettes with immobilised lymphocytes. This is contrary to reports by other workers [5, 6] who, using fluid phase systems with lower ratios of erythrocytes to lymphocytes, found no rosette formation when erythrocytes were sensitised with IgG1 Mabs. The minimum level of bound anti-D required to bring about rosette formation with IgG3-sensitised erythrocytes (500 molecules/cell) would indicate that this is a more sensitive technique than that employed by others [5], who found that at least 5000 IgG3 Mab molecules per cell were required, while rosette formation with IgG1 Mab-sensitised erythrocytes was not observed. Others were unable to obtain rosettes with either IgG1 or IgG3 anti-D Mabs [6].
It has been postulated that some property of the IgG1 Mabs may be lacking, such as amino acid sequence or oligosaccharide composition, or that the spatial orientation on the erythrocyte surface is such that they do not rosette [5]. The data contained in this report suggest that this is not the case, and that lack of rosette formation may be due to the insensitivity of the assays used. It was found here that erythrocytes sensitised comparably with IgG1 and IgG3 Mabs at 5000 molecules per cell or more gave similar levels of rosette formation, which is in agreement with one report using polyclonal anti-D [1] but not with another [4]. Slightly greater rosette formation at the higher sensitisation levels was achieved with Mab R1D7, an IgG1 anti-D of the Glm(3) allotype than with the other Mabs. It is of interest that R1D7 is the only one of these Mabs that promotes erythrocyte lysis in lymphocyte (K cell)-mediated ADCC (Ref. 10 and unpublished observations). Mabs H27, AB5 and 6D10 are not normally lytic in ADCC, but here they brought about rosette formation. The lymphocyte population used contained both K cells and B cells, bearing FcRIII and FcRII, respectively. The slightly greater number of rosettes formed with Mab R1D7 may have been due to additional rosetting with K cells, which did not take place with the other three Mabs. Considerable heterogeneity was found between individual lymphocyte donors in their ability to form rosettes, and this may reflect the number of FcR-bearing cells present or the density of the FcR expressed on the cells. The use of immobilised lymphocytes enables large numbers of antibodies to be tested in one assay, allows a high erythrocyte-to-lymphocyte ratio to be used resulting in great sensitivity and, as the rosettes are fixed and adhered to a solid support, the results are preserved, facilitating reading.
Acknowledgements The authors are grateful to Yvonne Richards (R.T.C., Manchester) for the provision of the radiolabelled antiglobulin reagent, Dr. A. Merry (B.G.R.L., Oxford) for the radiometric antiglobulin method, and Dr. R. Hancock for helpful discussion. Financial support from the Central Blood Laboratories Authority is acknowledged. 111
References [1] Gergely, J., Sarmay, G., Rozsnyay, Z., et al. (1986) Mol. lmmunol. 23, 1203. [2] Sarmay, G., Jefferis, R., Klein, E., et al. (1985) Eur. J. Immunol. 15, 1037. [31 Zupanska, B., Maslanka, K., van Loghem, E., et al. (1982) Vox. Sang. 43, 243. [4] Zupanska, B., Thomson, E. E. and Merry, A. H. (1986) Vox. Sang. 50, 97. [5] Merry, A. H., Brojer, E., Zupanska, B., et al. (1989) Vox. Sang. 56, 48. [6] Urbaniak, S. J. and Stewart, G. M. (1988) Rev. Ft. Transfus. Immunohematol. 31,237. [7] Rozsnay, Z., Sarmay, G., Walker, M., et al. (1989) Immunology 66, 491.
112
[8] Urbaniak, S. J. (1979) Br. J. Haematol. 42, 315. [9] Urbaniak, S. J. and Ayoub Greiss, M. (1980) Br. J. Haematol. 46, 447. [10] Kumpel, B. M., Leader, K. A. and Bradley, B. A. (1988) Biochem. Soc. Trans. 16, 733. [1l] Urbaniak, S. J., Ayoub Greiss, M., Crawford, R. J., et al. (1984) Vox Sang. 46, 323. [12] Pollack, W. (1984) in: Hemolytic Disease of the Newborn, pp. 53-66, American Association of Blood Banks, Arlington, VA. [13] Boyum, A. (1979) Scand. J. lmmunol. 5, 9. [14] Hancock, R. H. (1979) J. Immunol. Methods 27, 93. [151 Kumpel, B. M., Poole, G. D. and Bradley, B. A. (1989) Br. J. Haematol. 71, 125. [161 Leader, K. A., Kumpel, B. M., Poole, G. D., et al. (1989) Vox. Sang., in press. [17] Merry, A. H,, Thomson, E. E., Rawlinson, V. I., et al. (1982) Clin. Lab. Haematol. 4, 393.
