0013.7227/92/1315-2244$03.00/O Endocrinology Copyright 0 1992 by The Endocrine
Expression Receptors Cells
Vol. 131, No. 5 Printed in U.S.A.
Society
of Type I Insulin-Like on Human Peripheral
Growth Factor Blood Mononuclear
RON KOOIJMAN, MIA WILLEMS, CARLA J. C. DE HAAS, GER T. RIJKERS, ALEX L. G. SCHUURMANS, SYLVIA C. VAN BUUL-OFFERS, COB1 J. HEIJNEN, BEN J. M. ZEGERS
AND
Departments of Immunology (R. K., M. W., G. T. R., C.J. H., B. J. M.Z.) and Endocrinology (A. L.G.S., S.C.v.B.-O.), University Hospital for Children and Youth “Het Wilhelmina Kinderziekenhuis,” 3501 CA Utrecht, The Netherlands ABSTRACT Insulin-like growth factor-I and -II (IGF-I and IGF-II), both of which bind to type I IGF receptors, can modulate certain functions of the immune system. We, therefore, studied the expression of type I IGF receptors on purified subpopulations of peripheral blood mononuclear cells. Using two-color flow cytometry, specific binding of a monoclonal antibody directed against the type I IGF receptor (aIR3) was found in every subpopulation. Relatively high numbers of receptors were detected on monocytes, natural killer cells, and CD4’ T-helper cells, an intermediate number of receptors on CD8+ suppressor/cytotoxic T-cells, and a relatively low number of receptors on B-cells. The presence of these receptors was confirmed by specific binding of [1251]
IGF-I to purified subpopulations. otIR3 inhibited the binding of [lz51] IGF-I. The specific binding of [‘251]IGF-I to monocytes could be completely inhibited by IGF-II and insulin, but higher doses of these peptides were needed than of IGF-I. Scatchard analysis revealed the presence of 734 + 426 receptors/monocyte, with a Kd of 0.23 & 0.05 nM. A lower number of receptors (230 + 52), but with a higher affinity (Kd = 0.05 + 0.02 nM), was found on purified T-cells. The positive effect of IGF-I on natural killer cell cytotoxicity indicates that the type I IGF receptors on this cell type are functional. The possibility that IGF-I mediates hormonal effects on the immune system is discussed. (Endocrinology 131: 2244-2250, 1992)
T
ing of [‘251]IGF-I was inhibitable with aIR3. In addition, we show that the cytotoxic activity of NK cells can be enhanced by IGF-I.
HE INSULIN -like growth factors (IGF-I and IGF-II) comprise a family of peptides that promote the proliferation and differentiation of a variety of cells (1, 2). Specific receptors for both growth factors have been described; the type I IGF receptor binds both IGF-I and IGF-II with a high affinity, whereas the type II receptor binds IGF-II with a high affinity and IGF-I with a low affinity (3). IGF-I has been shown to augment chemotaxis (4) and in vitro T-cell proliferation (4-7). These studies and the findings that human macrophages(8), rat leukocytes (9), and human T-cell blasts (Kooijman, R., unpublished observation) produce IGFI indicate a possiblerole for IGF-I in immune function. Type I IGF receptors have been identified on human peripheral blood mononuclear cells (PBMC) (lo), but receptor studies on purified subpopulations thus far have only been performed for T-cells (4, 11). To obtain more insight into the potential role of IGF-I in the immune system, we investigated the expression of the type I IGF receptor in various subpopulations of human peripheral blood mononuclear cells. Using two-color flow cytometry, we found that natural killer (NK) cells, monocytes, B-cells, and both CD4+ and CD8+ T-cells possesstype I IGF receptors. We were able to confirm the presence of type I IGF receptors on all subpopulations by radioligand binding assayson purified subpopulations. The specific bindReceived May 15, 1992. Address all correspondence and requests for reprints to: Dr. R. Kooijman, Department of Immunology, Wilhelmina Kinderziekenhuis, P.O. Box 18009, 3501 CA Utrecht, The Netherlands,
Materials
and Methods
Reagents Human (h) IGF-I obtained from Cohn fraction IV of human plasma (12) was radioiodinated to a specific activity of 750-1500 Ci/mmol using the chloramine-T method. Recombinant hIGF-I and hIGF-II were kindly provided by Dr. Jeatran, Lilly Research Laboratories (Indianapolis, IN). Biotinylated goat antimouse immunoglobulin G (IgG) and (~1R3 (IgGlK) were purchased from Oncogene Science (Manhasset, NY). All other monoclonal antibodies (mAbs) and phycoerythrin (PE)-conjugated streptavidin were obtained from Becton Dickinson (Mountain View, CA). BSA and insulin were from Sigma (St. Louis, MO), and normal human serum was a pool from five healthy adults.
Cells Peripheral blood was drawn from healthy adult donors by venepuncture, using 7.5 U/ml heparin as an anticoagulant. PBMC were purified from the heparinized blood by centrifugation on Ficoll/Isopaque (Pharmacia, Uppsala, Sweden) density gradients (1.077 g/ml) at 1000 x g for 20 min at room temperature. For studies requiring large numbers of PBMC, these were purified from buffy coats (Red Cross Bloodbank, Utrecht, The Netherlands) with a citrate-phosphate-dextrose solution as an anticoagulant. Blood was depleted of thrombocytes by gently shaking for 10 min in the presence of glass beads and 10 mM CaCl?, and PBMC were subsequently isolated as described above. Monocytes were isolated from PBMC by centrifugation on a Percoll (Pharmacia, Uppsala, Sweden) density gradient (1.063 g/ml) at 400 x g for 30 min at room temperature. After this procedure, 92-95s of the
2244
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
I IGF RECEPTORS
cells at the Percoll interface were monocytes as assessed with antiCDllc (Leu M5) and flow cytometric analysis. For the isolation of T-cells and NK cells, PBMC were depleted of Bcells and monocytes by passage over a nylon wool column. Subsequently, the T-cells and NK cells were separated by rosetting the NK cells. To this end, cells were incubated with an excess of anti-CD16 (100 X lo6 cells in 100 ~1 Leu lib; 10 &ml) for 30 min at 5 C. Subsequently, the cells were washed twice in cold Minimum Essential Medium (MEM) and mixed with a 30.fold excess of goat antimouse &-coated ox erythrocytes and centrifuged for 7 min at 250 X g. Rosetted cells were separated from nonrosetted cells by density centrifugation on FicollIsopaque, as previously described (13). T-Cell preparations contained 95-98% T-cells, as determined by flow cytometric analysis using antiCD3 (Leu 4) as a lineage-specific marker. NK cell preparations were 9095% pure, as assessed with anti-CD16 (Leu lib and Leu 11~). B-Cells were isolated from a monocyte-depleted non-T-cell fraction, which was obtained by rosetting of T-cells with 2-aminoethylisothiouronium bromide-treated sheep red blood cells (14) and subsequent removal of monocytes by carbonyl iron treatment (15). The B-cells were separated from other cells by the rosetting technique described above, using anti-CD20 (100 X lo6 cells in 1 ml Leu 16; 5 Kg/ml) as a B-cellspecific antibody. Cell preparations were 90-97% pure, as assessed with anti-CD19 (Leu 12) and anti-CD20 (Leu 16). The cell viability of all preparations was determined by trypan blue exclusion and was always greater than 95%.
