HEMATOPATHOLOGY
Original Article
Flow Cytometric Measurement of
Glycosylphosphatidyl-inositol-linked Surface
Proteins on Blood Cells of Patients With
Paroxysmal Nocturnal Hemoglobinuria
Y.L. KWONG, M R C P A T H , 1 C.P. LEE, AILMS, 2 T.K. CHAN, FRCP, 1 AND L.C. CHAN, P H D 2
Paroxysmal nocturnal hemoglobinuria (PNH) is an acquired disorder
of hematopoiesis in which affected cells are deficient in glycosylphosphatidyl-inositol (GPI) anchored surface proteins. The authors used
flow cytometry to study 10 patients with PNH. They used a comprehensive panel of monoclonal antibodies against all nine currently known
GPI-linked surface proteins (CD14, CD16, CD24, CD48, CD55,
CD58, CD59, CD67, CD73) on cells of various lineages. Deficient cells
were identified in the granulocytic-monocytic and erythroid lineages in
all patients. However, the lymphoid lineage was affected in only eight
patients. The patterns of deficiency were variable, with deficient cells
constituting a part to all of the cells in the lineages tested. Certain
proteins, including CD16, CD58, and CD59, appeared to be preferentially expressed, despite severe deficiencies of other GPI-linked pro-
teins. Moreover, a trimodal pattern of expression of CD 16, CD48, and
CD59 was observed, in which a population of cells with intermediate
levels of expression were identified in addition to positive and deficient
cells. The authors' findings indicated a great degree of heterogeneity in
the patterns and levels of expression of the GPI-linked proteins in the
various cell types, as well as a possible heterogeneity in lineage involvement. The different patterns of expression of GPI-linked proteins
should be considered when using flow cytometry to diagnose PNH.
Finally, the clinical progression in some of the patients suggested a
possible link between PNH, aplastic anemia, and myelodysplasia. (Key
words: Flow cytometry; Glycosylphosphatidyl-inositol-linked proteins; Paroxysmal nocturnal hemoglobinuria) Am J Clin Pathol
1994;102:30-35.
Paroxysmal nocturnia hemoglobinuria (PNH) is an acquired
disorder of hematopoiesis, characterized by an unusual sensitivity of the abnormal red blood cells (RBCs) to complement
lysis. 'This unusual sensitivity to complement is due to absence
of molecules that regulate the complement cascade from the
plasma membrane. These molecules include CD55 (decay accelerating factor),2 which controls the activity of the "convertase" complex of C3bBb and C4b2a, 3 and CD59 (membrane
inhibitor of reactive lysis),4,5 which controls the activation of
the membrane attack complex C5b-9. 6 Red blood cells are susceptibile to the hemolytic action of complement, causing intravascular hemolysis and intermittent hemoglobinuria.
CD55 and CD59 are both attached to the plasma membrane
via a glycosyl-phosphatidylinositol (GPI) anchor. 7 ' 8 The deficiency of these molecules was demonstrable in the white cells
(WBCs) and platelets of patients with PNH. 9 " 11 Recently, other
GPI-linked membrane proteins, including CD 14, CD 16,
CD24, CD48, CD58, and CD67, have been proven deficient in
the WBCs of affected patients.'2""17 The common deficiency of
these GPI-linked proteins, as well as the demonstration of normal mRNA for CD55 in PNH cells,18 suggests that the abnor-
mality in PNH lies not in the production of these proteins, but
in the assembly of, or the linkage to, the GPI anchor. 19,20
The observation that the biochemical defect occurs in RBCs,
WBCs, and platelets of patients with PNH suggests that the
disease might be a disorder of a hematopoietic stem cell. In
female patients with PNH who are heterozygous for
glucose-6-phosphate dehydrogenase, the affected RBCs expresses a single isoenzyme.21,22 In a recent study involving five
patients with PNH, affected WBCs expressed a monoclonal
pattern of X chromosome inactivation, using the M27 /3 and
hypoxanthine phosphoribosyl transferase probes.23 These data
suggest that PNH may be a clonal disorder, arising out of somatic mutation of a hematopoietic stem cell.
In this study, we examined 10 patients with PNH for the
expression of nine currently known GPI-linked surface antigens (CD14, CD16, CD24, CD48, CD55, CD58, CD59, CD67,
and CD73) on RBCs, WBCs, and lymphocytes, to define the
patterns of expression of these proteins. We also investigated
the possible heterogeneity of lineage involvement in these
cases.