Elsevier IMLET 01317
Rosette formation between immobilised human lymphocytes and erythrocytes sensitised with monoclonal anti-D K. A. Leader, B. M. K u m p e l a n d B. A. Bradley United Kingdom Transplant Service, Southmead, Bristol, U.K.
(Received 14 August 1989; revision received 7 September 1989; accepted 8 September 1989)
1. Summary A rapid, reproducible and sensitive assay was developed to investigate the ability of human lymphocytes to form rosettes with erythrocytes sensitised with human monoclonal anti-D. Erythrocytes sensitised with a known number of anti-D molecules per cell were incubated with lymphocytes immobilised on plastic by poly(L-lysine), the resulting rosettes fixed, unbound erythrocytes removed by washing and the cell preparation stained. IgG1 and IgG3 antiD-coated erythrocytes gave similar rosette formation at sensitisation levels in the range of 5000-15000 molecules per cell, although at lower sensitisation levels IgG3 gave greater rosette formation than IgG1. A minimum of 500 IgG3 and 1000 IgG1 anti-D molecules per erythrocyte were required for rosetting.
2. Introduction The interaction of Fc receptor (FcR)-bearing human lymphocytes with cell-bound immunoglobulin (Ig) may be assessed by their ability to form rosettes with Ig-sensitised erythrocytes [1-7] or by measuring the extent of red cell lysis which may occur subsequently [6-10]. Rosette assays reflect the ability of cell-bound antibodies to bind to lymphocyte FcR, while antibody-dependent cell-mediated cytotoxici-
ty (ADCC) assays determine the ability of these antibodies to activate lytic responses. These cellular interactions may be of clinical significance. Lymphocyte-mediated ADCC has been shown to correlate with the clinical severity of haemolytic disease of the newborn (HDN) [11], and FcR-bearing lymphocytes may participate in antibody-mediated immune suppression (AMIS), which prevents maternal allo-immunisation to D-positive foetal erythrocytes following prophylactic administration of antiD [12]. It has been shown previously that some polyclonal and monoclonal IgGl anti-D may promote the lysis of D-positive erythrocytes by human lymphocytes [10, 11]. However, these findings are inconsistent with the failure of several workers to demonstrate rosette formation between human lymphocytes and erythrocytes sensitised with anti-D monoclonal antibodies (Mabs) of the IgG1 subclass [5, 6]. These assays were performed using lymphocytes and sensitised erythrocytes in fluid phase. We report here a lymphocyte rosette assay with erythrocytes sensitised with known levels of IgG1 and IgG3 monoclonal anti-D and lymphocytes attached to a solid phase, permitting a greater red cell-tolymphocyte ratio, resulting in increased sensitivity, and allowing preservation of the results to facilitate reading.
3. Materials and Methods Key words." Lymphocyte;Rosetting; Poly-(L-lysine);Monoclonal
anti-D Correspondence to: K. A. Leader, United Kingdom Transplant Service, Southmead Road, Bristol, BS105ND, U.K.
3.1. Preparation o f lymphocytes
Peripheral blood mononuclear cells (PBMC) were prepared by separation of whole blood over Lym-
0165-2478 / 89 / $ 3.50 © 1989 ElsevierScience Publishers B.V.(BiomedicalDivision)
109
phoprep (Nyegaard, Sweden) [13]. The PBMC were washed three times by repeated centrifugation (400 xg) and resuspension in RPMI-1640 (Imperial Laboratories). The cells were resuspended at 5 x 106 cells per ml in RPMI supplemented with 10°70 foetal calf serum (FCS) (RPMI/FCS). Monocytes were removed by adherence to plastic in tissue culture flasks (Nunclon) for 1 h at 37°C in RPMI/FCS. N o n adherent cells were removed, washed and resuspended at 1 x 105 cells per ml in PBS. Monocyte contamination was assessed by immunofluorescence using a CD14 antibody (Serotec) and was found to be less than 5°70.
3.2. Immobilisation of lymphocytes Ninety-six-well ELISA plates (Nunclon) were coated with 50/A per well of poly-(L-lysine) (0.1 mg/ml in PBS) for 1 h at room temperature. The plates were then washed three times in PBS and 100 #1 (1 x 104 cells) of the lymphocyte suspension added to each well. The plates were centrifuged at 8 5 0 x g for 10 min, the PBS was removed and 100 tzl of R P M I / F C S added to each well for 1 h to block the poly-(L-lysine) to which no cells had bound [14].
