Eur. J. Immunol. 1989.19: 469-474 Immunodeficiency and T cell malignancy 469

Syed Raziuddin', Hasan M- Ass& T cell malignancy in Richter's syndrome presenting as hyper IgM. Induction and characterization of a novel and Bayu Teklu'

Departments of Immunologyo,

King Saud University, College of

CD3+, CD4-, CD8+ T cell subset from phyto- hemagglutinin-stimulated patient's CD3+, CD4+, CD8' leukemic T cells

Pathology' and Internal Medicinea,

Medicine, Abha

A patient is described, having Richter's syndrome and immunodeficiency with hyper IgM, who developed suppressor T cell lymphoma (CD3+, CD4-, CD8') following untreated helper-suppressor T cell chronic lymphocytic leukemia (CD3+, CD4+, CD8+). The neoplastic T cells in both malignancies expressed interleukin (IL) 2 receptors but were deficient in typical CD2' and CD5' pan T antigens. Additionally, a large percentage of malignant lymph node T cells expressed HLA-DR+ activation antigens. In vitro immunoglobulin-production experiments demonstrated that the patient's leukemic blood T cells had an excess helper function for IgM synthesis but a suppressor function for IgG and IgA synthesis by normal B and T cells. The leukemic blood T cells demonstrated a poor response to phytohemagglutinin (PHA). A defect in IL 2 receptor expression was evident in PHA-stimulated leukemic blood T cells. Of interest was the observation that PHA stimulated the induction of a novel CD3+, CD4-, CD8+ T cell subset from patient's CD3+, CD4+, CD8' leukemic blood T cells. These PHA-induced CD3+, CD4-, CD8+ T cell subsets produced an elevated pro- liferative response to PHA and concanavalin A, had a helper cell function for IgM synthesis and produced highly elevated amounts of IL 2.

1 Introduction

The term Richter's syndrome was introduced for the develop- ment of malignant lymphoma in a patient with pre-existing chronic lymphocytic leukemia (CLL) [l, 21. With immunologi- cal marker studies, it was demonstrated that most CLL and lymphomas were in fact B cell malignancies, demonstrating a common clonal origin of the two malignancies in Richter's syndrome [ M I . Apart from several reports of B cell malig- nancies in Richter's syndrome, there is no report of a T cell malignancy in a patient with pre-existing CLL developing into a malignant lymphoma. The lymphocyte lineage of various T cell malignancies is identifiable by well-characterized mono- clonal antibodies [7-91. We report a patient with T cell CLL (T-CLL) whose leukemic cells in the blood had both helper and suppressor T cell phenotype (CD3+, CD4+, CD8'), and whose malignant lymphoma cells in lymph nodes (LN) had a characteristic suppressor T cell phenotype (CD3+, CD4-, CD8'). The patient's leukemic blood T cells were associated with hyper IgM syndrome. We demonstrated that phy- tohemagglutinin (PHA) stimulation of the patient's leukemic blood T cells with CD3+, CD4+, CD8+ phenotype induces a unique CD3+, CD4-, CD8' T cell subset in cultures in vitro. The characteristics of these CD3+, CD4-, CD8' T cell subsets are presented in this communication.

[I 72001

Correspondence: Syed Raziuddin, College of Medicine, P.O. Box 641, Abha, Saudi Arabia

Abbreviations: T-CLL: T-cell chronic lymphocytic leukemia PBL: Peripheral blood lymphocytes LN(C): Lymph node (cells) IL 2R: Interleukin 2 receptors PHA: Phytohemagglutinin Con A: Con- canavalin A PWM: Pokeweed mitogen (s)Ig: (Surface) immuno- globulin

