Abstract
Mitogenic stimulation in vitro of mouse T lymphocytes induces the production of the hematopoietic cytokines granulocyte–macrophage colony-stimulating factor and IL-3. The present experiments showed that simple crowding of murine spleen or lymph node cells was a sufficient inducing stimulus. Crowding did not have this consequence for thymus or marrow cells or spleen cells from nu/nu or Rag-1−/− mice lacking T lymphocytes. Crowding for as short a period as 24 h was sufficient to allow subsequent cytokine production in sparse cultures. Purified T lymphocytes also exhibited low levels of crowding induction of cytokine production and cytokine production was enhanced by IL-2 and IFN-γ. However, IFN-γ−/− spleen cells exhibited similar crowding-induced colony-stimulating factor production to that of control spleen cells. Excess cell crowding inhibited cytokine production. This inhibition was not caused by receptor internalization of cytokines but may contribute to the failure to observe IL-3 production in lymphoid organs in vivo. Coculture of allogeneic spleen or peritoneal cells was a strong inducing signal for colony-stimulating factor production but this parameter was unable to detect autoreactivity of T lymphocytes in mice that lack suppressor of cytokine signaling 1 and exhibit T lymphocyte activation.
Early studies showed that mitogenic stimulation of blood or spleen cells led to the production of cytokines active on hematopoiesis (1, 2). These cytokines eventually were identified as granulocyte–macrophage colony-stimulating factor (GM-CSF) and multi-CSF (IL-3), and analysis of fractionated populations indicated that T lymphocytes were necessary for CSF production in vitro (3, 4) and were likely to be the dominant source of this CSF in these cell populations, although an accessory function of other cells remained possible. These conclusions were supported by the ability of cloned normal T lymphocytes and T lymphocyte cell lines to produce both cytokines when mitogenically stimulated (5, 6). In early studies the activating stimulus used for T lymphocytes was a plant lectin, but more recent studies have used antibodies to cross-ligate T cell receptors to cause cellular activation (7).
Coculture of appropriate allogeneic spleen or blood cells represents another form of mitogenic stimulation that leads both to T lymphocyte activation and CSF production (8). What has not been studied in this context is whether CSF production in vitro is a sufficiently sensitive parameter to document autoreactivity of T lymphocytes against autologous tissues, either in classical autoimmune disease or inflammatory states such as that seen in suppressor of cytokine signaling 1 (SOCS-1) −/− mice, where activated T lymphocytes appear to be involved in disease development (9, 10).
The data on CSF production by mitogen-stimulated suspensions of lymphoid cells have proved difficult to reconcile with other studies on organ production of CSFs. In particular, studies on murine organs have shown that T lymphocytes are not the exclusive source of these CSFs and, further, that no organ produces detectable IL-3 or contains detectable IL-3 mRNA (11, 12). Although caution is clearly needed in interpreting in vitro data from lymphoid cell suspensions, the system is a robust one for producing CSFs and the induction systems involved remain to be characterized.
In the course of experiments to determine whether autoreactivity of murine cells might lead to detectable CSF production in vitro, unanticipated problems were encountered because of apparently spontaneous activation of CSF production in cultured normal spleen cells and the nature of this activation required characterization.
Materials and Methods
Production of CSF by Cell Cultures.
Dispersed cell suspensions were prepared of spleen, marrow, thymus, lymph node, and peritoneal cells from mice aged 2–3 months. Mouse strains used were C57BL, BALB/c, DBA, CBA, T cell-deficient nu/nu, Rag-1−/−, or NOD/SCID mice and genetically manipulated mice, βc−/− (13), IFN-γ−/−, or SOCS-1−/− (9), the SOCS-1−/− mice being aged 2–3 weeks. All mice were born and housed in protected animal rooms, and all studies were approved by the Animal Ethics Committee of The Walter and Eliza Hall Institute.
Cells were cultured in 1-ml vol of RPMI 1640 medium with 10% FCS in capped tubes at 37°C in a fully humidified atmosphere of 10% CO2 in air. In some experiments the medium also contained a final concentration of 5 × 10−5 M 2-mercaptoethanol. Where added to the cultures, IL-1, IL-2, or IFN-γ were purchased from PeproTech (Rocky Hill, NJ).
