Detection of B-lymphocyte(B-cell)-associated Antigens on Human Leukemic Lymphocytes Masking of Membrane Antigens CLEMENT C. S. HSU, M.D., AND ELAINE R. MORGAN, M.D.
S E V E R A L GROUPS OF I N V E S T I G A T O R S have reported preparation of xenogeneic antisera reacting with human leukemic cells in complement-dependent cytotoxicity in v/jro 1,3,11,16,25,27,28 or with the use of the Received June 22, 1977; received revised manuscript August 29, 1977; accepted for publication August 29, 1977. Supported by a grant from the Leukemia Research Foundation, Inc., and in part by NIH grant AI 12085. Address reprint requests to Dr. Hsu: Department of Medicine, Northwestern University Medical School, Ward Memorial Building 303 East Chicago Avenue, Chicago, Illinois 60611.
From the Section of Infectious Diseases-Hypersensitivity, Columbus-Cuneo-Cabrini Department of Medicine, Medical Center, and Section of Hematology, Children's Memorial Hospital, Department of Pediatrics, Northwestern University Medical School, Chicago, Illinois
immunofluorescent technic. 2,5,6,15,29 Many of these antisera2,5,29 have been shown to react with B-cellassociated antigens, which appears to be equivalent to la antigens on mouse B-cells. We have also prepared rabbit antisera (anti-EP) that reacted with human leukemic lymphocytes by complement-dependent cytotoxicity assay. 19 The anti-EP did not react with peripheral blood lymphocytes from normal individuals, with those from patients with nonlymphoproliferative disorders, or with leukemic cells from patients with acute myelogenous leukemia or chronic granulocytic leukemia. In order to characterize further the antigens detected by anti-EP, an indirect immunofluorescent technic was employed to examine the anti-EP reactivity. We found that the antigen detected was probably B-cell-associated antigen, present on both nonleukemic B-cells and leukemic lymphocytes. Further, we found that on leukemic cells from some patients there might be a "blocking components)" that interfered with the expression of membrane antigens on cell surfaces. We report here the results of our study of anti-EP reactivity with the use of the immunofluorescent technic. Materials and Methods Cells The cells examined included peripheral blood lymphocytes from clinically healthy laboratory personnel, circulating leukemic cells, and lymphocytes from patients with various leukemic or nonleukemic disorders (see Tables 2-5) diagnosed by standard clinical criteria, marrow cells from a patient with acute lymphocytic leukemia and a patient with lymphosarcomacell leukemia and lymphocytes isolated from chronically
0002-9173/78/1100/0741 $00.85 © American Society of Clinical Pathologists
741
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Hsu, Clement C. S., and Morgan, Elaine R.: Detection of B-lymphocyte(B-cell)-associated antigens on human leukemic lymphocytes. Masking of membrane antigens. Am J Clin Pathol 70: 741-747, 1978. An indirect immunofluorescent technic has been used for rabbit antisera (anti-EP) that demonstrated complement-dependent cytotoxicity against human leukemic lymphocytes but not against normal blood lymphocytes. With the immunofluorescent technic, the antisera were found to react with 2-23% of normal blood lymphocytes. Simultaneous staining of normal cells with anti-EP and for surface immunoglobulins (SIg) on bone-marrow-derived B-cells showed that the proportions of stained cells were similar to percentages of cells stained by anti-EP alone or for SIg alone. The percentage of anti-EP reactive cells also approximated the percentages of cells reactive to a known antiserum to human B-cell associated, or la-like, antigens. The anti-EP reacted with lymphoblastoid cells from two B-cell lines lacking the Epstein-Barr viral genome. The antigens detected by antiEP probably are B-cell-associated. The anti-EP intensely stained neoplastic cells of acute or chronic lymphocytic leukemia and lymphosarcoma-cell leukemia. Cells from two patients with chronic lymphocytic leukemia and from two patients with acute lymphocytic leukemia showed increased intensity of anti-EP staining and/or increased proportions of stained cells following overnight incubation in culture medium, compared with the preincubation samples. This observation suggests the presence of a "blocking component(s)" on cell surfaces, which interfered with anti-EP reactivity. After overnight incubation, the component might have been removed from the antigenic sites on cell surfaces. Further studies of leukemic lymphocytes using anti-EP for the cell-bound "blocking component" may reveal important pathogenetic mechanisms. (Key words: B-cell-associated antigens; Leukemia-associated antigens; Blocking component; Lymphocytic leukemias; la-like antigens.)
A.J.C.P. • November 1978
HSU AND MORGAN
742
Table I. Percentages of Normal Blood Lymphocytes Reactive with Anti-human Ig, anti-EP, or the Mixture of Anti-Ig and Anti-EP* Percentage Cells Reactive to Anti-Ig
Anti-EP
(%)
(%)
(%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
10
21 12 9 23 Br 12 8 8 9 5 4F 14 12 17 2 9K
ND ND ND ND ND ND ND ND ND ND ND ND ND 8 15 11 12 20 8 11 16
gBr
12 26Br 10 26 7 10 8 12 16Br 7 18Br 7 14 Il B r gBr
21 8 13 13
I3Br
13 17 4 5F 14
Anti-EP +
* The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint; ND = not done.
infected, surgically resected tonsils. Three B-lymphoid cell lines, TRU, which was established at our medical center and contains Epstein-Barr viral genome,19 and BJAB and RAMOS, which were provided by Dr. G. Klein and contain no Epstein-Barr viral genomes,35 were also examined. All cell samples were centrifuged on a Ficoll-Hypaque density gradient7 to remove erythrocytes, mature myeloid cells, and nonviable cells. Cells obtained were washed twice with Roswell Park Memorial Institute (RPMI) medium 1640 and resuspended in the same medium at 4 x 106 nucleated cells per milliliter. One-tenth milliliter of the cell suspension was then examined to determine the percentages of sheep erythrocyte (E)-rosetting thymusderived (T) lymphocytes (T-cells) and SIg-bearing B-cells as described previously,18 and for indirect immunofluorescent staining with various antisera. Leukocyte counts and differential counts were performed on peripheral blood samples to obtain absolute numbers of circulating lymphoid cells and percentages of lymphoblasts or myeloblasts. Antisera The antiserum used for direct immunofluorescent staining of B-cell SIg was purchased from Meloy
* Meloy Laboratories, Springfield, Virginia. t Miles Laboratories, Kankakee, Illinois.
Table 2. Percentages of T-cells, B-cells and Cells Reactive to Anti-EP in Blood Lymphocytes from Eleven Patients with Various Nonleukemic Diseases, or in Lymphocytes of Chronically Infected Tonsils*
Cell Donors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Diagnosis Euthyroid Graves' disease Iatrogenic hyperthyroidism Vogt-Koyanagi-Harada syndrome Stevens-Johnson syndrome Polymyositis Guillian-Barre' syndrome IgG paraproteinemia Hypogammaglobulinemia Hypogammaglobulinemia Hypogammaglobulinemia ?Sjogren's syndrome Tonsil Tonsil Tonsil Tonsil
T-cells (%)
B-cells (%)
Anti-EP Reactive Cells (%)
49 76
16 26
14Br 17
60 57 58 68 71 78 83 95 64 26 31 26 18
34Br
28Br
5 CJBr
3 6 0 0 0 6 24 49 52 42
24 16 6 3 23 2 0 8 64 49Br
51 18F
* The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Case No.
Laboratories.* It was fluorescein isothiocyanate(FITC)conjugated polyvalent antihuman Ig with specificity for gamma, alpha, mu, kappa, and lambda chains. An FITC-conjugated goat antirabbit 7S-Ig was also purchased from the Meloy Laboratories. The goat antirabbit 7S-Ig purchased was further absorbed with human Cohn fraction l i t (20 mg/ml) to remove antibodies cross-reactive with human Ig. The anti-EP used in the indirect immunofluorescent technic was prepared as detailed previously19 by immunizing two rabbits with leukemic lymphosarcoma cells from a patient (EP) whose leukemic cells carried both T- and B-cell markers.20,22 The antisera were absorbed with leukocytes from 16 healthy donors. A 1:64 dilution of this absorbed anti-EP was used in the study unless otherwise specified. The absorbed antiEP was ultracentrifuged at 100,000 x g for 90 minutes after the completion of absorption procedures to remove antigen-antibody complexes. An antiserum to human B-cell-associated antigen (anti-B-cell-associated antigen) prepared in rabbits by immunization with the antigen isolated from normal B-cells32 was supplied by Dr. M. F. Greaves.