2 Materials and methods

2.1 Case report

A 90-year-old Saudi man with a 6-month history of T-CLL (CD3+, CD4+, CD8+ T cell phenotype), and normal serum immunoglobulin (Ig) levels developed sudden generalized lymphadenopathy, fever, weight loss, malaise, and abdominal pain. The patient had received no chemotherapy or radiotherapy following the diagnosis of T-CLL. Laboratory studies reported the following peripheral blood values: hemoglobin 8.0 g/100 ml; white blood cell count 184 x 109/1 with 96% lymphocytes, 2% segmented neutro hils, 1% bands

sedimentation rate was 120 mmh. The total serum protein was 7.9 g/100 ml; serum IgG 22.6 mg/ml, IgM 187.4 mg/ml, IgE 86 IU/ml, IgA was undetectable. The spleen and liver were markedly enlarged. There was no serological evidence of cur- rent infection. The patient had massive enlargement of cervi- cal, axillary and inguinal LN, and a mass in the upper abdo- men. A biopsy specimen was taken from the left cervical LN. Part of it was used for routine histological examination and another portion was subjected to detailed immunophenotypic and functional analysis. In the biopsied LN, infiltration by large malignant lymphoid cells was observed by light micros- copy. Most of the malignant cells had large round nuclei and relatively rich cytoplasm. No Reed-Stemberg-like cells were present. The peripheral blood smear contained a predomi- nance of mature medium-sized lymphocytes with moderate amounts of cytoplasm, containing scattered azurophilic granules. The bone marrow was extensively infiltrated by malignant lymphoid cells. Histologically, a diffuse non-Hodg- kin's, well-differentiated lymphocytic lymphoma was diag- nosed in the cervical LN and bone marrow in this patient with pre-existing T-CLL. One month after the diagnosis of Rich- ter's syndrome and hyper IgM, the patient died of septic shock. An autopsy was not permitted.

and 1% monocytes; platelet count of 229 x 10 r /1. Erythrocyte

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1989 001 4-2980/89l0303-O469$02.50/0

470 S. Raziuddin, H. M. Assaf and B. TeWu Eur. J. Immunol. 1989.19: 469-474

2.2 Lymphocyte preparation

Human peripheral blood lymphocytes (PBL) were prepared by Histopaque-1077 (Sigma Chemical Co, St. Louis, MO) density centrifugation of heparinized peripheral blood obtained from patient and from normal healthy volunteers. LN cells (LNC) were prepared from patient's malignant LN by teasing apart the LN tissue,' and Histopaque-1077 (Sigma) density centrifugation [ 101.

2.3 Monoclonal antibodies and phenotypic analysis

OKT3, OKT4, OKT6, OKT8, OKT9, OKT10, OKT11, OKIal, OKDR, OK-CLL, OKMI antibodies were obtained from Ortho Diagnostics, Raritan, N.J. Anti-Leu-1, anti-Leu- 2a, and anti-Leu3a + 3b were obtained from Becton Dickin- son (Monoclonal Center, Mountain View, CA). Anti-Tac was kindly provided by Dr. Thomas A. Waldmann (NIH, Bethesda, MD). Analysis of surface phenotype with these antibodies was performed by indirect immunofluorescence technique as described [ll]. The lymphocytes were also assayed in a 2-aminoethylisothiouonium bromide (AET)- treated sheep erythrocyte (SRBC) rosetting technique [ 111. Surface immunoglobulin (sIg) were determined with fluores- cein isothiocyanate (F1TC)-labeled goat anti-human IgG, IgA and IgM polyvalent antibody (Ortho Diagnostics, Westwood, MA) *

either 2 pg/ml PHA, 5 pg/ml concanavalin A (Con A), 5 pg/ml PWM, or 1 pg/ml OKT3 antibody, for 72 h at 37 "C in com- plete RPMI 1640 medium, and incorporation of [3H]thymidine [lo].

2.8 Lymphocyte activation

The patient's leukemic blood T cells and purified normal blood T cells were stimulated with PHA (2 pg/ml) or Con A (5 pg/ml) for 72 h at 37 "C in complete RPMI 1640 medium. T cell phenotype with OKT4 and OKT8 antibody, and activation antigens HLA-DR and interleukin 2 receptors (IL 2R) were determined with OKIal and anti-Tac antibody, respectively, from these stimulated T cell cultures.