Routinely, supernatants were harvested from the cultures after 4 days of incubation and these were millipore-filtered before storage at −15°C for bioassay.
Bioassays for GM-CSF and IL-3.
Levels of IL-3 and GM-CSF in conditioned media were determined by using microwell proliferation assays with BaF3 cells to detect IL-3 and FDC-P1 cells to detect GM-CSF. FDC-P1 cells proliferate in response to both IL-3 and GM-CSF but the concentrations of these two factors were so different that essentially FDC-P1 assays were assays for GM-CSF. These interpretations were validated by the use of antisera to IL-3 that neutralized all activity on BaF3 cells and to GM-CSF that essentially neutralized all activity on FDC-P1 cells (refs. 11 and 12 and data not shown). Levels of IL-3 or GM-CSF in conditioned media were converted to pg/ml by using parallel titrations of purified recombined murine IL-3 or GM-CSF (PeproTech).
Preparation of T Lymphocyte Suspensions.
Suspensions of highly enriched T lymphocytes were prepared by using dispersed lymph node cells and murine T cell enrichment columns according to the manufacturer's instructions (R & D Systems, Minneapolis). Column-enriched preparations were analyzed by using a FACScan (Becton Dickinson) and found to contain 90–97% Thy-1-positive cells.
Results
In the present experiments, the deliberate use of mitogens or antireceptor antibodies to induce CSF production was avoided so that CSF production, possibly associated with autoreactivity, could be investigated. An immediate complication was encountered because simple crowding of cultured cells appeared to be sufficient to induce readily detectable GM-CSF and IL-3 production.
Effects of Cell Crowding.
As shown in Fig. 1A, progressive crowding of murine spleen cells during incubation resulted in a dramatic nonlinear rise in the production of both GM-CSF and IL-3. In various experiments, peak activity occurred with concentrations of 2–10 × 106 cells per ml. When cell concentrations were further increased, lower levels of GM-CSF and IL-3 were observed. As also shown in Fig. 1A, the inclusion of 5 × 10−5 M 2-mercaptoethanol in the medium during incubation did not alter the basic pattern of responses to cell crowding but consistently increased the levels of both GM-CSF and IL-3 in such cultures. The identity of the bioactivity detected in the spleen cell cultures was confirmed as GM-CSF and IL-3 by the use of specific neutralizing antibodies (data not shown).
Figure 1.
(A) Concentration-dependent rise and fall of GM-CSF and IL-3 in cultures of C57BL spleen cells with or without 2-mercaptoethanol (2ME). Means ± standard deviations from four replicate cultures are shown. (B) Concentration-dependent rise and fall in GM-CSF and IL-3 in cultures of βc−/− spleen cells lacking GM-CSF receptors versus control βc+/+ spleen cells. Mean values of duplicate cultures are shown. (C) Failure of GM-CSF production in cultures of nu/nu or Rag-1−/− spleen cells lacking T lymphocytes and relative inactivity of spleen cell cultures from NOD/SCID mice. Mean values of duplicate cultures are shown.
The lower cytokine concentrations in cultures containing very high cell concentrations might, in principle, have been caused by receptor-mediated endocytosis with degradation of internalized cytokine, even though the frequency of spleen cells expressing receptors for GM-CSF or IL-3 is low (14). However, as shown in Fig. 1B, the ultimate cell concentration-related fall in observed GM-CSF levels was also seen when using spleen cells from βc−/− mice that lack the common β receptor chain for GM-CSF and so have no high-affinity receptors for GM-CSF (13).
In contrast to these responses, crowding of thymus or marrow cells in culture did not lead to the production of detectable CSF. As shown in Table 1, thymus cells actually survived better in suspension culture than spleen cells. Addition of 2-mercaptoethanol had no consistent effect in increasing recovered cell numbers. In mitogen-stimulated lymphocyte cultures, the appearance of blast cells parallels cytokine production and is possibly the morphological expression of the required cellular activation. It was notable that blast cells were not present in thymus cell cultures and that, in spleen cell cultures, addition of 2-mercaptoethanol increased the frequency of blast cells as the possible basis for the observed increased CSF production. These findings were reproduced in three replicate experiments.