MATERIALS A N D METHODS
From the Departments ofl Medicine and 2Palhology, Queen Mary
Hospital, Pokfulam Road. Hong Kong.
Subjects
Manuscript received February 10, 1993; revision accepted June 1,
1993.
Address reprint requests to Dr. Kwong: University Department of
Medicine, Queen Mary Hospital, Pokfulam Road, Hong Kong.
Ten patients with paroxysmal nocturnal hemoglobinuria,
confirmed by positive acidifed serum lysis (Ham's test), were
evaluated. Healthy individuals were used as controls in each
experiment.
30
KWONG ET AL.
31
GPI-linked Proteins in PNH
TABLE 1. CLINICAL FEATURES OF 10 PNH PATIENTS
Blood counts*
Patient
No.
1
2
3
4
5
6
7
8
9
10
Sex
Age
(years)
Hb
(gldl)
White
blood cells
(X109/L)
M
M
F
F
M
F
M
M
M
F
36
40
38
22
37
39
56
38
38
46
10.5
5.8
8.3
7.4
9.3
5.7
5.6
9.4
3.0(10.3)
8.1
2.7 (5.5)
2.5 (4.0)
7.7
4.0
7.8
5.2
2.1 (4.2)
4.5
1.9(4.4)
2.5
Pit
(XltflL)
59 (298)
9(116)
313
28(100)
175
26(110)
5(140)
241
41 (117)
95
Duration
of Illness
(years)
Transfusion
Requirement
(units of
packed cells/year)
4
10
5
2
5
0
9
0
0
0
6
15
0
0
0
T
16
1
13
2
Presenting
Features
Anemia,
Anemia,
Anemia,
Anemia,
Anemia,
Anemia,
Anemia,
Anemia,
Anemia,
Anemia,
HT
HT
HT
HT
HT
HT
HT
HT
HT
HT
Bone Marrow
Histology
+
+
+
+
+
—
+
+
-
Erythroid hyperplasia
Hypoplasia
Erythroid hyperplasia
Erythroid hyperplasia
Erythroid hyperplasia
Erythroid hyperplasia
Hypoplasia
Erythroid hyperplasia
Erythroid hyperplasia
Trilineal myelodysplasia
HT = Ham's test; PNH = paroxysmal nocturnal hemoglobinuria.
* Blood counts at presentation. If the latest counts were significantly different, they were included in parentheses.
Materials
Blood samples were taken from the patient and control groups.
To minimize interference from transfused blood in patients who
required regular transfusion, the experiments were done immediately before the scheduled transfusion. For flow cytometric
analysis of surface antigen expression, monoclonal antibodies
against the following GPI-linked proteins were used: CD 14 (fluorescein conjugated; Coulter, Hialeah, FL), CD 16 (fluorescein
conjugated; Immunotech, Marseille, France), CD24 (immunoglobulin [Ig] G; gift from Dr. C.E. van der Schoot, Amsterdam,
the Netherlands), CD48 (IgG; gift from Dr. A.J. Henniker,
Westmead, Australia), CD55 (IgM, decay accelerating factor),
CD58 (IgG, leukocyte adhesion factor 3), CD59 (IgG, membrane
inhibitor of reactive lysis; BioProducts Laboratory, Herts, UK),
CD67 (IgG; gift from Dr. C.E. van der Schoot), and CD73
(ecto-5'-nucleotidase, IgG; gift from Dr. L. Thompson, Oklahoma City, Oklahoma). Monoclonal antibodies against CD 13
(phycoerythrin conjugated, Coulter), CD3 (fluorescein conjugated, Coulter), and glycophorin A (phycoerythrin conjugated,
Immunotech) were used to confirm the purity of the gated cell
types. The analysis of RBCs was performed on whole blood,
granulocytes, and monocytes on WBCs collected after 1:10
dextran 70 sedimentation, and on lymphocytes on mononuclear cells after Ficoll-hypague density gradient sedimentation.