2.9 Separation of T cell subsets

The patient's leukemic blood T cells were stimulated with PHA or Con A for 72 h as described above. Stimulated T cells were thereafter separated into OKT4' and OKT8' T cell sub- sets with OKT8 and OKT4 antibody, respectively, by a "pan- ning" procedure described by Wysocki and Sato [12]. OKT4' and OKT8' T cell preparations were then stained by indirect immunofluorescence technique with OKT4 and OKT8 anti- body, and percentage reactivity of OKT4' and OKT8' T cell preparations with both antibodies determined.

2.4 Serum immunoglobulin determination 2.10 Isolation of newly generated CD4-, CD8' T cell subsets

Serum IgG, IgA and IgM were determined by the radial immunodiffusion technique using commercially available "Diffu Gen" plates (Tago Inc., Burlingame, CA).

2.5 Preparation of T and Non-T Cells (B Cells)

PBL from normal healthy volunteers were separated into T and non-T cells with an AET-SRBC rosetting technique, and Histopaque-1077 density centrifugation [ll]. Analysis of the obtained cell population showed that the T cell preparation contained >95% OKT3' cells, and non-T cells contained about 93% sIg' cells (B cells).

CD4-, CD8' T cell subsets from the patient's PHA-activated cultures of CD4-, CD8' and CD4', CD8' T cells were sepa- rated by complement-mediated lysis of CD4', CD8' T cells employing OKT4 monoclonal antibody. In brief, 1 x lo7 cell mixtures were incubated with 100 pl OKT4 antibody on ice for 1 h. Thereafter, 25 p1 of rabbit complement (Flow Labs, McLean, VA) was added and a further incubation was carried out for 1 h at 37 "C. The cells were washed and the entire procedure was repeated. Analysis of the cells obtained showed that OKT4-treated population contained 99% OKT8' cells and <1% OKT4' cells. This procedure yielded purified CD4-, CD8' T cell subsets from the PHA-activated patient's CD4', CD8' leukemic blood T cells.

2.6 Helper suppressor function of T cells for Ig synthesis 2.11 IL 2 production

The helper and suppressor function which the patient's leukemic blood T cells may exert on normal B cell differentia- tion to secrete Ig was assayed by co-culturing the patient's T cells with normal B and/or T cells in the presence of pokeweed mitogen (PWM; 5 pg/ml) in a humidified 37 "C incubator with 5% C02 for 7 days. The secreted IgG, IgA and IgM in the cocultured supernatants was determined by enzyme-linked

IL 2 activity in supernatants of cells stimulated with PHA (2 yglml) for 48 h at 37 "C was measured in a standard prolif- eration assay [lo, 111, using an IL 2-dependent murine CTLL cell line. Units of IL 2 activity were defined as U (reciprocal titer of test sample at 30% of maximal counts/min standard) (reciprocal titer at 30% maximal counts/min) x 100.

immunosorbent assay (ELISA) with goat kti-human IgG, IgA and IgM antisera (Tago) [ 111. 3 Results

2.7 Proliferative response 3.1 T cell phenotype

Lymphocyte proliferative response was determined by cultur- ing patient's leukemic blood T cells, malignant LN T cells, and purified normal blood T cells (2.5 X 16 cells) in triplicate with

The immunophenotypic results of the patient's leukemic PBL and malignant LNC are shown in Table 1. The phenotype in both PBL and LNC was a mature T cell with CD3' pan T

Eur. J. Immunol. 1989.19: 469-474 Immunodeficiency and T cell malignancy 471

Table 2. Helper suppressor function of patient's leukemic blood T cellsa)

Table 1. Lymphocyte phenotyping of patient's PBL and LNCa)

Monoclonal antibody PBL LNC

OKT3 OKTl 1 Leu-1 OKT4 Leu3a + 3b OKT8 Leu-2a OKT6 OKT9 OKTlO OKIal OKDR OK-CLL OKMI Tac