Whole blood (95 ~1) was incubated with 5 ~1 cuIR3 (100 pg/ml) for 30 min at 5 C. The mAb (~1R3 has been shown to bind specifically to type I IGF receptors (16). As controls, an isotype-matched mAb (IgGlK) of irrelevant specificity (keyhole limpet hemocyanin) was used. Subsequently, erythrocytes were lysed by the addition of 1 ml lysing solution (Becton Dickinson) and incubation for 10 min at room temperature. After one wash in PBS, the cells were incubated at 5 C with biotinylated goat antimouse Ig (4 rg/ml) in MEM containing 1% BSA, 0.02% NaN3, and 20% normal human serum. After 30 min, the cells were washed twice in MEM containing 1% BSA and 0.02% NaN3. Labeling steps with PE-conjugated streptavidin (1:lO) and fluorescein-isothyocyanate (FITC)-conjugated mAbs against lineage-specific markers (1:lO) were performed according to the same procedure, except that normal human serum was omitted in the incubation medium. We used the same markers as those described above for isolated subpopulations. T-Helper cells and cytotoxic/suppressor T-cells were identified with anti-CD4‘(Leu 3) and anti-CD8 (Leu 2), respectivelv. This three-steu labeling method was 5 times more sensitive ihan a two-step method with a FITC-conjugated goat antimouse Ig. PBMC (0.5 X 106) were suspended in MEM containing 1% BSA, 0.02% NaNa, 20% normal human serum, and 5 pg/ml aIR3. After incubation for 30 min at 5 C, the cells were treated with biotinconjugated goat antimouse Ig, streptavidin-PE, and FITC-labeled cellspecific markers, as described above. Stained cells were analyzed on a FACStar+ flow cytometer equipped with an argon laser (488 nm excitation) and appropriate filter settings for FITC and FE fluorescence examination For each sample, forward light scatter, side scatter, green fluorescence (FITC), and red fluorescence (PE) signals were acquired in the list mode. The spectral overlap of FITC and PE fluorescence emission spectra was corrected by electronic compensation. Data analysis was performed using Lysys I software (Becton Dickinson). Based on the intensity of the FITC signal, an electronic gate was set on the subpopulation of cells staining for that uarticular mAb. The relative PE fluorescence intensities of these cells was compared with the FE fluorescence intensity of the same cells stained with the isotype-matched control antibody. The Wilcoxon signed rank test was used to determine the significance of differences in type I IGF receptor expression on different subpopulations of PBMC. A
IGF-I
binding
studies
Cells were washed twice in MEM containing were carried out with l-2 X lo6 cells in a final
1% BSA. Binding studies volume of 0.3 ml MEM-
ON MONONUCLEAR
CELLS
2245
1% BSA at 5 C for 16 h. There was no further increase in binding after 16 h of incubation (our unpublished results). For Scatchard analysis, cells were incubated with various concentrations of ‘251-labeled hIGF-I. Nonspecific binding for three concentrations of [‘Z”I]IGF-I was determined in the presence of at least a 500.fold excess of unlabeled recombinant hIGF-I. Calculated values for the other concentrations were derived by linear regression using Enzfitter software (Elsevier-Biosoft, Cambridge, United Kingdom). Competition binding studies were performed with 0.12-0.23 nM ‘251-labeled hIGF-I (80,000 cpm) and different concentrations of unlabeled recombinant hIGF-I, recombinant hIGF-II, and insulin. Cells were separated from the unbound ligand by two washing steps in 1 ml MEM-1% BSA at 5 C. Scatchard analysis of the binding data was performed with Enzfitter software, using a linear or nonlinear curve fit. Given values are the calculated means from at least three independent experiments + SD.
NK cytotoxicity
assay
NK cell activity was assessed by means of a 3-h 51Cr release cytotoxicity assay, with the K562 cell line used as the target. Monocytes were removed from PBMC by carbonyl iron treatment -‘-e remaining cells were incubated with or without IGF-I in round-b-nom microtiter wells (Nunc, Glostrup, Denmark) for 18 h at 37 C in 5% COZ. Culture medium consisted of RPM1 with 10% inactivated fetal calf serum (6, 17), 100 U/ ml penicillin, 100 pg/ml streptomycin, and 4 mM glutamine. Inactivation of fetal calf serum (FCS) was performed by treatment with dithiothreitol, a procedure that eliminates all detectable IGF-I and IGF-II (17). After incubation, the culture medium was replaced by RIM-5% FCS, 10 U/ ml interleukin-2 (Boerhinger, Mannheim, Germany), 100 U/ml penicillin, 100 Kg/ml streptomycin, 4 mM glutamine, and 0.5 PM 2-mercaptoethanol, After 2-h preincubation at 37 C in 5% CO?, the cells were added to 10,000 5’Cr-labeled target cells at various effecter/target cell ratios. Spontaneous and maximum release were determined by incubating labeled target cells in culture medium or 1% Triton X-100, respectively. Cytotoxicity was expressed by the formula: % cytotoxicity = (experimental release - spontaneous release)/(total release - spontaneous release).
Results Immunofluorescence The expression of type I IGF receptors on different subpopulations of PBMC was investigated using two-color flow cytometry. To avoid possible changes in receptor expression during the isolation of PBMC, the first labeling step with otIR3 was performed immediately after blood was taken. Type I receptors were detectable on all subpopulations, as depicted in a representative set of histograms (Fig. 1A). Large differences in the mean expression of the type I receptors on cells from different donors were not found (Table 1). Statistical analysis revealed that monocytes and CD4+ T-helper cells express significantly more receptors than CD%+ cytotoxic/suppressor T-cells and B-cells, and that B-cells express a lower number of receptors than all other cell types (P < 0.02). In all experiments, the T-cell subsets could be divided into two groups: one with a high fluorescence intensity, and the other with a low fluorescence intensity. The difference between the median fluorescence intensity of aIR3-labeled cells and the median intensity of control cells suggests a 5to 20-fold difference in type I IGF receptor expression between the two groups of cells. It should be noted that differential expression of type I IGF receptors may influence otIR3 reactivity and, thus, affect the validity of receptor number estimation. No differences in the amount of type I IGF receptors within other subsets were detected.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
2246
TYPE I IGF RECEPTORS
ON MONONUCLEAR
CELLS
Endo. 1992 Vol 131. No 5
0 Fluorescence
Fluorescence
intensity
intensity
FIG. 1. Expression of type I IGF receptors on different subpopulations of PBMC. Cells were incubated with olIR3 or an isotypic control in either whole blood (A) or a suspension of isolated PBMC (B). Gates were set based on characteristic forward and side-scatter properties of PBMC, followed by gating on cells positive for specific FITC-conjugated lineage-specific antibodies. Histograms of aIR3-stained cells (solid curues) are superimposed over histograms of cells stained with an irrelevant isotope-matched antibody (dotted curues). The data in A and B are representative of seven and five experiments, respectively (see also Table 1). TABLE 1. Relative specific subpopulations of PBMC
binding
of oIR3
nIR3 incubation whole blood (n = 7) T-Cells (CD3’) CD4+ T-cells CD8’ T-cells Monocytes (CD14+) B-Cells (CD20+) NK cells (CDl6’)
8.6 10.5 7.5 10.2 4.9 9.0
3~ 2.1 + 2.3 f 1.4 + 0.9 + 1.4 f 1.9
in
to different aIR3 incubation of isolated PBMC (n = 5) 5.9 7.2 5.1 6.3 3.4 7.9
f f f +
2.6 3.0 2.4 4.3
+ 1.1
+ 4.1
Several independent experiments using different donors were performed, as described in Fig. 1. The specific binding of aIR3 is expressed in arbitrary fluorescence intensity units (SD) which are obtained by subtracting the mean fluorescence of the control cells from the mean fluorescence of aIR3-treated cells.
We also studied the expression of type I IGF receptors on isolated PBMC. Figure 1B and Table 1 show that every subpopulation of isolated PBMC binds otIR3, but to a lesser extent than the same cells that were labeled in whole blood. However, these differences were not significant. IGF-I
binding
studies
Scatchard plot analysis of the results from binding studies on purified monocytes revealed linear plots consistent with the presenceof a single high affinity binding site (Fig. 2). We found 734 f 426 binding sites/cell, with a Kd of 0.23 + 0.05 nM (n = 5). Table 2 shows that binding of [‘251]IGF-I to this
0
0.2
0.4 [IGF-I]
0.6
0.8
1 .o
nM
FIG. 2. Binding of [‘*“I]IGF-I to monocytes. W, Specific binding; Cl, nonspecific binding. In-set, Scatchard plot. The Scatchard plot was calculated in a one-site model by using Enzfitter software. For this experiment, the calculated number of receptors per cell (*SD) was 363 t 11, and the Kd (*SD) was 0.25 t 0.02 nM. Data points are the means of duplicate incubations within a representative experiment.
site can be inhibited by the addition of olIR3, indicating that the high affinity site is a type I IGF receptor. The specificity of IGF-I-binding sites on monocytes was also confirmed by competition binding assays. IGF-II and insulin competed
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
I IGF RECEPTORS
TABLE 2. Characteristics of type I IGF receptors IGF-I to purified subpopulations from PBMC Specific binding (molecules/cell) at 0.08-0.16
nM
IGF-I
on purified
ON MONONUCLEAR
monocytes
% Inhibition of specific binding by aIR3
and CD3+
T-cells,
2247
CELLS and effects
of aIR3
on specific
Kd (nM)
No. of high affinity sites/cell
binding
No. of donors
Monocytes T-Cells
284 k 116 198 f 65
86 + 9 91 + 7
0.23 + 0.05” 0.048 + 0.018”~* 0.053/0.075” 0.037 + 0.010*
734 f 426 230 + 52 313/235 201 f 29
n=5 n=5 n=2 n=3
NK cells B-Cells
326 f 192 256 f 126
83 + 8 87 + 11
ND’ ND
ND ND
n=3 n=3
Scatchard CD3+ T-cells and presence * Scatchard b Scatchard L ND. Not
analysis was performed using a linear curve fit for monocytes and CD3’ T-cells from two donors, and a curvilinear obtained from three other donors. For inhibition studies, purified cells were incubated with 0.08-0.16 nM [‘*“I]IGF-I of 3 pg/ml aIR3. Nonspecific binding was determined by the addition of 80 nM unlabeled IGF-I. analysis using a linear curve fit. analysis using a nonlinear curve tit. determined.
