Studies on Thymocyte Subpopulations in Guinea Pigs VIICharacterization of Cell Populations Responsive to Guinea Pig Interleukin 1 and Interleukin 2

Immunobiol., vol. 170, pp. 448-459 (1985)

Department of Histology, Karolinska Institute, Stockholm, Sweden

Studies on Thymocyte Subpopulations in Guinea Pigs VII. Characterization of Cell Populations Responsive to Guinea Pig Interleukin 1 and Interleukin 2 G. SANDBERG, O. SODER, and]. TJERNBERG Received May 2, 1985 . Accepted in Revised Form September 10, 1985

Abstract The properties of thymocytes responding by proliferation to a mitogenic lectin (PHA), interleukin 1 (IL 1), or interleukin 2 (IL 2) were studied and compared to the properties of cells known to respond to a separate thymocyte growth factor (TGP), which has so far only been studied in guinea pigs. Thymocytes from guinea pigs were separated into subpopulations by density gradient centrifugation with Percoll and by rosette formation with rabbit erythrocytes. PHA-responsive cells were recovered exclusively in a non-rosetting, low-density population designated la, RFC-, constituting approximately 4 per cent of all thymocytes. Thus, according to the prevailing view of lymphocyte mitogenesis, IL I-producing, IL2-producing as well as IL 2-responding cells are all present in this population. The mitogen-responsive cells could be further stimulated by addition of IL 1 or IL 2, indicating that the magnitude of the mitogenic response was regulated by the production of these factors and was not restricted by the number of IL 2-responding cells. IL 1, to some extent, also enhanced the mitogen response in a non-rosetting, intermediate density population designated Ib, RFC-. None of the factors could affect the lack of mitogen responsiveness in the high-density population II, constituting approximately 85 % of the total population and probably including most small cortical thymocytes. We conclude that IL 2-responding thymocytes are present above all in the quantitatively small population la, RFC- (shown to contain also the mature, mitogen-responding cells) and to a smaller extent in Ib, RFC - (where a deficit in IL I-producing cells may explain the poor mitogen responsiveness), but not in the major, high-density population II. From our data, it is also evident that the mitogen, IL 1- and IL 2-reactive cells can be separated from a population of intensely proliferating thymic precursor cells which are stimulated to grow by TGP. Therefore, the growth of these immature cells does not seem to be regulated by IL 1 or IL 2 in the guinea pig.

Introduction

Considerable effort has been made during the last decades to gain a better insight into the intrathymic events behind the ontogenic development of cell-mediated immunity. Most studies have been performed in mice and rats Abbreviations: a-MEM = a-modified Eagle's Minimal Essential Medium; ConA = Concanavalin A; FCS = fetal calf serum; IL 1 = interleukin 1; IL 2 = interleukin 2; PHA = phytohemagglutinin; TGP = thymocyte growth peptide.

IL 1 and IL 2 Responsive Cells in Guinea Pig Thymus . 449

due to the availability in these species of several inbred strains and, more recently, due to the introduction of monoclonal antibodies to cell surface antigens defining several distinct subpopulations of thymocytes. Despite the recent progress, several basic issues concerning thymocyte differentiation remain unsolved. Among these are the pathway(s) of intrathymic cell migration and the mechanism and purpose of the intense proliferation and extensive cell death of cortical thymocytes in young animals. In a series of experiments, we examined the properties of guinea pig thymocytes. A population of large, low-density cells called la, constituting approximately 10-25 % of the thymocytes, has been shown to contain the majority of both immature proliferating cells and immunologically mature cells (1). These two populations were separated in our previous work on the basis of peanut agglutinin (PNA) binding, the immature cells being PNA + and the mature, mitogen-(PHA, ConA) responsive cells being PNA-. When the Ia population was instead subdivided by fractionation according to differential rosette formation with rabbit erythrocytes (a T cell and thymocyte marker in guinea pigs), mitogen-reactive cells could be distinguished from those responding to a thymocyte growth factor (TGP) (2). On the basis of those studies, the mitogen-responsive thymocytes could be characterized as la, PNA -, RFC-. The majority of thymocytes, small, high-density cells with low proliferative rate and corresponding to the typical cortical cells, responded neither to mitogenic lectins nor to TGP (2). The activation of lymphocytes with mitogenic lee tins involves the cooperation of several cell types (3). Therefore, the inability of some fractionated thymocyte populations to respond to mitogenic lectins might be due to the lack of any of these cell types, not only of the cell type eventually responding by blast transformation - the interleukin 2 (IL 2)-responsive cells. Conflicting data on IL 2-responsive thymocyte subpopulations have emerged from studies in mice. The present work was undertaken to study the nature of interleukin-responding cells in guinea pigs for comparison with those responding to TGP, known to stimulate proliferation of immature thymocytes. The possible involvement of these factors in regulation of normal thymocyte proliferation is discussed. Preliminary results of this investigation were presented at the 8th International Conference on Lymphatic Tissues and Immune Reactions (4) and the 16th International Leukocyte Culture Conference (Cambridge, U.K., 1984) (5).