E-rosettes sIg

(CD3, Pan T cells) (CD2, E-rosetting) (CD5, Pan T cells) (CD4, helper T cells) (CD4, helper T cells) (CD8, suppressor T cells) (CD8, suppressor T cells) (CD1, common thymocytes) (Transferrin receptor) (Activation antigens) (HLA-DR) (HLA-DR) (Leukemic cell antigens) (Monocytes) (CD25, IL 2R) (B cells)

9 4 9 7 7 2 8 1

8 4 9 76 3 91 88 79 81 0 0 1 0 6 2 0 6 8 2 56

81 7 2 0 67 73 3 0

11 6

a) Results are expressed as percentage of positive cells.

antigen, but were negative for the other CD2' and CD5' pan T antigens. In addition, the patient's PBL were shown to be of the T helper-suppressor phenotype (CD3', CD4', CD8') characterizing the disease as helper-suppressor T-CLL, whereas the malignant LNC had a typical T suppressor phenotype (CD3+, CD4-, CD8') characterizing the malignant lymphoma as a suppressor T cell lymphoma. A high percent- age of PBL and LNC had IL 2R on cells. The malignant LNC also expressed HLA-DR antigens. None of the cells from PBL and LNC showed reactivity with anti-x or anti-A Ig light chains (data not shown). The patient's serum Ig profile indicated an IgG of 276 mg/ml, an IgM of 1983 mg/ml and an absence of IgA, as determined by standard radial immunodiffusion and confirmed by ELISA technique.

3.2 In vitro Ig synthesis

The patient had an elevated serum IgM, low IgG and selective IgA deficiency. The capacity of the patient's leukemic blood T cells were evaluated for helper and suppressor function on PWM-induced normal B cell differentiations to secrete three major Ig (Table 2). We found that these leukemic T cells were deficient in IgG- and IgA-specific helper cell function, but had an excess helper cell activity for IgM synthesis by normal B cells. In addition, the patient's leukemic T cells suppressed IgG and IgA synthesis by normal B and T cells in co-culture experiments.

3.3 T cell proliferation

The in vitro proliferative response of the patient's leukemic blood T cells and malignant LN T cells to optimum doses of PHA, Con A, PWM and OKT3 is shown in Table 3. The patient's leukemic T cells produced a reduced response to all these mitogens. However, the malignant LN T cells produced a highly elevated response to Con A.

Cells added to normal Immunoglobulins (pglml) B cells (2.5 x Id) IgG IgA IgM

None 0 0 0.8 f 0.1 Normal T cells 3.4 k 0.8 1.3 f 0.2 8.3 k 1.7 Patient's T cells 0.2 k 0.1 0 12.8 2 2.1 Normal T cells plus Patient's T cells 0.3 f 0.1 0 17.4 f 2.6

a) Ig secretion after culturing normal B cells alone or B cells with equal amounts of normal and/or patient's leukemic blood T cells, stimulated with PWM (5 pgml) for 7 days. Mean f SEM of three separate determinations.

3.4 CD4-, CD8' T cell induction by PHA

The patient's leukemic blood T cells were stimulated with PHA and Con A for 72 h, and T cell phenotype as well as activation antigens were determined on these stimulated T cells (Table 4). Nonstimulated T cells (resting) were nega- tive for HLA-DR antigens, but 62% expressed IL 2R and 95%-98% both CD4 and CD8 T cell antigens. After stimula- tion with PHA, these leukemic T cells showed a higher per- centage of HLA-DR' cells but only 16% and 38% respec- tively, were IL 2R' and CD4'. However, after Con A stimu- lation, 73% of these T cells were IL 2R'; unstimulated and Con A-stimulated leukemic T cells were undistinguishable with respect to CD4 and CD8 expression. Both PHA- and Con A-stimulated patient's leukemic T cells expressed CD3 antigens, but were deficient in CD2 and CD5, pan T antigens (data not shown). These experiments were repeated three times with almost similar results as shown in Table 4. The OKT3 antibody stimulation of the patient's leukemic T cells was similar to Con A-stimulated T cells (data not shown). Thus, the data obtained indicated a major difference in acquir- ing CD4 T cell antigen and IL 2R between the patient's PHA- and Con A-stimulated leukemic T cells. This data also revealed that in the PHA-stimulated patient's leukemic T cells, cells of two different T cell phenotypes were pro- liferating. To test this, we separated CD4' and CD8" T cells from the patient's PHA- and Con A-stimulated T cells, and individual cell populations were stained with OKT4 and OKT8 antibody (Table 5) . CD4' and CD8' T cells obtained from Con A-stimulated cultures and CD4' T cells obtained from PHA-stimulated cultures stained equally positive with OKT4 and OKT8 antibody. Of interest, in CD8' T cells of PHA- stimulated cultures, a reduced percentage of OKT4 antibody (58%) reactivity was observed. These results suggested that the patient's leukemic blood T cells after stimulation with PHA, but not with Con A or OKT3, induced a small popula- tion of CD4-, CD8' T cell subsets in the in vitro cultures.