Methods
Surface antigen expression was measured by flow cytometry
using a Coulter EPICS Profile II flow cytometer (Coulter, Hialeah, FL) equipped with a 15 mW (488 nm) laser. Five to fifty
microliters of monoclonal antibodies were added to 100 /iL of
cells at a concentration of 1 X 10 7 /mL and incubated at 4 °C
for 30 minutes. After two washes with 2% bovine serum albumin (BSA) in phosphate-buffered saline (PBS), cells were analyzed if the primary antibody was already fluorescein conjugated. For unconjugated antibodies, 100 fih of a 1:40 diluted
secondary fluorescein-conjugated sheep antimouse Ig antibody
were added and incubated for 30 minutes in the dark. After two
washes with 2% BSA in PBS, the cells were analyzed immediately. Flow cytometric analysis of different cell types was performed with gating based on their respective characteristic
forward and side scatter properties under appropriate photomultiplertube voltage. Isotypically matched immunoglobulins
for each monoclonal antibody were used as negative controls in
all the experiments. A minimum of 5000 cells were analyzed.
To confirm the purity of gated cells, the expression of CD 13
was measured in cells in the graunlocyte/monocyte gate, CD3
in the lymphocyte gate, and glycophorin A in the RBC gate.
Cells were considered positive if their fluorescent intensity was
greater than the largest 2% of cells in the isotypic control.
RESULTS
Subjects and Clinical
Features
The patient group included six men and four women, aged
22-56 years (Table 1). Patients 2 and 7 had aplastic anemia;
their Ham's tests became positive 4 and 2 years after presentation, respectively. Patients 6 and 9 had varying degrees of pancytopenia. The Ham's tests were marginally positive at presentation, subsequently becoming strongly positive. All four
patients had gradual improvement of white cell and platelet
counts after the PNH phenotype appeared; patient 9 became
transfusion independent within 2 years. Patient 10 had myelodysplasia (refractory anemia, 2% ringed sideroblasts) and a negative Ham's test, which became weakly positive 1 year later.
(Table 1).
Expression
Individuals
of GPI-linked
Proteins in
Normal
Preliminary experiments were performed to define the types
of blood cells that expressed the various GPI-linked proteins at
levels optimal for flow cytometric analysis. These experiments
showed that CD 14 and CD48 were optimally expressed on
monocytes, and CD 16 and CD67 on granulocytes. CD24 was
measured on granulocytes; although it is also expressed on B
cells, its numbers are too small for optimal analysis. These
results agreed with previous observations.24"26 CD55, CD58,
and CD59 were measured on RBCs and granulocytes; they
were also measured on lymphocytes together with CD73 to
assess the involvement of the lymphoid lineage. CD73 is ex-
Vol. 102 •No. 1
32
HEMATOPATHOLOGY
Original Article
TABLE 2. E X P R E S S I O N OF GPI L I N K E D SURFACE A N T I G E N S I N CONTROLS A N D
Monocytes
(% positive)
Granulocytes
(°h positive)
PATIENTS
Red Blood Cells
(°/o positive)
Lymphocytes
(°/o positive)
Subjects
CD16*
CD24
CD67
CD55
CD58
CD59*
CD14
CD48*
CD55
CD58
CD59
CD73f
Controls
(n = 10)
1
2
3
4
5
6
7
8
9
10
92 ± 6
90/6
60/10
80/10
85/15
90/5
55/10
70/0
90/5
10/0
60/15
97 ± 3
15
13
15
30
10
12
0
28
0
40
95 + 5
15
13
15
15
10
10
0
24
0
40
95 ± 3
14
16
14
20
13
22
0
25
0
55
95 ± 6
98
97
98
98
98
96
97
99
85
98
98 ± 2
85/11
74/24
90/5
68/30
80/15
60/30
70/13
85/15
70/15
35/65
82 ± 13
6
12
5
15
6
6
0
15
10
25
93 ± 6
0/5
0/5
0/10
75/15
0/10
0/12
0/5
55/15
0
0/25
67 ± 14
38
16
40
42
50
47
30
45
30
50
57 ± 14
58
30
5
25
50
32
55
55
30
70
95 ± 3
53
34
58
40
80
51
65
78
50
80
100
5
60
7
10
40
50
20
100
10
95
CD55
CD58
CD59*
71 ± 2 0
10
30
15
15
20
15
15
30
10
50
87 ± 14
75
32
66
70
35
80
60
50
75
90
89 ± 5
75/0
25/25
60/0
5/50
0/15
0/80
20/20
0/70
75/0
45/20
* For anligens with trimodal distributions, thefirstvalue refers to percentage of cells with intermediate fluorescent intensity, and the last
value refers to cells with normal fluorescent intensity. Results of less than 5% were treated as zero.