5070 (approx 4x105 per ml) in RPMI. 3.5. Lymphocyte rosette assay The R P M I / F C S was removed from the prepared lymphocyte plates and 100/A (4x107 cells) of the sensitised erythrocyte suspension added to each well. The plates were centrifuged at 70 g for 3 min and incubated at 37 °C for 30 min. Without removing the medium the rosettes were fixed by the addition of 200 t~l of 1°70glutaraldehyde in PBS for 10 min. Supernatants were aspirated and unrosetted erythrocytes were gently resuspended in PBS and removed. The lymphocytes were stained with 0.1 °70trypan blue in PBS and the percentage of lymphocytes with 3 or more adherent erythrocytes was assessed. A minimum of 200 lymphocytes were counted per well. 4. Results
All the Mabs brought about rosette formation (Fig. 1), the extent of rosetting depending upon the erythrocyte sensitisation level. Each Mab was tested using lymphocytes from six different donors. Con-
Culture supernatants containing Mabs were diluted in R P M I / F C S and incubated with an equal volume of a 5°70 suspension of group O R1R2 erythrocytes for 2 h at 37 °C, followed by three washes in RPMI. Cell-bound IgG was quantified using a radiometric indirect antiglobulin method [17]. Sensitisation levels of 15 000, 10000, 5000, 1000 and 500 antibody molecules per erythrocyte were obtained by sensitising cells in supernatants appropriately diluted. After sensitisation, the cells were resuspended at 110
R1D7
25-
H27
25 2O
20-
The human Mabs with specificity for the Rh antigen D were derived from EBV-transformed Blymphoblastoid cell lines [10, 15, 16]. Four Mabs were used in this study, all of different isotype or Gm allotype, as follows: R1D7, IgG1 Glm(3); H27, IgG1 Glm(1,2,17); AB5, IgG1 Glm(1,17); 6D10, IgG3 G3m(5). 3.4. Sensitisation and quantitation of cell-bound anti-D
30~
30
3.3. Monoclonal antibodies
z5-~ 10-
15 I' :
10-~
i
i
i o
f5
I
~: : :::! fo
5
5 i
0:5
30
o
~, 1'5
1'0
5
i
015
30 AB5
E
25-
~
20-
20-
15-
15.
1 0 - - -
10--
6D10
25
5-
5 i . . . .
0
I
1'5
fo
I
•
5
t
0'5
0
i/::!iiiiiiiiiiiiiiiii!!i!L 1'5
fo
5
i
.....
0:5
Mab Molecules/Erythrocyte x 103
Fig. 1. Percentage of lymphocyte rosettes formed with erythrocytes sensitised with 15000, 10000, 5000, 1000 and 500 IgG1 and IgG3 Mab molecules per cell. Separate assays were performed with lymphocytes from six donors and the mean values are given here.
siderable heterogeneity was found between the ability of lymphocytes from different donors to form rosettes, but lymphocytes from any particular donor gave consistently high or low values with all the Mabs. For example, the rosetting obtained with erythrocytes sensitised with R1D7 at 10000 molecules per cell ranged between 5 %0and 29%. The mean of the results are expressed here and the standard deviation is not quoted because of the great variability between donors. The interassay standard deviation was determined by assessing the rosette formation between lymphocytes from one donor and erythrocytes sensitised with 10 000 molecules per cell of RID7 in six separate assays, and was found to be 1.8%. A control using unsensitised erythrocytes was included in each assay, and this gave no rosette formation. Each rosette contained too many erythrocytes (20 or more) for them to be counted. No clear relationship was found between IgG isotype or Gm allotype and rosette formation. At the highest sensitisation level tested (15 000 molecules/cell) R1D7 gave a greater mean percentage of rosetted lymphocytes (16%) than H27, AB5 or 6D10 (9%, 10% and 11% respectively). This was consistent with all six lymphocyte donors tested. IgG3 anti-D (Mab 6D10) promoted greater rosette formation than IgG1 at 1000 molecules per cell. A minimum of 500 IgG3 and 1000 IgG1 Mab molecules per erythrocyte were required for rosette formation to be observed.