Materials and Methods Preparation of thymocytes Male, 4- to 6-week-old guinea pigs (pigmented, rough-haired strain), weighing 250--300 g were used as cell donors. Suspensions of thymus cells were obtained as described previously (1).

450 . G. SANDBERG, O. SODER, and J. TJERNBERG Fractionation by density gradient centrifugation Nine vol Percoll® were mixed with 1 vol of 10-fold concentrated Hanks' buffer. Four ml of a thymocyte suspension were mixed with 7 ml of the Percoll-Hanks' mixture and were centrifugated at 15,000 X g for 11 min. Population Ia (large, low-density cells) was then collected from the upper 2 ml, Ib (intermediate-size and -density) from the next 3.5 ml, and population II (small, high-density cells) from the lower 5.5 ml of the gradient (1). Fractionation on the basis of rosette formation This procedure has been described in detail previously (2). In short, rosettes were prepared by incubating guinea pig thymocytes with rabbit erythrocytes for 1 h. Rosetted cells were separated by layering over a Percoll solution (density 1.0931 g/ml), centrifugation for 30 min at 400 X g, after which the RFC+ cells were obtained from the pellet, whereas the interphase cells were depleted of rosettes. The method was modified (higher density of Percoll) due to a slightly different behaviour of rosetted Ia cells as compared to rosetted normal thymocytes. After rosetting and separation on Percoll, 91 % of the normal thymocytes and 59 % of Ia cells were recovered as RFC + This corresponds to the fact that Ia cells have a lower rosette frequency than un separated thymocytes (6). All glass fiber filters containing the harvested cells were bleached by addition of one drop of 35 % H 20 2 to avoid quenching during liquid scintillation counting. This treatment resulted in approximately 40 % higher counts. Comparison of untreated and un separated thymocytes with rosetted, but likewise unseparated cells indicated no effect of rosette formation nor the presence of erythrocytes during culture on the comito genic effect of IL 2, but the direct effect seemed to have been inhibited. Both background proliferation and the effect of PHA were enhanced in the cultures containing erythrocytes. Preparation of guinea pig IL 1 Peritoneal exudate cells were induced in guinea pigs by instillation of 20 ml sterile thioglycollate (3 %). After 5 days, the peritoneal cavity was rinsed extensively with 40-50 ml a-MEM (37"C) containing Heparin (125 IU/ml). The obtained cells were washed twice and were resuspended to approximately 2 X 106 cells/ml in a-MEM (containing 100 IU penicillin, 100 [.tg streptomycin and 2 [.tmol glutamine per ml). The cells were incubated in 35-mm Petri dishes (3-4 ml per dish) for 1 h at 37°C in 5 % CO 2-humidified air. Nonadherent cells, probably including all lymphoid cells, were then removed by pouring off the medium and rinsing the dish with a-MEM repeatedly. New medium containing a-thioglycerol (7.5 X 10- 5 M), indomethacin (10- 5 M), phorbol myristate acetate (10 [.tg/ml) and 5 % fetal calf serum was then added, and the dishes were incubated for another 24 h at 37°C. The cell-free supernatant was then filtered (Micro flow 25, 0.2 [.tm), and ammonium sulphate precipitated (80 % saturation) at 4°C. After at least 1 h, the precipitate was centrifuged at 10,000 X g for 20 min, dissolved in 2 ml distilled water, and was dialyzed against 11 of distilled water followed by PBS for a total of 24 h and filtered as above. The material was stored at -20°C until use. The activity of IL 1 preparations was determined in a comito genic assay with guinea pig thymocytes as described below. Preparation of guinea pig IL 2 Guinea pig spleen cells were suspended in a-MEM (5 X 106/ml) containing 5 % fetal calf serum, a-thioglycerol (7.5 X 10- 5 M), and Concanavalin A (ConA) covalently bound to Sepharose beads (Con A-Sepharose; 10 % washed suspension, 30 [.tUml cell suspension) and were cultured in portions of 15 ml in tissue culture flasks (Lux, 50 ml) for 24 h. The cell-free supernatant (activation of spleen cells verified in replicate cultures at 48 h) was ammonium sulphate precipitated (80 % saturation) during 24 h at 4°C. The precipitate was centrifuged at 10,000 X g for 20 min, was dissolved in 2 ml distilled water and dialyzed against 11 distilled water overnight and then was dialyzed against the same volume of phosphate-buffered saline (PBS) for 12h. The material was stored at -20°C until use. To avoid the influence oftraces of Con A present in the supernatants, IL 2 preparations were supplemented with a-methyl- D-