Table 1.
Survival of cultured spleen and thymus cells
| Cells cultured | 2ME | Viable cells × 10−5 | Percent blast cells |
|---|---|---|---|
| 5 × 106 C57BL spleen cells | + | 6.1 ± 0.5 | 8.3 ± 2.1 |
| − | 5.9 ± 0.7 | 1.3 ± 0.6 | |
| 5 × 106 C57BL thymus cells | + | 29.3 ± 2.4 | 0 |
| − | 32.8 ± 4.7 | 0 |
Triplicate suspension cultures were analyzed after 3 days of incubation. Mean values ± SDs. 2ME = 5 × 10−5 M 2-mercaptoethanol.
The induction of nonlinear CSF production in crowded spleen cultures might have been a nonspecific metabolic consequence of cell crowding or have been caused by the achievement of effective concentrations of CSF-producing or CSF-inducing cells. To explore these possibilities, spleen cells were mixed with increasing concentrations of thymus or marrow cells. High thymus cell numbers (up to 10 × 106/ml) had only a weak enhancing effect on CSF production by spleen cells, suggesting that thymus populations may contain few cells with this type of accessory function. When increasing numbers of marrow cells were added to spleen cell suspensions, no enhancement was observed and often an actual inhibition of CSF production occurred (data not shown).
Effect of Lack of T Lymphocytes.
Spleen cells from nu/nu or Rag-1−/− mice that lack T lymphocytes uniformly failed to produce any detectable GM-CSF (Fig. 1C) or IL-3 (data not shown), regardless of the cell concentration used. Cells from immunodeficient NOD/SCID mice behaved in an intermediate fashion, producing some GM-CSF and IL-3 when crowded, but the concentrations were at least 10-fold below those produced by cells from control mice.
The absolute failure of cultured nu/nu and Rag-1−/− spleen cells to produce GM-CSF was in sharp contrast to the ability of organs from these mice to produce GM-CSF in cultures of minced organs. As shown by the examples in Table 2, organs from nu/nu mice had the ability to produce detectable GM-CSF, but no IL-3, and similar results were obtained by using organs from Rag-1−/− and NOD/SCID mice (data not shown).
Table 2.
CSF Production by minced nu/nu organs in vitro
| Mouse organ |
nu/nu 1
|
nu/nu 2
|
CBA
|
|||
|---|---|---|---|---|---|---|
| GM-CSF | IL-3 | GM-CSF | IL-3 | GM-CSF | IL-3 | |
| Thymus | — | — | — | — | 8,000 | 0 |
| Heart | 2,000 | 0 | 8,000 | 0 | 32,000 | 0 |
| Lung | 4,000 | 0 | 32,000 | 0 | 512,000 | 0 |
| Muscle | 500 | 0 | 8,000 | 0 | 2,000 | 0 |
| 4,000 | 0 | 4,000 | 0 | 2,000 | 0 | |
| 4,000 | 0 | 2,000 | 0 | 8,000 | 0 | |
| Bone shaft | 2,000 | 0 | 4,000 | 0 | 2,000 | 0 |
| Bladder | 120 | 0 | 0 | 0 | 1,000 | 0 |
| Testes | 500 | 0 | 500 | 0 | 1,000 | 0 |
nu/nu mice were on a CBA genetic background. Organs were incubated for 4 days in 2 ml of RPMI 1640 medium plus 10% FCS. Results shown are the levels of GM-CSF or IL-3 in pg per ml of conditioned media.
A T lymphocyte origin of CSF produced in crowded cultures was confirmed by culture of T lymphocytes purified from lymph nodes (Table 3). Levels of cytokine after incubation of the purified T lymphocytes were increased by addition of IL-2 or IFN-γ, cytokines likely to be produced by accessory cells in crowded cultures, although addition of IL-1 was not effective. Activity was also increased by coculture of T lymphocytes with syngeneic Rag-1−/− spleen cells, as was that of unfractionated C57BL spleen cells (data not shown), indicating that non-T lymphocyte cells in the spleen could produce enhancing or inducing signals, perhaps including those cytokines found to have enhancing activity.
Table 3.
