European Journal of Endocrinology (1998) ISSN

Transcription

1 European Journal of Endocrinology (1998) ISSN Elevated insulin-like growth factor (IGF) binding protein (IGFBP)-2 and IGFBP-4 expression of leukemic T-cells is affected by autocrine/paracrine IGF-II action but not by IGF type I receptor expression Martin W Elmlinger, Michael S Sanatani, Mathias Bell, Günther E Dannecker 1 and Michael B Ranke Paediatric Endocrinology and 1 Haematology Department, University Children s Hospital, Ruemelinstrasse 23, D Tuebingen, Germany (Correspondence should be addressed to M W Elmlinger) Abstract We recently found evidence indicating that the source of elevated serum insulin-like growth factor binding protein (IGFBP)-2 in leukemia was the leukemic T-cells. Here we report that locally produced IGF-II affects IGFBP-2 expression and growth of leukemic cells through the IGF type I receptor. We measured IGFBP-2, -4 and IGF type I receptor (IGF-I-R) mrna by RT-PCR, cell growth and IGFBP-2 secretion (per 10 6 cells). IGF-I-R binding sites were assessed by 125 I-IGF replacement studies. Inhibition using an IGF-II antibody showed that tumor cell-derived IGF-II accounts for a significant 25% (P<0.001) increase in IGFBP-2 secretion and enhanced growth (P<0.01) of leukemic T-cells after 7 days in culture. IGFBP-2 secretion, but not IGFBP-2 mrna was specifically increased by IGFs, while no specific effect of insulin was detectable. The addition of 100 ng/ml IGF-II enhanced the IGFBP-2 secretion 2.8-fold, while the use of IGF-I only enhanced IGFBP-2 secretion 1.7-fold, although IGF-I enhanced IGF-II action. Through inhibition using JB1, a peptide inhibiting the IGF signal transduction by blocking the IGF-I-R, we demonstrated the involvement of the IGF-I-R in IGFBP-2 and -4 expression and leukemic cell growth. However, only slight differences in the IGF-I-R mrna expression were seen for T- and B-cells compared with the differences found for the IGFBP-2 and -4 mrna or IGFBP-2 secretion. Thus, although IGF-I-R mediates the autocrine/paracrine effects of the IGFs, IGF-I-R mrna expression is most probably not involved in the differential IGFBP-2/IGFBP-4 expression in leukemic cells. European Journal of Endocrinology Introduction Insulin-like growth factors (IGFs) stimulate cell growth and differentiation (see (1) for review). IGF-I mediates growth hormone action in post-natal life, while IGF-II is highly expressed in fetal tissues. Cellular responses to IGFs are mediated by way of the IGF receptors and are affected by at least seven IGF binding proteins (IGFBP-1 to -6 (2); IGFBP-7 (3)). IGFBPs modulate IGF actions by altering their half-lives and by competing with IGF receptors (1, 4). IGF receptor expression, cell cycle and other growth factors can affect the tissue-specific production of IGFs and IGFBPs. Analogous to the pattern of expression in fetal tissues, high expression of IGF-II and IGFBP-2 was detectable in many tumors, e.g. leukemia (5 9). Furthermore, the growth of some tumors was found to be associated with disorders in IGF-II expression. In rhabdomyosarcoma, for instance, relaxation during imprinting of the maternal allele of the IGF-II gene (10, 11) led to IGF-II overexpression. This, in turn, could stimulate growth and possibly even IGFBP-2 expression in tumor cells. In Wilms tumor, a point mutation of the WT1 gene, encoding for a transacting factor of IGF receptor type I (IGF-I-R) transcription, caused deregulation of the IGF signal transduction by means of IGF-I-R overexpression (12). Elevated IGF-I-R expression could also be detected in some lymphoblastic cell lines (13), although a genetic predisposition has not yet been established in leukemia. Clinical studies (14, 15) have shown that the occasionally high IGFBP-2 levels in leukemic serum normalize after hematological remission following chemotherapy. We recently found evidence that leukemic T-cells could cause elevated IGFBP-2 concentrations in leukemic serum (6). In an earlier study (16), IGF-I-R involvement in the growth of leukemic B- and T-cells was presumed. In addition, there is evidence that IGFBP-2 and IGFBP-4, unlike IGFBP-1 and IGFBP-3 (to which mainly endocrine functions are ascribed (1)), modulate the action of IGFs locally. We therefore hypothesized that IGFBP-2 expression in leukemic T-cells is stimulated by IGF-II to modulate the mitogenic action of locally secreted IGF-II. Hence, we 1998 Society of the European Journal of Endocrinology

