IGF-1 down-regulates IFN-γR2 chain surface expression and desensitizes IFN-γ/STAT-1 signaling in human T lymphocytes

Interferon-γ (IFN-γ) produced by T and natural killer (NK) cells is considered the principal effector cytokine of cell-mediated immunity, and many studies have indicated that it also plays an important role in controlling T-cell homeostasis and apoptosis. Activated T cells, from mice in which the genes encoding for IFN-γ or for different components of the IFN-γ signaling pathway are knocked out, exhibit increased expansion and resistance to apoptosis.^1-6  Murine and human T cells activated in the presence of anti–IFN-γ or anti–IFN-γ receptor (IFN-γR) antibodies are resistant to activation-induced apoptosis.^7-9 

The response of T cells to IFN-γ is regulated by modulating the IFN-γR1 binding chain and IFN-γR2 signaling chain of its membrane receptor complex.^10,11  IFN-γR1 is expressed on the surface of both lymphoid and nonlymphoid cells,^12,13  whereas expression of IFN-γR2 is high in B and myeloid cells¹²  and low or absent in T cells.^8,13  In T cells that differentiate along the T helper 1 (Th1) pathway, down-modulation of IFN-γR2 acts as a negative regulatory mechanism that attenuates IFN-γ/signal transducer and activator of transcription-1 (STAT-1) signaling and limits the apoptotic effect of IFN-γ.^14-16 

Both IFN-γ–dependent and IFN-γ–independent mechanisms have been reported to down-regulate IFN-γR2 expression in T lymphocytes. During T-helper development, the loss of IFN-γR2 that makes mouse Th1 cells resistant to IFN-γ is induced by IFN-γ itself.¹⁴  We have reported that in resting and activated human T lymphocytes, IFN-γR2 expression is prevalently cytoplasmic.^9,13,17  Low surface IFN-γR2 results from fast, continuous recycling between surface and clathrin-coated vesicles.¹⁷  IFN-γR2 internalization is not affected in T cells from individuals genetically deficient for IFN-γR1,¹⁷  in established Th2 clones,⁹  and in T-cell lines unable to produce IFN-γ.¹³  However, the signals that play a major role in IFN-γ–independent down-regulation of IFN-γR2 in T cells have not yet been characterized.

Up-regulation of surface IFN-γR2 might occur following T-cell receptor (TCR) engagement^8,9  or following exposure of T cells to factors that negatively regulate growth, such as galectins¹⁸  and nitric oxide.¹⁹  Serum deprivation, such as interleukin-2 (IL-2) deprivation,²⁰  mimics the passive apoptosis induced by growth factor deprivation that T cells encounter in vivo. We have observed that serum deprivation increases surface IFN-γR2 in malignant T cells and renders them sensitive to apoptosis induced by IFN-γ.¹³  This suggests that serum factors or hormones may play a critical role in keeping IFN-γR2 expression low and thus limiting IFN-γ responsiveness in human T lymphocytes.

Among the factors present in serum, insulin-like growth factor-1 (IGF-1) has a profound effect on immune functions.^21,22  It promotes cord blood T-cell maturation,²³  proliferation, cytokine production,²⁴  and survival,²⁵  and inhibits their spontaneous and activation-induced apoptosis.²³  Blockade of IGF-1 receptor (IGF-1R) decreases the survival of T cells activated through the TCR and CD28, and enhances their susceptibility to Fas-induced apoptosis.²⁶ 

In this paper, the effect of IGF-1 or anti–IGF-1R mAb in regulating IFN-γR2 surface expression, STAT-1 activation, and apoptosis in human T lymphocytes was investigated. We show that IGF-1 delivers a signal for IFN-γR2 internalization and limits IFN-γ/STAT-1 signaling in human T cells. These results identify a new mechanism that desensitizes T lymphocytes to IFN-γ and may have implications for the regulation of apoptosis in normal and malignant T cells.