Production of CSF by purified T lymphocytes from C57BL lymph nodes
| Cells cultured | GM-CSF, pg/ml | IL-3, pg/ml |
|---|---|---|
| Lymph node T lymphocytes | ||
| 1 × 106/ml | 0 | 0 |
| 10 | 2 | |
| 2 × 106/ml | 0 | 1 |
| 60 | 4 | |
| 1 × 106/ml + IL-2 | 250 | 2 |
| 250 | 4 | |
| 2 × 106/ml + IL-2 | 2,000 | 8 |
| 2,000 | 8 | |
| 1 × 106/ml + IFN-γ | 1,000 | 2 |
| 1,000 | 4 | |
| 2 × 106/ml + IFN-γ | 2,000 | 4 |
| 2,000 | 4 | |
| Lymph node unfractionated cells | ||
| 1 × 106/ml | 120 | 1 |
| 0 | 1 | |
| 2 × 106/ml | 150 | 2 |
| 20 | 1 | |
| 1 × 106/ml + IL-2 | 500 | 8 |
| 0 | 0 | |
| 2 × 106/ml + IL-2 | 250 | 4 |
| 0 | 1 | |
| 1 × 106/ml + IFN-γ | 1,000 | 4 |
| 1,000 | 4 | |
| 2 × 106/ml + IFN-γ | 1,000 | 4 |
| 1,000 | 8 |
All cultures contained 2-mercaptoethanol. IL-2 was used at a final concentration of 100 units/ml. IFN-γ was used at a final concentration of 2 × 103 units/ml.
Although IFN-γ was a strong inductive stimulus for CSF production by spleen cells, it was not essential for crowding-induced CSF production. Spleen cells from IFN-γ−/− mice were identical to control mice in exhibiting comparable levels of crowding-induced CSF production, the inhibition of this production by excess cell concentrations and the responsiveness of low concentrations of spleen cells to induction by IFN-γ (data not shown).
In contrast to the regular production of IL-3 in unstimulated cultures of spleen cell suspensions, culture of minced spleen fragments results only in a very low and variable production of GM-CSF and never IL-3 (11), despite the fact that the spleen contains 200 × 106 cells and, on mincing, at least some cells must escape into the medium.
To investigate this strange discordance, cocultures were prepared of spleen fragments with or without spleen cell suspensions, using RPMI 1640 medium and 2-mercaptoethanol to optimize culture conditions for lymphocytes. Minced thymus tissue was used as a control organ. The results of a typical experiment are shown in Table 4. Minced spleen tissue produced low levels of GM-CSF and no IL-3, whereas minced thymus tissue usually produced high levels of GM-CSF, again with no IL-3. Mixing experiments failed to detect any inhibitory activity for IL-3 in such conditioned media.
Table 4.
Inhibition of C57BL spleen cell production of IL-3 by coculture with C57BL spleen fragments
| Cells cultured | GM-CSF, pg/ml | IL-3, pg/ml |
|---|---|---|
| 2 × 106/ml C57BL spleen cells | 2,000 | 32 |
| 4,000 | 64 | |
| 5 × 106/ml C57BL spleen cells | 4,000 | 125 |
| 4,000 | 125 | |
| Minced C57BL spleen tissue | 0 | 0 |
| 4 | 0 | |
| Minced C57BL spleen + 2 × 106/ml C57BL spleen cells | 4,000 | 0 |
| 8,000 | 0 | |
| Minced C57BL spleen + 5 × 106/ml C57BL spleen cells | 1,000 | 0 |
| 4,000 | 0 | |
| Minced C57BL thymus tissue | 128,000 | 0 |
| 64,000 | 0 | |
| Minced C57BL thymus + 2 × 106/ml C57BL spleen cells | 16,000 | 0 |
| 64,000 | 0 | |
| Minced C57BL thymus + 5 × 106/ml C57BL spleen cells | 32,000 | 0 |
| 64,000 | 0 |
All cultures were prepared by using 1 ml of RPMI 1640 medium with 10% FCS and contained a final concentration of 5 × 10−5 M 2-mercaptoethanol. One whole C57BL spleen or thymus was minced to produce organ fragments. Tubes were incubated for 4 days before harvesting the media.