Vol. 70 • No. 5
MEMBRANE ANTIGENS ON LEUKEMIC LYMPHOCYTES
Indirect Immunofluorescent
Staining
Examination of Fluorescence-stained
Cells
The prepared slides were examined within 24 hours under an ultraviolet microscope equipped with phasecontrast lenses, as described previously. 18 Two hundred lymphocytes, or nucleated cells in the case of marrow, acute myelogenous leukemia or chronic granulocytic leukemia cells, were counted, and proportions of fluorescence-stained cells were determined. Monocytes contaminated the density-gradient-separated cell suspensions, especially those from normal donors. The monocyte: lymphocyte ratios in suspensions from normal donors ranged from 0.02 to 0.78 (mean ± SD: 0.20 ± 0.14). More than 95% of cells in the suspension could be distinguished as lymphocytes or monocytes under phase-contrast lenses at x800 magnification. In density-gradient-separated chronic lymphocytic leukemia, acute lymphocytic leukemia, or acute myelogenous leukemia cells, the numbers of monocytes in the suspensions were usually negligible. Intensity of fluorescence was also recorded as " B r " for bright or " F " for faint staining. In several cases, the intensity of fluorescence was further examined by inserting a 10% light-transmitting filter in the light path of the ultraviolet microscope, and 200 more cells were counted for the fluorescent-staining cells. Intensely stainable cells
Table 3. Percentages of Blood Lymphocytes of Leukemic Cells from Various Patients and of Cells from a B-Lymphoblastoid Cell Line Reactive with Anti-Ig, Anti-EP, or Anti-B-cellAssociated Antigen Percentage of Cells Reactive to Diagnosis
(%)
(%)
Normal Stevens-Johnson syndrome Hypogammaglobulinemia Hypogammaglobulinemia T-cell lymphoma (LSLt) ALL§ AMLH B-cell line (TRU)
13
14
12
5 0 0 0 2 0 72
24 23 2 0 66K 59
20 23 1 0 45KK 35
100Br
(%)
98Br
* Supplied by Dr. M. F. Greaves. t The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint. $ LSL = lymphosarcoma-cell leukemia. § ALL = acute lymphocytic leukemia. " AML = acute myelogenous leukemia. ** BAA = B-cell-associated antigen.
remained fluorescence-positive, while the faint-staining cells became negative. Results Using the indirect immunofluorescent technic, antiEP was consistently found to stain a small fraction of normal lymphocytes. The reactivity of normal blood lymphocytes to polyvalent anti-Ig and to anti-EP was compared in samples from 21 normal individuals (Table 1). The proportions of SIg-bearing B-cells and anti-EP reactive cells were similar in most cases. Simultaneous staining of cells for SIg and with anti-EP in eight of them revealed the total proportion of stained cells to be similar to the percentage of B-cells or the percentage of anti-EP reactive cells when stained separately for each of these. Simultaneous E-rosetting and anti-EP staining in three blood samples also revealed that anti-EP reacted with the non-rosetting cells. In examination of blood lymphocytes from 11 patients who had nonneoplastic disorders and tonsillar cells from four donors (Table 2), the percentage of antiEP-reactive cells was generally found to parallel the percentage of SIg-bearing cells, although clear-cut disparity was observed (Donor 8). The above results suggested that anti-EP was reacting with normal SIgbearing B-cells and not Ig determinants. The percentage of anti-EP-reactive cells from var-
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
One-tenth milliliter of cell suspensions was incubated with 20 ytxl anti-EP, or anti-B-cell-associated antigen, at 4 C for 30 minutes and then at 37 C for 10 minutes. The cells were washed twice with 2 ml per wash of RPMI medium 1640 and resuspended in 0.1 ml of the same medium. Twenty ii\ of the FITC-conjugated goat antirabbit 7S-Ig was added to the suspension, which was then incubated at 4 C for 30 minutes. The cells were then washed twice, mounted on a glass slide, covered with a cover glass, and sealed with nail polish. To examine for the possibility of nonspecific attachment of goat antirabbit 7S-Ig or rabbit Ig to cell surfaces, controls were done simultaneously for most cell samples by incubating cells with FITC-conjugated goat antirabbit 7S-Ig alone (Control 1) and by incubating cells with normal rabbit serum followed by washing and staining with the FITC-conjugated goat antirabbit 7S-Ig (Control 2). In the latter part of the experiments the normal rabbit serum was absorbed with tonsillar cells and ultracentrifuged. No difference in results was seen with or without the absorption procedure. With a few exceptions, both Control 1 and Control 2 showed no fluorescent staining.
743
A.J.C.P. • November 1978
HSU AND MORGAN
744
detected by anti-EP were probably B-cell-associated antigens. Some of the monocytes in lymphocyte suspensions were also found to react with the antiserum. But a typical fluorescent capping phenomenon, as in SIg staining,18 was not observed. Reactivity of anti-EP to blood or marrow cells from patients with various leukemias or lymphomas other than acute lymphocytic leukemia are shown in Table 4 (Patients 1-7). Anti-EP stained the neoplastic lymphocytes in all specimens except that from Patient 5, whose circulating lymphocytes were probably neoplastic T-cells with scanty E-receptors (79% of the patient's cells bound at least one erythrocyte). The reactivity of anti-EP with nonlymphoid leukemic cells (Patients 8-16, Table 4) was difficult to assess, because of nonspecific staining observed in some specimens (Patients 10 and 15). It was clear, however, that acute myelogenous leukemia cells from Patient 8 did
Table 4. Blood Leukocyte Count and Percentages of Lymphocytes (L) and Lymphoblast Cells (Blasts) in the Blood, and of T-Cells, B-Cells and Cells Reactive to Anti-EP in Blood Leukemic Cells from Patients with CLL, LSL, AML, AMML, AUL, and CGL*tt 1Peripheral Blood Patient
Diagnosis
Density-gradient-separated Cells
Leukocyte (xlO 3 )
L(%)
Blasts (%)
T-cells (%)
B-cells (%)
Anti-EPReactive Cells (%)
1
CLL
19.8
82
0
18
79
92Br
2
CLL CLL
30.0 60.1
99 97
0 0
11 7
97 96
97Br
3
CLL
16.4
81
0
23
76
26
4
LSL LSL LSL
5.5 ND 4.2
36 ND 47
0 0 0
31 35 33
76 48 10F
59Br
LSL (T-cell lymphoma)
28.6
65
0
53
0
0
(Marrow)
96
96
4
90
65 Br
5.9
50
0
63
28
27
5 6 7
LSL (American Burkitt) American Burkitt lymphoma
93 F
75Br
53Br
8
AML
14.0
14
54
7
0
0
9
AML
14.7
20
78
ND
0
59
10§
AML
15.0
22
26
3
59"
83
11
AML (in remission)
6.3
52
0
67
0
25F
26.0
67
20
50
15
12
AMML
F
22Br
18F
13
AMML
5.6
14
70
1
100
14
AUL
3.9
10
38
3
0
3
15§
CGL
31.0
19
16
14
51
54
* CLL = chronic lymphocytic leukemia; LSL = lymphosarcoma-cel) leukemia; AML = acute myelogenous leukemia; AMML = acute myelomonocytic leukemia; CGL = chronic granulocytic leukemia; AUL = acute unclassifiable leukemia. t The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint; ND = not done.