The PHA-stimulated leukemic T cells when treated with OKT8 antibody and complement were lysed to 100%. How- ever, treatment with OKT4 antibody and complement yielded a T cell population with CD4-, CD8' phenotype. Thus, in cultures of PHA-stimulated patient's leukemic blood T cells (CD4', CD8') T cells with two different phenotypes CD4+, CD8' and CD4-, CD8' were produced. However, CD4+, CD8- were not produced. The CD4-, CD8' T cell subsets

472 S. Raziuddin, H. M. Assaf and B. Teklu Eur. J. Immunol. 1989.19: 469-474

Table 3. Proliferative response of patient's leukemic blood and malignant LN T cells

a) T cells from [3H]Thymidine incorporation, cpm x lo-"'

PHA Con A PWM OKT3

Normal blood 93 ? 8.5 87.2 f 5.2 11.9 f 2.3 82.4 It 7.8 Patient's blood 27.2 ? 3.7 45.2 k 3.2 9.6 k 2.2 52.5 k 5.3 Patient's LN 2.0 rf: 0.8 162.7 f 14.5 2.4 f 0.8

were isolated from these PHA-stimulated cell cultures by lysis of CD4 antigens (CD4', CD8+) using OKT4 antibody and complement. The data on detailed T cell phenotype of these CD4-, CD8' T cell subsets are shown in Table 6. These CD4-, CD8' T cell subsets had CD3 antigens but were defi- cient in CD2 and CD5 pan T antigens. They also expressed 57%-61% HLA-DR activation antigens but only 11% of IL 2R.

3.5 CM-, CDS' T cell function

The functional properties of these newly generated CD4-, CD8+ T cell preparations are shown in Table 7. They pro-

Table 4. Expression of T cell phenotype and activation antigens after PHA and Con A stimulation

T cells Stimulation"' Reactivity with monoclonal antibodyb' from OKT4 OKT8 OKIal Anti-Tac

Normal None 63 37 0 0 PHA 71 29 33 24

Con A 73 27 27 19 Patientc' None 98 95 0 61-

16 PHA - 38 95 48 Con A 93 98 41 73

-

a) T cells (2.5 X lo6 cells) in complete RPMI 1640 medium were stimulated with PHA (2 yglml) or Con A (5 yglml) at 37 "C for 72 h, and reactivity of stimulated cells with various monoclonal antibodies examined.

b) Results are expressed as percentage of stained cells. c) Patient's leukemic blood T cells.

Table 5. Distribution of T cell phenotype in patient's PHA- and Con A-stimulated CD4' and CD8' T cell subsets")

Stimulation Separated cell Percentage staining populations with

OKT4 OKT8

PHA

Con A

CD4' 96 97

CD4' 95 93 CD8' 95 96

CD8' - 58 95

a) Patient's leukemic blood T cells were stimulated with PHA (2 ygl ml) or Con A (5 yglml) at 37 "C for 72 h. Stimulated T cells were separated into CD4+ and CDS' T cell subsets. CD4+ and CD8' separated T cell subsets were then stained with OKT4 and OKT8 monoclonal antibody.