with [‘251]IGF-I for the binding sites with 4 and 833 times lower potencies than that of IGF-I, respectively (Fig. 3). Scatchard analysis on these binding data revealed the same high affinity site as those observed using different concentrations of [‘251]IGF-I (Fig. 3, inset). Scatchard plot analysis of purified (CD3+) T cells revealed both a high and a low affinity site in three donors (Fig. 4A) and only one high affinity binding site in two donors (Fig. 4B). When the amount of receptors was calculated using a two-site and a one-site model, respectively, we found a comparable number of high affinity binding sites.However, the mean Kd was higher for the two-binding site model (0.037 us. 0.064 nM, respectively; Table 2). The number of receptors on T-cells (230 + 52) should be considered as the average number on several distinct subsets (Fig. 1A). As depicted in Table 2, the specific binding of [iZ51]IGF-I on T-
fit was used for in the absence
cells can be inhibited by 91 + 7% with (~1R3,which identifies the high affinity binding site as a type I IGF receptor. The presence of type I receptors on B-cells and NK cells was confirmed by binding studiesto purified cell fractions. Table 2 also shows that the specific binding of [‘251]IGF-I to both B-cells and NK cellscan be inhibited by the addition of otIR3, indicating that all subsetspossessspecific binding sites for IGF-I and that at a concentration of about 0.1 nM the majority of IGF-I is bound to type I IGF receptors. NK cytotoxicity assay Type I IGF receptors on T-cells have been shown to be functional by the finding that IGF-I can modulate in vitro Tcell function (4-7). NK cells are lymphocytes involved in immunosurveillance for viral infections and neoplastic dis-
140 120 binding assay for [“L”I]IGF-I on monocytes. Purified monocytes were incubated with 0.18 nM [““I]IGF-I and increasing concentrations of unlabeled IGF-I (Ml, IGF-II (01, or insulin (7). Binding of [““I]IGF-1 is expressed as a percentage of maximum binding. Data pointsare the means of duplicate incubation within a representative experiment. The mean relative potencies (*SD) of IGF-II and insulin to compete for binding of [““I]IGF-I were 4.0 c 0.8 and 833 + 236 times that of IGF-I (n = 3). Inset, Scatchard plot showing a high and a low affinity binding site (the calculated values + SD are 758 + 102 high affinity sites, with a Kd of 0.121 + 72 nM, and 7000 f 4500 low affinity sites, with a Kd of 85 + 77 nM1.
of [“‘I]
FIG. 3. Competition
100 80 60 40 20 0 1
10
100
r&l unlabelled peptide
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
2248
I IGF RECEPTORS
ON MONONUCLEAR 90
$-2 4 a,a
." g
2
z
200
-
70 -
600
400
1992 No 5
80 -
800
7 0' a
Endo. Voll31.
CELLS
g
0
0.2
0.4
0.8
0.6 [IGF-I]
1.0
1.2
60 -
0 ;;: .- 50 fii :: 40 au 30
nM
20 10 1 3:1
6:l
12.5:1
effecter/target
z q 5
50
1
0 0
0.1
0.2
0.3 [IGF-I]
0.4
0.5
0.6
0.7
0.8
nM
FIG. 4. Binding of [““I]IGF-1 to CD3’ T-cells. A, Binding curves (W, specific binding; Cl, nonspecific binding) and Scatchard plot (inset) for T-cells with two binding sites. The calculated number of high affinity sites per cell (&SD) was 234 t 105, and the Kd (*SD) was 0.034 t 0.024 nM. The calculations for the low affinity site were inaccurate (SD, >lOO%). B, Binding curves (W, specific binding; 0, nonspecific binding) and Scatchard plot (inset) for T-cells with one binding site. The calculated number of high affinity sites per cell (&SD) was 235 + 7, and the Kd (&SD) was 0.075 + 0.007 nM. Data points are the means of duplicate incubations within a representative experiment.
eases.To study the functional role of type I IGF receptors on NK cells, we investigated the effects of IGF-I on NK cell cytotoxicity. Cells were cultured for 18 h in RPM1 supplemented with inactivated FCS. Inactivated FCS has been shown previously to be suitable for measuring the effects of IGF-I on T-cell proliferation (6). Cell viability and the percentage of NK cells were assessedby trypan blue exclusion and flow cytometry using anti-CD16 (Leu llb), respectively. Neither parameter was affected by inactivation of FCS or addition of IGF-I to cells cultured in inactivated FCS (data not shown). As depicted in Fig. 5A, the cytotoxic activity of NK cells was reduced when the cells were cultured in inactivated FCS compared to cells cultured in nontreated FCS.
25:l
cell
so:1
ratio
* Ii..b
70
1B
60
-
2
50
-
- iit .-0 ‘c
40
-
**
T
tii ::
30-
bp
20
-
10
-
o-
*
*
xx
drl
1
2
3
6:l
4
E/T
1
cell
ratio
2
3
4
donor
#
3:1
FIG. 5. Effects of IGF-I on NK cell cytotoxicity in monocyte-depleted PBMC. A, Effects of IGF-I on cytotoxicity at various effecter/target (E/T) cell ratios. The cytotoxicity was tested after culture in FCS (+) and inactivated FCS (0, without IGF-I; A, with lo-‘” M IGF-I; 0, with lo-” M IGF-I). Data points are the means of triplicate incubations (k SD) within a representative experiment. Significant effects of IGF-I in inactivated FCS are indicated (*, P < 0.02). B, Effects of IGF-I on cytotoxicity at effecter/target cell ratios of 3:l and 6:l for four different donors (0, without IGF-I; W, with lo-“’ M IGF-I). Significant differences are indicated (*, P < 0.02; **, P < 0.001).
The addition of 10-i’ and lOmEM IGF-I partially restored the NK cell activity. The strongest effects on cytotoxicity were observed at lower effecter/target cell ratios. When NK cells from four different donors were tested at effecter/target cell ratios of 3:l and 6:1, we found positive effects of IGF-I, although one observation was not significant (Fig. 58). The small effect of IGF-I in donor 1 may be a consequenceof the high cytotoxic activity of untreated cells. Discussion
This is the first study demonstrating the presence of type I IGF receptors on isolated monocytes, B-cells, and NK cells
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
I IGF RECEPTORS
using radioligand binding studies. Furthermore, we partially characterized the receptors on monocytes by competition binding assays using purified monocytes. Evidence for the presence of type I IGF receptors on resting T-cells has been presented by Tapson et al. (4), who also demonstrated the presence of two IGF-I-binding sites. They found 45 high affinity sites/cell, with a K,, of 0.12 nM, which is comparable to our results. Further analysis of IGF-I-binding sites was performed by Kozak et al. (11) on phytohemagglutininactivated T-cells. They demonstrated two binding sites and showed that the relative potencies of unlabeled IGF-I, IGFII, and insulin to compete for binding to the high affinity site are consistent with the binding of IGF-I to the type I IGF receptor. The low affinity site for IGF-I, which in our study was present on T-cells in three of 5 donors has not been identified. Membrane bound IGF-binding proteins or crossreactivity of IGF-I with type II IGF receptors or with insulin receptors may be responsible for low affinity binding. Both, our flow cytometric and binding data contradict the findings of Stuart et al. (18) that only monocytes and, to a lesser extent, B cells are responsible for the binding of IGF-I to PBMC. Our results, however are in accordance with the absence of a correlation between the binding of IGF-I and the monocyte concentration in monocyte enriched or depleted cell preparations as observed by Thorsson et al. (10). Furthermore, we confirmed the presence of type I IGF receptors using radioligand binding studies on purified subpopulations. The discrepancy between our results and the findings of Stuart et al. (18) might be explained by a difference in loss of receptors during isolation of PBMC and by different sensitivities of the immunofluorescence assays. Since the antibody we used for flow cytometry was added to whole blood within five minutes of venepuncture, the binding pattern of this antibody probably reflects the in uivo expression of type I IGF receptors on PBMC. Effects of heparin on the receptor expression are not likely, because the same binding pattern of otIR3 was found when 4 mM EDTA was used as an anticoagulant (data not shown). Binding of otIR3 to isolated PBMC was lower than that in whole blood. Since the reduction in binding was not the same in all subpopulations, it is unlikely that this effect is due to different conditions of incubation with the antibody. In addition, binding of cuIR3 sometimes decreased (up to 60%) during the purification of T-cells from PBMC, although the same labeling procedure was used before and after purification (data not shown). These results imply that the number of receptors estimated by Scatchard analysis on purified cells does not represent the number of receptors present in isolated PBMC or whole blood. We estimate that monocytes in whole blood possess about 500-2000 type I IGF receptors/cell, whereas the average number of receptors on T-cells is approximately 200-600. Furthermore, our flow cytometric analysis shows that NK cells possess about the same number of receptors as monocytes and T-cells and that B-cells express about 45% less receptors. Type I IGF receptors on T-cells have been shown to be functional in T-cell proliferation (4-7) and chemotaxis (4). In addition, the EDs0 of IGF-I for T-cell proliferation (0.12 nM) is proportional to the affinity of the type I receptor on these cells (6). The low EDso and the presence of high affinity type