5. Discussion It was found that both IgG1 and IgG3 monoclonal anti-D-sensitised erythrocytes could form rosettes with immobilised lymphocytes. This is contrary to reports by other workers [5, 6] who, using fluid phase systems with lower ratios of erythrocytes to lymphocytes, found no rosette formation when erythrocytes were sensitised with IgG1 Mabs. The minimum level of bound anti-D required to bring about rosette formation with IgG3-sensitised erythrocytes (500 molecules/cell) would indicate that this is a more sensitive technique than that employed by others [5], who found that at least 5000 IgG3 Mab molecules per cell were required, while rosette formation with IgG1 Mab-sensitised erythrocytes was not observed. Others were unable to obtain rosettes with either IgG1 or IgG3 anti-D Mabs [6].
It has been postulated that some property of the IgG1 Mabs may be lacking, such as amino acid sequence or oligosaccharide composition, or that the spatial orientation on the erythrocyte surface is such that they do not rosette [5]. The data contained in this report suggest that this is not the case, and that lack of rosette formation may be due to the insensitivity of the assays used. It was found here that erythrocytes sensitised comparably with IgG1 and IgG3 Mabs at 5000 molecules per cell or more gave similar levels of rosette formation, which is in agreement with one report using polyclonal anti-D [1] but not with another [4]. Slightly greater rosette formation at the higher sensitisation levels was achieved with Mab R1D7, an IgG1 anti-D of the Glm(3) allotype than with the other Mabs. It is of interest that R1D7 is the only one of these Mabs that promotes erythrocyte lysis in lymphocyte (K cell)-mediated ADCC (Ref. 10 and unpublished observations). Mabs H27, AB5 and 6D10 are not normally lytic in ADCC, but here they brought about rosette formation. The lymphocyte population used contained both K cells and B cells, bearing FcRIII and FcRII, respectively. The slightly greater number of rosettes formed with Mab R1D7 may have been due to additional rosetting with K cells, which did not take place with the other three Mabs. Considerable heterogeneity was found between individual lymphocyte donors in their ability to form rosettes, and this may reflect the number of FcR-bearing cells present or the density of the FcR expressed on the cells. The use of immobilised lymphocytes enables large numbers of antibodies to be tested in one assay, allows a high erythrocyte-to-lymphocyte ratio to be used resulting in great sensitivity and, as the rosettes are fixed and adhered to a solid support, the results are preserved, facilitating reading.
Acknowledgements The authors are grateful to Yvonne Richards (R.T.C., Manchester) for the provision of the radiolabelled antiglobulin reagent, Dr. A. Merry (B.G.R.L., Oxford) for the radiometric antiglobulin method, and Dr. R. Hancock for helpful discussion. Financial support from the Central Blood Laboratories Authority is acknowledged. 111
References [1] Gergely, J., Sarmay, G., Rozsnyay, Z., et al. (1986) Mol. lmmunol. 23, 1203. [2] Sarmay, G., Jefferis, R., Klein, E., et al. (1985) Eur. J. Immunol. 15, 1037. [31 Zupanska, B., Maslanka, K., van Loghem, E., et al. (1982) Vox. Sang. 43, 243. [4] Zupanska, B., Thomson, E. E. and Merry, A. H. (1986) Vox. Sang. 50, 97. [5] Merry, A. H., Brojer, E., Zupanska, B., et al. (1989) Vox. Sang. 56, 48. [6] Urbaniak, S. J. and Stewart, G. M. (1988) Rev. Ft. Transfus. Immunohematol. 31,237. [7] Rozsnay, Z., Sarmay, G., Walker, M., et al. (1989) Immunology 66, 491.
112
[8] Urbaniak, S. J. (1979) Br. J. Haematol. 42, 315. [9] Urbaniak, S. J. and Ayoub Greiss, M. (1980) Br. J. Haematol. 46, 447. [10] Kumpel, B. M., Leader, K. A. and Bradley, B. A. (1988) Biochem. Soc. Trans. 16, 733. [1l] Urbaniak, S. J., Ayoub Greiss, M., Crawford, R. J., et al. (1984) Vox Sang. 46, 323. [12] Pollack, W. (1984) in: Hemolytic Disease of the Newborn, pp. 53-66, American Association of Blood Banks, Arlington, VA. [13] Boyum, A. (1979) Scand. J. lmmunol. 5, 9. [14] Hancock, R. H. (1979) J. Immunol. Methods 27, 93. [151 Kumpel, B. M., Poole, G. D. and Bradley, B. A. (1989) Br. J. Haematol. 71, 125. [161 Leader, K. A., Kumpel, B. M., Poole, G. D., et al. (1989) Vox. Sang., in press. [17] Merry, A. H,, Thomson, E. E., Rawlinson, V. I., et al. (1982) Clin. Lab. Haematol. 4, 393.