IL 1 and IL 2 Responsive Cells in Guinea Pig Thymus . 451 mannoside (0.2M in dialyzed IL2 stock solutions). If this step was omitted, IL2 displayed a strong direct mitogenic effect with un separated thymocytes. IL 2 preparations were assayed in a «maintenance assay» by first culturing Con A-activated spleen cells (5 x 106 cells/ml, 2 [.lg/ml) for 72 h in the presence of 5 % FCS. The activated spleen cells were then washed and recultured (5 x 105 cells/ml in a-MEM with a-thioglycerol (7.5 X 10- 5 M) and 5 % FCS) in the presence of serial dilutions of IL2 and were harvested at 48 h following a 2-h-pulse of tritiated thymidine. Tests of response

to

IL 1 and IL 2

Both interleukin preparations were tested in combination with PHA (5 [.lg/ml) in a comitogenic assay with guinea pig thymocytes. IL 1 or IL 2 were added in doses found optimal for comitogenic stimulation of DNA synthesis in unseparated thymocytes. The cell populations to be tested were cultured (5 X 105 cells in 100 [.ll a-MEM supplemented with penicillin, streptomycin and a-thioglycerol (7.5 X 10- 5 M)) in flat-bottomed Linbro plates for 48 h. Before harvest, a 2-h-pulse with tritiated thymidine (0.5 [.lCi in 10 [.ll) was given. The cells were collected on glass fiber filters with a Skatron harvester and were washed extensively with distilled water. In experiments with rosetted cells, the filters were bleached by HzO z as described above. Incorporated radioactivity was determined by liquid scintillation counting and was expressed as mean cpm of duplicate cultures. Effects of added substances were given as dcpm, that is difference in cpm as compared to the appropriate controls. In the case of the interleukins, the comito genic effect was obtained by comparing with cultures that were given only mitogen, whereas a direct effect on DNA synthesis was detected by comparing with cells cultured only in medium. Manufacturers of materials and drugs

The following materials and drugs were used: a-MEM (a-modified Eagle's Minimal Essential Medium, Flow Laboratories, England), a-methyl-D-mannoside (grade III) and athioglycerol (Monothioglycerol) (Sigma, St. Louis, MO, U.S.A.), cell harvester (Skatron, Flow Laboratories), ConA and ConA-Sepharose (Pharmacia Fine Chemicals, Uppsala, Sweden), FCS (fetal bovine serum, Flow Laboratories), Heparin (5000 !E/ml without preservative, KabiVitrum, Sweden), indomethacin (Sigma), Microflow filters (Microflow 25, Flow Laboratories), Percoll (Pharmacia Fine Chemicals), Petri dishes (Falcon 60 X 15 mm, Becton Dickinson, U.S.A.), PHA (PHA-P, Difco Laboratories, Detroit, U.S.A.), phorbol (phorbol12-myristate-13-acetate, Sigma), thioglycollate (Brewer thioglycollate medium, Difco Laboratories), thymidine (TRA 120, Amersham International pic, Amersham, England), tissue culture flasks (LUX 5350, Lab-Tek Division, Miles Laboratories, Inc., Naperville, U.S.A.), tissue culture plates (Titertek, 96-flat bottom wells, Flow Laboratories).

Results

Characterization of cells responding to PHA Separation of thymocytes according to their density resulted in recovery of mitogen-responsive cells among the low-density cells la. Of these, the non-rosetting cells included almost all responding cells (Table 1). Thus, a distinct population of low-density, non-rosetting thymocytes, constituting approximately 4 % of all thymocytes, contained all mitogen reactive cells.

Characterization of cells responding to IL 1 The low-density cell population la responded most vigorously to the comitogenic effect of IL 1, but there was also some effect on population lb.