ON MONONUCLEAR
CELLS
2249
I IGF receptors open up the possibility that leukocyte-derived IGF-I, which has been shown to be produced in very low amounts by rat leukocytes (9), plays a role as an autocrine factor in the immune system. Our observation that resting CD4’ and CDB+ T-cells express type I IGF receptors together with the binding data described by Johnson et al. (7) using purified CD4+ and CDB+ subsets of activated T-cells indicate that IGF-I might influence the actions of both T-helper cells (CD4+) and cytotoxic/suppressor T-cells (CDB+). This implies that there are potentially two mechanisms by which IGF-I can augment T-cell proliferation. Both modulation of Thelper cells, which secrete growth and differentiation factors, and modulation of T-suppressor cells can be involved. The presence of monocytes reduced the effects of IGF-I on T-cell proliferation (6) and NK cell cytotoxicity (data not shown). Whether these negative effects of monocytes are mediated by IGF-I remains to be elucidated. At the moment, the effects of IGF-I on the secretion of cytokines and prostaglandins that are known to affect T-cell and NK cell functions are being studied. Our observation that IGF-I can modulate the cytotoxic activity of NK cells suggests that the type I IGF receptors are also functional on this cell type. There are indications that IGF-I might increase NK cell cytotoxicity in viva. In women with impaired endogenous GH secretion and in healthy adults that were treated with GH, the NK cell cytotoxicity correlated ‘iyith the circulating levels of IGF-I (19, 20). Furthermore, GH deficiency results in a lower level of IGF-I and is associated with an impairment in NK cell cytotoxicity in both experimental animals and humans (21-24). Long term treatment of GH-deficient children increased NK cell cytotoxicity (22). Further studies are required to assess whether there is a causal relation between IGF-I levels and NK cell function. Since IGF-I secretion in different tissues can be regulated by several hormones (25), it would be interesting to address the role of IGF-I as a mediator of hormonal action on the immune system, such as the effects of GH on antibody production (26) and the effects of ACTH on antibody secretion, cytokine production, and NK cell function (27). References 1. Froesch ER, Schmid C, Schwander J, Zapf J 1985 Actions of insulin-like growth factors. Annu Rev Physiol 47:443-467 2. Baxter RC 1986 The somatomedins: insulin-like growth factors. Adv Clin Chem 25:49-115 3. Rechler MM, Nissley NP 1985 The nature and regulation of the receptors for insulin-like growth factors. Annu Rev Physiol47:425442 4. Tapson VF, Boni Schnetzler M, Pilch PF, Center DM, Berman JS 1988 Structural and functional characterization of the human T lymphocyte receptor for insulin-like growth factor I in vitro. J Clin Invest 82:950-957 5. Schimpff RM, Repellin AM, Salvatoni A, Thieriot Prevost G, Chatelain P 1983 Effect of purified somatomedins on thymidine incorporation into lectin-activated human lymphocytes. Acta Endocrinol (Copenh) 102:21-26 6. Kooijman R, Willems M, Rijkers GT, et al 1992 Effects of insulinlike growth factors and growth hormone on the ir! vitro proliferation of lymphocytes. J Neuroimmunol 38:95-104 7. Johnson EW, Jones LA, Kozak RW 1992 Expression and function of insulin-like growth factors on anti-CD3-activated human T lymphocytes. J Immunol 148:63-71 8. Rom WN, Basset P, Fells GA, Nukiwa T, Trapnell BC, Crysal RG
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
2250
9.
10.
11,
12.
13.
14.
15.
16.
17.
TYPE
I IGF RECEPTORS
1988 Alveolar macrophages release an insulin-like growth factor Itype molecule. J Clin Invest 82:1685-1693 Baxter JB, Blalock JE, Weighent DA 1991 Characterization of immunoreactive insulin-like growth factor from leukocytes and its regulation by growth hormone. Endocrinology 129:1727-1734 Thomson AV, Hintz RL 1977 Specific ‘Z51-somatomedin receptors on circulating human mononuclear cells. Biochem Biophys Res Commun 74:1566-1573 Kozak RW, Haskell JF, Greenstein LA, Rechler MM, Waldmann TA, Nissley SP 1987 Type I and II insulin-like growth factor receptors on human phytohemagglutinin-activated T lymphocytes. Cell 1mmun01109:318-331 Van Schravendijk CFH, Hooghe-Peters EL, Van den Brande JL, Pipeleers DG 1986 Receptors for insulin-like growth factors and insulin on murine fetal cortical brain cells. Biochem Biophys _ . Res Commun 135:228-238 Mudde GC. Verberne Cl, Groeneveld K, De Gast GC 1984 Human tonsil B lymphocyte function. I. The proliferative response to Staphylococcus aureus and pokeweed mitogen in relation to surface heavy mu and delta. Clin Exp Immunol 56:709-715 Gmelig Meyling F, Uytdehaag ACGM, Ballieux RE 1977 Human B activation i?z vitro. T Dependent pokeweed mitogen-induced differentiation of blood B lymphocytes. Cell Immunol 33:156-169 Perlmann H, Perlmann I’, Pape GR, Hallden G 1976 Purification, fractionation and assay of antibody-dependent lymphocyte effector cells (K-cells) in human blood. Stand J Immunol [Suppl] 5:57-68 Ku11 FC, Jacobs S, Su YF, Svoboda ME, Van Wijk JJ, Cuatrecasas P 1983 Monoclonal antibodies to receptors for insulin and somatomedin-C. J Biol Chem 258:6561-6566 Van Zoelen EJ, van Oostwaard TM, van der Saag PT, de Laat SW 1985 Phenotypic transformation of normal rat kidney cells in a growth-factor-defined medium: induction by a neuroblastoma-derived transforming growth factor independently of the EGF receptor. J Cell Physiol 123:151-160
ON MONONUCLEAR
CELLS
Endo. Voll31.
1992 No 5
18. Stuart CA, Meehan RT, Neale LS, Cintron NM, Furlanetto RW 1991 Insulin-like growth factor-l binds selectively to human peripheral blood monocytes and B-lymphocytes. J Clin Endocrinol Metab 72:1117-1122 19. Crist DM, Peake GT, Mackinnon LT, Sibbit WL, Kraner JC 1987 Exogenous growth hormone treatment alters body composition and increases natural killer cell activity in women with impaired endogenous growth hormone secretion. Metabolism 36:1115-l 117 20. Crist DM, Kraner JC 1990 Supplemental growth hormone increases the tumor cytotoxic activity of natural killer cells in healthy adults with normal growth hormone secretion. Metabolism 39:1320-1324 21. Saxena GB, Saxena RK, Adler WH 1982 Regulation of natural killer activity in vim. III. Effect of hypophysectomy and growth hormone treatment on the natural killer cell activity of the mouse spleen cell population. Int Arch Allergy Appl Immunol 67:169-174 22. Bozzola M, Valtorta A, Moretta A, Cisternino M, Biscaldi I, Schimpff RM 1990 In vitro and in vivo effect of growth hormone on cytotoxic activity. J Pediatr 117:596-599 23. Kiess W, Malozowski S, Gelato M, et al 1988 Lymphocyte subset distribution and natural killer activity in growth hormone deficiency before and during short-term treatment with growth hormone releasing hormone. Clin Immunol Immunopathol 48:85-94 24. Kiess W, Doerr H, Butenandt 0, Belohradsky BH 1986 Lymphocyte subsets and natural killer activity in growth hormone deficiency. N Engl J Med 314:321 25. Holly JMP, Wass JAH 1989 Insulin-like growth factors; autocrine, paracrine or endocrine? New perspectives of the somatomedin hypothesis in the light of recent developments. J Endocrinol 122:611618 26. Bozzola M, Cisternino M, Valtorta A, et al 1989 Effect of biosynthetic methionyl growth hormone (GH) therapy on the immune function in GH-deficient children. Horm Res 31:153-156 27. Bateman A, Singh A, Kral T, Solomon S 1992 The immunehypothalamic-pituitary-adrenal axis. Endocr Rev lo:92
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
Expression Receptors Cells
Vol. 131, No. 5 Printed in U.S.A.