Coculture of minced spleen or thymus with syngeneic spleen cell suspensions did not consistently alter observed levels of GM-CSF but no IL-3 was detectable in the media from such cocultures. Viable cell counts on the spleen cell suspensions in the cocultures during the coincubation indicated that massive death occurred within 24 h of initiation of the cocultures and few or no surviving cells from the spleen cell suspensions were present by day 2. Toxicity of the organ fragments for lymphoid cells may be the basis for the failure to find evidence of IL-3, as a lymphocyte product, in such organ fragment cocultures or in conditioned media produced by minced lymphoid organs.
Duration of Cell Crowding Required for Induction of CSF Production.
Experiments were performed to determine whether cell crowding was needed only as an initiating event for CSF production or was needed throughout the incubation period.
Crowded (5 × 106 cells per ml) spleen cell suspensions were incubated then, after 24 or 48 h, cells were removed and maintained as uncrowded cultures of 0.5 and 1.0 × 106 cells per ml until completion of the 4-day incubation period. Assays showed no detectable CSF in the culture fluid at 24 and 48 h, indicating that little preexisting CSF could have been transferred during the cell dilution step.
The results of a typical experiment are shown in Fig. 2. CSF was not detectable in control cultures of 0.5 × 106 or 1 × 106 cells per ml, but both GM-CSF and IL-3 were produced in cultures of 0.5 or 1 × 106 cells per ml diluted after initial culture for 1 or 2 days at a concentration of 5 × 106 cells per ml. CSF production by cells initially cultured under crowded conditions for 2 days was no higher than that by cells cultured under crowded conditions for only 1 day.
Figure 2.
Failure of C57BL spleen cells to produce GM-CSF or IL-3 at 0.5 or 1.0 × 106 cells per ml but activity of cells at these concentrations after initial incubation for 24 h at a concentration of 5 × 106 cells per ml. Each column represents assays from an independent culture.
The induction of CSF production by cell crowding appeared able to be achievable by only a relatively short period of cell crowding and could then persist in cells subsequently cultured at low concentrations.
CSF Production in Allogeneic Cultures.
Fig. 3 shows an example of the consequences of allogeneic coculture of C57BL and BALB/c spleen cells in cultures containing 2-mercaptoethanol. There are several features to note. The rise in CSF concentrations in the allogeneic cultures required time and therefore is unlikely to represent release of preformed CSF. Second, in both syngeneic and allogeneic cultures, levels of IL-3 were 10-fold lower than those of GM-CSF. Third, levels of GM-CSF and IL-3 in allogeneic cultures were higher than those in syngeneic cultures initiated with twice the number of cells used in the mixture. Possibly the levels in the allogeneic cultures were also more sustained than in the syngeneic cultures. A similar pattern was observed with cocultures of C57BL with DBA cells and of BALB/c cells with DBA cells.
Figure 3.
Increased production of GM-CSF and IL-3 in cocultures of 1 × 106 C57BL and 1 × 106 BALB/c spleen cells versus cultures of 2 × 106 C57BL or BALB/c cells. Shown are mean values of duplicate cultures.
A probable one-sided allogeneic activation of CSF production was observed when spleen cells lacking T lymphocytes (nu/nu or Rag-1−/− cells) were cocultured with normal allogeneic spleen cells. Coculture of two lymphocyte-deficient allogeneic spleen populations did not result in CSF production.
Parallel studies using thymus cells indicated in most experiments that no CSF production occurred in cocultures of allogeneic thymus cells. Coculture of thymus cells with allogeneic spleen cells resulted in low levels of CSF production. Cocultures of allogeneic marrow cells also produced no detectable CSF but when allogeneic spleen cells were cocultured with marrow cells some increase in CSF production was observed. Both marrow and thymus cells seemed able to act as allogeneic priming cells to elicit responses by allogeneic spleen cells that were not large but were in excess of those attributable to cell crowding.
Parallel experiments using peritoneal cells indicated that the cells had some capacity to produce GM-CSF and IL-3 and coculture of allogeneic peritoneal cells moderately enhanced CSF production. Coculture of peritoneal cells with allogeneic spleen cells also resulted in some enhancement of CSF production.
