Biotherapy 4: 53-63, 1992. © 1992 Kluwer Academic Publishers. Printed in the Netherlands.

Polyclonai activation of human lymphocytes and induction of cytotoxic lymphocytes by streptococcal preparations Toshihiro Sato, Masahiko Fujii, Hideki Satoh, Tetsuro Itoh, Hidemi Rikiishi & Katsuo Kumagai Department of Microbiology, Tohoku University, School of Dentistry, Sendai 980, Japan Received 6 April 1991; accepted 16 April 1991 Key words: helper T cells, lymphocyte activation, Streptococcus, y6 T cells Abstract

Polyclonal activation of human peripheral blood lymphocytes (PBLs) in vitro by preparations of Streptococcus pyogenes Su strain (OK-432) and other heat-killed strains was investigated. The streptococcal preparations tested induce a proliferative response of PBLs via interleukin-2 (IL-2)-independent pathways. The proliferative response is accompanied by the generation of lymphoblastic cells (LBCs), which consist of heterologous lymphocyte populations: CD4 + helper type of T cells, and C D 4 - C D 8 - double-negative (DN) lymphocytes, including both CD3 + TcRy6 + T cells and CD2 +CD3immature type of T or non-T cell type of lymphocytes. Almost all the LBCs express Leu19, TfR (transferrin receptor), LFA-1 and CD38 (OKT10) antigens, which are expressed on activated T cells, NK cells and some other lymphocytes. The proliferative response of human PBLs is also accompanied by the generation of potent cytotoxic activity against NK-sensitive and -resistant targets. C-dependent cytolysis and cell sorting experiments of OK-432-activated LBCs revealed that both CD3 + and CD3types of CD4 CD8- DN lymphocytes, but not CD4 ÷ helper T cells, may be major populations responsible for the cytotoxicity induced. On the other hand, CD4-CD8- T cells may be required for the proliferation of PBLs and generation of cytotoxic effector cells. These results suggest that the OK-432 and other streptococcal preparations stimulate the human PBLs in vitro to induce the proliferation/activation of CD4 + T cells, mediating the following generation of DN cytotoxic effector lymphocytes. Abbreviations: DN T cells: double negative T cells; LBCs: lymphobtastoid cells; mAb: monoclonal antibody; MLR: mixed lymphocyte reaction; PBLs: peripheral blood lymphocytes; TcR: T cell receptor. Introduction

Gram-positive and -negative bacteria and their products of surface components possess a strong adjuvant effect on the induction of delayed hypersensitivity or on the production of circulatory antibodies directed against certain specific antigens [1]. Recently, some of these microbial agents have also been used successfully against a variety of malignant animal tumors [2, 3]. These bacterial agents, when incubated in vitro with murine or human lymphoid cells, induce poly-

clonal activation of lymphocytes, although there is disagreement as to whether T cells, B cells or both respond to bacterial stimulation [4-8]. This polyclonal activation of lymphocytes induced by bacterial stimulation has also been found to be associated with the development of cytotoxic effector cells such as activated macrophages, NK cells or unidentified cytotoxic cells [9-12]. Thus, this polyclonal activation associated with the generation of cytotoxic cells has been thought to be a useful common antitumor effect of these agents, although the mechanisms involved in this

54 polyclonal activation of lymphocytes and characterization of the effector cells remain unclear. The Streptococcus species, as well as other gram-positive bacteria such as Corynebacteria or Mycobacteria, are some of the microbial agents most extensively used for their augmenting effects on humoral, cell-mediated or tumor-associated immunity [13, 14]. Picibanil (OK-432) [15] is a lyophilized preparation of penicillin- and heat-treated Streptococcus (Str.) pyogenes Su strain, which has been developed as an immunomodulating agent and is widely used in the study of immunochemotherapy against experimental tumors [16-18] and cancer in humans [19, 20]. The administration of this agent to tumor bearing mice augments or induces cytotoxicities against the tumor cells and shows potent antitumor activity in certain transplantable or autochthonous tumors. This agent has also been shown to have a much wider immunological effect as an immunoadjuvant in both humoral and cell-mediated immune responses in relation to viral or bacterial infections [21,22]. The present study was undertaken to investigate the mechanisms of polyclonal activation of human peripheral blood lymphocytes (PBLs) which are induced by Str. pyogenes preparations of OK-432 and other heat-killed strains. We have demonstrated that these streptococcal preparations induce potent cytotoxicity and the proliferation of heterologous T cell populations via an IL-2-independent mechanism. These cells include CD4 + helper type T cells and populations of C D 4 - C D 8 - DN lymphocytes including CD3+TcRy/6 + T cells and CD3- TcR "),16lymphocytes, which may be responsible for the cytotoxicity induced.

physiological saline solution and kept in a refrigerator before use. Other strains of Str. pyogenes, e.g., Sv (type 3, ATCC 21059), T3 (type 3) and two clinical isolates maintained in our laboratories, were killed by exposure to H 2 0 2 , penicillin and heating as described above or heated at 100°C for 5 min and then used. Both treatments substantially resulted in similar mitogenic activities against human lymphocytes.