t Expressed as a percentage of the positive cells in the control (15-30% of lymphocytes).
pressed in approximately 30% of CD3-positive lymphocytes,27
but it is age dependent, with the level decreasing after age 40. 28
The level of CD73 in patients was therefore expressed as a
percentage of the level of an age-matched control included in
the same experiment. Finally, measurement of CD 13 in the
granulocyte-monocyte gate, CD 3 on the lymphocyte gate, and
glycophorin A in the red cell gate gave results consistently
above 90%.
The mean levels of expression of GPI-linked surface antigens
on different cell types of 10 normal donors are summarized in
Table 2.
*N
l^^w.
w.,A
A
GRANULOCYTES
.CD 24
1.Isotypic control
2.Normal
3.Patient 8
4.Patient 9
b .CD 16
1.Isotypic control
2.Normal
3.Patient 8
4.Patient 7
C. CD 58
1 .Isotyoic control
2.Normal
3.Patient 3
.^"v...^^.--.
MONOCYTES
a . C D 14
FIG. I. Patterns of expression of GPI-linked proteins on granulocytes.
X axis: log green fluorescence. Y axis: relative cell counts. A, CD24.
Note bimodal distribution in patient 8, and no positive cells in patient
9. B, CDI6. Note trimodal distribution of negative cells, cells of intermediate fluorescence, and cells of normal fluorescence intensities in
normal and patient 8. The population of negative cells had very low
meanfluorescenceand was outside the plot on the left side. Patient 7
showed only positive cells of intermediatefluorescence.Expression of
CD59 showed similar patterns. C, CD58. Note majority of cells in
patient 3 were positive at a reduced mean fluorescence.
b . CD 48
1.Isotypic control
1.Isotypic control
2.Normal
2.Normal
3.Patient 1
3.Patient 8
FIG. 2. Patterns of expression of GPI-linked proteins on monocytes. X
axis: log greenfluorescence.Yaxis: relative cell counts. A, CD 14. Note
bimodal distribution in patient 1. B, CD48: Note trimodal distribution,
with the presence of cells of intermediatefluorescence.The third population of negative cells were outside the plot on the left side.
A.J.C.P.-July 1994
33
KWONG ET AL.
GPI-linked Proteins in PNH
a diminished population expressing the antigens at the same
fluorescent intensity as in the control group. The second pattern was seen in CD48 in patients 4 and 8 (Fig. 2B). This pattern showed a trimodal distribution with one negative population, one positive population at a reduced fluorescent intensity,
and one positive population at the same intensity as in the
control group. The third pattern was total absence of positive
cells, found in CD 14 in patient 7 and in CD48 in patient 9.
Lymphocytes. Eight patients (1-6, 9) showed only one pattern of expression of CD55, CD58, CD59, and CD73: a bimodal distribution with the presence of negative cells together
with a variably diminished population showing expression at
the same fluorescent intensity as in the control group (Table 2
and Fig. 3A). Patients 8 and 10 showed almost normal expression of all of these antigens (Fig. 3B).
Red blood cells. All patients showed a single pattern of expression of CD55 and CD58: a bimodal distribution of negative cells and another population of cells positive at the same
fluorescent intensity as in the control group, although the expression of CD58 was discordantly higher than CD55 (Table 2
and Fig. 4A and B). However, the expression of CD59 showed
three patterns (Fig. 4C). The first pattern (patients 2, 4, 7, 10)
was trimodal, showing one negative population, one positive
LYMPHOCYTES
CD 55
b . CD 58
1.Isotypic control
1.Isotypic control
2.Normal
2.Normal
3.Patient 5
3.Patient 8
4.Patient 2
4.Patient 3
FIG. 3. Patterns of expression of GPI-linked proteins on lymphocytes.
X axis: log green fluorescence. Y axis: relative cell counts. A, CD55.
Note diminished number of cells with the same fluorescence intensities
as control. Expression of CD59 showed similar patterns. B, CD58.
Note that patient 8 had cells of normal fluorescence.