Society
of Type I Insulin-Like on Human Peripheral
Growth Factor Blood Mononuclear
RON KOOIJMAN, MIA WILLEMS, CARLA J. C. DE HAAS, GER T. RIJKERS, ALEX L. G. SCHUURMANS, SYLVIA C. VAN BUUL-OFFERS, COB1 J. HEIJNEN, BEN J. M. ZEGERS
AND
Departments of Immunology (R. K., M. W., G. T. R., C.J. H., B. J. M.Z.) and Endocrinology (A. L.G.S., S.C.v.B.-O.), University Hospital for Children and Youth “Het Wilhelmina Kinderziekenhuis,” 3501 CA Utrecht, The Netherlands ABSTRACT Insulin-like growth factor-I and -II (IGF-I and IGF-II), both of which bind to type I IGF receptors, can modulate certain functions of the immune system. We, therefore, studied the expression of type I IGF receptors on purified subpopulations of peripheral blood mononuclear cells. Using two-color flow cytometry, specific binding of a monoclonal antibody directed against the type I IGF receptor (aIR3) was found in every subpopulation. Relatively high numbers of receptors were detected on monocytes, natural killer cells, and CD4’ T-helper cells, an intermediate number of receptors on CD8+ suppressor/cytotoxic T-cells, and a relatively low number of receptors on B-cells. The presence of these receptors was confirmed by specific binding of [1251]
IGF-I to purified subpopulations. otIR3 inhibited the binding of [lz51] IGF-I. The specific binding of [‘251]IGF-I to monocytes could be completely inhibited by IGF-II and insulin, but higher doses of these peptides were needed than of IGF-I. Scatchard analysis revealed the presence of 734 + 426 receptors/monocyte, with a Kd of 0.23 & 0.05 nM. A lower number of receptors (230 + 52), but with a higher affinity (Kd = 0.05 + 0.02 nM), was found on purified T-cells. The positive effect of IGF-I on natural killer cell cytotoxicity indicates that the type I IGF receptors on this cell type are functional. The possibility that IGF-I mediates hormonal effects on the immune system is discussed. (Endocrinology 131: 2244-2250, 1992)
T
ing of [‘251]IGF-I was inhibitable with aIR3. In addition, we show that the cytotoxic activity of NK cells can be enhanced by IGF-I.
HE INSULIN -like growth factors (IGF-I and IGF-II) comprise a family of peptides that promote the proliferation and differentiation of a variety of cells (1, 2). Specific receptors for both growth factors have been described; the type I IGF receptor binds both IGF-I and IGF-II with a high affinity, whereas the type II receptor binds IGF-II with a high affinity and IGF-I with a low affinity (3). IGF-I has been shown to augment chemotaxis (4) and in vitro T-cell proliferation (4-7). These studies and the findings that human macrophages(8), rat leukocytes (9), and human T-cell blasts (Kooijman, R., unpublished observation) produce IGFI indicate a possiblerole for IGF-I in immune function. Type I IGF receptors have been identified on human peripheral blood mononuclear cells (PBMC) (lo), but receptor studies on purified subpopulations thus far have only been performed for T-cells (4, 11). To obtain more insight into the potential role of IGF-I in the immune system, we investigated the expression of the type I IGF receptor in various subpopulations of human peripheral blood mononuclear cells. Using two-color flow cytometry, we found that natural killer (NK) cells, monocytes, B-cells, and both CD4+ and CD8+ T-cells possesstype I IGF receptors. We were able to confirm the presence of type I IGF receptors on all subpopulations by radioligand binding assayson purified subpopulations. The specific bindReceived May 15, 1992. Address all correspondence and requests for reprints to: Dr. R. Kooijman, Department of Immunology, Wilhelmina Kinderziekenhuis, P.O. Box 18009, 3501 CA Utrecht, The Netherlands,
Materials
and Methods
Reagents Human (h) IGF-I obtained from Cohn fraction IV of human plasma (12) was radioiodinated to a specific activity of 750-1500 Ci/mmol using the chloramine-T method. Recombinant hIGF-I and hIGF-II were kindly provided by Dr. Jeatran, Lilly Research Laboratories (Indianapolis, IN). Biotinylated goat antimouse immunoglobulin G (IgG) and (~1R3 (IgGlK) were purchased from Oncogene Science (Manhasset, NY). All other monoclonal antibodies (mAbs) and phycoerythrin (PE)-conjugated streptavidin were obtained from Becton Dickinson (Mountain View, CA). BSA and insulin were from Sigma (St. Louis, MO), and normal human serum was a pool from five healthy adults.
Cells Peripheral blood was drawn from healthy adult donors by venepuncture, using 7.5 U/ml heparin as an anticoagulant. PBMC were purified from the heparinized blood by centrifugation on Ficoll/Isopaque (Pharmacia, Uppsala, Sweden) density gradients (1.077 g/ml) at 1000 x g for 20 min at room temperature. For studies requiring large numbers of PBMC, these were purified from buffy coats (Red Cross Bloodbank, Utrecht, The Netherlands) with a citrate-phosphate-dextrose solution as an anticoagulant. Blood was depleted of thrombocytes by gently shaking for 10 min in the presence of glass beads and 10 mM CaCl?, and PBMC were subsequently isolated as described above. Monocytes were isolated from PBMC by centrifugation on a Percoll (Pharmacia, Uppsala, Sweden) density gradient (1.063 g/ml) at 400 x g for 30 min at room temperature. After this procedure, 92-95s of the
2244
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
I IGF RECEPTORS
cells at the Percoll interface were monocytes as assessed with antiCDllc (Leu M5) and flow cytometric analysis. For the isolation of T-cells and NK cells, PBMC were depleted of Bcells and monocytes by passage over a nylon wool column. Subsequently, the T-cells and NK cells were separated by rosetting the NK cells. To this end, cells were incubated with an excess of anti-CD16 (100 X lo6 cells in 100 ~1 Leu lib; 10 &ml) for 30 min at 5 C. Subsequently, the cells were washed twice in cold Minimum Essential Medium (MEM) and mixed with a 30.fold excess of goat antimouse &-coated ox erythrocytes and centrifuged for 7 min at 250 X g. Rosetted cells were separated from nonrosetted cells by density centrifugation on FicollIsopaque, as previously described (13). T-Cell preparations contained 95-98% T-cells, as determined by flow cytometric analysis using antiCD3 (Leu 4) as a lineage-specific marker. NK cell preparations were 9095% pure, as assessed with anti-CD16 (Leu lib and Leu 11~). B-Cells were isolated from a monocyte-depleted non-T-cell fraction, which was obtained by rosetting of T-cells with 2-aminoethylisothiouronium bromide-treated sheep red blood cells (14) and subsequent removal of monocytes by carbonyl iron treatment (15). The B-cells were separated from other cells by the rosetting technique described above, using anti-CD20 (100 X lo6 cells in 1 ml Leu 16; 5 Kg/ml) as a B-cellspecific antibody. Cell preparations were 90-97% pure, as assessed with anti-CD19 (Leu 12) and anti-CD20 (Leu 16). The cell viability of all preparations was determined by trypan blue exclusion and was always greater than 95%.