Cell preparation and cultivation. PBLs were obtained from volunteers by using the FicollIsopaque density centrifugation technique [23, 24]. To avoid any stimulation by xenogeneic proteins other than the bacterial proteins tested, the isolated PBLs were washed three times with serum-free RPMI1640 medium (Gibco Laboratories, Grand Island, NY) and finally suspended in a medium containing 20% autologous human serum (AHS) (RPMI/AHS)~ as previously described [23]. The cells (2 x 10/ml) were cultured both with and without OK-432 at varying concentrations at 37°C for 1 to 8 days in a CO 2incubator. Triplicate cultures were prepared in plastic microplates with 24 to 96 flat-bottomed wells (Corning Inc., Corning, NY) [23, 24]. DNA synthesis assay. To test the proliferative response of lymphocytes, a plastic microwell culture plate with 96 flat-bottomed wells (Nunc, Denmark) was used for the cell culture. A total of 200/xl/well of cells (1 x 106/ml) was suspended in the medium and triplicate cultures were continued for the desired periods. Eighteen h before the end of incubation, the cells were pulsed with 0.5 /xCi of 3H-TdR. The extent to which the 3H-TdR was incorporated into the D N A was determined by using triplicates with a liquid scintillation counter [23, 24].

Materials and methods

Streptococcal preparations. Picibanil (OK-432) is a lyophilized preparation of the Su strain of

Streptococcus pyogenes (type 3), which was treated with a short exposure to H 2 0 2 followed by inactivation with a combination of exposure to 1 x 10 4 units/ml of penicillin for 30 h at 37°C and heating for 20 min at 45°C [15-22]. It was supplied by Chugai Pharmaceutical Co. Ltd., Tokyo. One tenth mg of dried streptococci (5 x 1 0 7 - - 1 x 108) in a vial was suspended in 1 ml of

Cytotoxicity assay. The cytotoxic function of the whole PBLs and CD3 ÷CD4- CD8-, C D 3 - C D 4 - C D 8 - cells and other cells sorted by a FACStar (Becton-Dickinson Inc.), which were stimulated by streptococcal preparations, was examined by a 4 h specific chromium assay. Target cells for the cytotoxicity were NK-sensitive myeloid cell line, K562 and NK-resistant human cell line, FL. The methods and formula for calculating specific cytolysis have been previously described [24, 25].

55

Monoclonal antibodies (mAb ). Purified, fluorescein isothiocyanate (FITC)-and phycoerythrin (PE)-coupled mAbs, were used in this study. The differentiation antigens recognized by the mAb are summarized in Table 1 [25]. Antibodies O K T l l (anti-CD2), OKT3 (anti-CD3), OKT8 (anti-CD8) and OKT10 (anti-CD38) and their conjugates were purchased from Ortho Pharmaceutical Corp. Raritan, NJ; antibodies Leu4 (anti-CD3), Leu3a (anti-CD4), LeuM1 (antiCD15), L e u l l (anti-CD16), Leul9, and anti-IL2R (anti-CD25) were purchased from BectonDickinson, Inc., Immunochemistry System Inst., Mountain View, CA; mAb TcR-1, which reacts to an epitope on the framework of the T-cell receptor ce//3 heterodimer, was also purchased from Becton Dickinson, Inc.; mAb TcR61, which reacts to the 6 protein expressed on the cell surface in association with the TcRy-chain, was purchased from T Cell Science Inc. Cambridge, MA [26]. Immunofluorescence test. Phenotype of lymphocytes was analyzed by mAb in conjugation with single or two-color flowcytometry. The staining method and enumeration method of fluorescein positive cells by a FACS analyzer (BectonDickinson Inc.) have been previously described [26]. Briefly, freshly isolated or cultured PBLs were washed extensively with PBS and were allowed to react for 30 min on ice with optimal amounts of directly conjugated mAb at 1.0 x 107 cells/ml in a total volume of 100 ~1. Cells were then washed twice in cold PBS and resuspended

Table 1. Specificitiesof mAbs used. mAb

CD cluster

Majorcellreactivity

OKTI1 OKT3, Leu4 Leu3a OKT8 anti-LFA-1 LeuM1 Leull anti-IL-2R OKT10 Leu19 anti-Nu-TfRl anti-TcR-I anti-TcR 61 anti-SIg

CD2 CD3 CD4 CD8 CD11a CD15 CD16 CD25 CD38 CD56 -

T & NK subset T (antigenreceptorcomplex) T subset T & NK subsets T, B, monocyte,granulocyte Macrophage NK, PMN NK & T subsets Early thymoeyte NK & T subsets Lympbobtasts& macrophage T (TcRc~//3beterodimers) T (TcRy/6 heterodimers) B

in cold 1% paraformaldehyde in PBS, pH7.4. Samples were analyzed on a dual laser FACScan flowcytometry (Becton-Dickinson Inc.). Data analysis was performed on a Hewlett Packard computer by using the consort 30 program (Becton-Dickinson Inc.).

Cell sorting. LBCs stained with FITC-labeled mAbs as described above were purified by a fluorescence-activated cell sorter (FACStar; Becton-Dickinson Inc.) into positive and negative fractions as described [27]. IL-2 and IL-2 assays. Human recombinant IL-2 (rIL-2) with a specific activity of 1.0 x 107 U/mg [23] was supplied by Shionogi Pharmaceutical Co., Ltd. (Osaka, Japan) and used as a standard IL-2. The IL-2-dependent CTLL-2 cell line used in routine assays for the detection of IL-2 activity was maintained in RPMI1640 containing 10% FCS and 50 U/ml of rIL-2. Using this cell line, the assays of IL-2 activity were carried out [23]. Results are expressed as the mean and SD of specific 3H-TdR incorporation (cpm) in triplicate cultures [23-25]. IL-2 levels of 0.05U/ml or higher could be assayed by this method. Murine monoclonal antibody against human rIL-2 was supplied by Shionogi Pharmaceutical Co. Ltd. One unit of anti-rIL-2 was defined as the dose required to neutralize one unit of rIL-2. TU27 mAb [28], which reacts to the p75 subunit of IL-2 receptors expressed on the human lymphoid cells, was supplied by Dr. Sugamura (Tohoku University School of Medicine).