452 . G. SANDBERG, O. SODER, and

J. TJERNBERG

Table 1. Effect of PHA on thymocyte subpopulations Cell type

No. of experiments

RFC+ RFC Ia Ia,RFC+ la, RFC Ib

4 4 10 6 6 13 13

II

~cpm

5533 ± 25028 ± 12450 ± 3799 ± 68049 ± 3405 ± 77±

2278' 10 550 2300 876 17067 660 28

Cell number (% of all thymocytes) 91 9

9 5 4 6 85

PHA (5 ~g/ml) was added to cultures of fractionated thymus cells (5 x 105 in 100 ~l) at the start of incubation. Incorporation of tritiated thymidine was measured at 48 h following a 2-hpulse. The effect of PHA is expressed as the difference in cpm (~cpm) when compared to control cultures. The frequency of each subpopulation in the intact thymus was estimated from results on distribution after the various separation steps. , Mean±SEM.

The high-density population II did not respond at all (Table 2). When la cells were further subdivided into RFC+ and RFC- populations, only the RFC- cells were responding in the assay (Table 2). In addition to the strong comitogenic effect of IL 1, there was also a smaller direct effect (Table 2). Also in this case, it was the la, Ib, and RFCcells which were most responsive. In all six experiments performed the RFC+ cells were inhibited. Table 2. Direct and comito genic effects of IL 1 on thymocyte subpopulations Cell type

No. of experiments

~cpm

A. Direct effect Ia la, RFC+ la, RFC Ib

II

7 6 6 10 10

1530 ± -3105 ± 680 ± 888 ± 55 ±

454' 797 175 238 28

B. Comito genic effect Ia la, RFC+ la, RFC Ib II

7 6 6 10 10

27023 ± -63 ± 20726 ± 5294 ± 217 ±

5825 1281 5902 2396 79

Optimal amounts of IL 1 were added to thymocyte subpopulations (5 x 105 in 100 ~l) at the start of incubation. Incorporation of tritiated thymidine was measured at 48 h following a 2-hpulse. The direct effect of IL 1 is expressed as difference in cpm (~cpm) when compared to control cultures without IL 1 or mitogen, whereas the comitogenic effect is obtained by comparison with mitogen- (PHA) treated cultures. , Mean ± SEM.

IL 1 and IL 2 Responsive Cells in Guinea Pig Thymus . 453

5000

"\"

2500

Fig. 1. Effect of IL 1 in different dilutions on thymocytes cultured in the presence (-) or absence (- - -) of PHA. The comito genic effect is expressed as activity (~cpm) in comparison with cultures treated with PHA only.

[~

",. ..... - ..... 0

"<\. "-

+-..,.......-r--"T-....,: ~~,.......p1

2

4

8

16 32 64 128

Reciprocal dilution

I 40000 " \

"\

20000

Fig. 2. Effect of IL 2 in maintenance assay with preactivated spleen lymphoblasts. Means of two separate IL 2 preparations tested on different occaSIOns.

\ ~.

E

fr ~

~.

o+--.....--~-~--~--; 2

4

8

16

32

Reciprocal dilution

""oj .~ 5000

.~ .~

Fig. 3. Direct and comito genic effects of different dilutions of IL 2 on thymocytes. The IL 2 preparations were supplemented with a-methyl-D-mannoside. The comito genic effect was studied in PHA-treated cultures.

E

c..

U

4

.----.

O:----:----~----i----' 2

4

8

Reciprocal dilution

16

454 . G. SANDBERG, O. SODER, and

J. TJERNBERG

Table 3. Direct and comitogenic effects of IL 2 on thymocyte subpopulations Cell type

Direct effect ticpm

Co-mitogenic effect ticpm

Cell number (% of all thymocytes)

Ia Ib II

2136 ± 616' 725 ± 333 72 ± 38

14152±3031 3481 ± 1410 217± 139

12 6 82

Optimal amounts of IL 2 with a-methyl-D-mannoside were added to the separated thymocyte subpopulations at the start of incubation. Incorporation of tritiated thymidine was measured at 48 h following a 2-h-pulse. The direct effect of IL 2 is expressed as difference in cpm (ticpm) when compared to cultures treated without IL 2, whereas the comito genic effect is obtained by comparison with mitogen-(PHA) treated cultures. The frequency of each subpopulation in the intact thymus was estimated from results on distribution after fractionation by density gradient centrifugation. 'Mean ± SEM (n = 3).