Whole blood (95 ~1) was incubated with 5 ~1 cuIR3 (100 pg/ml) for 30 min at 5 C. The mAb (~1R3 has been shown to bind specifically to type I IGF receptors (16). As controls, an isotype-matched mAb (IgGlK) of irrelevant specificity (keyhole limpet hemocyanin) was used. Subsequently, erythrocytes were lysed by the addition of 1 ml lysing solution (Becton Dickinson) and incubation for 10 min at room temperature. After one wash in PBS, the cells were incubated at 5 C with biotinylated goat antimouse Ig (4 rg/ml) in MEM containing 1% BSA, 0.02% NaN3, and 20% normal human serum. After 30 min, the cells were washed twice in MEM containing 1% BSA and 0.02% NaN3. Labeling steps with PE-conjugated streptavidin (1:lO) and fluorescein-isothyocyanate (FITC)-conjugated mAbs against lineage-specific markers (1:lO) were performed according to the same procedure, except that normal human serum was omitted in the incubation medium. We used the same markers as those described above for isolated subpopulations. T-Helper cells and cytotoxic/suppressor T-cells were identified with anti-CD4‘(Leu 3) and anti-CD8 (Leu 2), respectivelv. This three-steu labeling method was 5 times more sensitive ihan a two-step method with a FITC-conjugated goat antimouse Ig. PBMC (0.5 X 106) were suspended in MEM containing 1% BSA, 0.02% NaNa, 20% normal human serum, and 5 pg/ml aIR3. After incubation for 30 min at 5 C, the cells were treated with biotinconjugated goat antimouse Ig, streptavidin-PE, and FITC-labeled cellspecific markers, as described above. Stained cells were analyzed on a FACStar+ flow cytometer equipped with an argon laser (488 nm excitation) and appropriate filter settings for FITC and FE fluorescence examination For each sample, forward light scatter, side scatter, green fluorescence (FITC), and red fluorescence (PE) signals were acquired in the list mode. The spectral overlap of FITC and PE fluorescence emission spectra was corrected by electronic compensation. Data analysis was performed using Lysys I software (Becton Dickinson). Based on the intensity of the FITC signal, an electronic gate was set on the subpopulation of cells staining for that uarticular mAb. The relative PE fluorescence intensities of these cells was compared with the FE fluorescence intensity of the same cells stained with the isotype-matched control antibody. The Wilcoxon signed rank test was used to determine the significance of differences in type I IGF receptor expression on different subpopulations of PBMC. A
IGF-I
binding
studies
Cells were washed twice in MEM containing were carried out with l-2 X lo6 cells in a final
1% BSA. Binding studies volume of 0.3 ml MEM-
ON MONONUCLEAR
CELLS
2245
1% BSA at 5 C for 16 h. There was no further increase in binding after 16 h of incubation (our unpublished results). For Scatchard analysis, cells were incubated with various concentrations of ‘251-labeled hIGF-I. Nonspecific binding for three concentrations of [‘Z”I]IGF-I was determined in the presence of at least a 500.fold excess of unlabeled recombinant hIGF-I. Calculated values for the other concentrations were derived by linear regression using Enzfitter software (Elsevier-Biosoft, Cambridge, United Kingdom). Competition binding studies were performed with 0.12-0.23 nM ‘251-labeled hIGF-I (80,000 cpm) and different concentrations of unlabeled recombinant hIGF-I, recombinant hIGF-II, and insulin. Cells were separated from the unbound ligand by two washing steps in 1 ml MEM-1% BSA at 5 C. Scatchard analysis of the binding data was performed with Enzfitter software, using a linear or nonlinear curve fit. Given values are the calculated means from at least three independent experiments + SD.
NK cytotoxicity
assay
NK cell activity was assessed by means of a 3-h 51Cr release cytotoxicity assay, with the K562 cell line used as the target. Monocytes were removed from PBMC by carbonyl iron treatment -‘-e remaining cells were incubated with or without IGF-I in round-b-nom microtiter wells (Nunc, Glostrup, Denmark) for 18 h at 37 C in 5% COZ. Culture medium consisted of RPM1 with 10% inactivated fetal calf serum (6, 17), 100 U/ ml penicillin, 100 pg/ml streptomycin, and 4 mM glutamine. Inactivation of fetal calf serum (FCS) was performed by treatment with dithiothreitol, a procedure that eliminates all detectable IGF-I and IGF-II (17). After incubation, the culture medium was replaced by RIM-5% FCS, 10 U/ ml interleukin-2 (Boerhinger, Mannheim, Germany), 100 U/ml penicillin, 100 Kg/ml streptomycin, 4 mM glutamine, and 0.5 PM 2-mercaptoethanol, After 2-h preincubation at 37 C in 5% CO?, the cells were added to 10,000 5’Cr-labeled target cells at various effecter/target cell ratios. Spontaneous and maximum release were determined by incubating labeled target cells in culture medium or 1% Triton X-100, respectively. Cytotoxicity was expressed by the formula: % cytotoxicity = (experimental release - spontaneous release)/(total release - spontaneous release).
Results Immunofluorescence The expression of type I IGF receptors on different subpopulations of PBMC was investigated using two-color flow cytometry. To avoid possible changes in receptor expression during the isolation of PBMC, the first labeling step with otIR3 was performed immediately after blood was taken. Type I receptors were detectable on all subpopulations, as depicted in a representative set of histograms (Fig. 1A). Large differences in the mean expression of the type I receptors on cells from different donors were not found (Table 1). Statistical analysis revealed that monocytes and CD4+ T-helper cells express significantly more receptors than CD%+ cytotoxic/suppressor T-cells and B-cells, and that B-cells express a lower number of receptors than all other cell types (P < 0.02). In all experiments, the T-cell subsets could be divided into two groups: one with a high fluorescence intensity, and the other with a low fluorescence intensity. The difference between the median fluorescence intensity of aIR3-labeled cells and the median intensity of control cells suggests a 5to 20-fold difference in type I IGF receptor expression between the two groups of cells. It should be noted that differential expression of type I IGF receptors may influence otIR3 reactivity and, thus, affect the validity of receptor number estimation. No differences in the amount of type I IGF receptors within other subsets were detected.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
2246
TYPE I IGF RECEPTORS
ON MONONUCLEAR
CELLS
Endo. 1992 Vol 131. No 5
0 Fluorescence
Fluorescence
intensity
intensity
FIG. 1. Expression of type I IGF receptors on different subpopulations of PBMC. Cells were incubated with olIR3 or an isotypic control in either whole blood (A) or a suspension of isolated PBMC (B). Gates were set based on characteristic forward and side-scatter properties of PBMC, followed by gating on cells positive for specific FITC-conjugated lineage-specific antibodies. Histograms of aIR3-stained cells (solid curues) are superimposed over histograms of cells stained with an irrelevant isotope-matched antibody (dotted curues). The data in A and B are representative of seven and five experiments, respectively (see also Table 1). TABLE 1. Relative specific subpopulations of PBMC
binding
of oIR3
nIR3 incubation whole blood (n = 7) T-Cells (CD3’) CD4+ T-cells CD8’ T-cells Monocytes (CD14+) B-Cells (CD20+) NK cells (CDl6’)
8.6 10.5 7.5 10.2 4.9 9.0
3~ 2.1 + 2.3 f 1.4 + 0.9 + 1.4 f 1.9
in
to different aIR3 incubation of isolated PBMC (n = 5) 5.9 7.2 5.1 6.3 3.4 7.9
f f f +
2.6 3.0 2.4 4.3
+ 1.1
+ 4.1
Several independent experiments using different donors were performed, as described in Fig. 1. The specific binding of aIR3 is expressed in arbitrary fluorescence intensity units (SD) which are obtained by subtracting the mean fluorescence of the control cells from the mean fluorescence of aIR3-treated cells.
We also studied the expression of type I IGF receptors on isolated PBMC. Figure 1B and Table 1 show that every subpopulation of isolated PBMC binds otIR3, but to a lesser extent than the same cells that were labeled in whole blood. However, these differences were not significant. IGF-I
binding
studies
Scatchard plot analysis of the results from binding studies on purified monocytes revealed linear plots consistent with the presenceof a single high affinity binding site (Fig. 2). We found 734 f 426 binding sites/cell, with a Kd of 0.23 + 0.05 nM (n = 5). Table 2 shows that binding of [‘251]IGF-I to this
0
0.2
0.4 [IGF-I]
0.6
0.8
1 .o
nM
FIG. 2. Binding of [‘*“I]IGF-I to monocytes. W, Specific binding; Cl, nonspecific binding. In-set, Scatchard plot. The Scatchard plot was calculated in a one-site model by using Enzfitter software. For this experiment, the calculated number of receptors per cell (*SD) was 363 t 11, and the Kd (*SD) was 0.25 t 0.02 nM. Data points are the means of duplicate incubations within a representative experiment.
site can be inhibited by the addition of olIR3, indicating that the high affinity site is a type I IGF receptor. The specificity of IGF-I-binding sites on monocytes was also confirmed by competition binding assays. IGF-II and insulin competed
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
I IGF RECEPTORS
TABLE 2. Characteristics of type I IGF receptors IGF-I to purified subpopulations from PBMC Specific binding (molecules/cell) at 0.08-0.16
nM
IGF-I
on purified
ON MONONUCLEAR
monocytes
% Inhibition of specific binding by aIR3
and CD3+
T-cells,
2247
CELLS and effects
of aIR3
on specific
Kd (nM)
No. of high affinity sites/cell
binding
No. of donors
Monocytes T-Cells
284 k 116 198 f 65
86 + 9 91 + 7
0.23 + 0.05” 0.048 + 0.018”~* 0.053/0.075” 0.037 + 0.010*
734 f 426 230 + 52 313/235 201 f 29
n=5 n=5 n=2 n=3
NK cells B-Cells
326 f 192 256 f 126
83 + 8 87 + 11
ND’ ND
ND ND
n=3 n=3
Scatchard CD3+ T-cells and presence * Scatchard b Scatchard L ND. Not
analysis was performed using a linear curve fit for monocytes and CD3’ T-cells from two donors, and a curvilinear obtained from three other donors. For inhibition studies, purified cells were incubated with 0.08-0.16 nM [‘*“I]IGF-I of 3 pg/ml aIR3. Nonspecific binding was determined by the addition of 80 nM unlabeled IGF-I. analysis using a linear curve fit. analysis using a nonlinear curve tit. determined.