Percoll-discontinuous

gradient

centrifugation.

PBLs stimulated with streptococcal preparations were applied to a discontinuous density gradient of Percoll (Pharmacia Fine Chemicals), according to a slightly modified procedure of the methods previously described [29]. Briefly, seven different gravities of Percoll in PBS, ranging from d = 1.077 g and graded by 0.03 g concentration diminutions to d = 1.049g, were made with PBS. After carefully layering 1.5 ml of each solution into a 15-ml plastic tube, 4 × 10 7 cells in 2 ml of medium were placed on the top of the gradient and then centrifuged at 550g for 45 rain at room temperature. Cells from seven layers were collected from the top (fraction I) with a

56 Pasteur pipette and were washed two times with medium. The lowest density fraction (Fr.I), in which a few cells were recovered, contained 30% to 40% dead cells. Therefore, the other six fractions, Fr.II to Fr.VII, were generally used for the experiments. The recovery of the cells from Fr.II to Fr. VII was more than 90% of input, and the viability was 95% by the trypan blue exclusion method.

Complement-dependent cytolysis. Complement dependent cytolysis was applied to separate freshly isolated PBLs or low density cells (Fr.II or III) into populations depleted either of CD4 ÷ or CD8 ÷ cells or both. mAbs used were murine complement-fixing mAbs, OKT4 and OKT8. The cells which survived after these treatment were washed three times with RPMI1640 and used after adjustment of the cell numbers. Purity of the separated fractions was examined by a FACS analyzer after staining with fluorescein antibody against murine Ig. The complement used was low-toxicity rabbit C (Cedarlane Lab., Canada) and was used at a final concentration of 1/10.

[

i'°i o 0,ol

o.1

1.o

lo pg/m~

Fig. I. Dose response of human PBL cultures to varying concentrations of streptococcal preparations of OK-432 ( 0 ) and Sv-strains (O). Symbols represent mean cpm of 3H-TdR incorporation in PBL (2 × 106/ml) cultures from three normal donors ( - standard deviation, SD), cultures being terminated after 5 days.

sured on days 2, 4, 6 or 8. As shown in the results, 3H-TdR incorporation induced by either preparation was at the peak level on day 6, although strong reactivity was also observed on day 8. Repeated experiments using other preparations of T3 strain or clinical isolates of Str.

Results

Dose-response kinetics and time course of DNA replication induced by streptococcal preparations in human PBLs First, we determined the dose-response kinetics for streptococcal preparations of Su-strain (OK432) and Sv-strain in induction of D N A replication of human PBLs in culture. PBLs isolated from three individuals were incubated with the bacterial preparations at concentrations ranging from 0.01 to 10/zg/ml and assayed for DNA replication of the cells by 3H-TdR incorporation on day 5 (Figure 1). Maximum DNA replication was induced by either preparation at concentrations higher than 0.5/~g/ml. Next, the time course kinetics of D N A replication of PBLs induced by OK-432 and Sv-strain was examined. Typical results are illustrated in Figure 2, in which PBLs from a healthy donor were stimulated with either preparation at 1/xg/ ml. 3H-TdR incorporation by the cells was mea-

104

a +i o

g 0

.~_

10~

Q.

/

o

2

6 Culture days

Fig. 2. Time course of proliferative response of human PBL stimulated with OK-432 or Sv strain of heat-killed Str. pyogenes at 1/zg/ml. Human PBLs from a healthy donor were cultured at 2 × 106/ml in the presence of OK-432 or Sv strain for 8 days. At the indicated days, cells incubated with OK-432 ( 0 ) or Sv strain (©) were assayed for DNA replication by 3H-TdR incorporation pulsed for the final 18 h of incubation.

57

pyogenes yielded similar results: the maximum mitogenic response was obtained with all preparations tested at 0.5 to 5/xg/ml on days 6 or 7 (data not shown).

4O

10

A x

q

No involvement of IL-2 in OK-432-induced DNA replication

5

.a

g o

The time course kinetics of D N A replication of PBLs induced by streptococcal preparations is quite distinct from those induced by conventional T cell mitogens PHA, ConA or staphylococcal enterotoxins, which under these same culture conditions reached the peak on days 2 or 3 [30]. These D N A replications were mediated by IL-2, which was produced by the T cells stimulated by such mitogens. In a previous report [24], we also demonstrated that the stimulation of PBLs by OK-432 in a RPMI/AHS having the potent antitumor cytotoxicity, is not accompanied by IL-2 production during entire culture periods, suggesting that the induction of antitumor cytotoxicity by OK-432 may be a phenomenon mediated in an IL-2- independent manner. Therefore, we examined whether the proliferative response of human PBLs to OK-432 and other streptococcal preparations is also independent of IL-2 production. For this purpose, PBLs were cultured in the presence of OK-432 at 1/~g/ml in the RPMI/ AHS for varying times and assayed for IL-2 activity in the culture fluids by using a CTLL-2 cell line. PBLs stimulated with staphylococcal enterotoxin B (SEB) served as a control for IL-2 production. A representative result shown in Figure 3 indicates that no minimal levels of IL-2 (0.05 U/ml) could be detected in PBL cultures stimulated with OK-432 throughout the entire culture periods, although an active D N A replication with a peak on day 6 was induced. However, a large amount (9.5 U/ml) of IL-2 was detected in the PBL cultures stimulated with SEB. Repeated experiments with different individuals revealed that no IL-2 at the definite levels could generally be detected in the OK-432-stimulated PBLs. However, such IL-2 was detected in some cases at low but significant levels (0.2 U/ ml), particularly one or two days after incubation. Therefore, we examined the effect of mAbs reactive to human IL-2 or IL-2 receptors on the