2 338 M W Elmlinger and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (1998) 138 investigated the regulation of tumor cell growth and IGFBP-2 and -4 expression in leukemic T- and B-cell lines (6). Our aim was to establish whether IGF-II exerts an autocrine/paracrine effect and whether the IGF-I-R is involved in this regulation. Materials and methods Cell culture and growth factor treatment Leukemic B-cell (Raji) and T-cell (Molt-4) lines were cultured in serum-free (SFM) RPMI 1640 medium (6) for up to seven days. Cells were counted using a Coulter counter (Cobas Minos STE, Hoffmann LaRoche, Basel, Switzerland). In order to study the effects of growth factors, the leukemic cells were cultured in SFM containing different final concentrations (10, 50 or 100 ng/ml) of IGF-I (a kind gift from Pharmacia and Upjohn, Stockholm, Sweden), IGF-II (Mediagnost, Tübingen, Germany) and insulin (Sigma, Deisenhofen, Germany). Seventy nanograms per milliliter IGF-II were used as a high physiological concentration for elaboration of the time course of IGFBP-2 secretion. In a subset of experiments, 1.0 mg/ml of the IGF analog JB1 (16) was added to the SFM to block IGF-I-R binding sites. High concentrations (>200 ng/ml) of JB1 could replace IGF-I, thus demonstrating specifity of IGF-I binding for the IGF-I-R binding sites. Since high amounts of IGF-II (6) are known to be secreted by Molt-4 cells, we also added an excess amount (700 ng/ml) of a monoclonal IGF-II antibody (DPC Biermann, Bad Nauheim, Germany) with the aim of suppressing the possible autocrine/paracrine effect of IGF-II. Reverse transcriptase (RT)-PCR analysis of specific mrnas RT-PCR analysis was carried out to study the effects of growth factors on IGFBP-2, -4 and IGF-I-R mrna expression. First, total RNA was isolated from >10 6 cells. Semi-quantitative PCR was performed using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression of each sample (set to 100 units), which served as an internal control of mrna expression as described (6). Criteria for primer choice were the following: regarding the protein structure of the aligned IGFBP-1 to -6 sequences, we chose primers (with no intra- and intermolecular homologies), referring to the middle domain of the mature IGFBP-2 and -4, i.e. the respective cdna, where minimum homology is found between IGFBPs (1). As far as possible, the exonic primers span at least one intron to avoid false positive signals due to genomic DNA, although DNAse treatment has been carried out. PCR procedures, i.e. annealing temperature, time and Mg 2þ concentration were optimized for each primer pair. Primer pairs were (a) IGFBP-2 (6), nucleotides , (b) IGFBP-4 (sense) sn5 0 - ATCGAGGCCAT CCAGGAA-3 0 ; (antisense) asn5 0 -AAA GCTGTCA GCCAGCTG-3 0, nucleotides , (c) IGF type I receptor IGF-I-R: sn5 0 -ACAGAGAACCC-CAA GACTGAGG-3 0 ;asn5 0 -TGATGTTGTAGGTGTCTGCGGC-3 0, nucleotides After electrophoresis, PCR products were quantified densitometrically (CS1, Cybertech, Berlin, Germany) by using a specific software package (WinCam 2.1, Cybertech). Specific radioimmunoassay for IGFBP-2 IGFBP-2 secretion by the cells was measured in conditioned medium by specific radioimmunoassay (RIA) as described previously (6). No difference in IGFBP-2 antibody binding was found between noncomplexed IGFBP-2 and IGFBP-2 complexed with IGF-I or IGF-II. Intra- and interassay variation of the IGFBP-2 RIA (sensitivity: 0.2 ng/ml) was 8.8 and 10.7% respectively. IGFBP-2 secretion is shown for 10 6 cells. Receptor binding of 125 I-labeled IGF-I on leukemic cells The