% No leukemic cells were seen in the blood of Patient 7. Tests were repeated for Patients 2 and 4. § Control slide 2 of Patients 10 and IS showed 16% and 85% fluorescence-positive cells, respectively.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
ious sources was then compared with the proportion of cells reactive to a known antiserum to B-cell-associated antigen and with the percentage of SIg-bearing cells (Table 3). In many instances there was a disparity between the percentage of anti-B-cell-associated antigenreactive cells and the percentage of SIg-bearing cells. In all of these cases the percentage of anti-EP-reactive cells tended to parallel the percentage of anti-B-cellassociated antigen-reactive cells, although the intensities of fluorescence varied. In Case 1 (normal donor) simultaneous staining of cells with anti-EP and antiB-cell-associated antigen yielded 16%fluorescentcells, which is comparable to the 14% anti-EP-reactive cells or 12% anti-B-cell-associated antigen-reactive cells. Furthermore, two B-cell lines lacking the Epstein-Barr viral genome, BJAB and RAMOS, were also stained with anti-EP, with 78% and 99% reactivity, respectively. All of these results suggest that the antigens
Vol. 70 • No. 5
MEMBRANE ANTIGENS ON LEUKEMIC LYMPHOCYTES
cells, and acute lymphocytic leukemia cells also abolished anti-EP reactivity towards normal and acute lymphocytic leukemia cell values. It should be noted that the anti-EP staining of the leukemic lymphocytes, including chronic lymphocytic leukemia cells and lymphosarcoma-cell leukemia cells, was far more intense than that of brightly-staining normal B-cells. However, in several cases anti-EP reacted only weakly or reacted with only small fractions of leukemic cells. Four of these cell samples, two from chronic lymphocytic leukemia (Patients 2 and 3) and two from acute lymphocytic leukemia (Patients 17 and 18), were re-examined after the cells were incubated overnight at room temperature in RPMI medium 1640. The cells were found to react intensely with antiEP and to produce clearly higher proportions of reactive cells, or both (Table 6). To cells from Patient 18, a 10% light-transmitting filter was applied, and the fluorescent cells were counted again. The fluorescencepositive cells increased from 15 to 50% after the incubation. The increased reactivity was not due to selective
Table 5. Blood Leukocyte Count and Percentages of Lymphocytes (L) and Lymphoblasts (Blasts) in the Blood, and of T-Cells, B-Cells and Cells Reactive to Anti-EP in Blood Leukemic Cells or Bone Marrow Cells from 18 Patients with Acute Lymphocytic Leukemia* Density-gradient-separated Cells
Peripheral Blood
Patient 1
Leukocytes (xlO 3 )
L(%)
Blasts (%)
T-cells (%)
B-cells (%)
Anti-EP Reactive Cells (%)
8.0 (Marrow)
46 0
41 96
24 11
3 0
68Br 97 nr
2
12.7
50
49
14
4
86Br
3
264.0
3
96
3
1
97Br
4
18.6
8
84
13
1
95Br
5
97.0
23
75
8
1
93 Br
6
16.0
35
61
10
2
90Br
7
4.9
54
43
3
4
94"r
8t
8.2
14
68
4
0
94"'
9
97.1
1
98
0
2
66K
10
2.3
68
18
40
27
11
3.4
54
2
38
8
25
12t
6.8
39
44
16
0
80F
13
5.2
67
6
37
0
12
14
5.6 21.6
21 14
0 4
ND 34
ND 11
4 12
15
27.9
19
73
11
0
16
13.5
51
48
12
5
10F
17t
24.0
78
20
5
0
40K
18
45.9
40
56
11
1
71 f "
t
* The intensity of the immunofluorescenl staining is indicated by "Br" for bright and "F" for faint; NO = not done. t Patients in relapse.
53 Br
93Br
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
not react with anti-EP, whereas the acute myelogenous leukemia cells in Patient 9 did. Table 5 shows results from 18 patients with acute lymphocytic leukemia. The acute lymphocytic leukemia cells of these cases appeared to be null cells without T- or B-cell markers, i.e., E receptors or SIg. Anti-EP intensely stained proportions of cells that approximated the percentages of cells not forming E rosette. To demonstrate that the antigens on normal blood B-cells were cross-reactive with those on tonsillar and leukemic lymphocytes, 3 /xl of absorbed anti-EP(190 (JL\ of 1:64 dilution) were further absorbed with 108 normal blood lymphocytes, i.e., approximately four times more lymphocytes than were originally used for absorption. The anti-EP had been shown to react only with SIg-bearing B-cells in these normal blood lymphocytes. This absorbed anti-EP failed to stain tonsillar, acute lymphocytic leukemia or chronic lymphocytic leukemia cells, each obtained from two donors. Absorption of a sample of anti-EP with similar proportions of tonisllar cells, chronic lymphocytic leukemia
745
HSU AND MORGAN
746
Table 6. Proportions of Anti-EP-Reactive Cells from Two CLL (2 and 3) and Two ALL (17 and 18) Patients before and after the Overnight Incubation*ft Patient 2 Before incubation After incubation Cell recovery after incubation
93 F 98 Br 92%
Patient 3
Patient 17
Patient 18
26
40 77
71 F (15%) 84 (50%)
70%
72%
71
Br
98%
• CLL = chronic lymphocytic leukemia; ALL = acute lymphocytic leukemia. t The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint. t Numbers in parentheses represent percentages of fluorescence-stained cells after a 10% light-transmitting Alter was interposed in the light path of the ultraviolet microscope.
Discussion Our results suggested that the antigens detected by anti-EP were probably B-cell-associated antigens since the antiserum reacted with normal blood B-cells, neoplastic B-cells, and neoplastic null cells, all of which are known to carry B-cell-associated antigens.13 The relationship of our B-cell-associated antigen to B-cell alloantigens described by others41213,24,26,32 is unclear. Since a fraction of nearly all our nonleukemic donors' cells reacted with the anti-EP, the antigens detected by this antiserum could be the cross-reactive antigen of all B-cell alloantigens.26 Because of the nonspecific attachment of rabbit Ig to cells of myelomonocytic series, which occurred quite frequently (Table 4), we could not ascertain any association of B-cell-associated antigens with chronic granulocytic leukemia, acute myelogenous leukemia, or acute myelomonocytic leukemia cells. Others have reported that B-cell-associated antigens may be present on cells of myelomonocytic series and on concanavallin-A stimulated T lymphoblasts.4,12,13,24,26,33 We also observed more intense immunofluorescent staining of leukemic lymphocytes compared with normal lymphocytes, indicating the presence of a larger quantity of B-cell-associated antigens on leukemic cell surfaces. This finding could explain the lack of anti-EP reactivity with normal cells when using the less sensitive cytotoxicity assay.19 It has generally been assumed that la-like antigens are responsible for the mitogenic activity in the mixed lymphocyte reaction.12 Studies by Tsukimoto and
associates31 revealed that neoplastic T lymphoblasts had less mitogenic effect on allogeneic lymphocytes than the neoplastic null-lymphoblasts. Our finding that B-cell-associated antigens was not detectable on neoplastic T-cells from Patient 5 (Table 4) is consistent with the above data. The use of the immunofluorescent technic for detection of lymphocyte surface antigens raises certain questions. In particular, it has been stated that fluoresceinated goat antibodies or rabbit antibodies may aggregate spontaneously and nonspecifically attach to Fc-receptor-bearing lymphocytes.34 Our control slides, however, revealed no cell samples that were stained nonspecifically by goat antirabbit 7S-Ig preabsorbed with human Ig, and only a few samples showed significant attachment of rabbit Ig. A larger Ig aggregate formed by heating or by cross-linking with anti-Ig may be needed for the attachment of aggregates to Fcreceptors on lymphocytes.23 Our observation in two chronic lymphocytic leukemia and two acute lymphocytic leukemia samples that overnight incubation of cells resulted in increased reactivity to anti-EP deserves special attention. One explanation of the finding is that these leukemic cells lacked B-cell-associated antigens in the cell membrane because of inhibition of membrane antigen synthesis in vivo. By incubating the cells in vitro, leukemic lymphocytes synthesized and expressed B-cell-associated antigens on the cell surfaces. Although this possibility cannot be ruled out, we feel that it is unlikely that cells will produce membrane antigens in vitro at room temperature without serum supplementation of the culture medium. Another explanation is that the incubation allowed either rearrangement in the direction of the antigenic molecules or cleavage of peptide bonds of the molecules by membrane protease activity,10 thus making available more antigenic sites. Another alternative explanation, which we favor, is that initially some component(s) on the cell surface sterically blocked anti-EP interaction with surface antigens. By overnight incubation, this "blocking component" may have been removed, either by a membrane conformational change or by shedding,10 thus unmasking antigenic sites that reacted with anti-EP. The cell-bound "blocking component" is not likely to be an anti-leukemic-cell antibody, since, in chronic lymphocytic leukemia cells, staining for SIg also became brighter after the overnight incubation. Also, no SIg was detectable on acute lymphocytic leukemia cells. In support of the observation that leukemia-associated antigens could be masked by a surface component is a report8 stating that after slight trypsinization of rat lymphoma cells, there was a threefold increase in Gross leukemia virusassociated antigenic sites. There have also been reports
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
cell death, because recovery of viable cells after the overnight incubation was 70-98% (Table 6). In chronic lymphocytic leukemia cells, staining for SIg also became more intense. A similar overnight incubation of blood lymphocytes from five nonleukemic individuals produced no appreciable change in the percentages of anti-EP-reactive cells.