2O

10

B

"7"

g .4

g s _a

10

I

g g

Ix:

0

2

4

6

8

10

Days after incubation

Fig. 3. DNA replication and IL-2 production of PBLs stimulated with staphylococcal enterotoxin B (SEB) (A) and OK432 (B). PBLs were cultured for 10 days in the presence of SEB at 0.1 ixg/ml (Sigma Chemical Co. St. Louis, Mo) or OK-432 at 1.0 p.g/ml. The values represent the mean value of 3H-TdR incorporation, cpm (O) and IL-2, units/ml ( 0 ) of triplicate specimens.

proliferation. The PBLs were stimulated with OK-432 at 1.0/~g/ml in the presence of anti-IL-2 mAb at 10 and 1 U / m l or mAb against p75 subunit of IL-2 receptors at 10 and 1U/ml. PBLs stimulated with SEB at 0.1/zg/ml were served as a control (Table 2). As shown in the table, OK-432-induced proliferation of PBLs is not suppressed by either mAb, whereas proliferation of PBLs stimulated by SEB is prominently inhibited by the antibodies. These results indicate that SEB and OK-432 extend their mitogenic effects on human PBLs via an IL-2dependent and an IL-2-independent pathway, respectively.

Generation of lymphoblastoid cells (LBCs) with heterologous surface markers and cytotoxicity against NK-sensitive and -insensitive targets Next, we characterized the cells growing in an

58

Table 2. Effect of anti-fL-2 and-fL-2 receptor sera on the mitogenic response of human PBLs to OK-432. 413

3H-TdR incorporation (cpm +-SD) Antiserum

PBL cultured with OK-432

PBL cultured with SEB

a-rhlL-2

10 U/mt 1 U/ml

24,200 -+ 1,800 29,200_+3,100 20,110-+2,563

29,040 -+4,080 4,400 _+2,750 ND

TU27

50/zg/ml 5/zg/ml

25,010 -+2,100 23,200-+3,200 20,810+ 900

29,534 -+5,470 6,040 +- 790 I5,050 -+ 1,690

One million/ml of PBLs were cultured with 1.0 fxg/ml of OK-432 for 5 days in the presence of anti-IL-2 serum or anti-IL-2 receptor serum, TU27 at the indicated concentrations. PBLs were also cultured with 0.1/xg/ml of SEB for 72 h in the presence of anti-IL-2 or TU27 sera. Both were then tested for 3H-TdR incorporation as described in Materials and methods. Data represent the mean -+SD of trioplicate cultures. OK-432-stimulated culture. PBLs from a donor were incubated with 1 . 0 / z g / m l of OK-432 and examined for D N A replication, the changes in the surface markers of cells and their cytotoxicity against an NK-insensitive target, FL cells, during culture. It was found that large lymphoblastoid cells (LBCs) with lobulated nuclei and abundant cytoplasm containing granules, which were smaller than those found in the h u m a n L G L [25], a p p e a r to increase in number. As shown in Figure 4, generation of these LBCs was associated with the induction of cytotoxicity against FL targets, which was also acc o m p a n i e d by an increase in the n u m b e r of cells bearing L e u l 9 that is an antigen expressed on activated N K and T cells. This is in contrast to a decrease in the cells bearing an N K cell antigen, CD16. These results strongly suggest the possibility that OK-432 induces cytotoxic effector cells, which are morphologically similar to L G L , but might be different from the N K cells with a typical p h e n o t y p e of C D 2 + C D 3 - C D 1 6 +.

Partial purification of LBCs and immunofluorescence analysis of their surface phenotype T o further characterize the LBCs and to determine the correlation between the LBCs and the cytotoxicity induced by OK-432, we applied the cells cultured with OK-432 to Percoll-discon-

40

-

tl

.¢

g 2C

20 ._o

u

0 0

2

4

6

8

Culturedays

Fig. 4. Correlation between inductions of cytotoxicity against NK resistant FL cells and Leul9 + lymphoblastoid cells (LBC) in OK-432-stimulated PBL. PBLs were cultured with OK-432 at 1/xg/ml for 8 days, and examined for cytotoxicity against NK-resistant target FL cells at an E/T ratio of 10:1 (N), and % of LBC-(O), Leul9-(&) and Leull-(ll) positive cells in culture. The data represent the mean of triplicate cultures. tinuous density gradient centrifugation to separate the LBCs from other cellular elements in the cultures. As shown in Figure 5, the LBCs are distributed in low-density fractions with peaks at F r . I I and at Fr. III, are coincident with a cytotoxicity against FL targets, and are separated f r o m the other cellular c o m p a r t m e n t s with higher densities. These results suggest that the LBCs may be responsible for the cytotoxicity induced. These blasts, detected in the OK-432-stimulated cultures, could be distinguished by FACS fluorometry on the basis of increased forward light scattering. T h e r e f o r e , low-density F r . I I and F r . I I I cells were collected and stained with various monoclonal antisera and analyzed for specific phenotypes of the blasts by specific immunofluorescence for each antiserum and cell size monitored by forward light scattering. Representative results on the surface phenotype of lymphoblasts stimulated with OK-432 are shown in Figures 6A and B. Figure 6A shows that the OK-432-induced lymphoblasts contain 45% CD3 + T cells, which may consist of 24% C D 4 + C D 8 -, 3% C D 4 - C D 8 + T cells, and 16% T cell receptor (TcR) y / 6 T cells. The results also show that the OK-432 blasts contain 36.3% C D 2 + C D 3 - lymphocytes. On the other hand, Figure 6B show that more than 60% of the blasts express L e u l 9 antigen, which is expressed on

59

Characterization of effector cells for the cytotoxicity induced by 0K-432

o ' ~c" 50

e Ft. ~ Ft. n

Fr.m Fr.IV

Fr, V

Fr.VI

Fr~Vll Unit.