»>>-A^
Expression
PNH
of GPI-linked
Proteins in Patients
With
Granulocytes. Four different patterns of expression of GPIlinked surface antigens were found. (Table 2 and Fig. 1 A) The
first pattern was found in CD24, CD67, and CD55 (patients
1-6, 8, 10). It showed a bimodal distribution with a predominantly negative population and a minor, positive population at
approximately the same fluorescent intensity as in the control
group. The second pattern was found in CD 16 and CD59 (patients 1-6, 8, 10; Fig. IB) and showed a trimodal distribution
with one negative population, one positive population at a reduced fluorescent intensity, and one positive population at the
same intensity as in the control group. The third pattern was
seen in CD58 (all patients; Fig. 1C), which was expressed by the
majority of cells but at a lower fluorescent intensity than in the
control group. The fourth pattern was seen in CD 16, CD24,
CD55, and CD67 in patients 7 and 9 only, in whom no cells
expressing these antigens could be found.
Monocytes. Three different patterns were found for CD 14
and CD48 (Table 2 and Fig. 2A). The first pattern (patients
1-3, 5, 6, 10) was bimodal with a majority of negative cells and
K
\^_
tw,
y\^
JK
k
A.
RED BLOOD CELLS
a . CD 55
b . CD 58
C . C D 59
1.Isotypic control
1.Isotypic control
1.Isotypic control
2.Normal
2.Normal
3.Patient 4
3.Patient 7
2.Normal
3.Patient 7
4.Patient 1
S.Patient 5
FIG. 4. Patterns of expression of GPI-linked proteins on red blood cells.
A'axis: log green fluorescence, yaxis: relative cell counts. A, CD55. B,
CD58. Note that expression of CD58 was discordingly higher than
CD55. C, CD59. Note trimodal distribution in patient 7. Patient 1 had
only positive cells of intermediate fluorescence, whereas patient 5 had
only positive cells of normal fluorescence.
Vol. 102-No. 1
34
HEMATOPATHOLOGY
Original Article
population at a reduced fluorescent intensity, and one positive
population at normal intensity. The second pattern (patients 1,
3, 9) was bimodal, showing one negative population and one
population with reduced fluorescent intensity. The third pattern (patients 5,6, 8) was bimodal, showing one negative population and one population with normal fluorescent intensity.
DISCUSSION
This study is the first to report the use of a comprehensive
panel of antibodies against all currently known GPI-linked surface proteins in the examination of patients with PNH. It provides important guidance for using flow cytometry in the laboratory diagnosis of this disorder. Several patterns of expression
of GPI-linked surface antigens were found in cells of different
lineages. In granulocytes, the bimodal distribution for CD24,
CD67, and CD55 can be explained by the positive cells with
normal fluorescent intensity being residual cells of the normal
clone, and negative cells reflecting the PNH clone. For CD 16
and CD59, however, a population of intermediately positive
cells was also found, indicating a preferential preservation of
expression of these antigens at a reduced level in some of the
cells in the PNH clone. This phenomenon has been reported in
the expression of CD16 in PNH polymorphs, 7 and we have
shown that a similar mechanism occurs in CD59. Interestingly,
the expression of CD58 occurred in the majority of cells but at
a slightly reduced level, perhaps because CD58 exists in a transmembrane form in addition to the GPI-linked form in
WBCs.29"31 Finally, in patients 7 and 9, cells expressing CD24,
CD55, and CD67 were undetectable, showing absence of the
normal clone and predominance of the PNH clone.
The three different patterns of expression of CD 14 and
CD48 in monocytes can be explained along similar lines as
those in granulocytes. A good concordance was found between
the size of the PNH clone, as reflected by CD 14 and CD48
negative monocytes, and that reflected by CD24, CD55, and
CD67 negative granulocytes. This is to be expected, as granulocytes and monocytes are derived from the same progenitor
cells.