A.J.C.P. • November 1978
Vol. 70 • No. 5
MEMBRANE ANTIGENS ON LEUKEMIC LYMPHOCYTES
Acknowledgments. Dr. M. F. Greaves, of University College, London, made suggestions and provided B-cell-associated antigens. Dr. G. Klein, of Karolinska Institutet, Sweden, supplied various cell lines. Dr. Juanita Rivera-Arcilla provided technical assistance.
References 1. Baker MA, Ramachander K, Taub RN: Specificity of heteroantisera to human acute leukemia-antigens J Clin Invest 54:1273-1278, 1974 2. Billing R, Rafizadeh B, Drew I, et al: Human B-lymphocyte antigens expressed by lymphocytic and myelocytic leukemia cells. I. Detection by rabbit antisera. J Exp Med 144: 167-178, 1976 3. Billing R, Terasaki PI: Human leukemia antigen. I. Production and characterization of antisera. J Natl Cancer Inst 53:16351638, 1974 4. Billing RJ, Terasaki PI, Honig R, et al: The absence of B-cell antigen B2 from leukemia cells and lymphoblastoid cell lines. Lancet 1:1365-1367, 1976 5. Billing R, Ting A, Terasaki PI: Human B-lymphocyte antigens expressed by lymphocytic and myelocytic leukemia cells. II. Detection by human anti-B-cell alloantisera. J Natl Cancer Inst 58:199-203, 1977 6. Borella L, Sen L, Casper JT: Acute lymphoblastic leukemia (ALL) antigens detected with antisera to E-rosette-forming and non-E-rosette forming ALL blasts. J Immunol 118: 309-315, 1977 7. Boyum A: Isolation of mononuclear cells and granulocytes from human blood. Scand J Clin Lab Invest 21 (suppl 97):77-89, 1968 8. Brandchaft PB, Boone CW: Increase in Gross (G) antigen sites on the surface of AKR virus-induced rat lymphoma cells after treatment with trypsin. J Immunol 113:94-102, 1974 9. Davis S: Circulating lymphocyte subpopulations in chronic lymphocytic leukemia. New Engl J Med 294:1150-1153, 1976 10. Doljanski F, Kapeller M: Cell surface shedding. The phenomenon and its possible significance. J Theor Biol 62:253-270, 1976
11. Durantez A, Zighelboim J, Gale RP: Leukemia-associated antigens detected by heterologous antisera. J Natl Cancer Inst 56:1217-1219, 1976 12. Editorial: B-lymphocyte alloamigens in man. Lancet 2:240-242, 1976 13. Fu SM, Winchester RJ, Hunkel HG: The occurrence of the HL-B alloantigens on the cells of unclassified acute lymphoblastic leukemia. J Exp Med 142:1334-1337, 1975 14. Gibofsky A, Terasaki PI: Trypsinization of lymphocytes for HL-A typing. Transplantation 13:192-194, 1972 15. Greaves MF, Brown G, Rapson NJ, et al: Antisera to acute lymphoblastic leukemia cells. Clin Immunol Immunopathol 4:67-84, 1975 16. Halterman RH, Leventhal BG, Mann DL: An acute leukemia antigen: Correlation with clinical status. N Eng J Med 287: 1272-1274, 1972 17. Hellstrom KE, Hellstrom I: Immunity to neuroblastomas and melanomas. Annu Rev Med 23:19-38, 1972 18. Hsu CCS, Chen Y, Patterson R: Peripheral blood B-lymphocyte. abnormalities associated with hyperthyroidism of Graves' disease. Clin Exp Immunol 26:431-440, 1976 19. Hsu CCS, Marti GE, Mittal KK: Antisera against leukemiaassociated antigens on human lymphocytes. Clin Exp Immunol 27:487-496, 1977 20. Hsu CCS, Marti GE, Schrek R, et al: Lymphocytes bearing B- and T-cell markers in patients with lymphosarcoma cell leukemia. Clin Immunol Immunopathol 3:385-395, 1975 21. Klein G: Immunological surveillance against neoplasia. Harvey Lect 69:71-102, 1974 22. Lin PS, Hsu CCS: Human leukemic T-cells with complement receptors. Clin Exp Immunol 23:209-213, 1976 23. Linthicum DS, Bahu RM, Sell S: Differences in ultrastructural appearances and modulation of endogenous and exogenous surface immunoglobulin of rabbit lymphocytes. J Immunol 117:1179-1188, 1976 24. Mann DL, Abelson L, Harris S, et al: Detection of antigens specific for B-lymphoid cultured cell lines with human alloantisera. J Exp Med 142:84-89, 1975 25. Mann DL, Halterman R, Leventhal B: Acute leukemia associated antigens. Cancer 34 (suppl): 1446-1451, 1974 26. McDevitt HO, Delovitch TL, Press JL, et al: Genetic and functional analysis of the la antigens: Their possible role in regulating the immune response. Transplant Rev 30:197-235, 1976 27. Metzgar RS, Mohanakumar T, Miller DS: Antigens specific for human lymphocytic and myeloid leukemic cells: detection by non-human primate serum. Science 178:986-988, 1972 28. Mohanakumar T, Metzgar RS, Miller DS: Human leukemia cell antigens: serologic characterization with xenoantisera. J Natl Cancer Inst 52:1435-1444, 1974 29. Schlossman SF, Chess L, Humphreys RE, et al: Distribution of la-like molecules on the surface of normal and leukemic human cells. Proc Natl Acad Sci USA 73:1288-1292, 1976 30. Tarro G: Appearance in trypsinized normal cells of reactivity with antibody presumably specific for malignant cells. Proc Natl Acad Sci USA 70:325-327, 1973 31. Tsukimoto I, Wong KY, Lampkin BC: Surface markers and prognostic factors in acute lymphoblastic leukemia. New Engl J Med 294:245-248, 1976 32. Welsh K, Turner NJ: Preparation of antiserum specific to human B-cells by immunization of rabbits with immune complexes. Tissue Antigen 8:197-205, 1976. 33. Wernet P: Human la-type alloantigens: Methods of detection, aspects of chemistry and biology, markers for disease states. Transplant Rev 30:271-289, 1976. 34. Winchester RJ, Fu SM, Hoffman T, Kunkel HG: IgG on lymphocyte surfaces: technical problems and the significance of a third cell population. J Immunol 114:1210-1212, 1975 35. Yefenof E, Klein G: Difference in antibody induced redistribution of membrane IgM in EBV-genome free and EBV positive human lymphoid cells. Exp Cell Res 99:175-178, 1976
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
that previously undetected cell anitgens may appear after trypsin treatment of the cells.14™ The "blocking component" appears to coat the leukemic cell surfaces nonspecifically, because expression of both B-cell-associated antigens and SIg on chronic lymphocytic leukemia cells was masked. On acute lymphocytic leukemia cells, in addition to B-cellassociated antigens, another membrane antigen detectable with rabbit anti-acute lymphocytic leukemia cells prepared as described by Greaves and colleagues15 has also been found to be masked by the "blocking component" (Hsu and Morgan, in preparation). Theoretically, the presence of such a component might influence the clinical course of the disease by interfering with the host's immunologic recognition of the leukemic cells.17-21 It is relevant that a recent study9 revealed progressively less SIg on chronic lymphocytic leukemia cells as the disease advanced. It would be important to determine whether this decrease in SIg on chronic lymphocytic leukemia cells is due to a "blocking component." The appearance of a cellbound "blocking component" may be a way for leukemic cells to evade the host's immunologic surveillance.