Phytohemagglutinin (PHA) was obtained from Invitrogen (Milan, Italy); rabbit anti–phospho-tyr (701)-STAT-1, anti–STAT-1 polyclonal antibodies, and horseradish peroxidase–conjugated goat anti–rabbit immunoglobulin G (IgG) were from Cell Signaling (Beverly, MA); isotype negative control mouse IgG1, fluorescein isothiocyanate (FITC)–conjugated rabbit anti–mouse Ig, and FITC-conjugated anti-CD3, anti-CD2, and anti-CD25 monoclonal antibodies (mAbs) were from Dako (Glostrup, Denmark). The recombinant human (r-h) IGF-1 and blocking anti–IGF-1R antibody were from R&D (Minneapolis, MN); r-h insulin was from Novo Nordisk A/S (Bagsvaerd, Denmark); r-h IL-2 was from EuroCetus (Milan, Italy); mouse IgG1 antihuman IFN-γR2 mAb was from PBL Biomedical Laboratories (New Brunswick, NJ); r-h IFN-γ was kindly provided by Dr M. Brunda (Hoffman-La Roche, Nutley, NJ); and mouse IgG1 antihuman IFN-γR1 γR99 mAb was kindly provided by Dr G. Garotta (Hoffmann-La Roche, Basel, Switzerland).

The culture medium was RPMI 1640 (BioWhittaker, Walkersville, MD) supplemented with penicillin, streptomycin, gentamycin, 2.5 × 10^–5 M 2-mercaptoethanol (2-ME), and 10% fetal calf serum (FCS; Invitrogen) and is referred to hereafter as complete medium. All the in vitro cultures were performed at 37°C in a 5% CO₂ humidified atmosphere.

Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized venous blood obtained from 5 healthy donors by Ficoll-Type 400 (Pharmacia, Uppsala, Sweden) gradient centrifugation and stimulated (1 × 10⁶ cells/mL) with PHA (2.5 μg/mL). After 3 days, PHA-activated T lymphocytes (T lymphoblasts) were cultured in complete medium containing r-h IL-2 (20 U/mL). ST4 T cells were derived from a convoluted-type T-cell lymphoma stabilized in vitro and in nu/nu mice starting from biopsy material.²⁷  Molt-4 (American Type Culture Collection [ATCC, Rockville, MD]; CRL1582) and Jurkat (ATCC; CRL8161) are human T cells from acute lymphoblastic leukemia. U937 (ATCC; CRL1593) is a promonocytic cell line, and Raji (ATCC; CCL86) cells are human B lymphocytes from a patient with Burkitt lymphoma.

To evaluate IFN-γR1 and IFN-γR2 expression, malignant T, B, and myeloid cells were cultured in the presence or absence of serum, with or without IGF-1 (100 ng/mL), anti–IGF-1R mAb (10 μg/mL), or insulin (3600 ng/mL). In some experiments, T lymphoblasts were cultured in the presence of IGF-1, anti–IGF-1R mAb, or insulin in complete medium, with or without IL-2 (20 U/mL). After 24 hours, cells were recovered, washed twice in cold phosphate-buffered saline (PBS), supplemented with 0.2% bovine serum albumin (BSA) and 0.1% sodium azide, and stained for surface protein with unconjugated anti–IFN-γR1 γR99 or anti–IFN-γR2 mAb, followed by FITC-conjugated rabbit anti–mouse Ig. All labeling steps were followed by incubation for 30 minutes at 4°C and separated by 2 washes with cold PBS supplemented with 0.2% BSA and 0.1% sodium azide. In some experiments, the expression of IFN-γR2 was evaluated in CD3⁺ peripheral blood lymphocytes (PBLs), either freshly isolated or cultured for 48 hours in complete medium alone, in complete medium containing 10 μg/mL anti–IGF-1R, or in medium without serum containing 100 ng/mL IGF-1. Cells were simultaneously stained with FITC-conjugated anti-CD3 mAb and biotin-conjugated anti–IFN-γR2, followed by phycoerythrin (PE)–conjugated streptavidin (Dako). To determine apoptosis, the Annexin-V–FITC Apoptosis Detection Kit was used (Oncogene, Boston, MA). Briefly, control cells or cells pretreated for 24 hours with IGF-1 (100 ng/mL) or anti–IGF-1R mAb (10 μg/mL) were cultured for a further 48 hours in the presence of IFN-γ (1000 U/mL), resuspended, and stained with FITC-conjugated Annexin-V and propidium iodide (PI) according to the manufacturer's instructions. Membrane expression and apoptosis were determined with a FACScan flow cytometer (Becton Dickinson, Milan, Italy). Each plot represents the results from 10 000 events.