Fig. 5. Correlation between the cytotoxicity against FL targets and the % of LBCs distributed in the different fractions of the Percoll-discontinuous density gradient centrifugation. PBLs stimulated with OK-432 at 1/zg/ml for 6 days were subjected to Percoll discontinuous density gradient centrifugation, and each fraction was assayed for cytotoxicity against FL target cells by 4 h 51Cr release assay at an E / T ratio of 10:1 (R), and % of positive LBCs (1~) determined with forward light scattering of FACScan.

active lymphocytes. In addition, almost all (95100%) of the LBCs also express TfR, LFA-1 and CD38 antigen, which are expressed in a variety of lymphocytes such as early thymocytes, NK cells or activated T cells. However, they express no CD16, which is an antigen associated with a typical population of NK cells. Neither surface Ig (B cells) nor CD15 (monocyte) was found in LBCs. It is also interesting to note that 96% of LBCs express antigen for IL-2 receptors, CD25. These results indicate that the LBCs induced by OK-432 contain CD3 ÷ T cells, particularly CD4 + helper / inducer type T cells, C D 2 ÷ C D 3 + C D 4 - C D 8 - T c R y / 8 + DN T cells and a C D 2 + C D 3 - C D 4 - C D 8 - C D 1 6 lymphocyte population(s). The third population may include early precursors of DN T cells, in which neither TcR~//3 nor TcR~//~ T cell receptors are expressed [31-33], and the other non-T lymphocyte populations such as NK cells, from which CD16 antigen expression was lost [33].

We characterized the phenotypic profiles of the effector cells in lymphoblastoid cells by cell separation in the complement dependent cell tysis by mAbs against CD4 and CD8. The results show that the fractions from which both CD4 + and CD8 + cells were depleted, exhibited much more active cytotoxicity against K562 and FL targets than CD4- or CD8- depleted fractions or unseparated LBCs (data not shown). Furthermore, the separation of CD3 + and CD3- DN lymphocytes by immunofluorescence cell sorting revealed that both fractions exhibit cytotoxicity against K562 and FL targets, although the extent of the activity is dependent on the experiment (Table 3). The results indicate that the populations responsible for the cytotoxicity may be both CD3+CD4-CD8 - DN T cells and CD2+CD3 DN lymphocytes.

Roles of CD4 + T cells in DNA replication and induction of the cytotoxicity As shown in Figure 6, CD4 ÷ helper type of T cell was one of the major populations in PBLs, which responded to stimulation by OK-432. It was also shown that these cells appeared not be responsible for the cytotoxicity generated in OK432-activated PBLs. It has, however, been shown that the CD4 + helper T cells may be necessary for the generation of cytotoxic effector cells, because depletion of CD4 + cells from the PBLs reduced the proliferation of activated PBLs with a marked reduction of induced cytotoxicity (Table 4).

Discussion

In the present study, we have demonstrated that incubation of human PBLs with OK-432 and other heat-killed preparations of Str. pyogenes results in a unique lymphoproliferation, which seems to reach the peak on day 6. This proliferative response of PBLs is in contrast with what has been observed with plant lectins such as P H A or ConA or microbial lectins, staphylococcal enterotoxins [30]. Those mitogenic responses

60

A

45.2

24.1

tO

16.3

.: t, :!

,i

2.9

:~ ~ , ~

E.L

........

I

..... ........

B

1

........

69.2

Leu19

TfR

98.1

LFA-1

y\ c 0

0

i

......

CD8

CD2

0.6

~CD16

CD15

3.2

TcRal

)

99.9

CD38

94.1

0.8

CD25

96.3

Sig

#,~ ., ++ i,

li L

if' !

I i

i

.......

/

.......

i

. . . . . . . . .

, ,-,,,+

........ ........

r

+ ...... i ,+

........

Log fluorescence intensity Fig. 6. Surface phenotype of partially purified LBCs with FACScan. A. The two color immunofluorescence test revealed surface phenotype of partially purified LBCs with Percoll discontinuous density gradient centrifugation. LBCs were stained with PE-anti CD3 mAb and FITC-anti CD2 mAb, PE-anti CD4 mAb and FITC-anti CD8 mAb, PE-anti CD3 mAb and FITC-anti TcR61. B. Single-color immunofluorescence test revealed surface phenotype of partially purified LBCs stained various mAbs. The numbers represented the % of cells positive for each mAb.

have a time course with a peak on day 2 or 3 different from those induced by Str. pyogenes. The reaction is also well known to be IL-2dependent. The proliferative response induced

by streptococcal preparations is not accompanied by IL-2 production and has also been found not to be inhibited by the addition of a specific antiserum against the human rlL-2 or IL-2 re-

Table 3. Cytotoxicity of OK-432 induced LBCs.

Table 4. Abrogation of proliferative response of PBLs by OK-432 after depletion of CD4, CD8 or both.