with [‘251]IGF-I for the binding sites with 4 and 833 times lower potencies than that of IGF-I, respectively (Fig. 3). Scatchard analysis on these binding data revealed the same high affinity site as those observed using different concentrations of [‘251]IGF-I (Fig. 3, inset). Scatchard plot analysis of purified (CD3+) T cells revealed both a high and a low affinity site in three donors (Fig. 4A) and only one high affinity binding site in two donors (Fig. 4B). When the amount of receptors was calculated using a two-site and a one-site model, respectively, we found a comparable number of high affinity binding sites.However, the mean Kd was higher for the two-binding site model (0.037 us. 0.064 nM, respectively; Table 2). The number of receptors on T-cells (230 + 52) should be considered as the average number on several distinct subsets (Fig. 1A). As depicted in Table 2, the specific binding of [iZ51]IGF-I on T-
fit was used for in the absence
cells can be inhibited by 91 + 7% with (~1R3,which identifies the high affinity binding site as a type I IGF receptor. The presence of type I receptors on B-cells and NK cells was confirmed by binding studiesto purified cell fractions. Table 2 also shows that the specific binding of [‘251]IGF-I to both B-cells and NK cellscan be inhibited by the addition of otIR3, indicating that all subsetspossessspecific binding sites for IGF-I and that at a concentration of about 0.1 nM the majority of IGF-I is bound to type I IGF receptors. NK cytotoxicity assay Type I IGF receptors on T-cells have been shown to be functional by the finding that IGF-I can modulate in vitro Tcell function (4-7). NK cells are lymphocytes involved in immunosurveillance for viral infections and neoplastic dis-
140 120 binding assay for [“L”I]IGF-I on monocytes. Purified monocytes were incubated with 0.18 nM [““I]IGF-I and increasing concentrations of unlabeled IGF-I (Ml, IGF-II (01, or insulin (7). Binding of [““I]IGF-1 is expressed as a percentage of maximum binding. Data pointsare the means of duplicate incubation within a representative experiment. The mean relative potencies (*SD) of IGF-II and insulin to compete for binding of [““I]IGF-I were 4.0 c 0.8 and 833 + 236 times that of IGF-I (n = 3). Inset, Scatchard plot showing a high and a low affinity binding site (the calculated values + SD are 758 + 102 high affinity sites, with a Kd of 0.121 + 72 nM, and 7000 f 4500 low affinity sites, with a Kd of 85 + 77 nM1.
of [“‘I]
FIG. 3. Competition
100 80 60 40 20 0 1
10
100
r&l unlabelled peptide
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
2248
I IGF RECEPTORS
ON MONONUCLEAR 90
$-2 4 a,a
." g
2
z
200
-
70 -
600
400
1992 No 5
80 -
800
7 0' a
Endo. Voll31.
CELLS
g
0
0.2
0.4
0.8
0.6 [IGF-I]
1.0
1.2
60 -
0 ;;: .- 50 fii :: 40 au 30
nM
20 10 1 3:1
6:l
12.5:1
effecter/target
z q 5
50
1
0 0
0.1
0.2
0.3 [IGF-I]
0.4
0.5
0.6
0.7
0.8
nM
FIG. 4. Binding of [““I]IGF-1 to CD3’ T-cells. A, Binding curves (W, specific binding; Cl, nonspecific binding) and Scatchard plot (inset) for T-cells with two binding sites. The calculated number of high affinity sites per cell (&SD) was 234 t 105, and the Kd (*SD) was 0.034 t 0.024 nM. The calculations for the low affinity site were inaccurate (SD, >lOO%). B, Binding curves (W, specific binding; 0, nonspecific binding) and Scatchard plot (inset) for T-cells with one binding site. The calculated number of high affinity sites per cell (&SD) was 235 + 7, and the Kd (&SD) was 0.075 + 0.007 nM. Data points are the means of duplicate incubations within a representative experiment.
eases.To study the functional role of type I IGF receptors on NK cells, we investigated the effects of IGF-I on NK cell cytotoxicity. Cells were cultured for 18 h in RPM1 supplemented with inactivated FCS. Inactivated FCS has been shown previously to be suitable for measuring the effects of IGF-I on T-cell proliferation (6). Cell viability and the percentage of NK cells were assessedby trypan blue exclusion and flow cytometry using anti-CD16 (Leu llb), respectively. Neither parameter was affected by inactivation of FCS or addition of IGF-I to cells cultured in inactivated FCS (data not shown). As depicted in Fig. 5A, the cytotoxic activity of NK cells was reduced when the cells were cultured in inactivated FCS compared to cells cultured in nontreated FCS.
25:l
cell
so:1
ratio
* Ii..b
70
1B
60
-
2
50
-
- iit .-0 ‘c
40
-
**
T
tii ::
30-
bp
20
-
10
-
o-
*
*
xx
drl
1
2
3
6:l
4
E/T
1
cell
ratio
2
3
4
donor
#
3:1
FIG. 5. Effects of IGF-I on NK cell cytotoxicity in monocyte-depleted PBMC. A, Effects of IGF-I on cytotoxicity at various effecter/target (E/T) cell ratios. The cytotoxicity was tested after culture in FCS (+) and inactivated FCS (0, without IGF-I; A, with lo-‘” M IGF-I; 0, with lo-” M IGF-I). Data points are the means of triplicate incubations (k SD) within a representative experiment. Significant effects of IGF-I in inactivated FCS are indicated (*, P < 0.02). B, Effects of IGF-I on cytotoxicity at effecter/target cell ratios of 3:l and 6:l for four different donors (0, without IGF-I; W, with lo-“’ M IGF-I). Significant differences are indicated (*, P < 0.02; **, P < 0.001).
The addition of 10-i’ and lOmEM IGF-I partially restored the NK cell activity. The strongest effects on cytotoxicity were observed at lower effecter/target cell ratios. When NK cells from four different donors were tested at effecter/target cell ratios of 3:l and 6:1, we found positive effects of IGF-I, although one observation was not significant (Fig. 58). The small effect of IGF-I in donor 1 may be a consequenceof the high cytotoxic activity of untreated cells. Discussion
This is the first study demonstrating the presence of type I IGF receptors on isolated monocytes, B-cells, and NK cells
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
TYPE
I IGF RECEPTORS
using radioligand binding studies. Furthermore, we partially characterized the receptors on monocytes by competition binding assays using purified monocytes. Evidence for the presence of type I IGF receptors on resting T-cells has been presented by Tapson et al. (4), who also demonstrated the presence of two IGF-I-binding sites. They found 45 high affinity sites/cell, with a K,, of 0.12 nM, which is comparable to our results. Further analysis of IGF-I-binding sites was performed by Kozak et al. (11) on phytohemagglutininactivated T-cells. They demonstrated two binding sites and showed that the relative potencies of unlabeled IGF-I, IGFII, and insulin to compete for binding to the high affinity site are consistent with the binding of IGF-I to the type I IGF receptor. The low affinity site for IGF-I, which in our study was present on T-cells in three of 5 donors has not been identified. Membrane bound IGF-binding proteins or crossreactivity of IGF-I with type II IGF receptors or with insulin receptors may be responsible for low affinity binding. Both, our flow cytometric and binding data contradict the findings of Stuart et al. (18) that only monocytes and, to a lesser extent, B cells are responsible for the binding of IGF-I to PBMC. Our results, however are in accordance with the absence of a correlation between the binding of IGF-I and the monocyte concentration in monocyte enriched or depleted cell preparations as observed by Thorsson et al. (10). Furthermore, we confirmed the presence of type I IGF receptors using radioligand binding studies on purified subpopulations. The discrepancy between our results and the findings of Stuart et al. (18) might be explained by a difference in loss of receptors during isolation of PBMC and by different sensitivities of the immunofluorescence assays. Since the antibody we used for flow cytometry was added to whole blood within five minutes of venepuncture, the binding pattern of this antibody probably reflects the in uivo expression of type I IGF receptors on PBMC. Effects of heparin on the receptor expression are not likely, because the same binding pattern of otIR3 was found when 4 mM EDTA was used as an anticoagulant (data not shown). Binding of otIR3 to isolated PBMC was lower than that in whole blood. Since the reduction in binding was not the same in all subpopulations, it is unlikely that this effect is due to different conditions of incubation with the antibody. In addition, binding of cuIR3 sometimes decreased (up to 60%) during the purification of T-cells from PBMC, although the same labeling procedure was used before and after purification (data not shown). These results imply that the number of receptors estimated by Scatchard analysis on purified cells does not represent the number of receptors present in isolated PBMC or whole blood. We estimate that monocytes in whole blood possess about 500-2000 type I IGF receptors/cell, whereas the average number of receptors on T-cells is approximately 200-600. Furthermore, our flow cytometric analysis shows that NK cells possess about the same number of receptors as monocytes and T-cells and that B-cells express about 45% less receptors. Type I IGF receptors on T-cells have been shown to be functional in T-cell proliferation (4-7) and chemotaxis (4). In addition, the EDs0 of IGF-I for T-cell proliferation (0.12 nM) is proportional to the affinity of the type I receptor on these cells (6). The low EDso and the presence of high affinity type