ST4 cells (1 × 10⁶ cells/mL) were cultured in complete medium with or without anti–IGF-1R mAb (10 μg/mL), or in the absence of serum, with or without IGF-1 (100 ng/mL). After 24 hours, IFN-γR2 expression was evaluated by RT-PCR on the recovered cells as previously described.¹⁹  Total cellular RNA was extracted with the Trizol (Invitrogen). Specific glyceraldehyde-3-phosphate dehydrogenase (G3PDH) primer pairs were obtained from Clontech (Palo Alto, CA). PCR product (15 μL) was electrophoresed in a 2% agarose gel in Tris (tris(hydroxymethyl)aminomethane)/boric acid/EDTA (ethylenediaminetetraacetic acid) buffer. Gels were stained with ethidium bromide (Sigma Chemical, St Louis, MO) and photographed.

Biotin-conjugated anti–IFN-γR2 mAb was used at concentrations of 10 to 20 μg/mL. Briefly, 1 mg/mL anti–IFN-γR2 mAb was dialyzed against 0.1 M carbonate buffer (pH 8.5) and conjugated to biotin-N-hydroxysuccinimide ester (1 mg/mL; Sigma Chemical) in dimethyl sulfoxide (DMSO) for 4 hours at room temperature and dialyzed against PBS. ST4 cells cultured for 24 hours in the presence or absence of serum were recovered and incubated with biotin-conjugated IFN-γR2 or isotype-matched mouse IgG1 control mAb for 4 hours at 37°C or at 4°C in the absence or presence of IGF-1 (100 ng/mL). Cell surface–associated mAb was removed by treating twice with acid pH (2 minutes at pH 3.0) as described.¹⁷  Then cells were fixed and permeabilized as described elsewhere¹⁷  and incubated for 30 minutes at 4°C with PE-conjugated streptavidin. In parallel endocytosis experiments, serum-deprived ST4 cells were incubated in the absence or presence of scalar doses of IGF-1 (from 1 to 100 ng/mL) with unconjugated IFN-γR2 mAb or isotype-matched mouse IgG1 control mAb as described in “Flow cytometry.” After fixation and permeabilization, cells were incubated with a rabbit F(ab′)₂ FITC-conjugated anti–mouse Ig (Dako). IFN-γR2 endocytosis was measured as cell-associated specific fluorescence by flow cytometry.

Treated cells (5 × 10⁶) were washed twice in cold PBS and then collected by centrifugation. Nuclear proteins (25 or 30 μg protein) were extracted as previously described¹³  and separated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis at 140 V on 8% protein mini-gels. Gels were electroblotted onto polyvinylidene fluoride membranes (Bio-Rad, Hercules, CA) at 100 V for one hour, and the equality of the amount of protein analyzed was checked by nonspecific staining with Ponceau S (Sigma Chemical). The membranes were blocked with TTBS (20 mM Tris-HCl [pH 7.5], 0.5 M NaCl, and 0.05% Tween 20) and 5% nonfat dry milk for 3 hours and then incubated overnight with a 1:1000 dilution of anti–phospho-tyr (701)-STAT-1, or anti–STAT-1 rabbit polyclonal antibodies. After washing with TTBS, blots were incubated with 1:2000 horseradish peroxidase–conjugated goat anti–rabbit IgG antibody. Antibody reactions were visualized with enhanced chemiluminescence reagents according to the manufacturer's instructions (ECL plus; Amersham International, Bucks, United Kingdom). Nonphosphorylated STAT-1 was used as a control for equal protein loading. Fold increase of treated cells relative to untreated cells was quantitated after normalization with nonphosphorylated STAT-1. Density scanning was performed using the University of Texas Health Science Center (San Antonio, TX) ImageTools for Windows 2.0.

Following pretreatment with medium, IGF-1, or anti–IGF-1R mAb for 48 hours, cells (1 × 10⁷ cells/mL) were incubated with 1000 U/mL IFN-γ at 37°C. At the appropriate time interval, 100 μL was removed and used for EMSA. Preparation of cell lysates, EMSAs, and all data analyses were performed as previously described.¹³  EMSAs were performed with a 22-bp sequence containing a STAT-1α binding site corresponding to the sis inducible element (SIE) from the IRF-1 gene promoter (5′-GATCG ATTTCCCCGAAATCATG-3′).¹³ 