% Cytotoxicity b Effector ~ ( E / T ) 20:1 WholeLBCs CD4 C D 8 CD3-CD4-CD8 CD3+CD4-CD8 -

3H-TdR incorporation

K562

FL

10:1 5:1 20:1

10:1 5:1

57.4 43.4 96.7 78.4 84.7 82.3 I00 100

18.1 39.5 68.4 90.2 60.0 83.4 70.5 100

26.9 65.3 69.6 57.1

19.3 40.6 42.8 35.1

"PBLs were cultured with OK-432 (I.0/xg/ml) for 6 days before sorting. ~% cytotoxicity was determined with 4 h SaCr release assay against NK sensitive target K562 and NK resistant target FL cells.

Cell populations

cpm (mean +- SD)

% Reduction

Untreated + C c~CD4 + C aCD8 + C aCD4, a C D 8 + C

34,540 - 3,884 11,594 - 2,538 40,698 -+ 3,475 16,035 - 2,916

66.4% -17.8% 55.6%

The C-treated or C puls anti-CD4, or - C D 8 , or - C D 4 and - C D 8 treated PBLs were reconstituted to 1 × 106/ml. 3HTdR incorporation was measured on day 6 of culture. Depletion of CD4 +, CD8 + or both was confirmed by immunofluorescence test. Each value is the mean -+SD of triplicate cultures.

61 ceptor. In contrast, the mitogenic response to the microbial T-cell mitogen, SEB, tested in the present study is accompanied by IL-2 production. Some cases of human PBL cultures with OK-432 show the production of IL-2 at tow but significant levels, particularly in those incubated with medium fortified with FCS. These DNA replications are, however, inhibited only partially by the antisera (data not shown). These results indicate that the DNA replication induced by streptococcal preparations may not be dependent on IL-2. It has been reported that all the stimuli that induce proliferation of T cells initiate the synthesis of IL-2 and the formation of IL-2R on the cell surface. However, recent work, including studies demonstrating IL-2-independent proliferation in some murine T cell clones [34], suggests the existence of an additional non-IL-2-independent pathway of proliferation. The existence of an IL-2-independent pathway of human T cells proliferation has been demonstrated in PBL culture stimulated with mAb UCTH1 [35]. The existence of IL-2 and IL-4 independent pathway has also been demonstrated in PBL culture stimulated with soluble mAb against OKT3 alone, although the PBL cultures stimulated with anti-OKT3 plus PMA produce IL-2 [36]. We showed that DNA replication in murine syngeneic or human autologous mixed lymphocyte reaction (MLR) may not be IL-2-dependent [23]. In the present study, we have also demonstrated that this OK-432-induced proliferative response of human PBLs is accompanied by the generation of lymphoblastoid cells (LBCs) consisting of heterologous T cell populations. Partial purification by Percoll-discontinuous density centrifugation and immunofluorescence tests revealed that the LBCs contained CD4 ÷ helper type of T cells, CD3+CD4-CD8 - T c R 7 / 6 ÷ DN T lymphocytes and CD2+CD3-CD4-CD8 - DN lymphocytes, in which neither TcRa//3 nor TcRT/6 T cell receptors were expressed. Almost all of these cell populations expressed the activated lymphocyte antigens, TfR, LFA-1 and CD38, but neither CD16 NK cell antigens nor B cell markers. The proliferative response of these LBCs was also accompanied by induction of cytotoxicity against both NK-sensitive and NK-insensitive

tumor targets. The present results indicate that although the streptococcal preparations stimulate the PBLs to induce activation/proliferation of CD4 ÷ helper T cells, they may not be responsible for the cytotoxicity induced. It has, however, been shown that the removal of CD4 ~ T cells before stimulation by OK-432 results in the reduced proliferation of PBLs and induction of cytotoxicity, suggesting that the CD4 ÷ T cells may be required for the generation of effector cells in the OK-432-induced cytotoxicity. On the other hand, complement-dependent lysis by CD4 and CD8 mAbs and separation of CD3 ÷ and CD3- DN lymphocytes by cell sorting indicate that both the DN lymphocytes of CD3 ÷ and CD3- may be responsible for the cytotoxicity. The vast majority of CD3+4-8 - T cells found in the thymus lack the conventional TcR a//3 complex but express TcR y/6-CD3 receptor complex [37]. It has been reported that about 3% of normal peripheral blood T lymphocytes belong to these DN TcRT/6 T cells [38]. These TcR,//6 T cells, which are found in the thymus and peripheral lymphoid tissues, can mediate an NK-like cytotoxicity [39]. The present study has indicated that streptococcal preparations stimulate, directly or indirectly, the proliferation of a small number of CD3 + TcR7/6 + DN T cells present in the PBLs. Recent studies have also shown that the early thymocytes or prothymocytes can be divided into discrete subsets by using a variety of other markers, although the developmental relationships of the various subsets have not been established [31, 32]. Among them, CD2+3-4-8 - T cells, which express neither a//3 nor ,//6 TcR, may belong to the most phenotypically immature population. They can be found in the thymus and in the fetal liver, and may probably be found in the peripheral lymphoid organs in small numbers, in addition to CD2+3+4-8 - DN T cells [31,32]. It may be possible that OK-432 also stimulates the proliferation of a small number of these immature population of T cells. It has, however, been presumed that activated NK cells from which CD16 expression is lost, might be induced by OK-432. The present results also show that these cells, which are induced by streptococcal preparations, regardless of being positive or negative for CD3