The dose-response of direct and comito genic effects of IL 1 is shown in Figure 1. Characterization of cells responding to IL 2

The effect of IL 2 on maintenance of Con A-activated splenic blasts is shown in Figure 2. IL 2 supplemented with a-methyl-D-mannoside displayed a dose-dependent comito genic effect in combination with PHA but Table 4. Direct and comito genic effects of IL 2 on rosetting (RFC+) and non-rosetting (RFC-) thymocyte subpopulations ticpm Dose of IL 2

Cell type 1 :1

1 :4

A. Direct effect Control RFc+ RFC

-120 ± 103' -724 ± 395 2181 ± 378

53 ± 20 -426 ± 235 2252 ± 245

B. Comito genic effect Control RFC+ RFC

6948 ± 1 231 3352 ± 171 14242 ± 4837

3052 ± 789 1 057 ± 262 16518 ± 6 166

IL 2 with a-methyl-D-mannoside was added to rosetting (RFC+) or non-rosetting (RFC-) thymocyte populations at the start of incubation. Control cells were the whole thymus populations after addition of rabbit erythrocytes but before separation into RFC+ and RFCcells. Incorporation of tritiated thymidine was measured at 48 h following a 2-h-pulse. The direct effect of IL 2 is expressed as difference in cpm (ticpm) when compared to cultures treated without IL 2 or mitogen, whereas the comito genic effect is obtained by comparison with mitogen-(PHA) treated cultures. After separation, 91 % of the cells were recovered as RFC+, 9% as RFC. , Mean ± SEM (n = 4).

IL 1 and IL 2 Responsive Cells in Guinea Pig Thymus . 455

lacked direct mitogenic effect on unseparated thymocytes (Fig. 3). It was evident that the comitogenic effect of IL 2 was restricted principally to population la (Table 3). Separation of cells on the basis of rosette formation gave an enrichment of responding cells in the RFC- population (Table 4). A small, direct mitogenic effect of IL 2 was seen on RFC-, whereas RFC+ cells were inhibited.

Background proliferation The magnitude of the effects of mitogens and interleukins is better appreciated when a comparison can be made with untreated cultures. These background activities (mean cpm) are therefore summarized here. The uptake of 3H-thymidine was measured after 48 h of incubation in a-MEM and a 2-h-pulse of the isotope. Normal thymocytes: 185; RFC+: 2260; RFC-: 1245; la: 1897; Ib: 144; II: 57; la, RFC+: 5002; la, RFC-: 2327. The higher values after rosetting are due to enhancement of background proliferation by the presence of erythrocytes (see «Materials and Methods»).

Discussion Mitogen-reactive cells in the thymus are usually considered to represent a minority of immunologically mature, T cell-like, medullary, PNA - thymocytes with the capacity to develop IL 2 receptors and to respond to IL 2. The IL 2-producing cells may belong to another or partly identical mature population (7-9). The IL 2 production, in turn, seems to depend on a factor (IL 1) produced by non-lymphoid cells (10). Upon fractionation of thymus cells, the dissociation of this co-operating entity may result in an inability to respond to a mitogenic stimulus. We have previously described the reactivity of normal and fractionated thymus cell populations in guinea pigs to mitogenic lectins, and we reported that this ability was restricted to, and was enriched in population la, PNA -, RFC- (1, 2). We also found that the la, RFC- population contained the IL 2-reactive cells (4). The present results indicate that the potency of this population to respond to mitogen was limited by both the IL I-producing and IL 2-producing capacity, since addition of either factor augmented the response. Of the non-responding populations, one (Ib, RFC-) was evidently in part restricted by lack of IL 1 production. The quantitatively large population II lacked IL 2-responding cells, and possibly also one or both of the other co-operating cell types. Comparison of these results with those obtained in other species is of interest. In accordance with our results, NISHIMURA et al. (11) reported that IL 2-responding cells in mouse thymus are low-density cells. There has been disagreement about whether the inability of «cortical», PNA + thymocytes to respond to mitogens can be circumvented by addition of IL 2 (for