number of IGF binding sites per cell of the leukemic T- and B-cell lines was assessed by means of replacement experiments. Molt-4 and Raji cells were grown to confluency in 12-well plates, washed twice with cold PBS (Ca 2þ - and Mg 2þ -free), counted and incubated with 125 I-labeled ( counts/0.012 ng) IGF-I in binding buffer (0.1 mmol/l HEPES, 1.2 mmol/l Mg 2 SO 4, 2.5 mmol/l KCl, 120 mmol/l NaCl, 10 mmol/l glucose, 1 mmol/l EDTA, 1% (w/v) BSA, ph 7.4). Varying concentrations of non-radioactive recombinant human (rh) IGF-I ( nmol/l) were used to replace the 125 I-IGF-I. One hundred and sixty-five nanomol/liter IGF-I were used to detect non-specific binding. The cells were incubated overnight at 4 C with gentle shaking. After the cells were transferred to small centrifuge tubes and carefully washed (twice) with PBS, the cell pellet-associated radioactivity was measured using a g-counter (Berthold, Wildbad, Germany). Curve-fitting and data analyses were conducted using the PC- LIGAND 3.1 program (developed by P Munson, GraphPad Software Inc., San Diego, CA, USA). Statistical methods Data were analyzed using standard statistical methods, including a paired Student s t-test. Results are reported as the mean S.E.M. of three or more independent experiments and significance was assigned a value of P<0.05, for differences between values of growth factor-treated cells and controls. Results Dose-dependent effects of IGF-I, IGF-II and insulin on IGFBP-2 expression Figure 1 shows the influence of treatment with different concentrations of IGF-I, IGF-II and insulin on secretion

3 EUROPEAN JOURNAL OF ENDOCRINOLOGY (1998) 138 IGFBP-2 and IGFBP-4 regulation by IGF-II in leukemic cells 339 (Fig. 1a) and mrna expression of IGFBP-2 (Fig. 1b, Table 1) in the leukemic T-cell line Molt-4. In an earlier report, Molt-4 cell lines in contrast to leukemic B-cell lines showed high expression of IGFBP-2 (6). We observed significant stimulation of IGFBP-2 mrna expression only when using 50 ng/ml IGF-II, and not with any IGF-I or insulin concentration. Higher doses (50 and 100 ng/ml) of IGF-II led to minor, but insignificant, stimulation of IGFBP-2, IGFBP-4 and IGF-I-R mrna. However, in contrast to insulin, IGF-I and IGF-II led to a marked, dose-dependent specific increase in the secretion of IGFBP-2 in leukemic T-cells. The increase in IGFBP-2 observed in insulin-treated cells was fully compensated by the increased cell number. IGF-II was 2.5 to 2.8 times more effective than IGF-I (P<0.001), and its action could be enhanced by 25% (P<0.01) in combination with IGF-I. Obviously, the IGFs regulate IGFBP-2 secretion in leukemic T-cells post-transcriptionally, rather than at the transcriptional level. Autocrine/paracrine IGF-II action on IGFBP-2 synthesis and cell growth Figure 2 shows the time course of secretion of IGFBP-2 (Fig. 2a) and cell growth (Fig. 2b) in Molt-4 cells under the influence of IGF-II for 7 days. Statistical significances of the differences are given in Table 2. IGFBP-2 concentrations increased time-dependently in the conditioned medium of cells which were not treated with IGFs (control). A highly significant (Table 2) further increase in IGFBP-2 secretion (P<0.001) and cell growth (P<0.01) resulted after addition of 70 ng/ml IGF-II. In contrast, the effect of IGF-I was less pronounced (see also Fig. 1a). In previous experiments [ 3 H]thymidine incorporation and Figure 1 Dose-dependent effects of IGF-I, IGF-II and insulin on (a) IGFBP-2 secretion and (b) IGFBP- 2 mrna accumulation in the leukemic T-cell line, Molt-4. Cells were grown in SFM with 10, 50 or 100 ng/ml IGF-I, IGF-II, or an equimolar mixture of both, or insulin over 4 days, or without IGFs (no treatment, control). IGFBP-2 mrna was measured by quantitative RT-PCR. The amount of IGFBP-2 per 10 6 cells in the conditioned medium was measured by specific RIA. Data are expressed as mean values of 3 5 experiments S.E.M. n.s., not significant; n.d., not determined; *P<0.05, **P<0.01, ***P<0.001 compared with control.