The observation that serum deprivation up-regulates IFN-γR2 on human T lymphocytes¹³  prompted us to evaluate whether IGF-1 (present in serum) was involved in IFN-γR2 internalization. We used 3 T-cell lines (ST4, Jurkat, and Molt-4) that do not produce IFN-γ (data not shown) and display low surface expression of IFN-γR2 when grown in the presence of serum.¹³  The fact that IFN-γR2 protein is prevalently accumulated in the cytoplasm of these malignant T-cell lines¹³  underlies an IFN-γ–independent IFN-γR2 internalization mechanism.¹⁷ 

Since these T-cell lines express high surface levels of IGF-1R and do not secrete IGF-1 (Schillaci et al²⁸  and data not shown), we evaluated the effect of IGF-1 (100 ng/mL) on IFN-γR2 internalization when they were cultured in the absence of serum. Since insulin, which is also present in serum,²⁹  activates IGF-1R, although at 10-fold higher concentration than IGF-1,³⁰  a 36-fold higher concentration of insulin was used to assess the specificity of action of IGF-1 on IFN-γR2 expression. In addition, the effect of preventing IGF-1 from binding to its specific receptor with a blocking anti–IGF-1R mAb was evaluated on T-cell lines cultured in the presence of serum.

Flow cytometry showed that malignant T cells cultured for 24 hours in the presence of serum displayed low-level surface IFN-γR2. Only 22% to 29% of cells were IFN-γR2 positive with a mean fluorescence intensity (MFI) between 7 to 23 (Table 1). IGF-1 slightly decreased this percentage in all 3 lines, although the MFI of Molt-4 and ST4 cells was reduced more than 2-fold (Table 1). Conversely, anti–IGF-1R mAb markedly enhanced both the percentage of IFN-γR2–positive cells and the MFI in all 3 cell lines (Table 1). No differences in percent positivity or MFI were observed in the presence of insulin (Table 1).

When IGF-1 was depleted by serum deprivation, the percentage of IFN-γR2–positive cells was markedly enhanced in all 3 cell lines, though enhancement of the MFI was appreciable only in Jurkat and Molt-4 cells (Table 1). The enhancement of IFN-γR2 expression was not affected by the presence of insulin. In contrast, IGF-1 down-regulated IFN-γR2 expression in all 3 cell lines. In the presence of IGF-1, the percentage of IFN-γR2–positive cells and the MFI of ST4 and Molt-4 cells fell to levels lower than those found in the presence of serum (Table 1).

The effect of IGF-1, anti–IGF-1R mAb, and insulin on IFN-γR2 expression was also evaluated in B (Raji) and myeloid (U937) cells, which constitutively express high levels of both chains.¹³  Their high percent positivity and MFI for IFN-γR2 were not affected by the presence or lack of serum (Table 1).

Unlike IFN-γR2, the high surface expression of IFN-γR1 on ST4 cells was not affected by IGF-1, anti–IGF-1R mAb (Figure 1), or insulin (data not shown). Similar results were observed in Jurkat, Molt-4, Raji, and U937 cells (data not shown).

Thus, it seems that IGF-1 provides a signal that mediates IFN-γR2 internalization. This signal is restricted to T cells and does not affect the expression of IFN-γR1.

RT-PCR analysis revealed that in ST4 cells cultured in the presence of serum IFN-γR2 mRNA is constitutively expressed (Figure 2). This was not modified by 24-hour culture in the absence of serum, in the absence of serum in the presence of IGF-1, or in complete medium in the presence of anti–IGF-1R mAb (Figure 2). At this time point, cell viability of ST4 cells was increased by 6% to 7% in the presence of serum or the absence of serum in the presence of IGF-1, and reduced by 3% to 4% in the absence of serum or in complete medium in the presence of anti–IGF-1R mAb. These results indicate that modulation of IFN-γR2 surface expression by IGF-1 or anti–IGF-1R mAb did not affect T lymphocyte IFN-γR2 transcription and was not influenced by variations in cell viability due to long-term culture. Similar results were obtained with Jurkat and Molt-4 (data not shown).