747
S E V E R A L GROUPS OF I N V E S T I G A T O R S have reported preparation of xenogeneic antisera reacting with human leukemic cells in complement-dependent cytotoxicity in v/jro 1,3,11,16,25,27,28 or with the use of the Received June 22, 1977; received revised manuscript August 29, 1977; accepted for publication August 29, 1977. Supported by a grant from the Leukemia Research Foundation, Inc., and in part by NIH grant AI 12085. Address reprint requests to Dr. Hsu: Department of Medicine, Northwestern University Medical School, Ward Memorial Building 303 East Chicago Avenue, Chicago, Illinois 60611.
From the Section of Infectious Diseases-Hypersensitivity, Columbus-Cuneo-Cabrini Department of Medicine, Medical Center, and Section of Hematology, Children's Memorial Hospital, Department of Pediatrics, Northwestern University Medical School, Chicago, Illinois
immunofluorescent technic. 2,5,6,15,29 Many of these antisera2,5,29 have been shown to react with B-cellassociated antigens, which appears to be equivalent to la antigens on mouse B-cells. We have also prepared rabbit antisera (anti-EP) that reacted with human leukemic lymphocytes by complement-dependent cytotoxicity assay. 19 The anti-EP did not react with peripheral blood lymphocytes from normal individuals, with those from patients with nonlymphoproliferative disorders, or with leukemic cells from patients with acute myelogenous leukemia or chronic granulocytic leukemia. In order to characterize further the antigens detected by anti-EP, an indirect immunofluorescent technic was employed to examine the anti-EP reactivity. We found that the antigen detected was probably B-cell-associated antigen, present on both nonleukemic B-cells and leukemic lymphocytes. Further, we found that on leukemic cells from some patients there might be a "blocking components)" that interfered with the expression of membrane antigens on cell surfaces. We report here the results of our study of anti-EP reactivity with the use of the immunofluorescent technic. Materials and Methods Cells The cells examined included peripheral blood lymphocytes from clinically healthy laboratory personnel, circulating leukemic cells, and lymphocytes from patients with various leukemic or nonleukemic disorders (see Tables 2-5) diagnosed by standard clinical criteria, marrow cells from a patient with acute lymphocytic leukemia and a patient with lymphosarcomacell leukemia and lymphocytes isolated from chronically
0002-9173/78/1100/0741 $00.85 © American Society of Clinical Pathologists
741
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Hsu, Clement C. S., and Morgan, Elaine R.: Detection of B-lymphocyte(B-cell)-associated antigens on human leukemic lymphocytes. Masking of membrane antigens. Am J Clin Pathol 70: 741-747, 1978. An indirect immunofluorescent technic has been used for rabbit antisera (anti-EP) that demonstrated complement-dependent cytotoxicity against human leukemic lymphocytes but not against normal blood lymphocytes. With the immunofluorescent technic, the antisera were found to react with 2-23% of normal blood lymphocytes. Simultaneous staining of normal cells with anti-EP and for surface immunoglobulins (SIg) on bone-marrow-derived B-cells showed that the proportions of stained cells were similar to percentages of cells stained by anti-EP alone or for SIg alone. The percentage of anti-EP reactive cells also approximated the percentages of cells reactive to a known antiserum to human B-cell associated, or la-like, antigens. The anti-EP reacted with lymphoblastoid cells from two B-cell lines lacking the Epstein-Barr viral genome. The antigens detected by antiEP probably are B-cell-associated. The anti-EP intensely stained neoplastic cells of acute or chronic lymphocytic leukemia and lymphosarcoma-cell leukemia. Cells from two patients with chronic lymphocytic leukemia and from two patients with acute lymphocytic leukemia showed increased intensity of anti-EP staining and/or increased proportions of stained cells following overnight incubation in culture medium, compared with the preincubation samples. This observation suggests the presence of a "blocking component(s)" on cell surfaces, which interfered with anti-EP reactivity. After overnight incubation, the component might have been removed from the antigenic sites on cell surfaces. Further studies of leukemic lymphocytes using anti-EP for the cell-bound "blocking component" may reveal important pathogenetic mechanisms. (Key words: B-cell-associated antigens; Leukemia-associated antigens; Blocking component; Lymphocytic leukemias; la-like antigens.)
A.J.C.P. • November 1978
HSU AND MORGAN
742
Table I. Percentages of Normal Blood Lymphocytes Reactive with Anti-human Ig, anti-EP, or the Mixture of Anti-Ig and Anti-EP* Percentage Cells Reactive to Anti-Ig
Anti-EP
(%)
(%)
(%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
10
21 12 9 23 Br 12 8 8 9 5 4F 14 12 17 2 9K
ND ND ND ND ND ND ND ND ND ND ND ND ND 8 15 11 12 20 8 11 16
gBr
12 26Br 10 26 7 10 8 12 16Br 7 18Br 7 14 Il B r gBr
21 8 13 13
I3Br
13 17 4 5F 14
Anti-EP +
* The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint; ND = not done.
infected, surgically resected tonsils. Three B-lymphoid cell lines, TRU, which was established at our medical center and contains Epstein-Barr viral genome,19 and BJAB and RAMOS, which were provided by Dr. G. Klein and contain no Epstein-Barr viral genomes,35 were also examined. All cell samples were centrifuged on a Ficoll-Hypaque density gradient7 to remove erythrocytes, mature myeloid cells, and nonviable cells. Cells obtained were washed twice with Roswell Park Memorial Institute (RPMI) medium 1640 and resuspended in the same medium at 4 x 106 nucleated cells per milliliter. One-tenth milliliter of the cell suspension was then examined to determine the percentages of sheep erythrocyte (E)-rosetting thymusderived (T) lymphocytes (T-cells) and SIg-bearing B-cells as described previously,18 and for indirect immunofluorescent staining with various antisera. Leukocyte counts and differential counts were performed on peripheral blood samples to obtain absolute numbers of circulating lymphoid cells and percentages of lymphoblasts or myeloblasts. Antisera The antiserum used for direct immunofluorescent staining of B-cell SIg was purchased from Meloy
* Meloy Laboratories, Springfield, Virginia. t Miles Laboratories, Kankakee, Illinois.
Table 2. Percentages of T-cells, B-cells and Cells Reactive to Anti-EP in Blood Lymphocytes from Eleven Patients with Various Nonleukemic Diseases, or in Lymphocytes of Chronically Infected Tonsils*
Cell Donors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Diagnosis Euthyroid Graves' disease Iatrogenic hyperthyroidism Vogt-Koyanagi-Harada syndrome Stevens-Johnson syndrome Polymyositis Guillian-Barre' syndrome IgG paraproteinemia Hypogammaglobulinemia Hypogammaglobulinemia Hypogammaglobulinemia ?Sjogren's syndrome Tonsil Tonsil Tonsil Tonsil
T-cells (%)
B-cells (%)
Anti-EP Reactive Cells (%)
49 76
16 26
14Br 17
60 57 58 68 71 78 83 95 64 26 31 26 18
34Br
28Br
5 CJBr
3 6 0 0 0 6 24 49 52 42
24 16 6 3 23 2 0 8 64 49Br
51 18F
* The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Case No.
Laboratories.* It was fluorescein isothiocyanate(FITC)conjugated polyvalent antihuman Ig with specificity for gamma, alpha, mu, kappa, and lambda chains. An FITC-conjugated goat antirabbit 7S-Ig was also purchased from the Meloy Laboratories. The goat antirabbit 7S-Ig purchased was further absorbed with human Cohn fraction l i t (20 mg/ml) to remove antibodies cross-reactive with human Ig. The anti-EP used in the indirect immunofluorescent technic was prepared as detailed previously19 by immunizing two rabbits with leukemic lymphosarcoma cells from a patient (EP) whose leukemic cells carried both T- and B-cell markers.20,22 The antisera were absorbed with leukocytes from 16 healthy donors. A 1:64 dilution of this absorbed anti-EP was used in the study unless otherwise specified. The absorbed antiEP was ultracentrifuged at 100,000 x g for 90 minutes after the completion of absorption procedures to remove antigen-antibody complexes. An antiserum to human B-cell-associated antigen (anti-B-cell-associated antigen) prepared in rabbits by immunization with the antigen isolated from normal B-cells32 was supplied by Dr. M. F. Greaves.