456 . G. SANDBERG, O. SODER, and J. TJERNBERG

refs, see ref. 12). It was recently reported that a population of mouse thymocytes consisting of mostly spontaneously proliferating, PNA + cells responded to mitogen if IL 2 was also added (13). Similarly, the immunoincompetence (cytotoxic assay) of PNA + thymocytes was explained by their lack of IL 2-producing helper cells (14). Later, the need for additional soluble signals was suggested (15). According to DRABER and KISIELOW (16), only a minority of PNA +, high-density thymocytes in mice were rendered Con A-responsive in the presence of IL 2. These cells were PNA +, but in other aspects they were phenotypically mature. CONLON et al. (17) reported that thymus cells responding to IL 2 are PNA +, whereas IL 2-producing cells are PNA -. In addition, in a recent report concerning human thymus, isolated cells of «cortical type» could respond to Con A, whereas medullary type cells could not (18). In contrast, PIANTELLI et al. (19) found no evidence of mitogen, IL 1- or IL 2-responding cells among human PNA + cortical thymocytes. The presence of IL 2 receptors on most fetal thymocytes and on post-natal subcapsular thymocytes has been demonstrated (20, 21). Different opinions have been presented regarding the possibility of inducing proliferation in these cells by IL 2 (21, 22). Thus, there is at present no clear picture of the properties of IL 2-producing and -responding cells in the thymus. The thymus is characterized by an intense proliferation, at least tenfold higher than in other lymphoid organs. The bulk of cell divisions takes place in the thymic cortex, but a small number of mitotic cells are found also in the medulla (23). It is not known whether the proliferating cortical and medullary thymocytes represent two independently developing cell lineages or whether they are merely representative of different maturational stages of one cell lineage. The former hypothesis has some support from studies in chick, demonstrating that cortical and medullary thymocytes arise independently from separate precursor cells (24). Studies in guinea pigs show that dividing cortical thymocytes are concentrated in two distinct areas in the thymic cortex, subcapsular and juxtamedullary (25, 26), indicating that different proliferative subpopulations may exist also within the cortex. How cell proliferation in the thymus is initiated and maintained is one of the central questions in thymocyte differentiation that still remains unanswered. Interleukins have been suggested to playa role in the regulation of thymocyte growth. One hypothesis has been that IL 2, secreted by medullary thymocytes, diffuses from the medulla to the peripheral regions of the thymic lobuli and stimulates the intense proliferative activity in the subcapsular thymic cortex (27, 28). However, it has been reported that the intense cortical proliferation starts before the appearance of medullary IL 2producing cells during thymus ontogeny (29). IL 2 was initially described as a lymphokine which has an effect on peripheral immunocompetent lymphocytes, and it is therefore tempting to speculate that this growth factor might act in the thymus with the mature medullary population as target. The results presented in the present work support such an interpretation.

IL 1 and IL 2 Responsive Cells in Guinea Pig Thymus . 457

Thus, the IL 2-responsive cells were found in a distinct subpopulation of thymocytes with functional (i.e. PHA-responsive) capability. The picture is complicated by a recent report describing a small population (1-3 % of all thymocytes) that is scattered in the thymic cortex including cytotoxic T cell precursors responding to IL 2 in limiting dilution cultures (30). Furthermore, it is not known whether these cells are proliferating in situ in the thymic cortex; however, it is clear that the size and location of this population discriminates it from the rapidly dividing thymic precursor population. As mentioned, data on IL 2 responsiveness of immature cortical thymocytes are conflicting. In the present work, no indication was found that interleukins could stimulate the proliferation of immature cortical thymocytes.

In a series of recent papers, we described a thymic factor (TGP) that stimulates the proliferation of immature guinea pig thymocytes (31, 32). This factor has been characterized as a low molecular weight acidic peptide which has neither a chemical nor a functional relationship with thymic maturation hormones or interleukins. The target cells were la, PNA +, RFC+ thymocytes (2, 33) in contrast to the IL 1- and IL 2-responding thymocytes which are RFC-. Thus, our results clearly demonstrate that the cells responding to TGP are different from those responding to IL 1 and IL 2. Both IL 2 and TGP are progression growth factors, i.e. the growthstimulating effect requires preactivation with another class of growth factors inducing «competence» (see discussion in ref. 34). For immunocompetent cells, mitogenic lectins and antigen act as competence factors (35). No competence factors have yet been identified for the immature thymic precursor cells, although thymic hormones and membrane-bound thymic antigens have been proposed as candidates (32). In our experiments, and as previously described, both IL 1 and IL 2 had a weak direct stimulatory effect on thymocytes in addition to the much stronger comitogenic effect. Since the direct and comito genic effects were observed on the same populations, it is possible that the direct effect reflects a lowlevel of auto activation of thymocytes. However, our experimental protocol does not allow us to totally exclude that some immature cells occur in this population, which respond to the direct effect of the interleukins. In that case, these immature cells would have to differ from those responding to TGF. Elucidation of the precise physiological role of IL 1, IL 2, and TGP in thymocyte proliferation will have to await further studies.