Thus spleen and, to a lesser degree, peritoneal cells exhibited allogeneic responses as detected by CSF production and to a limited degree thymus and marrow cells could act as allogeneic stimuli for spleen cells, although apparently not able themselves to produce CSF.
Autoreactivity of SOCS-1−/− Cells.
To determine whether CSF production could be used to detect autoreactivity in SOCS-1−/− T lymphocyte populations, cultures of neonatal SOCS-1−/− and SOCS-1+/+ thymus and spleen cells were prepared together with cocultures of SOCS-1−/− and SOCS-1+/+ cells.
The results shown in Table 5 are typical of five separate experiments. At the cell concentrations able to be prepared from neonatal organs, little or no CSF production was detectable with −/− or +/+ spleen cells. Furthermore, no CSF production was detectable in cocultures of these cells. This finding suggested that, if SOCS-1−/− T lymphocytes in spleen cell suspensions were autoreactive, the level of reactivity to +/+ cells was too low to be detectable by CSF production. In contrast, SOCS-1−/− thymus cells consistently produced GM-CSF and IL-3 (Table 5), an activity not detectable with SOCS-1+/+ thymus cells. This observation supports other evidence of the activated state of SOCS-1−/− T lymphocytes (9, 10, 15, 16).
Table 5.
Production of CSF by SOCS-1−/− thymus cells but no evidence of autoreactivity against SOCS-1+/+ cells
| Cells cultured | Cell per ml × 10−6 | GM-CSF, pg/ml | IL-3, pg/ml |
|---|---|---|---|
| SOCS-1−/− thymus cells | 1 | 20 | 0 |
| 90 | 1 | ||
| 120 | 0 | ||
| 2 | 500 | 2 | |
| 500 | 4 | ||
| 1,000 | 2 | ||
| SOCS-1−/− spleen cells | 1 | 0 | 0 |
| 0 | 0 | ||
| 0 | 0 | ||
| 2 | 0 | 0 | |
| 0 | 0 | ||
| 0 | 0 | ||
| SOCS-1+/+ thymus cells | 1 | 0 | 0 |
| 0 | 0 | ||
| 0 | 0 | ||
| 2 | 0 | 0 | |
| 0 | 0 | ||
| 0 | 0 | ||
| SOCS-1+/+ spleen cells | 1 | 0 | 0 |
| 0 | 0 | ||
| 0 | 0 | ||
| 2 | 0 | 0 | |
| 0 | 0 | ||
| 0 | 0 | ||
| SOCS-1−/− thymus cells + SOCS-1+/+ spleen cells | 1 + 1 | 160 | 4 |
| 0 | 0 | ||
| 250 | 4 | ||
| SOCS-1−/− spleen cells + SOCS-1+/+ spleen cells | 1 + 1 | 0 | 0 |
| 0 | 0 | ||
| 0 | 0 | ||
| SOCS-1+/+ thymus cells + SOCS-1−/− spleen cells | 1 + 1 | 0 | 0 |
| 0 | 0 | ||
| 0 | 0 | ||
| SOCS-1−/− thymus cells + SOCS-1+/+ thymus cells | 1 + 1 | 50 | 0 |
| 80 | 0 | ||
| 0 | 0 |
Triplicate cultures were prepared in 1 ml of RPMI 1640 medium with 10% FCS and 5 × 10−5 M 2-mercaptoethanol by using 2-week-old SOCS-1−/− and littermate SOCS-1+/+ mice. Media were harvested and analyzed after 4 days of incubation.
Discussion
An early report (17) noted that mouse spleen cells in unstimulated suspension cultures could produce GM-CSF and IL-3, removal of T lymphocytes prevented this activity, and the FCS in the medium might act as a foreign antigenic stimulus to induce this activity by spleen cells. The present studies confirmed the capacity of murine spleen cells to produce GM-CSF and IL-3 in cultures to which no mitogen had been added and again indicated that T lymphocytes were necessary for the GM-CSF and IL-3 production and were a cellular source of these cytokines. The possibility that the FCS in the medium might have served as a foreign antigen to activate the cultured T lymphocytes was not further addressed in the present studies.