ON MONONUCLEAR
CELLS
2249
I IGF receptors open up the possibility that leukocyte-derived IGF-I, which has been shown to be produced in very low amounts by rat leukocytes (9), plays a role as an autocrine factor in the immune system. Our observation that resting CD4’ and CDB+ T-cells express type I IGF receptors together with the binding data described by Johnson et al. (7) using purified CD4+ and CDB+ subsets of activated T-cells indicate that IGF-I might influence the actions of both T-helper cells (CD4+) and cytotoxic/suppressor T-cells (CDB+). This implies that there are potentially two mechanisms by which IGF-I can augment T-cell proliferation. Both modulation of Thelper cells, which secrete growth and differentiation factors, and modulation of T-suppressor cells can be involved. The presence of monocytes reduced the effects of IGF-I on T-cell proliferation (6) and NK cell cytotoxicity (data not shown). Whether these negative effects of monocytes are mediated by IGF-I remains to be elucidated. At the moment, the effects of IGF-I on the secretion of cytokines and prostaglandins that are known to affect T-cell and NK cell functions are being studied. Our observation that IGF-I can modulate the cytotoxic activity of NK cells suggests that the type I IGF receptors are also functional on this cell type. There are indications that IGF-I might increase NK cell cytotoxicity in viva. In women with impaired endogenous GH secretion and in healthy adults that were treated with GH, the NK cell cytotoxicity correlated ‘iyith the circulating levels of IGF-I (19, 20). Furthermore, GH deficiency results in a lower level of IGF-I and is associated with an impairment in NK cell cytotoxicity in both experimental animals and humans (21-24). Long term treatment of GH-deficient children increased NK cell cytotoxicity (22). Further studies are required to assess whether there is a causal relation between IGF-I levels and NK cell function. Since IGF-I secretion in different tissues can be regulated by several hormones (25), it would be interesting to address the role of IGF-I as a mediator of hormonal action on the immune system, such as the effects of GH on antibody production (26) and the effects of ACTH on antibody secretion, cytokine production, and NK cell function (27). References 1. Froesch ER, Schmid C, Schwander J, Zapf J 1985 Actions of insulin-like growth factors. Annu Rev Physiol 47:443-467 2. Baxter RC 1986 The somatomedins: insulin-like growth factors. Adv Clin Chem 25:49-115 3. Rechler MM, Nissley NP 1985 The nature and regulation of the receptors for insulin-like growth factors. Annu Rev Physiol47:425442 4. Tapson VF, Boni Schnetzler M, Pilch PF, Center DM, Berman JS 1988 Structural and functional characterization of the human T lymphocyte receptor for insulin-like growth factor I in vitro. J Clin Invest 82:950-957 5. Schimpff RM, Repellin AM, Salvatoni A, Thieriot Prevost G, Chatelain P 1983 Effect of purified somatomedins on thymidine incorporation into lectin-activated human lymphocytes. Acta Endocrinol (Copenh) 102:21-26 6. Kooijman R, Willems M, Rijkers GT, et al 1992 Effects of insulinlike growth factors and growth hormone on the ir! vitro proliferation of lymphocytes. J Neuroimmunol 38:95-104 7. Johnson EW, Jones LA, Kozak RW 1992 Expression and function of insulin-like growth factors on anti-CD3-activated human T lymphocytes. J Immunol 148:63-71 8. Rom WN, Basset P, Fells GA, Nukiwa T, Trapnell BC, Crysal RG
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
2250
9.
10.
11,
12.
13.
14.
15.
16.
17.
TYPE
I IGF RECEPTORS
1988 Alveolar macrophages release an insulin-like growth factor Itype molecule. J Clin Invest 82:1685-1693 Baxter JB, Blalock JE, Weighent DA 1991 Characterization of immunoreactive insulin-like growth factor from leukocytes and its regulation by growth hormone. Endocrinology 129:1727-1734 Thomson AV, Hintz RL 1977 Specific ‘Z51-somatomedin receptors on circulating human mononuclear cells. Biochem Biophys Res Commun 74:1566-1573 Kozak RW, Haskell JF, Greenstein LA, Rechler MM, Waldmann TA, Nissley SP 1987 Type I and II insulin-like growth factor receptors on human phytohemagglutinin-activated T lymphocytes. Cell 1mmun01109:318-331 Van Schravendijk CFH, Hooghe-Peters EL, Van den Brande JL, Pipeleers DG 1986 Receptors for insulin-like growth factors and insulin on murine fetal cortical brain cells. Biochem Biophys _ . Res Commun 135:228-238 Mudde GC. Verberne Cl, Groeneveld K, De Gast GC 1984 Human tonsil B lymphocyte function. I. The proliferative response to Staphylococcus aureus and pokeweed mitogen in relation to surface heavy mu and delta. Clin Exp Immunol 56:709-715 Gmelig Meyling F, Uytdehaag ACGM, Ballieux RE 1977 Human B activation i?z vitro. T Dependent pokeweed mitogen-induced differentiation of blood B lymphocytes. Cell Immunol 33:156-169 Perlmann H, Perlmann I’, Pape GR, Hallden G 1976 Purification, fractionation and assay of antibody-dependent lymphocyte effector cells (K-cells) in human blood. Stand J Immunol [Suppl] 5:57-68 Ku11 FC, Jacobs S, Su YF, Svoboda ME, Van Wijk JJ, Cuatrecasas P 1983 Monoclonal antibodies to receptors for insulin and somatomedin-C. J Biol Chem 258:6561-6566 Van Zoelen EJ, van Oostwaard TM, van der Saag PT, de Laat SW 1985 Phenotypic transformation of normal rat kidney cells in a growth-factor-defined medium: induction by a neuroblastoma-derived transforming growth factor independently of the EGF receptor. J Cell Physiol 123:151-160
ON MONONUCLEAR
CELLS
Endo. Voll31.
1992 No 5
18. Stuart CA, Meehan RT, Neale LS, Cintron NM, Furlanetto RW 1991 Insulin-like growth factor-l binds selectively to human peripheral blood monocytes and B-lymphocytes. J Clin Endocrinol Metab 72:1117-1122 19. Crist DM, Peake GT, Mackinnon LT, Sibbit WL, Kraner JC 1987 Exogenous growth hormone treatment alters body composition and increases natural killer cell activity in women with impaired endogenous growth hormone secretion. Metabolism 36:1115-l 117 20. Crist DM, Kraner JC 1990 Supplemental growth hormone increases the tumor cytotoxic activity of natural killer cells in healthy adults with normal growth hormone secretion. Metabolism 39:1320-1324 21. Saxena GB, Saxena RK, Adler WH 1982 Regulation of natural killer activity in vim. III. Effect of hypophysectomy and growth hormone treatment on the natural killer cell activity of the mouse spleen cell population. Int Arch Allergy Appl Immunol 67:169-174 22. Bozzola M, Valtorta A, Moretta A, Cisternino M, Biscaldi I, Schimpff RM 1990 In vitro and in vivo effect of growth hormone on cytotoxic activity. J Pediatr 117:596-599 23. Kiess W, Malozowski S, Gelato M, et al 1988 Lymphocyte subset distribution and natural killer activity in growth hormone deficiency before and during short-term treatment with growth hormone releasing hormone. Clin Immunol Immunopathol 48:85-94 24. Kiess W, Doerr H, Butenandt 0, Belohradsky BH 1986 Lymphocyte subsets and natural killer activity in growth hormone deficiency. N Engl J Med 314:321 25. Holly JMP, Wass JAH 1989 Insulin-like growth factors; autocrine, paracrine or endocrine? New perspectives of the somatomedin hypothesis in the light of recent developments. J Endocrinol 122:611618 26. Bozzola M, Cisternino M, Valtorta A, et al 1989 Effect of biosynthetic methionyl growth hormone (GH) therapy on the immune function in GH-deficient children. Horm Res 31:153-156 27. Bateman A, Singh A, Kral T, Solomon S 1992 The immunehypothalamic-pituitary-adrenal axis. Endocr Rev lo:92
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 September 2015. at 03:33 For personal use only. No other uses without permission. . All rights reserved.