Acknowledgements This study was supported by grants from Ellen, Walter and Lennart Hesselmans Foundation for Scientific Research, the Swedish Medical Research Council (project No. 06239), and Karolinska Institutet. We wish to thank Miss CAROLA FAHLGREN and Miss MARGARETA WIDING for excellent technical assistance.

458 . G. SANDBERG, O. SODER, and

J. TJERNBERG

References 1. SANDBERG, G., O. SODER, S. KOLARE, and U. ERNSTROM. 1983. Studies on thymocyte subpopulations in guinea pigs. III. Physical and functional characteristics of six subpopulations separated by density gradient centrifugation and PNA binding. Exp. Cell. BioI. 51: 257. 2. SANDBERG, G., and O. SODER. 1984. Studies on thymocyte subpopulations in guinea pigs. 5. Rosette formation as a tool to separate thymocyte growth factor responsive and mitogen reactive cells. Int. Archs Allergy appl. Immun. 74: 297. 3. LARSSON, E. I.. 1982. Functional heterogeneity of helper T cells. Two distinct helper T cells are required for the production of T cell growth factor. J. Immunol. 128: 742.

4. SODER, 0., G. SANDBERG, U. ERNSTROM, and J. LOTHWALL. 1985. Characterization of T cell growth factor responsive thymocytes in guinea pigs. In: Microenvironments in the Lymphoid System. Ed. G. G. B. KLAUS, Plenum Press, New York, pp. 693-700.

5. SANDBERG, G., and O. SODER. 1984. Studies on proliferating thymocyte subpopulations in guinea pigs, separated on the basis of density and rosette formation (abstract). Immunobiol. 167: 133. 6. SANDBERG, G. 1983. Studies on thymocyte subpopulations in guinea pigs. 2. Three subpopulations differing in rosette forming ability as detected by both visual and electronic counting of rosettes. Int. Archs Allergy app!. Immun. 71: 328. 7. ANDRUS, 1.., A. GRANIELLl-PIPERNO, and E. REICH. 1984. Cytotoxic T cells both produce and respond to interleukin 2. J. Exp. Med. 59: 647. 8. BENDTZFN, K. 1983. Biological properties of Interleukins. Allergy 38: 219. 9. GULLBERG, M., and E.-I.. LARSSON. 1983. Con A-induced TCGF-reactivity is selectively acquired by Lyt-2-positive T cell precursors. J. Immuno!. 131: 19. 10. OPPENHEIM, J..J., and I. GERY. 1982. Interleukin 1 is more than an interleukin. Immuno!. Today 3: 113. 11. NISHIMURA, T., H. KOZUTSUMI, and Y. HASHIMOTO. 1984. Mitogenic effect of IL 2 on non-thymic and thymic lymphocytes of the mouse. Thymus 6: 225. 12. VATTERONI, M. 1.., and M. PAPIERNIK. 1984. Thymic lymphocytes. II. Phenotypic modifications of thymocytes after Concanavalin A stimulation in the presence of Interleukin 2: Early modifications of Lyt 1 +2+ subset and later proliferation of cells with more mature phenotypes. Cell. Immuno!. 83: 124. 13. NISHIMURA, T., H. KOZUTSUMI, and Y. HASHIMOTO. 1984. Different response of mouse thymocyte subpopulations to Interleukin 2 and Concanavalin A. Thymus 6: 235. 14. WAGNER, H., C. HARDT, R. BARTLETT, M. ROLLINGHOFF, and K. PFIZENMAIER. 1980. Intrathymic differentiation of cytotoxic T lymphocyte (CTL) precursors. I. The CTL immunocompetence of peanut agglutinin-positive (cortical) and negative (medullary) Lyt 1,2,3 thymocvtes. J. Immul1o!' 125: 2532. 15. CONLON, P. ]., c. A. RAMTHUN, C. S. HENNEY, S. GILLIS. 1982. Cytokine-dependent thymocyte responses. II. Generation of cytotoxic T lymphocytes from immature thymocytes. J. Immunol. 129: 11. 16. DRABER, P., and P. KISIELOW. 1981. Identification and characterization of immature thymocytes responsive to T cell growth factor. Eur.]. Immuno!. 11: 1. 17. CONLON, P. J. 1982. Cytokine-dependent thymocyte responses: Characterization ofIL-I and IL-2 target subpopulations and mechanism of action.]. Immuno!. 128: 797. 18. GELIN, c., I.. BOUMSELL, J. DAUSSET, and A. BERNARD. 1984. The heterogeneity and functional capacities of human thymocyte subpopulations. Proc. Nat!. Acad. Sci. USA 81: 4912. 19. PIANTELLl, M., I.. LAURIOLA, G. B. FERRARA, and P. MUSIANI. 1982. Human thymocyte subsets. Interleukins activity on mitogen responsiveness. In: Immunology and Ageing, Ed. N. FABRIS, Martinus Nijhoff Publishers, Hague/Boston/London. 20. TAKACS, 1.., H. OSAWA, and T. DIAMANTSTEIN. 1984. Detection and localization by the monoclonal anti-interleukin 2 receptor antibody AMT -13 of IL 2 receptor-bearing cells in