In lymphocytes, the presence of CD55, CD58, CD59, and
CD73 negative cells indicated involvement of the lymphoid
lineage. However, the proportion of CD55, CD58, and CD59
negative lymphocytes was considerably smaller than that
shown by analysis of the granulocyte-monocyte lineage. This
discrepancy between the granulocyte-monocyte and lymphoid
clone sizes probably reflects a difference in the proliferation of
the PNH clone along the myeloid and lymphoid lineages.16'32
In patients 7 and 9, expression of CD55, CD58, CD59, and
CD73 appeared normal, implying that the lymphoid lineage
might not be part of the PNH clone. These results therefore
indicate a heterogeneity of lineage involvement, with some
cases affecting both the myeloid and the lymphoid lineages (a
multipotential stem cell), and others involving only a myeloid,
but not the lymphoid, progenitor. This is consistent with previous studies that used a more restricted panel of antibodies
against GPI-linked proteins." 1532 - 34
The expression of CD55 and CD58 on RBCs followed the
bimodal pattern observed in granulocytes and monocytes, although CD58 again appeared discordantly higher. CD58 only
existed in the GPI-linked form in RBCs,35 indicating a preferential preservation of expression of CD58 on some of the RBCs of
the PNH clone, similar to CD 16 and CD59 in granulocytes.
However, the expression of CD59 was more heterogenous,
showing a variable combination of negative, intermediate, and
positive cells. Although CD55 and CD59 act together to control complement activation, CD59 is the more important protein, as RBCs with CD55 inhibited by antibody blocking were
partially sensitive to acidified serum lysis, whereas those with
CD59 inhibited were completely sensitive.36 Thus, the three
different populations of CD59 negative, intermediate, and positive cells probably correspond to PNH III (markedly sensitive
to complement), PNH II (intermediate sensitivity), and PNH I
(normal sensitivity) RBCs.37"39 Rosse and colleagues40 have recently demonstrated the direct relationship between CD59 expression and complement sensitivity, cells with intermediate
expression of the protein corresponding to PNH II cells. We
have shown that these populations of cells can be easily and
accurately distinguished using flow cytometry.
In addition to demonstrating the heterogenous patterns of
expression of GPI-linked proteins in cells in different patients,
this study illustrated two other interesting points. First, a preferentially preserved expression of some GPI-linked proteins apparently exists. This might indicate a hierarchy of access of
different protein molecules to available GPI anchors, as previously suggested.17 Second, the presence of cells of intermediate
expression showed that the GPI defect was likely to be quantitative and not strictly qualitative.
Recent studies have addressed the relationship between
PNH and aplastic anemia. Both PNH and aplastic anemia may
arise from a damaged marrow, with the PNH clone having a
relative proliferative advantage, so that the development of the
PNH anomaly in patients with aplastic anemia often results in
hematologic improvement. 39 ' 41 This clinical progression is illustrated in at least two of our patients (2 and 7) who had
cytopenia. Considerable improvement was noted after the development of PNH. In two other patients (6 and 9), who had
marginally positive Ham's tests, improvement of cytopenias
actually occurred with expansion of the PNH clone, as evidenced by the Ham's tests becoming strongly positive. Finally,
in one of our patients, PNH developed from an antecedent
myelodysplasia. These observations further support the possible nosologic link between aplastic anemia, PNH, and myelodysplasia-leukemia, first proposed by Dameshek 42 and
strengthened by later observations. 43
REFERENCES
1. Rosse W, Dacie JV. Immune lysis of normal human and paroxysmal nocturnal hemoglobinuria (PNH) red blood cells. I. The
sensitivity of PNH red cells to lysis by complement and specific
antibody. J Clin Invest 1966;45:736-757.
2. Nicholson-Weller A, March JP, Rosenfeld SI, Austen K.F. Affected
erythrocytes of patients with paroxysmal nocturnal hemoglobinuria are deficient in the complement regulatory protein, decay accelerating factor. Proc Natl AcadSci USA 1980; 80:50665070.
3. Medof ME, Kinoshita T, Silber R, Nussenzweig V. Amelioration
of the lytic abnormalities of paroxysmal nocturnal hemoglobinuria with decay accelerating factor. Proc Natl Acad Sci USA
1985;82:2980-2984.
4. Holguin MH, Fredrick LR, Bernshaw NJ, Parker CJ. Isolation and
characterization of a membrane protein from normal human
erythrocytes that inhibits reactive lysis of the erythrocytes of
paroxysmal nocturnal hemoglobinuria. J Clin Invest
1989;84:7-17.
5. Holguin MH, Wilcox LA, Bernshaw NJ, Rosse WF, Parker CJ.
Relationship between the membrane inhibitor of reactive lysis
A.J.C.P.-July 1994
KWONG ET AL.
GPI-linked Proteins in PNH
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
and the erythrocyte phenotypes of paroxysmal nocturnal hemoglobinuria. J Clin Invest 1989;84:1387-1394.