62 complex, express IL-2 receptors on the cell surface. The same function has been reported in DN T cells which are isolated from the PBLs, grow in response to IL-2, and generate cloned cells with a DN T phenotype and cytotoxicity [40, 41]. In another study [42], we have also shown that induction of TcRy/6 + DN T cells by streptococcal preparations can be detected by Northern hybridization using cDNA for ,/and 6 chains of TcR. It has also been found that the factor responsible for the activity of the bacterial preparation stimulating y6 T cells is a heat-stable 10-15 kDa protein(s) of cytoplasmic membranes, which is antigenicaUy different from the heat shock proteins (Itoh et al. in preparation). The same activity inducing polyclonal proliferation of DN T cells has been found in other gram-positive bacteria such as Propionibacterium acnes or Mycobacterium tuberculosis, in which adjuvanticity and anti-tumor effects can also be seen [43]. Further investigations on this polyclonal activation of PBLs by microbial agents, inducing the activation of TcR 7/6 cells and other lymphocytes including early precursors of DN T cells or NK cells, may shed light upon the adjuvanticity and anti-tumor effects of these agents, and may also provide valuable information on the pathogenesis of inflammatory reactions and tissue injury triggered by infections induced by grampositive bacteria such as hemolytic streptococci and the immunological disorders induced following such infection.

Acknowledgement This work was supported in part by Grant-in-Aid for Cancer Research from the Ministry of Education, Science and Culture, Japan.

References 1. Hibbs JB Jr, Remington JS, Stewart CC. Modulation of immunity and host resistance by microorganisms. Pharmacol Ther 1980; 8: 37-69. 2. Bast RC, Bast BS, Rapp HJ. Critical review of previously reported animal studies of tumor immunotherapy with nonspecific immunostimulants. Ann N Y Acad Sci 1976; 277: 60-93.

3. Scott MT, Milas L. The distribution and persistence in vivo of Corynebacterium parvurn in relation to its antitumor activity. Cancer Res 1977; 37: 1673-9. 4. Bank G, Forsgren A. Many bacteria species are mitogenic for human blood B lymphocytes. Scand J Immunol 1978; 8: 347-54. 5. Riis~inen L, Karhum~iki E, Marjuri R, Arvilommi H. Bacteria induce lymphokine synthesis potyclonally in human B lymphocytes. J Immunol 1978; 121: 418-20. 6. R~is~inen L, Karhum/iki E, Majuri R, Arvilommi H. Polyclonal activation of human lymphocytes by bacteria. Infect Immun 1980; 28: 368-72. 7. Clagett JA, Engle D. Polyclonat activation: a form of primitive immunity and its possible role in the pathogenesis of inflammatory diseases. Dev Comp Immunol 1978; 2: 235-42. 8. M611er E, Str6m H, AI-Balagi S. Role of polyclonal activation in specific immune responses. Scand J Immunol 1980; 12: 177-82. 9. Kasahara T, Harada H, Shioiri-Nakano K, Wakasugi H, Imai M, Mayumi M, Sano T, Sugiura A. Potentiation of natural killer cell activity of human lymphocytes in vitro: the participation of interferon in stimulation with Staphylococcus aureus Cowan I bacteria but not with Protein A. Immunology 1982; 45: 687-95. 10. Lapham C, John PA, Tomar RH. The mechanism of enhancement of natural killer cell activity by soluble streptococcal products. Clin Immunol Immunopathol 1986; 40: 335-46. 11. Buck GE, Kelly MT. Investigation of the component of Propionibacteriurn acnes (Corynebacteriurn parvum) responsible for macrophage activation. Infect lmmun 1980; 27: 620-7. 12. Kotani S, Watanabe Y, Shimono T, Narita T, Kato K, Stewart-Tull DES, Kinoshita F, Yokogawa K, Kawata S, Shiba T, Kusumoto S, Tarumi Y. Immunoadjuvant activities of cell walls, their water-soluble fractions and peptidoglycan subunits prepared from various gram-positive bacteira and of synthetic N-acetylmuramyl peptides. Z lmmun Allergieforsch 1975; 149: 302-19. 13. Wilder RL, Allen JB, Wahl LM, Calandra GB, Wahl SM. The pathogenesis of group A streptococcal cell walt-induced polyarthritis in the rat: comparative studies in arthritis resistant and susceptible inbred rat strains. Arthritis Rheum 1983; 26: 1442-51. 14. Miller GA, Urban J, Jackson RW. Effects of a streptococcal lipoteichoic acid on host responses in mice. Infect Immun 1976; 13: 1408-17. 15. Okamoto H, Shoin S, Koshimura S, Shimizu R. Studies on the anticancer and streptolysin S-forming activities of hemolytic streptococci. Japan J Microbiol 1967; 11: 3236. 16. Saito M, Nanjo M, Aonuma E, Noda T, Nakadate I, Ebina T, Ishida N. Activated macrophages are responsible for the tumor inhibitory effect in mice receiving intravenous injection of OK-432. Int J Cancer 1987; 33: 271-6. 17. Oshimi K, Kano S, Takaku F, Okumura K. Augumentation of mouse natural killer cell activity by a streptococ-