3.9 f 1.6

PHA 2yglml; ConA 5yglml; PWM 5yg/ml; OKT3 1 yglml. Cells (2.5 X 1 e ) were cultured for 72 h at 37 "C with mitogens or medium alone. In- corporation of 1 yCi (= 37 kBq) [3H]thymidine was assessed during the last 18 h of the culture. The data shown are cpm f SD of triplicate cultures, after subtraction of the background counts (medium alone).

Table 6. Cell surface staining of CD4-, CD8' T cells")

Reagents used CD4-, CD8' cells ("/.I

OKT3 OKTl1 Leu-1 OKT4 OKT8 OKIal OKDR Tac Sk?

100 0 0 3

98 61 57 11 0

a) Purified CD4-, CD8' T cells generated from a 72-h culture of PHA-activated CD4', CD8' patient's leukemic blood T cells were stained with various monoclonal antibodies. Results are expressed as percentage positivity.

duced an elevated proliferative response to both PHA and Con A mitogens and had a helper cell function for IgM synthe- sis by normal B cells, but were however, deficient in suppres- sor cell activity for Ig synthesis by normal B and T cells. They produced a highly elevated amount of IL 2, as compared to patient's blood leukemic T cells, malignant LN T cells and normal control T cell preparations (Table 8).

4 Discussion

This report describes a unique T cell malignancy and immunodeficiency with hyper IgM in a patient with Richter's syndrome. The patient's leukemic blood T cells expressed CD2-, CD3', CD4+, CDY, CD8' (helper-suppressor) phenotype, whereas, malignant LN T cells expressed a CD2-, CD3', CD4-, CDT, CD8' (suppressor) phenotype. The simultaneous occurrence of helper and suppressor (CD4+, CD8') phenotype in T cell malignancy has been reported ear- lier [13-161. The increased propensity for the development of secondary lymphomas has also been reported following the treatment of Hodgkin's disease, leukemia, mycosis fungoides and other solid tumors [16-181. However, many of these reports were incomplete because of the lack of immunological studies. The present case report appears to represent the first immunological documentation of suppressor T cell lymphoma and hyper IgM immunodeficiency developing in a nontreated patient with helper-suppressor T cell leukemia.

Immunodeficiency with hyper IgM is a syndrome character- ized by the absence of IgG or IgA in serum and concomitant elevated serum levels of IgM. Clinically, patients with this syndrome are similar to patients with other immunodeficien-

Eur. J. Immunol. 1989.19: 469-474 Immunodeficiency and T cell malignancy 473

Table 7. Functional properties of newly generated CD4-, CD8' T cells

(A) Proliferation of cells (3H]Thymidine incorporation (cpm x 10.') PHA Con A PWM

CD4-, CD8' T cells") 124.8 f 18.2 98.1 f 11.6 2.5 i 0.8 CD4-, CD8' T cells (normal) 53.5 f 7.3 31.7 f 6.4 8.4 i 2.6

(B) Ig synthesis in co-cultures

CD4-, CD8' T cells Normal B cells CD4-, CD8' T cells + ') Normal B cells Normal T cells + Normal B cells CD4-. CD8' T cells + Normal T cells + Normal B cells

0 0 0 0 0 0 0 0 0 0 0.2 0.3

0 0 0 0 8.2 9.6

2.6 1.9 1.2 1.4 7.8 6.3 b)

2.8 2.3 1.5 1.3 9.7 11.4

cies, with frequent infections and neoplasms [19-241. Our patient's leukemic blood T cells were able to induce increased IgM secretion in B cells from normal donors, and possessed a suppressor activity for IgG and IgA synthesis. This was consis- tent with a T cell defect, clinically observed hyper IgM immunoglobulinemia. Earlier, Mayer et al. [23] demonstrated that in some patients a pure T cell abnormality was pathogenic in hyper IgM syndrome.