Rollins SA, Zhao JI, Ninomiya H, Sims PJ. Inhibition of homologous complement by CD59 is mediated by a species selective
recognition conferred through binding to C8 within C5b-8 or
C9 within C5b-9. J Immunol 1991; 146:2345-2351.
Rosse WF. Phosphatidylinositol-linked proteins and paroxysmal
nocturnal hemoglobinuria. Blood 1991;75:1595-1601.
Cross GA. Glycolipid anchoring of plasma membrane proteins.
Annu Rev Cell Biol 1990;6:1-39.
Aster RH, Enright SE. A platelet and granulocyte membrane defect in paroxysmal nocturnal hemoglobinuria: Usefulness for
the detection of platelet antibodies. J Clin Invest 1969;48:11991210.
Stern M, Rosse WF. Two populations of granulocytes in paroxysmal nocturnal hyemoglobinuria. Blood 1979;53:928-934.
Nicholson-Weller A, Apicer DB, Austen KF. Deficiency of the
complement regulatory protein, "decay-accelerating factor," on
membranes of granulocytes, monocytes, and platelets in paroxysmal nocturnal hemoglobinuria. N Engl J Med 1985;
312:1091-1097.
Selvaraj P, Dustin ML, Silber R, Low MG, Springer TA. Deficiency of lymphocyte function-associated antigen 3 (LFA-3) in
paroxysmal nocturnal hemoglobinuria. Functional correlates
and evidence for a phosphatidylinositol membrane anchor. J
Exp Med 1987; 166:1011-1025.
Simmons DL, Tan S, Tenen DG, Nicholson-Weller A, Seed B.
Monocyte antigen CD 14 is a phospholipid anchored membrane
protein. Blood 1989;73:284-289.
Okuda K, Kanamaur A, Ueda E, Kitani T, Nagai K. Membrane
expression of decay-accelerating factor on neutrophils from
normal individuals and patients with paroxysmal nocturnal hemoglobinuria. Blood 1990; 75:1186-1191.
van der Schoot E, Huizinga TWJ, van't Veer-Korthof ET, et al.
Deficiency of glycosyl-phosphatidylinositol-linked membrane
glycoproteins of luekocytes in paroxysmal nocturnal hemoglobinuria, description of a new diagnostic cytofluorometric assay.
Blood 1990;76:1853-1859.
Schubert J, Uciechowski P, Delany P, et al. The PIG-anchoring
defect in NK lymphocytes of PNH patients. Blood
1990;76:1181-1187.
Edberg JC, Salmon JE, Whitlow M, Kimberly RP. Preferential
expression of human FCTRIII P M N (CD16) in paroxysmal nocturnal hemoglobinuria. J Clin Invest 1991;87:58-67.
Stafford HA, Tykocinski ML, Lublin DM, et al. Normal polymorphic variations and transcription of the decay accelerating factor gene in paroxysmal nocturnal hemoglobinuria. Proc Natl
AcadSci USA 1988;85:880-884.
Carothers DJ, Hazra SV, Anderson SW, Medof ME. Synthesis of
aberrant decay-accelerating factor proteins by affected paroxysmal nocturnal hemoglobinuria leucocytes. J Clin Invest
1990;85:47-54.
Mahoney JF, Urakaze M, Hall S, et al. Defective glycosylphosphatidylinositol anchor synthesis in paroxysmal nocturnal hemoglobinuria granulocytes. Blood 1992;79:1400-1403.
Oni SB, Osunkoya BO, Luzzato L. Paroxysmal nocturnal hemoglobinuria: Evidence for monoclonal origin of abnormal red
cells. Blood 1970;36:145-152.
Rotoli B, Robledo R, Scarpato N, Luzzato L. Two populations of
erythroid cell progenitors in paroxysmal nocturnal hemoglobinuria. Blood 1984;64:847-851.
Josten KM, Tooze JA, Borthwick-Clarke C, Gordon-Smith EC,
Rutherford TR. Acquired aplastic anemia and paroxysmal nocturnal hemoglobinuria: Studies on clonality. Blood 1991;
78:3162-3167.
35
24. Fleit HB, Wright SD, Unkeless JC. Human neutrophil FCT receptor distribution and structure. Proc Natl Acad Sci USA
1982;79:3275-3279.