To confirm that IGF-1 induces internalization of IFN-γR2, we analyzed its effects on the accumulation of biotin-conjugated anti–IFN-γR2 mAb, at 37°C or at 4°C, in ST4 cells cultured without serum. We have previously shown that in human T lymphoblasts, internalization of IFN-γR2 induces at least a 2-fold increase of cell-associated specific fluorescence of anti–IFN-γR2 mAb uptake after 4-hour incubation at 37°C.¹⁷  In ST4 cells cultured with serum for 24 hours, an about 2-fold increase of cell-associated specific fluorescence of anti–IFN-γR2 mAb uptake was observed after 4-hour incubation at 37°C (Figure 3A). In contrast, no anti–IFN-γR2 mAb uptake was observed on ST4 cells cultured for 24 hours in the absence of serum (Figure 3A), indicating that IFN-γR2 internalization was inhibited. Similar inhibition of IFN-γR2 uptake was observed when ST4 cells were incubated for 4 hours at 37°C in potassium-free medium (data not shown), which prevents endocytosis of receptors that use clathrin for internalization.¹⁷  However, when IGF-1 was added to serum-deprived ST4 cells during the 4-hour incubation, the uptake of anti–IFN-γR2 mAb was completely restored (Figure 3A). IGF-1–induced IFN-γR2 internalization in serum-deprived ST4 cells was found to be dose dependent (Figure 3B). These data indicated that, in the absence of serum, IFN-γR2 internalization in T cells is inhibited and can be restored by IGF-1.

To determine whether IGF-1–induced IFN-γR2 internalization is confined to malignant T cells, the effect of anti–IGF-1R mAb on surface IFN-γR2 in normal human T lymphocytes was evaluated. In resting CD3⁺ PBLs, IFN-γR2 surface expression is barely detectable (Novelli et al⁸  and data not shown). IFN-γR2 internalization was not modified in these cells after 24- to 48-hour culture in the presence or absence of serum, nor in the presence of serum and anti–IGF-1R mAb (data not shown). This indication that IGF-1–induced IFN-γR2 internalization is restricted to cycling T cells was confirmed by stimulating PBMCs from healthy donors to proliferate with PHA. After 5 days, PHA-activated T lymphoblasts were cultured for a further 48 hours in complete medium containing IL-2, with or without anti–IGF-1R mAb. Flow cytometry showed that T lymphoblasts cultured in complete medium expressed low levels of surface IFN-γR2 (mean IFN-γR2 percent positivity, 11 ± 1; mean MFI, 7 ± 2) (Figure 4) but, in the presence of anti–IGF-1R mAb, IFN-γR2 expression was enhanced (mean IFN-γR2 percent positivity, 44 ± 4; mean MFI, 82 ± 2) (Figure 4). These data indicated that IGF-1 plays a general role in inducing IFN-γR2 internalization in cycling human T cells.

The same pattern of IFN-γR2 internalization in the presence or absence of anti–IGF-1R mAb was also observed in T lymphoblasts cultured in the absence of IL-2 (data not shown). Thus, although IL-2 constitutes the main growth and survival factor for T cells, it is not involved in the negative regulation of the IFN-γR2.

Interaction of IFN-γ with its receptor phosphorylates Janus kinase 1 (Jak-1) and Jak-2, which activate STAT-1.^31,32  Since surface IFN-γR2 levels critically control the rate of IFN-γ–induced STAT-1 activation,¹³  the effect of IGF-1 or anti–IGF-1R mAb on the ability of IFN-γ to induce activation of STAT-1 on ST4 cells was evaluated.

ST4 cells were cultured in complete medium in the absence or presence of IGF-1 or anti–IGF-1R blocking mAb. After 48 hours, each culture was recovered and cultured for different times in the presence of IFN-γ. Nuclear proteins were extracted at different time points and analyzed by Western blot with a specific anti–phospho-tyr (701)-STAT-1 mAb (Figure 5A) and by EMSA with a high-affinity STAT-1α binding oligonucleotide (Figure 5B).

In cells cultured in complete medium, pretreated with either IGF-1 or anti–IGF-1 mAb and not exposed to IFN-γ, no phosphorylation of STAT-1 was observed (Figure 5A). Addition of IFN-γ to cells cultured in complete medium induced phosphorylation of STAT-1, which was detectable after 15 minutes and then decreased. It was still detectable after 1 hour but not after 4 hours. In cells treated with IGF-1 prior to IFN-γ treatment, phosphorylation of STAT-1 was almost completely abrogated (Figure 5A). Compared with ST4 cells cultured in medium alone, IFN-γ–induced STAT-1 phosphorylation was approximately 3-fold lower after 15 minutes and almost disappeared after one hour (Figure 5A). In cells pretreated with anti–IGF-1R mAb, the high-level phosphorylation of STAT-1 induced by IFN-γ was much more sustained, being still detectable after 6 hours (Figure 5A).