Vol. 70 • No. 5
MEMBRANE ANTIGENS ON LEUKEMIC LYMPHOCYTES
Indirect Immunofluorescent
Staining
Examination of Fluorescence-stained
Cells
The prepared slides were examined within 24 hours under an ultraviolet microscope equipped with phasecontrast lenses, as described previously. 18 Two hundred lymphocytes, or nucleated cells in the case of marrow, acute myelogenous leukemia or chronic granulocytic leukemia cells, were counted, and proportions of fluorescence-stained cells were determined. Monocytes contaminated the density-gradient-separated cell suspensions, especially those from normal donors. The monocyte: lymphocyte ratios in suspensions from normal donors ranged from 0.02 to 0.78 (mean ± SD: 0.20 ± 0.14). More than 95% of cells in the suspension could be distinguished as lymphocytes or monocytes under phase-contrast lenses at x800 magnification. In density-gradient-separated chronic lymphocytic leukemia, acute lymphocytic leukemia, or acute myelogenous leukemia cells, the numbers of monocytes in the suspensions were usually negligible. Intensity of fluorescence was also recorded as " B r " for bright or " F " for faint staining. In several cases, the intensity of fluorescence was further examined by inserting a 10% light-transmitting filter in the light path of the ultraviolet microscope, and 200 more cells were counted for the fluorescent-staining cells. Intensely stainable cells
Table 3. Percentages of Blood Lymphocytes of Leukemic Cells from Various Patients and of Cells from a B-Lymphoblastoid Cell Line Reactive with Anti-Ig, Anti-EP, or Anti-B-cellAssociated Antigen Percentage of Cells Reactive to Diagnosis
(%)
(%)
Normal Stevens-Johnson syndrome Hypogammaglobulinemia Hypogammaglobulinemia T-cell lymphoma (LSLt) ALL§ AMLH B-cell line (TRU)
13
14
12
5 0 0 0 2 0 72
24 23 2 0 66K 59
20 23 1 0 45KK 35
100Br
(%)
98Br
* Supplied by Dr. M. F. Greaves. t The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint. $ LSL = lymphosarcoma-cell leukemia. § ALL = acute lymphocytic leukemia. " AML = acute myelogenous leukemia. ** BAA = B-cell-associated antigen.
remained fluorescence-positive, while the faint-staining cells became negative. Results Using the indirect immunofluorescent technic, antiEP was consistently found to stain a small fraction of normal lymphocytes. The reactivity of normal blood lymphocytes to polyvalent anti-Ig and to anti-EP was compared in samples from 21 normal individuals (Table 1). The proportions of SIg-bearing B-cells and anti-EP reactive cells were similar in most cases. Simultaneous staining of cells for SIg and with anti-EP in eight of them revealed the total proportion of stained cells to be similar to the percentage of B-cells or the percentage of anti-EP reactive cells when stained separately for each of these. Simultaneous E-rosetting and anti-EP staining in three blood samples also revealed that anti-EP reacted with the non-rosetting cells. In examination of blood lymphocytes from 11 patients who had nonneoplastic disorders and tonsillar cells from four donors (Table 2), the percentage of antiEP-reactive cells was generally found to parallel the percentage of SIg-bearing cells, although clear-cut disparity was observed (Donor 8). The above results suggested that anti-EP was reacting with normal SIgbearing B-cells and not Ig determinants. The percentage of anti-EP-reactive cells from var-
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
One-tenth milliliter of cell suspensions was incubated with 20 ytxl anti-EP, or anti-B-cell-associated antigen, at 4 C for 30 minutes and then at 37 C for 10 minutes. The cells were washed twice with 2 ml per wash of RPMI medium 1640 and resuspended in 0.1 ml of the same medium. Twenty ii\ of the FITC-conjugated goat antirabbit 7S-Ig was added to the suspension, which was then incubated at 4 C for 30 minutes. The cells were then washed twice, mounted on a glass slide, covered with a cover glass, and sealed with nail polish. To examine for the possibility of nonspecific attachment of goat antirabbit 7S-Ig or rabbit Ig to cell surfaces, controls were done simultaneously for most cell samples by incubating cells with FITC-conjugated goat antirabbit 7S-Ig alone (Control 1) and by incubating cells with normal rabbit serum followed by washing and staining with the FITC-conjugated goat antirabbit 7S-Ig (Control 2). In the latter part of the experiments the normal rabbit serum was absorbed with tonsillar cells and ultracentrifuged. No difference in results was seen with or without the absorption procedure. With a few exceptions, both Control 1 and Control 2 showed no fluorescent staining.
743
A.J.C.P. • November 1978
HSU AND MORGAN
744
detected by anti-EP were probably B-cell-associated antigens. Some of the monocytes in lymphocyte suspensions were also found to react with the antiserum. But a typical fluorescent capping phenomenon, as in SIg staining,18 was not observed. Reactivity of anti-EP to blood or marrow cells from patients with various leukemias or lymphomas other than acute lymphocytic leukemia are shown in Table 4 (Patients 1-7). Anti-EP stained the neoplastic lymphocytes in all specimens except that from Patient 5, whose circulating lymphocytes were probably neoplastic T-cells with scanty E-receptors (79% of the patient's cells bound at least one erythrocyte). The reactivity of anti-EP with nonlymphoid leukemic cells (Patients 8-16, Table 4) was difficult to assess, because of nonspecific staining observed in some specimens (Patients 10 and 15). It was clear, however, that acute myelogenous leukemia cells from Patient 8 did
Table 4. Blood Leukocyte Count and Percentages of Lymphocytes (L) and Lymphoblast Cells (Blasts) in the Blood, and of T-Cells, B-Cells and Cells Reactive to Anti-EP in Blood Leukemic Cells from Patients with CLL, LSL, AML, AMML, AUL, and CGL*tt 1Peripheral Blood Patient
Diagnosis
Density-gradient-separated Cells
Leukocyte (xlO 3 )
L(%)
Blasts (%)
T-cells (%)
B-cells (%)
Anti-EPReactive Cells (%)
1
CLL
19.8
82
0
18
79
92Br
2
CLL CLL
30.0 60.1
99 97
0 0
11 7
97 96
97Br
3
CLL
16.4
81
0
23
76
26
4
LSL LSL LSL
5.5 ND 4.2
36 ND 47
0 0 0
31 35 33
76 48 10F
59Br
LSL (T-cell lymphoma)
28.6
65
0
53
0
0
(Marrow)
96
96
4
90
65 Br
5.9
50
0
63
28
27
5 6 7
LSL (American Burkitt) American Burkitt lymphoma
93 F
75Br
53Br
8
AML
14.0
14
54
7
0
0
9
AML
14.7
20
78
ND
0
59
10§
AML
15.0
22
26
3
59"
83
11
AML (in remission)
6.3
52
0
67
0
25F
26.0
67
20
50
15
12
AMML
F
22Br
18F
13
AMML
5.6
14
70
1
100
14
AUL
3.9
10
38
3
0
3
15§
CGL
31.0
19
16
14
51
54
* CLL = chronic lymphocytic leukemia; LSL = lymphosarcoma-cel) leukemia; AML = acute myelogenous leukemia; AMML = acute myelomonocytic leukemia; CGL = chronic granulocytic leukemia; AUL = acute unclassifiable leukemia. t The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint; ND = not done.