The presence of IL 2R and HLA-DR antigens on resting T cells are considered as activation antigens, which appear sequentially when normal T cells are stimulated by lectin or antigen in vitro and in vivo [25]. Our patient's leukemic blood T cells after stimulation with PHA, but not with Con A, expressed fewer IL 2R than unstimulated T cells. This data may indicate that PHA not only failed to induce IL 2R, but that it suppressed the prior IL 2R expression on these leukemic T cells. It is unclear whether this is an actual regula- tion of IL 2R or represents blocking. However, another acti- vation antigen, HLA-DR was expressed on PHA-stimulated patient's leukemic T cells almost equally to Con A-stimulated T cells.

Table 8. IL 2 production")

Source of cells Cell phenotype IL 2 activity (Unitdml)

Patient Leukemic blood T cells CD3', CD4+, CD8+ 67 74 LN T cells CD3t. CD4-. CD8t 18 23 PHA-generated T cells CD3+, CD4-, CD8+ 118 132 Normal PBL - 94 81 Purified CD4+ T cells CD3 '. CD4'. CD8- 115 97 Purified CD8+ T cells CD3', CD4-, CD8+ 36 28

a) IL 2 activity was determined in supernatants of cells cultured for 48 h at 37 "C with PHA (2 pg/ml) in a proliferation assay using CTLL cell line.

CD4-, CD8' T cells (2.5 X 16 cells) were cultured in triplicate with PHA (2 pg/ml), Con A (5 &ml) or PWM (5 pg/ml) for 72 h at 37 "C, and proliferative response was de- termined by incorporation of [3H]Thy- midine. The data shown are cpm after sub- traction of background counts. CD4-, CD8' T cells (2.5 x lo5 cells) were cultured with equal numbers of purified nor- mal B cells and/or T cells for 7 days at 37 "C in the presence of PWM (5 @nl) and Ig (pg/ml) secreted in the supernatants were determined by an ELISA.

Further characterization of this patient's leukemic blood T cells revealed that, after stimulation with PHA, these leukemic T cells were differenting in culture into cells of two different phenotypes, CD3+, CD4', CD8+ and CD3+, CD4-, CD8'. The experimental conditions allowed us to purify only one subset, CD3+, CD4-, CD8' from these stimulated cell mixtures. These CD3+, CD4-, CD8' T cell subsets produced a good proliferative response to PHA and Con A, and posses- sed IgM-specific helper function on normal B cell differentia- tions. Although, they had a typical suppressor T cell phenotype, but were deficient in suppressor cell function in Ig synthesis. The ability to produce an elevated amount of IL 2 upon further stimulation with PHA inspite of the low IL 2R expression was an interesting property of these CD3+, CD4-, CD8' T cell subsets.

Since, PHA stimulation of patient's leukemic blood T cells (CD4+, CD8+) resulted in the proliferation of a minor subset of T cells with CD4-, CD8' phenotype, it could be argued that secondary T cell malignancy (malignant lymphoma) with CD4-, CD8' phenotype in our .patient probably originated from a clone developed from patient's CD4+, CD8' leukemic blood T cells. Thus, immunological studies described here pro- vide a clinical example that activated leukemic T cells with helper-suppressor bi-phenotype mediate the pathogenesis of suppressor T cell lymphoma, and immunodeficiency with hyper IgM.

The authors thank Dr. Gideon Goldstein, Ortho Diagnostics, Raritan, N . J . , for OKT6, OKT9 and OKTlO, Dr. Thomas A . Waldmann, NIH, Bethesda, for anti-Tac, and Becton Dickinson, Mountain View, CA, for anti-Leu-1, anti-Leu-2a and anti-Leu 3a + 36 monoclonal anti- bodies. W e are grateful to Dr. Verner Paetkau, University of Alberta, Edmonton, Canada, for the interleukin 2-dependent murine CTLL cell line. We are also grateful to Dr. Abdul Ar iz Alsaigh of the Department of Surgery for his cooperation and for LN biopsy done for these immu- nological studies.

Received October 18, 1988.