In cells cultured in complete medium, pretreated with either IGF-1 or anti–IGF-1R mAb and not exposed to IFN-γ, no DNA binding activity of STAT-1 was observed (Figure 5B). In cells cultured in complete medium, IFN-γ treatment induced STAT-1 DNA binding activity that was detected after 15 minutes, after which it decreased and was mostly lost after 1 hour. Compared with cells cultured in complete medium, the IFN-γ–induced STAT-1 DNA binding activity of cells pretreated with IGF-1 was reduced in amplitude and delayed, as it was observed only after 30 minutes and was no longer detectable after 1 hour (Figure 5B). In contrast, compared with cells cultured in medium only, the IFN-γ–induced STAT-1 DNA binding activity of cells pretreated with anti–IGF-1R mAb was 2-fold and 3.5-fold increased after 15 and 30 minutes, respectively (Figure 5B).

Pretreatment of Raji B and U937 myeloid cells with IGF-1 did not modify the ability of IFN-γ to induce strong and sustained STAT-1 activation (data not shown).

These data indicate that, by decreasing IFN-γR2 expression, IGF-1 plays a role in extinguishing the IFN-γ/STAT-1 pathway in human T lymphocytes. By contrast, IGF-1R functional blockade strongly enhances and sustains STAT-1 activation.

T lymphocytes express low surface levels of IFN-γR2 and are resistant to STAT-1–dependent apoptosis induced by IFN-γ.^8,13,17  Since pretreatment with IGF-1 decreased IFN-γR2 and IFN-γ–dependent STAT-1 activation in malignant T cells (Figures 1,5), we evaluated whether anti–IGF-1R mAb pretreatment reinstated their sensitivity to the apoptotic signal of IFN-γ.

ST4 cells were cultured with or without either 100 ng/mL IGF-1 or 10 μg/mL anti–IGF-1R mAb. After 48 hours, each culture was recovered, split, and cultured for a further 24 hours in the presence or absence of 1000 U/mL IFN-γ. This was followed by staining cells simultaneously with PI and FITC-conjugated Annexin-V to evaluate apoptosis (Figure 6).

When cultured in both medium and IGF-1, ST4 displayed very low percentages of apoptotic cells, irrespective of the presence of IFN-γ (Figure 6, top and middle panels), whereas prior treatment with anti–IGF-1R mAb led to a dramatic increase in apoptosis after exposure to IFN-γ (Figure 6, lower panels).

These data indicate that IGF-1 blockade reinstates the IFN-γ–dependent apoptotic pathway in malignant T cells.

This study shows that IGF-1 is a critical environmental factor that selectively keeps IFN-γR2 expression low and thus limits activation of the IFN-γ/STAT-1 pathway in T lymphocytes. By using 3 non–IFN-γ–producing malignant T-cell lines that express IFN-γR2 prevalently in their cytoplasm, we have demonstrated that depletion of IGF-1, by serum deprivation or by a mAb that blocks its receptor, increases surface expression of IFN-γR2. Addition of IGF-1 down-regulates surface IFN-γR2 in serum-deprived T cells. This down-regulation is the result of internalization since, in malignant T cells cultured in the absence of serum, IGF-1 induces dose-dependent enhancement of anti–IFN-γR2 mAb uptake but it does not have any effect on IFN-γR2 mRNA expression. B-, myeloid, and T-cell lines used in this study express similar IGF-1R levels and respond to IGF-1 (data not shown). However, IGF-1 regulates IFN-γR2 in T cells only. This implies that B and myeloid cells expressing high surface levels of IFNγ-R2 (Badovinac et al⁵  and this study) are not sensitive to the internalization signal mediated by IGF-1. Moreover, our data indicate that T-cell–specific IFN-γR2 internalization is induced by IGF-1 but not by insulin.

Our data indicate that IGF-1 induces the internalization of IFN-γR2 but not that of IFN-γR1. This provides further evidence of the independent regulation of these 2 chains in human T cells.¹⁷  In effect, T-cell surface expression is down-regulated by IFN-γ,^9,11  and in IFN-γ–stimulated Jurkat T cells IFN-γR1 is endocytosed and translocated to the nucleus from plasma membrane lipid microdomains or “rafts,”³³  which constitute a clathrin-independent endocytic pathway.³⁴  In contrast, in human T cells IFN-γR2 is constitutively internalized in clathrin-coated vescicles.¹⁷  Thus, it would seem that distinct pathways of endocytosis are involved in ligand-dependent IFN-γR1 and IGF-1–dependent IFN-γR2 internalization in T cells. This hypothesis is endorsed by the observation that filipin, a cavaeolae/lipid microdomain inhibitor,³³  did not affect the ability of IGF-1 to induce IFN-γR2 internalization (data not shown).