The most striking aspect of the present cultures was that the production of CSF was nonlinear with respect to cultured cell numbers, implying that, even if serum was functioning as a mitogen, cell crowding played a critical role in inducing CSF production. Cell crowding appeared to be an initiating event whose persistence was not needed throughout the incubation period. Cell crowding might exert a variety of nonspecific metabolic effects on cell behavior but, in control studies, thymus cells were not particularly effective and marrow cells were inactive as crowding populations. Cell populations able to induce CSF production by crowding included spleen populations from mice lacking T lymphocytes and CSF induction seemed to involved a special cell–cell interaction, perhaps involving the production of agents like IL-2 and IFN-γ.
CSF production was enhanced by addition of 2-mercaptoethanol to the medium. Previous studies have reported that 2-mercaptoethanol has a major impact on the behavior of cultured lymphocytes (18, 19) because it promotes cystine transport and enhances lymphocyte survival and proliferation, possibly by increasing IL-2 production (19–21). In the present studies, viable cell numbers were not substantially increased by the use of 2-mercaptoethanol although the percentage of blast cells was increased and blast cell formation has been noted as a key part of a mitogen-induced response by T lymphocytes in which cytokine production occurs.
There is a puzzling discordance for CSF production between the data obtained from whole organs or large organ fragments and the data obtained from spleen cell suspensions. This difference is most evident for IL-3 that has never been detected in whole organs or organ fragments (11, 12). The nonlinear behavior of cell suspensions clearly indicates that accessory cells assist, or are necessary for, CSF production but an excess of such cells appears to be inhibitory. If this process was to become more evident with the extremely high cell concentrations in an intact organ, this might be the reason for complete failure of IL-3 production to occur. The present study reinforces earlier data (11, 12) that T lymphocytes are not a major cellular source of organ CSFs, nonetheless their curious functional responses to cell crowding are intriguing and may be relevant for cell behavior in other organs.
Coculture of allogeneic spleen cells reproducibly increased CSF production above that expected on a simple cell crowding basis from the combined cell concentrations used. This is likely to represent genuine antigenic/mitogenic stimulation of T lymphocytes of the type well documented by many earlier studies. The question posed in the present study was whether CSF production was a sufficiently sensitive response to detect the likely autoreactivity in SOCS-1−/− mice where tissue damage and disease appears to involve aberrant activity of T lymphocytes in an IFN-γ-dependent process (9, 10, 16). The coculture studies using SOCS-1−/− and SOCS-1+/+ tissues, of necessity, had to be performed by using neonatal populations that, for SOCS-1−/− mice, also had some depletion of T lymphocytes (9, 15). The results were disappointing in that CSF production was not convincingly triggered in cocultures of −/− with +/+ spleen cells. Conversely, however, CSF production was detectable in cultures of SOCS-1−/− thymus cells, which do exhibit activation markers (9, 10) and possibly are already reacting against self-antigens within the thymus itself. However, thymus cell production of CSF was weak and the detection system did not appear suitable for more detailed studies on the potential autoaggression of T lymphocytes in SOCS-1−/− mice.
Acknowledgments
We are indebted to Dr. Warren Alexander for supplying the SOCS-1−/− and IFN-γ−/− mice used in these experiments and Drs C. Drinkwater and L. Robb for originally producing the βc−/− mice. All other mice were supplied by the Department of Animal Services of this institute. This work was supported by the Carden Fellowship Fund of the Cancer Council (Victoria, Australia), the National Health and Medical Research Council (Canberra, Australia), the Cooperative Research Centre for Cellular Growth Factors (Melbourne), and National Institutes of Health Grant CA22556.