474 S. Raziuddin, H. M. Assaf and B. Teklu Eur. J. Immunol. 1989.19: 469-474

5 References

1 Richter, M. N., Am. J . Pathol. 1928. 4: 285. 2 Lortholary, P. P., Boiron, M., Ripault, P., Levy, J., Manus, A.

and Bernard, J., Nouv. Rev. Franc. Hematol. 1964. 4: 621. 3 EMO, A., Catovsky, D., O’Brian, M., Cherchi, M., Kumaran, T.

and Galton, D. A. G., Br. J. Haematol. 1979. 41: 9. 4 Harousseau, J. L., Flandrin, G., Tricot, G., Brout, J., Seligmann,

M. and Bernard, J., Cancer 1981. 48: 1302. 5 Von Dougen, J., Hooijkan, H., Michiels, J., Grosveld, G., De-

Klein, A., Vanderkwast, Th., PMs, M., Abels, J. and Hagemei- jer, A., Blood 1984. 64: 571.

6 Suster, S. and Rwylin, A. M., Cancer 1987. 59: 1412. 7 Reinhen, E. L., Kung, P. C., Goldstein, G., Levy, R. H. and

Schlossman, S . F., Proc. Natl. Acad. Sci. USA 1979. 76: 4061. 8 Pandolfi, F., Semenzato, G., De-Rossi, G., Strong, D., Quinti, I.,

Pezzuto, A., Mandelli, F. and Aiuti, A., Clin. Irnmunol. Imrnunopathol. 1982.24: 330.

9 Knowles, D. M., 11, Dodson, L., Raab, R., Mittler, R., Talle, M. A. and Goldstein, G., Diagn. Immunol. 1983. 1: 142.

10 Raziuddin, S., Kibler, R. F. and Morrison, D. C., J . Irnmunol. 1982.128: 2073.

11 Raziuddin, S., Madan, A. andDanial, H. B., Scand. J . Immunol. 1987.26: 487.

12 Wysocki, L. and Sato, V., Proc. Natl. Acad. Sci. USA 1978. 75: 2844.

13 Nadler, L. N., Reinhen, E. L., Weinstein, T., Oris, C. J. and Schlossmann, S. F., Blood 1980.55: 806.

14 Greaves, M. F., Rao, J., Hariri, G., Verbi, D., Datovsky, D. and Goldstein, G., Leu&. Res. 198I. 5: 281.

15 Habeshaw, J. A., Bailey, D., Stansfeld, A. G. and Greaves, M. F., Br. J. Cancer 1983. 47: 327.

16 Yanagihara, E. T., Parker, J. W., Meyer, P. R., Cab, M. J., Hofman, F. and Lukes, R. J., Am. J . Clin. Pathol. 1984.81: 249.

17 Scheen, S., Banks, R. and Winkelmann, R., Mayo. Clin. Proc. 1984.59: 95.

18 Samlowski, W., Conrath, F., Piepkorn, M. and Kjeldsberg, C., Arch. Pathol. Lab. Med. 1985. IO9: 864.

19 Rosen, F. S. and Janeway, C. A., Jr., N. Engl. J. Med. 1966.275: 709.

20 Goldman, A., Ritzman, S., Houston, E., Sidwells, S., Brachter, R. and Levin, W. C., J. Pediatr. 1967. 70: 16.

21 Geha, R. S., Hyslop, N., Alami, S., Farah, F., Schneeberger, E. E. and Rosen, F. S . , J . Clin. Invest. 1979. 64: 385.

22 Levitt, D., Haber, P., Rich, K. and Cooper, M. D., J. Clin. Invest. 1983. 72: 1650.

23 Mayer, L., Kwan, P., Thompson, C., KO, H., Chiorazzi, N., Waldmann, T. A. and Rosen, F. S . , N . Engl. J . Med. 1986. 314: 409.

24 Mayer, L., Posnett, D. N. and Kunkel, H. G., J. Exp. Med. 1985. 161: 134.

25 Yachie, A., Miyawaki, T., Uwadana, N., Ohzeki, S. and Tani- guchi, N., J . Imrnunol. 1983. 131: 731.