Negative regulation of IFN-γR2 by IGF-1 also occurs in normal activated T cells. We observed that in the presence of anti–IGF-1R mAb the expression of IFN-γR2 is enhanced in PHA-activated lymphoblasts but not in resting CD3⁺ PBLs. Thus, IGF-1 regulates internalization only in activated, cycling T cells. This is consistent with previous observations that the IGF-1R is expressed in polyclonally activated T cells³⁵  but not in resting T cells.³⁶  However, our data indicate that IGF-1–mediated down-regulation of IFN-γR2 expression is not influenced by IL-2 since IFN-γR2 up-regulation was observed in the presence of anti–IGF-1R mAb irrespective of the presence of IL-2. IFN-γR1 was controlled only by IL-2 withdrawal.¹⁰  In effect, both IFN-γR1 and IL-2Rβ share the same clathrin-independent pathway of endocytosis.^33,35 

Our data clearly show that IGF-1 desensitizes IFN-γ/STAT-1 signaling in T cells since IFN-γ–dependent STAT-1 activation was strongly inhibited by IGF-1 pretreatment. Thus, by limiting the availability of IFNγ-R2 at the T-cell surface, IGF-1 is a factor that directly prevents IFN-γ from triggering a sufficient number of heterodimeric receptors.⁴  The fact that IGF-1 and anti–IGF-1R mAbs appear to have a greater effect on the IFN-γ–induced activation of STAT-1 than on expression of IFN-γR2 would suggest that other IGF-1–dependent mechanisms concur in IFN-γ/STAT-1 desensitization. It is conceivable that IGF-1, by activating STAT-3,³⁷  might suppress the STAT-1 pathway³⁸  and enhance the ability of IFN-γ to activate STAT-3.^39,40  The reciprocal influence of IFN-γ and IGF-1 signaling is further suggested by the observation that IFN-γ up-regulates the expression of IGF-1R in T lymphocytes (data not shown). Studies to dissect the role of cross-talk through the IGF-1R and IFN-γR in regulating the STAT-1/STAT-3 balance in T lymphocytes are currently in progress in our laboratory.

The ability of IFN-γ to induce apoptosis in T cells is dependent on STAT-1 activation.^6,13  Our data show that blockade of IGF-1R reinstates IFN-γ–induced apoptosis of T cells. Revival of strong IFN-γ–induced apoptosis by this blockade may be of importance in regulation of IFN-γ/STAT-1–dependent apoptosis of T cells in both physiologic and pathologic scenarios. IFN-γ plays a central role in cancer immunoediting,^41,42  and human tumors develop a selective insensitivity to IFN-γ signaling.⁴³  Our data suggest that in addition to being a growth and antiapoptotic factor,⁴⁴  IGF-1 acts as a negative regulator of IFN-γ signaling and thus contributes to the escape of developing tumor cells from the immunoregulatory (eg, antigen [Ag] presentation) and antiproliferative effects of IFN-γ.⁴⁵  The resistance of malignant T cells to IFN-γ–induced apoptosis may be a consequence of the down-regulation of IFN-γR2.^13,46  Our data suggest that by reinstating IFN-γ–induced STAT-1 activation and apoptosis, IGF-1R blockade may offer a way of overcoming the resistance of malignant T cells to IFN-γ. In vivo experiments are in progress to ascertain the effect of combined administration of anti–IGF-1R mAb and IFN-γ on the growth of malignant T cells in severe combined immunodeficiency (SCID) mice.

During the early steps of Th1 differentiation, the IFN-γ/STAT-1 pathway is critical for the decision of naive T cells to develop along the Th1 lineage.⁴⁷  Since our data show that IGF-1 is a critical factor for tuning IFN-γ induction of STAT-1 in T lymphocytes, it would be interesting to evaluate the possibility of modulating or inhibiting Th1 development by IGF-1.

We thank Drs J. Iliffe, E. Ferrero, and S. Pellegrini for critically reading the manuscript.