25. Hadam MR. Cluster report: CD48. In: Knapp W, Dorken B, Gilks
WR, et al, eds. Leucocyte Typing IV. 2nd ed. Oxford, UK: Oxford University Press, 1989, pp. 661-664.
26. Tetteroo PAT, Bos MJE, Visser FJ, von dem Borne AEGKr. Neutrophil activation detected by monoclonal antibodies. J Immunol 1986; 136:3427-3432.
27. Thompson LF, Ruedi JM, Low MG, Clement LT. Distribution of
ecto-5'-nucleotidase on subsets of human T and B lymphocytes
as detected by indirect immunofluorescence using goat antibodies. J Immunol 1987;139:4042-4048.
28. Boss GR, Thompson LF, Spiegelberg HL, Pichler WJ, Seegmiller
JE. Age-dependency of lymphocyte ecto-5'-nucleotidase activity. J Immunol 1980; 125:679-682.
27. Plesner T, Hansen NE, Carlsen K. Estimation of Pi-bound proteins on blood cells from PNH patients by quantitative flow
cytometry. Br J Haematol 1990;75:585-590.
30. Hillman P, Hows JM, Luzzatto L. Two distinct patterns of glycosylphosphatidylinositol (GPI) linked protein deficiency in the
red cells of patients with paroxysmal nocturnal haemoglobinuria. Br J Haematol 1992;80:399-405.
31. Dustin ML, Selvaraj P, Mattaliano RJ, Springer TA. Anchoring
mechanism for LFA-3 cell adhesion glycoprotein at membrane
surface. Nature 1987;329:846-848.
32. Schubert J, Alvarado M, Uciechowski P, et al. Diagnosis of paroxysmal nocturnal haemoglobinuria using immunophenotyping
of peripheral blood cells. Br J Haematol 1991;79:487-492.
33. Cooper MR, Currie MS, Rustagi PK, Logue GL. T lymphocytes
escape the membrane defect in paroxysmal nocturnal haemoglobinuria. Br J Haematol 1983;55:263-271.
34. Kinoshita T, Medof ME, Silker R, Nussenzweig V. Distribution of
decay-accelerating factor in peripheral blood of normal individuals and patients with paroxysmal nocturnal hemoglobinuria. J
Exp Med 1985;162:75-92.
35. Selvaraj P, Dustin ML, Silber R, Low MG, Springer TA. Deficiency of lymphocyte function-associated antigen 3 (LFA-3) in
paroxysmal nocturnal hemoglobinuria. J Exp Med 1987;
166:1011-1025.
36. Wilcox LA, Ezzell JL, Bernshaw NJ, Parker CJ. Molecular basis of
the enhanced susceptibility of the erythrocytes of paroxysmal
nocturnal hemoglobinuia to hemolysis in acidified serum.
Blood 1991;78:820-829.
37. Rosse WF, Adams JP, Thorpe AM. The population of cells in
paroxysmal nocturnal haemoglobinuria of intermediate sensitivity to complement lysis: Significance and mechanism of increased immune lysis. Br J Haematol 1974;28:181-190.
38. Shichishima T, Terasawa T, Hashimoto C, et al. Heterogenous
expression of decay accelerating factor and CD59/membrane
attack complex inhibition factor on paroxysmal nocturnal haemoglobinuria (PNH) erythrocytes. Br J Haematol 1991;
78:545-550.
39. Hillmen P, Hows JM, Luzzatto L. Two distinct patterns of glycosylphosphatidylinositol (GPI) linked protein deficiency in the
red cells of patients with paroxysmal nocturnal haemoglobinuria. Br J Haematol 1992;80:399-405.
40. Rosse WF, Hoffman S, Campbell M, et al. The erythrocytes in
paroxysmal nocturnal haemoglobinuria of intermediate sensitivity to complement lysis. Br J Haematol 1991;79:99-107.
41. Rotoli B, Luzzatto L. Paroxysmal nocturnal hemoglobinuria. Semin Hematol 1989;26:201-207.
42. Dameshek W. Riddle: What do aplastic anemia, paroxysmal nocturnal hemoglobinuria (PNH) and "hypoplastic" leukemia
have in common? Blood 1967;30:251-254.
43. Young NS. The problem of clonality in aplastic anemia: Dr. Dameshek's riddle, restated. Blood 1992;79:1385-1392.
Vol. 102-No. 1