Abbreviations
- CSF
colony-stimulating factor
- GM-CSF
granulocyte–macrophage CSF
- SOCS-1
suppressor of cytokine signaling 1
References
- 1.Parker J W, Metcalf D. J Immunol. 1974;112:502–510. [PubMed] [Google Scholar]
- 2.Cline M J, Golde D W. Nature. 1974;248:703–704. doi: 10.1038/248703a0. [DOI] [PubMed] [Google Scholar]
- 3.Staber F G, Hültner L, Marcucci F, Krammer P H. Nature. 1982;298:79–82. doi: 10.1038/298079a0. [DOI] [PubMed] [Google Scholar]
- 4.Kelso A, Metcalf D. Adv Immunol. 1990;48:69–105. doi: 10.1016/s0065-2776(08)60752-x. [DOI] [PubMed] [Google Scholar]
- 5.Schreier M H, Iscove N N. Nature. 1980;287:228–230. doi: 10.1038/287228a0. [DOI] [PubMed] [Google Scholar]
- 6.Kelso A, Metcalf D. J Cell Physiol. 1985;123:101–110. doi: 10.1002/jcp.1041230115. [DOI] [PubMed] [Google Scholar]
- 7.Maraskovsky E, Troutt A B, Kelso A. Int Immunol. 1992;4:475–485. doi: 10.1093/intimm/4.4.475. [DOI] [PubMed] [Google Scholar]
- 8.Kelso A, Troutt A B, Maraskovsky E, Gough N M, Morris L, Pech M H, Thomson J A. Immunol Rev. 1991;123:85–114. doi: 10.1111/j.1600-065x.1991.tb00607.x. [DOI] [PubMed] [Google Scholar]
- 9.Starr R, Metcalf D, Elefanty A G, Brysha M, Willson T A, Nicola N A, Hilton D J, Alexander W A. Proc Natl Acad Sci USA. 1998;95:14395–14399. doi: 10.1073/pnas.95.24.14395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Marine J-C, Topham D J, McKay C, Wang D, Parganas E, Stravopodis D, Yoshimura A, Ihle J N. Cell. 1999;98:609–616. doi: 10.1016/s0092-8674(00)80048-3. [DOI] [PubMed] [Google Scholar]
- 11.Metcalf D, Willson T A, Hilton D J, Di Rago L, Mifsud S. Leukemia. 1995;9:1556–1564. [PubMed] [Google Scholar]
- 12.Ryan P J, Willson T, Alexander W S, Di Rago L, Mifsud S, Metcalf D. Leukemia. 2001;15:1248–1255. doi: 10.1038/sj.leu.2402173. [DOI] [PubMed] [Google Scholar]
- 13.Robb L, Drinkwater C C, Metcalf D, Li R, Köntgen F, Nicola N A, Begley C G. Proc Natl Acad Sci USA. 1995;92:9565–9569. doi: 10.1073/pnas.92.21.9565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Metcalf D, Nicola N A. The Hemopoietic Colony-Stimulating Factors: From Biology to Clinical Applications. Cambridge, U.K.: Cambridge Univ. Press; 1995. [Google Scholar]
- 15.Naka T, Matsumoto T, Narazaki M, Fujimoto M, Morita Y, Ohsawa Y, Saito H, Nagasawa T, Uchiyama Y, Kishimoto T. Proc Natl Acad Sci USA. 1998;95:15577–15582. doi: 10.1073/pnas.95.26.15577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Naka T, Tsutsui H, Fujimoto M, Kawazoe Y, Kohzaki H, Morita Y, Nakagawa R, Narazaki M, Adachi K, Yoshimoto T, et al. Immunity. 2001;14:535–545. doi: 10.1016/s1074-7613(01)00132-7. [DOI] [PubMed] [Google Scholar]
- 17.Davignon J L, Kimoto M, Kindler V, De Kossodo S, Vassalli P, Izui S. Eur J Immunol. 1988;18:1367–1372. doi: 10.1002/eji.1830180910. [DOI] [PubMed] [Google Scholar]
- 18.Broome J D, Jeng M M. J Exp Med. 1973;138:574–592. doi: 10.1084/jem.138.3.574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Metcalf D. J Immunol. 1976;116:635–638. [PubMed] [Google Scholar]
- 20.Sugama K, Namba Y, Hatanaka M, Hanaoka M. Microbiol Immunol. 1987;31:691–700. doi: 10.1111/j.1348-0421.1987.tb03130.x. [DOI] [PubMed] [Google Scholar]
- 21.Chen G, Wang S-H, Converse C A. Int J Immunopharmacol. 1994;16:755–760. doi: 10.1016/0192-0561(94)90095-7. [DOI] [PubMed] [Google Scholar]