IL 1 and IL 2 Responsive Cells in Guinea Pig Thymus . 459

21.

22.

23.

24.

25.

26. 27. 28.

29.

30.

31.

32. 33. 34. 35.

the developing thymus of the mouse embryo and in the thymus of cortisone-treated mice. Eur. J. Immunol. 14: 1152. HARDT, c., T. DIAMANTSTEIN, and H. WAGNER. 1985. Developmentally controlled expression of IL 2 receptors and of sensitivity to IL 2 in a subset of embryonic thymocytes. J. Immunol. 134: 3891. VON BOEHMER, H., A. CRISANTI, P. KISIELOW, and W. HAAS. 1985. Absence of growth by most receptor-expressing fetal thymocytes in the presence of interleukin-2. Nature 314: 539. GOLDSCHNEIDER, I., K. SHORTMAN, D. MCPHEE, S. LINTHICUM, G. MITCHELL, F. BATTYE, and F. BOLLUM. 1982. Identification of subsets of proliferating low Thy 1 cells in the thymus cortex and medulla. Cell. Immunol. 69: 59. JOUTERAU, F., and N. LE DOUARIN. 1982. Demonstration of a cyclic renewal of the lymphocyte precursor cells in the quail thymus during embryonic and perinatal life. J. Immunol. 129: 1869. POSTE, M. E., and I. A. OLSON. 1973. An investigation of the sites of mitotic activity in the guinea-pig thymus using autoradiography and colcemid-induced mitotic arrest. Immunology 24: 691. KOLARE, S., and G. SANDBERG. 1985. Studies on thymocyte subpopulations in guinea pigs. VI. Differentiation of precursor cells in vivo and in vitro. Immunobiol., 170: 338. JORDAN, R., and J. ROBINSON. 1981. T lymphocyte differentiation. In: The Thymus Gland, Ed. M. Kendall, Academic Press, London. p. 151. WAGNER, H., C. HARDT, K. HEEG, K. PFIZENMAIER, W. SOLBACH, R. BARTLETT, H. STOCKINGER, and M. ROLLINGHOFF. 1980. T-T cell interactions during cytolytic T lymphocyte (CTL) responses: T cell derived helper factor (interleukin 2) as a probe to analyze CTL responsiveness and thymic maturation of CTL progenitors. Immunol. Rev. 51: 215. CEREDIG, R., D. DRALYNAS, F. FITCH, and H. MAC DONALD. 1983. Precursors of T cell growth factor producing cells in the thymus: ontogeny, frequency, and quantitative recovery in a subpopulation of phenotypically mature thymocytes defined by antibody GK-1.5. J. Exp. Med. 158: 1654. FINK, P. ]., W. M. GALLATIN, R. A. REICHERT, E. C. BUTCHER, and I. L. WEISSMAN. 1985. Homing receptor-bearing thymocytes, an immunocompetent cortical subpopulation. Nature 313: 233. SODER, 0., and U. ERNSTROM. 1984. Purification to apparent homogeneity of a thymocyte specific growth factor from calf thymus. Biochem. Biophys. Res. Commun. 120: 420. SODER, O. 1985. Thymocyte growth factor, a progression growth factor for immature cortical thymocytes. Compo Immuno!. Microbio!' Infect. Dis. To appear in vo!' 8. SODER, 0., and U. ERNSTROM. 1983. Characterization of the target cell for a thymocyte specific growth factor in guinea pigs. Thymus 5: 141. SODER, 0., and U. ERNSTROM. 1984. Recruitment of thymocytes from G j into S phase by a thymic factor. Int. Archs Allergy app!. Immun. 74: 186. KLAUS, c., and C. HAWRYOLOWICZ. 1984. Cell cycle control in lymphocyte stimulation. Immuno!. Today 5: 16.

Dr. G. SANDBERG, Department of Histology, Karolinska Institute, P.O.B. 60400, S-104 01 Stockholm, Sweden