4 340 M W Elmlinger and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (1998) 138 Table 1 Expression of IGFBP-2, IGFBP-4 and IGF type I receptor mrna in a leukemic T-(Molt-4) and B-(Raji) cell line after introduction of growth factors and inhibitors. The amount of specific mrnas was measured by means of quantitative RT-PCR using total RNA. GAPDH mrna expression was set to 100 units and used as a standard for IGFBP-2, -4 and IGF-I-R mrna expression, given in relative units per 10 6 cells. Cells were grown over four days in SFM as a control, with SFM comprising 70 ng/ml IGF-I or IGF-II, 1 mg/ml IGF antagonist JB1 or 0.7 mg/ml anti-igf-ii antibody. Data (relative units) are expressed as mean values of 3 4 experiments S.E.M. Molt-4 Raji Culture conditions IGFBP-2 IGFBP-2 IGF type I receptor IGFBP-2 IGFBP-4 IGF type I receptor SFM control IGF-I IGF-II JB * * Anti-IGF-II 42 6* 7 2* * * P < 0:05 compared with control. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) proliferation assay have revealed a stimulating effect of IGFs on the proliferation of lymphocytes (not shown). The use of a monoclonal IGF-II antibody strongly inhibited the action of locally-secreted IGF-II on both IGFBP-2 secretion (P<0.001) and cell growth. The inhibition caused by the IGF antagonist JB1, a peptide which closely binds to IGF-I-R (17), demonstrated the involvement of IGF-I-R in IGFBP-2 regulation and cell growth. After an initial phase (four days) of insignificant inhibition of IGFBP-2, the secretion came to a complete halt, while partial inhibition of cell growth began on the first day. This suggests that IGF-I-R is strongly involved in the mediation of the autocrine/ paracrine action of IGF-II. Expression of IGF type I receptor, IGFBP-2 and IGFBP-4 mrna Table 1 compares the mrna expression of IGFBP-2, IGFBP-4 and IGF type I receptor (IGF-I-R) in the leukemic T-cell line, Molt-4 and in the B-cell line, Raji. Cells were cultured for four days in SFM comprising IGFs and inhibiting agents. Both IGFBP-2 and IGFBP-4 mrna were highly expressed in the T-cell line, Molt-4, but were only weakly expressed in the B-cell line, Raji. IGF-I and IGF-II did not significantly influence either the expression of IGFBP-2 mrna (see also Fig. 1a), or IGFBP-4 mrna in leukemic B- and T-cells, although minor inhibition was observed by the anti-igf-ii antibody (P<0.05). The IGF antagonist, JB1, did not Figure 2 Time course of (a) IGFBP-2 secretion and (b) cell growth in the leukemic T-cell line, Molt-4. Cells were grown in the presence of 70 ng/ml IGF-II, or 1 mg/ml IGF antagonist JB1 or 0.7 mg/ml anti-igf-ii antibody. IGFBP-2 was assayed by RIA and the cell number was determined using a Coulter blood cell counter. Data are expressed as means of 3 4 experiments S.E.M. (for statistical significances see Table 2).