63

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

cal preparation, OK-432. J Natl Cancer Inst 1980; 65: 1265-9. Hojo H, Hashimoto Y. Cytotoxic cells induced in tumorbearing rats by a Streptococcus preparation (OK-432). Gann 1981; 72: 692-9. Torisu M, Katano M, Kimura Y, Itoh H, Takesue M. New approach to management of malignant ascites with a streptococcal preparation, OK-432. I. Improvement of host immunity and prolongation of survival. Surgery 1983; 93: 357-64. Watanabe Y, Iwa T. Clinical value of immunotherapy for lung cancer by the streptococcal preparation, OK-432. Cancer 1984; 53: 248-53. Ebihara K, Minamishima Y. Protective effect of biological response modifiers on murine cytomegalovirus infection. J Virology 1984; 51: 117-22. Saito H, Watanabe T, Kitagawa T, Asano K. Protective effect of bacterial immunopotentiators, OK432 and LC9018 on Pseudomonas aeruginosa infection in tumor bearing mice. Hiroshima J Medical Science 1985; 34: 459-62. Suzuki R, Suzuki S, Takahashi T, Kumagai K. Production of a cytokine with interleukin 3-tike properties and cytokine-dependent proliferation in human autologous mixed lymphocyte reaction. J Exp Med 1986; 164: 168299. Fujii M, Abo T, Kumagai K. Cytokines produced by blood mononuclear cells stimulated with the streptococcal preparation OK-432: effect on production by supplementing the medium with xenogeneic serum. Cancer Immunol Immunother 1988; 27: 97-102. Itoh K, Tilden AB, Kumagai K, Batch CM. Leull ÷ lymphocytes with natural killer (NK) activity are precursors or recombinant interleukin 2 (rlL2)-induced activated killer (AK) cells. J Immunol 1985; 134: 802-7. Faure F, Jitsukawa S, Triebel F, Hercend T. Characterization of human lymphocytes expressing the CD3-T/6 complex with anti-receptor monoclonal antibody. J lmmunol 1988; 141: 3357-60. Abo T, Cooper MD, Balch CM. Characterization of HNKI ÷ (Leu-7) human lymphocytes. I. Two distinct phenotypes of human NK cells with different cytotoxic capability. J Immunol 1982; 129: 1752-7. Takeshita T, Goto Y, Tada K, Nagata K, Asao H, Sugamura K. Monoclonal antibody defining a molecule possibly identical to the p75 subunit of interleukin 2 receptor. J Exp Med 1989; 169: 1323-32. Arai S, Yamamoto H, Itoh K, Kumagai K. Suppressive effect of human natural killer cells on pokeweed mitogen-induced B cell differentiation. J Immunol 1983; 131: 651-7. Langford MP, Stanton GL, Johnson HM. Biological effects of staphylococcal enterotoxin A on human peripheral lymphocytes. Infect Immun 1978; 22: 62-8. Haynes BF, Martin ME, Kay HH, Kurtzberg J. Early

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

events in human T cell ontogeny. Phenotypic characterization and immunohistological localization of T cell precursors in early human fetal tissues. J Exp Med 1988; 168: 1061-80. Denning SM, Kurtzberg J, Le PT, Tuck DT, Singer KH, Haynes BF. Human thymic epithelial cells directly induce activation of autologous immature thymocytes. Proc Natl Acad Sci USA 1988; 85: 3125-9. Nagler A, Lanier LL, Cuira S, Phillips JH. Comparative studies of human FcRllI-positive and negative natural killer cells. J Immunol 1989; 143: 3183-91. Moldwin RL, Lancki DW, Herold KC, Fitch FW. An antigen receptor-driven interleukin 2-independent pathway for proliferation of murine cytolytic T lymphocyte clones. J Exp Med 1986; 163: 1566-82. Van Wauwe JP, Goossens JG, Beverley PCL. Human T lymphocyte activation by monoclonal antibodies; OKT3, but not UCHT1, triggers mitogenesis via an interleukin 2-dependent mechanism. J Immunol 1984; 133: 129-32. Laing TJ, Weiss A. Evidence for IL-2 independent proliferation in human T cells~ J Immunol 1988; 140: 105662. Brenner MB, McLean J, Dialynas DP, Strominger JL, Smith JA, Owen FL, Seidman JG, Stephen Ip, Rosen F, Krangel MS. Identification of a putative second T-cell receptor. Nature 1986; 322: 145-9~ Suzuki R, Suzuki S, Igarashi M, Kumagai K. Induction of interleukin 3 but not interleukin 2 or interferon production in the syngeneic mixed lymphocyte reaction. J Immunol 1986; 137: 1564-72. Lanier LL, Ruitenberg J J, Phillips JH. Human CD3 ÷ T lymphocytes that express neither CD4 nor CD8 antigens. J Exp Med 1986; 164: 339-44. Brenner MB, McLean J, Scheft H, Riberdy J, Ang S, Seidman JG, Devtin P, Krangel MS. Two forms of the T cell receptor y protein found on peripheral blood cytotoxic T lymphocytes. Nature 1987; 325: 689-94. Ramsdell FJ, Golub SH. Generation of lymphokineactivated killer cell activity from human thymocytes. J Immunol 1987; 139: 1446-53. Abo T, Sugawara S, Seki S, Fujii M, Rikiishi H, Takeda K, Kumagai K. Induction of human TcRT6 + and T c R y 6 - CD2+3 - double negative lymphocytes by bacterial stimulation. Int Immunol 1990; 2: 775-85. Thole JER, Hindersson P, DeBrugn J, Cremers F, van der Zee J, de Cock H, Tommassen J, van Eden W, van Embden JDA. Antigenic relatedness of a strongly immunogenic 65 kDa mycobacterial protein antigen with a similar sized ubiquitous bacterial common antigen. Microbiol Pathogenesis 1988; 4: 71-83.

Address for correspondence: K. Kumagai, Department of Microbiology, Tohoku University School of Dentistry, Sendai 980, Japan.