% No leukemic cells were seen in the blood of Patient 7. Tests were repeated for Patients 2 and 4. § Control slide 2 of Patients 10 and IS showed 16% and 85% fluorescence-positive cells, respectively.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
ious sources was then compared with the proportion of cells reactive to a known antiserum to B-cell-associated antigen and with the percentage of SIg-bearing cells (Table 3). In many instances there was a disparity between the percentage of anti-B-cell-associated antigenreactive cells and the percentage of SIg-bearing cells. In all of these cases the percentage of anti-EP-reactive cells tended to parallel the percentage of anti-B-cellassociated antigen-reactive cells, although the intensities of fluorescence varied. In Case 1 (normal donor) simultaneous staining of cells with anti-EP and antiB-cell-associated antigen yielded 16%fluorescentcells, which is comparable to the 14% anti-EP-reactive cells or 12% anti-B-cell-associated antigen-reactive cells. Furthermore, two B-cell lines lacking the Epstein-Barr viral genome, BJAB and RAMOS, were also stained with anti-EP, with 78% and 99% reactivity, respectively. All of these results suggest that the antigens
Vol. 70 • No. 5
MEMBRANE ANTIGENS ON LEUKEMIC LYMPHOCYTES
cells, and acute lymphocytic leukemia cells also abolished anti-EP reactivity towards normal and acute lymphocytic leukemia cell values. It should be noted that the anti-EP staining of the leukemic lymphocytes, including chronic lymphocytic leukemia cells and lymphosarcoma-cell leukemia cells, was far more intense than that of brightly-staining normal B-cells. However, in several cases anti-EP reacted only weakly or reacted with only small fractions of leukemic cells. Four of these cell samples, two from chronic lymphocytic leukemia (Patients 2 and 3) and two from acute lymphocytic leukemia (Patients 17 and 18), were re-examined after the cells were incubated overnight at room temperature in RPMI medium 1640. The cells were found to react intensely with antiEP and to produce clearly higher proportions of reactive cells, or both (Table 6). To cells from Patient 18, a 10% light-transmitting filter was applied, and the fluorescent cells were counted again. The fluorescencepositive cells increased from 15 to 50% after the incubation. The increased reactivity was not due to selective
Table 5. Blood Leukocyte Count and Percentages of Lymphocytes (L) and Lymphoblasts (Blasts) in the Blood, and of T-Cells, B-Cells and Cells Reactive to Anti-EP in Blood Leukemic Cells or Bone Marrow Cells from 18 Patients with Acute Lymphocytic Leukemia* Density-gradient-separated Cells
Peripheral Blood
Patient 1
Leukocytes (xlO 3 )
L(%)
Blasts (%)
T-cells (%)
B-cells (%)
Anti-EP Reactive Cells (%)
8.0 (Marrow)
46 0
41 96
24 11
3 0
68Br 97 nr
2
12.7
50
49
14
4
86Br
3
264.0
3
96
3
1
97Br
4
18.6
8
84
13
1
95Br
5
97.0
23
75
8
1
93 Br
6
16.0
35
61
10
2
90Br
7
4.9
54
43
3
4
94"r
8t
8.2
14
68
4
0
94"'
9
97.1
1
98
0
2
66K
10
2.3
68
18
40
27
11
3.4
54
2
38
8
25
12t
6.8
39
44
16
0
80F
13
5.2
67
6
37
0
12
14
5.6 21.6
21 14
0 4
ND 34
ND 11
4 12
15
27.9
19
73
11
0
16
13.5
51
48
12
5
10F
17t
24.0
78
20
5
0
40K
18
45.9
40
56
11
1
71 f "
t
* The intensity of the immunofluorescenl staining is indicated by "Br" for bright and "F" for faint; NO = not done. t Patients in relapse.
53 Br
93Br
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
not react with anti-EP, whereas the acute myelogenous leukemia cells in Patient 9 did. Table 5 shows results from 18 patients with acute lymphocytic leukemia. The acute lymphocytic leukemia cells of these cases appeared to be null cells without T- or B-cell markers, i.e., E receptors or SIg. Anti-EP intensely stained proportions of cells that approximated the percentages of cells not forming E rosette. To demonstrate that the antigens on normal blood B-cells were cross-reactive with those on tonsillar and leukemic lymphocytes, 3 /xl of absorbed anti-EP(190 (JL\ of 1:64 dilution) were further absorbed with 108 normal blood lymphocytes, i.e., approximately four times more lymphocytes than were originally used for absorption. The anti-EP had been shown to react only with SIg-bearing B-cells in these normal blood lymphocytes. This absorbed anti-EP failed to stain tonsillar, acute lymphocytic leukemia or chronic lymphocytic leukemia cells, each obtained from two donors. Absorption of a sample of anti-EP with similar proportions of tonisllar cells, chronic lymphocytic leukemia
745
HSU AND MORGAN
746
Table 6. Proportions of Anti-EP-Reactive Cells from Two CLL (2 and 3) and Two ALL (17 and 18) Patients before and after the Overnight Incubation*ft Patient 2 Before incubation After incubation Cell recovery after incubation
93 F 98 Br 92%
Patient 3
Patient 17
Patient 18
26
40 77
71 F (15%) 84 (50%)
70%
72%
71
Br
98%
• CLL = chronic lymphocytic leukemia; ALL = acute lymphocytic leukemia. t The intensity of the immunofluorescent staining is indicated by "Br" for bright and "F" for faint. t Numbers in parentheses represent percentages of fluorescence-stained cells after a 10% light-transmitting Alter was interposed in the light path of the ultraviolet microscope.
Discussion Our results suggested that the antigens detected by anti-EP were probably B-cell-associated antigens since the antiserum reacted with normal blood B-cells, neoplastic B-cells, and neoplastic null cells, all of which are known to carry B-cell-associated antigens.13 The relationship of our B-cell-associated antigen to B-cell alloantigens described by others41213,24,26,32 is unclear. Since a fraction of nearly all our nonleukemic donors' cells reacted with the anti-EP, the antigens detected by this antiserum could be the cross-reactive antigen of all B-cell alloantigens.26 Because of the nonspecific attachment of rabbit Ig to cells of myelomonocytic series, which occurred quite frequently (Table 4), we could not ascertain any association of B-cell-associated antigens with chronic granulocytic leukemia, acute myelogenous leukemia, or acute myelomonocytic leukemia cells. Others have reported that B-cell-associated antigens may be present on cells of myelomonocytic series and on concanavallin-A stimulated T lymphoblasts.4,12,13,24,26,33 We also observed more intense immunofluorescent staining of leukemic lymphocytes compared with normal lymphocytes, indicating the presence of a larger quantity of B-cell-associated antigens on leukemic cell surfaces. This finding could explain the lack of anti-EP reactivity with normal cells when using the less sensitive cytotoxicity assay.19 It has generally been assumed that la-like antigens are responsible for the mitogenic activity in the mixed lymphocyte reaction.12 Studies by Tsukimoto and
associates31 revealed that neoplastic T lymphoblasts had less mitogenic effect on allogeneic lymphocytes than the neoplastic null-lymphoblasts. Our finding that B-cell-associated antigens was not detectable on neoplastic T-cells from Patient 5 (Table 4) is consistent with the above data. The use of the immunofluorescent technic for detection of lymphocyte surface antigens raises certain questions. In particular, it has been stated that fluoresceinated goat antibodies or rabbit antibodies may aggregate spontaneously and nonspecifically attach to Fc-receptor-bearing lymphocytes.34 Our control slides, however, revealed no cell samples that were stained nonspecifically by goat antirabbit 7S-Ig preabsorbed with human Ig, and only a few samples showed significant attachment of rabbit Ig. A larger Ig aggregate formed by heating or by cross-linking with anti-Ig may be needed for the attachment of aggregates to Fcreceptors on lymphocytes.23 Our observation in two chronic lymphocytic leukemia and two acute lymphocytic leukemia samples that overnight incubation of cells resulted in increased reactivity to anti-EP deserves special attention. One explanation of the finding is that these leukemic cells lacked B-cell-associated antigens in the cell membrane because of inhibition of membrane antigen synthesis in vivo. By incubating the cells in vitro, leukemic lymphocytes synthesized and expressed B-cell-associated antigens on the cell surfaces. Although this possibility cannot be ruled out, we feel that it is unlikely that cells will produce membrane antigens in vitro at room temperature without serum supplementation of the culture medium. Another explanation is that the incubation allowed either rearrangement in the direction of the antigenic molecules or cleavage of peptide bonds of the molecules by membrane protease activity,10 thus making available more antigenic sites. Another alternative explanation, which we favor, is that initially some component(s) on the cell surface sterically blocked anti-EP interaction with surface antigens. By overnight incubation, this "blocking component" may have been removed, either by a membrane conformational change or by shedding,10 thus unmasking antigenic sites that reacted with anti-EP. The cell-bound "blocking component" is not likely to be an anti-leukemic-cell antibody, since, in chronic lymphocytic leukemia cells, staining for SIg also became brighter after the overnight incubation. Also, no SIg was detectable on acute lymphocytic leukemia cells. In support of the observation that leukemia-associated antigens could be masked by a surface component is a report8 stating that after slight trypsinization of rat lymphoma cells, there was a threefold increase in Gross leukemia virusassociated antigenic sites. There have also been reports
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
cell death, because recovery of viable cells after the overnight incubation was 70-98% (Table 6). In chronic lymphocytic leukemia cells, staining for SIg also became more intense. A similar overnight incubation of blood lymphocytes from five nonleukemic individuals produced no appreciable change in the percentages of anti-EP-reactive cells.
A.J.C.P. • November 1978
Vol. 70 • No. 5
MEMBRANE ANTIGENS ON LEUKEMIC LYMPHOCYTES
Acknowledgments. Dr. M. F. Greaves, of University College, London, made suggestions and provided B-cell-associated antigens. Dr. G. Klein, of Karolinska Institutet, Sweden, supplied various cell lines. Dr. Juanita Rivera-Arcilla provided technical assistance.
References 1. Baker MA, Ramachander K, Taub RN: Specificity of heteroantisera to human acute leukemia-antigens J Clin Invest 54:1273-1278, 1974 2. Billing R, Rafizadeh B, Drew I, et al: Human B-lymphocyte antigens expressed by lymphocytic and myelocytic leukemia cells. I. Detection by rabbit antisera. J Exp Med 144: 167-178, 1976 3. Billing R, Terasaki PI: Human leukemia antigen. I. Production and characterization of antisera. J Natl Cancer Inst 53:16351638, 1974 4. Billing RJ, Terasaki PI, Honig R, et al: The absence of B-cell antigen B2 from leukemia cells and lymphoblastoid cell lines. Lancet 1:1365-1367, 1976 5. Billing R, Ting A, Terasaki PI: Human B-lymphocyte antigens expressed by lymphocytic and myelocytic leukemia cells. II. Detection by human anti-B-cell alloantisera. J Natl Cancer Inst 58:199-203, 1977 6. Borella L, Sen L, Casper JT: Acute lymphoblastic leukemia (ALL) antigens detected with antisera to E-rosette-forming and non-E-rosette forming ALL blasts. J Immunol 118: 309-315, 1977 7. Boyum A: Isolation of mononuclear cells and granulocytes from human blood. Scand J Clin Lab Invest 21 (suppl 97):77-89, 1968 8. Brandchaft PB, Boone CW: Increase in Gross (G) antigen sites on the surface of AKR virus-induced rat lymphoma cells after treatment with trypsin. J Immunol 113:94-102, 1974 9. Davis S: Circulating lymphocyte subpopulations in chronic lymphocytic leukemia. New Engl J Med 294:1150-1153, 1976 10. Doljanski F, Kapeller M: Cell surface shedding. The phenomenon and its possible significance. J Theor Biol 62:253-270, 1976
11. Durantez A, Zighelboim J, Gale RP: Leukemia-associated antigens detected by heterologous antisera. J Natl Cancer Inst 56:1217-1219, 1976 12. Editorial: B-lymphocyte alloamigens in man. Lancet 2:240-242, 1976 13. Fu SM, Winchester RJ, Hunkel HG: The occurrence of the HL-B alloantigens on the cells of unclassified acute lymphoblastic leukemia. J Exp Med 142:1334-1337, 1975 14. Gibofsky A, Terasaki PI: Trypsinization of lymphocytes for HL-A typing. Transplantation 13:192-194, 1972 15. Greaves MF, Brown G, Rapson NJ, et al: Antisera to acute lymphoblastic leukemia cells. Clin Immunol Immunopathol 4:67-84, 1975 16. Halterman RH, Leventhal BG, Mann DL: An acute leukemia antigen: Correlation with clinical status. N Eng J Med 287: 1272-1274, 1972 17. Hellstrom KE, Hellstrom I: Immunity to neuroblastomas and melanomas. Annu Rev Med 23:19-38, 1972 18. Hsu CCS, Chen Y, Patterson R: Peripheral blood B-lymphocyte. abnormalities associated with hyperthyroidism of Graves' disease. Clin Exp Immunol 26:431-440, 1976 19. Hsu CCS, Marti GE, Mittal KK: Antisera against leukemiaassociated antigens on human lymphocytes. Clin Exp Immunol 27:487-496, 1977 20. Hsu CCS, Marti GE, Schrek R, et al: Lymphocytes bearing B- and T-cell markers in patients with lymphosarcoma cell leukemia. Clin Immunol Immunopathol 3:385-395, 1975 21. Klein G: Immunological surveillance against neoplasia. Harvey Lect 69:71-102, 1974 22. Lin PS, Hsu CCS: Human leukemic T-cells with complement receptors. Clin Exp Immunol 23:209-213, 1976 23. Linthicum DS, Bahu RM, Sell S: Differences in ultrastructural appearances and modulation of endogenous and exogenous surface immunoglobulin of rabbit lymphocytes. J Immunol 117:1179-1188, 1976 24. Mann DL, Abelson L, Harris S, et al: Detection of antigens specific for B-lymphoid cultured cell lines with human alloantisera. J Exp Med 142:84-89, 1975 25. Mann DL, Halterman R, Leventhal B: Acute leukemia associated antigens. Cancer 34 (suppl): 1446-1451, 1974 26. McDevitt HO, Delovitch TL, Press JL, et al: Genetic and functional analysis of the la antigens: Their possible role in regulating the immune response. Transplant Rev 30:197-235, 1976 27. Metzgar RS, Mohanakumar T, Miller DS: Antigens specific for human lymphocytic and myeloid leukemic cells: detection by non-human primate serum. Science 178:986-988, 1972 28. Mohanakumar T, Metzgar RS, Miller DS: Human leukemia cell antigens: serologic characterization with xenoantisera. J Natl Cancer Inst 52:1435-1444, 1974 29. Schlossman SF, Chess L, Humphreys RE, et al: Distribution of la-like molecules on the surface of normal and leukemic human cells. Proc Natl Acad Sci USA 73:1288-1292, 1976 30. Tarro G: Appearance in trypsinized normal cells of reactivity with antibody presumably specific for malignant cells. Proc Natl Acad Sci USA 70:325-327, 1973 31. Tsukimoto I, Wong KY, Lampkin BC: Surface markers and prognostic factors in acute lymphoblastic leukemia. New Engl J Med 294:245-248, 1976 32. Welsh K, Turner NJ: Preparation of antiserum specific to human B-cells by immunization of rabbits with immune complexes. Tissue Antigen 8:197-205, 1976. 33. Wernet P: Human la-type alloantigens: Methods of detection, aspects of chemistry and biology, markers for disease states. Transplant Rev 30:271-289, 1976. 34. Winchester RJ, Fu SM, Hoffman T, Kunkel HG: IgG on lymphocyte surfaces: technical problems and the significance of a third cell population. J Immunol 114:1210-1212, 1975 35. Yefenof E, Klein G: Difference in antibody induced redistribution of membrane IgM in EBV-genome free and EBV positive human lymphoid cells. Exp Cell Res 99:175-178, 1976
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
that previously undetected cell anitgens may appear after trypsin treatment of the cells.14™ The "blocking component" appears to coat the leukemic cell surfaces nonspecifically, because expression of both B-cell-associated antigens and SIg on chronic lymphocytic leukemia cells was masked. On acute lymphocytic leukemia cells, in addition to B-cellassociated antigens, another membrane antigen detectable with rabbit anti-acute lymphocytic leukemia cells prepared as described by Greaves and colleagues15 has also been found to be masked by the "blocking component" (Hsu and Morgan, in preparation). Theoretically, the presence of such a component might influence the clinical course of the disease by interfering with the host's immunologic recognition of the leukemic cells.17-21 It is relevant that a recent study9 revealed progressively less SIg on chronic lymphocytic leukemia cells as the disease advanced. It would be important to determine whether this decrease in SIg on chronic lymphocytic leukemia cells is due to a "blocking component." The appearance of a cellbound "blocking component" may be a way for leukemic cells to evade the host's immunologic surveillance.
747