5 EUROPEAN JOURNAL OF ENDOCRINOLOGY (1998) 138 IGFBP-2 and IGFBP-4 regulation by IGF-II in leukemic cells 341 Table 2 Statistical significance of differences between cell growth and IGFBP-2 secretion of growth factor-treated (as described in Material and methods) and nontreated leukemic cells after three, five and seven days in culture, corresponding to Fig. 2a and b. Values are means S.E.M. of secreted IGFBP-2 (ng/10 6 cells) and cell number ( 1000 cells). Days in culture IGFBP-2 (ng/10 6 cells) IGF-II 82 6* 126 7*** 136 8*** Anti-IGF-II 38 3*** 46 5*** 78 6*** JB n:s: 57 9** 54 7*** Cell number ( 1000 cells) IGF-II *** *** ** Anti-IGF-II * *** n:s: JB * *** *** * P < 0:05, ** P < 0:01, *** P < 0:001 compared with control; n.s., not significant. significantly influence IGFBP-2 expression in Molt-4 cells, but completely inhibited the low expression of IGFBP-2 and -4 in Raji cells. Thus, the mrna expression of IGFBP-2 and -4 appears to be regulated in the same way in leukemic cells. The expression of IGF-I-R mrna was of the same order of magnitude in Molt-4 and Raji cells, although the expression of IGFBP-2 and -4 and the levels of IGFBP-2 secretion were of an order of magnitude higher in Molt-4 cells. This premise was validated by receptor binding studies (Fig. 3) which revealed a relatively low number of IGF-I-R binding sites in Raji (850 sites) and Molt-4 (600 Figure 3 Scatchard plot of the binding of 125 I-IGF-I to the IGF-I-R of Molt-4 and Raji cells. Curve fitting was carried out by assuming one receptor binding site for IGF-I (1-site-fit). The dissociation constant (K d ) and the number of specific binding sites were estimated by employing a specific software package (see Materials and methods). The amount of bound IGF-I was up to 2 10 ¹11 mol/l (2E-11). Each point represents the mean of between three and five independent binding experiments; the S.E.M. was 9.8%. sites) cells. IGF-II neutralization through anti-igf-ii and the use of the antagonist JB1, however, led to a significant increase in IGF-I-R mrna expression in both cell lines (P<0.05 Molt-4; P<0.05 Raji). Discussion The IGFs are known to stimulate the growth of tumor cells (7, 8). As IGF-II is predominantly a fetal growth factor (1), elevated expression of IGF-II may be a sign for partial re-establishment of an embryonal genetic program in the tumor cells. We had previously shown that leukemic T-, but not B-cell lines secreted IGF-II and expressed high IGFBP-2 mrna and protein (6). Since the lymphoblastic IGFBP-2 expression appears to be about an order of magnitude higher than in normal B- and T-lymphocytes (6, 18, J L Föll, L Dannecker, C Zehrer, S Hettmer, M W Elmlinger, D Niethammer, M B Ranke & G E Dannecker, unpublished observations), an IGFBP-2 overexpression is likely in leukemic cells. The IGF-II/mannose-6-phosphate receptor in lymphoblastic cells was found to be mainly regulated by developmental factors (19). Therefore, we focused on the IGF-I-receptor expression. The present study revealed two major findings. First, we demonstrated a specific effect of IGF-II, produced by the tumor cells, as an autocrine/paracrine stimulator of IGFBP-2 production and enhanced growth of leukemic T-cells. Since the concentrations of IGFBP-2 and -4 were adjusted to the cell number, all changes shown can be considered as specific, i.e. independent of cell growth. A significant (P<0.001), strong decrease in IGFBP-2 secretion, and a slight decrease in IGFBP-2 and IGFBP-4 mrna (P<0.05) were seen after neutralization of locally produced IGF-II by the use of the IGF-II antibody. The inhibition by the IGF-II antibody also revealed an autocrine/paracrine effect of IGF-II on leukemic cell growth. The addition of IGF-I and IGF-II stimulated IGFBP-2 secretion, but IGFBP-2 and IGFBP-4

6 342 M W Elmlinger and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (1998) 138 mrna accumulation in leukemic T-cells was only poorly stimulated. IGF-I was a weaker stimulator of IGFBP-2 secretion, but could enhance IGF-II action. Even low doses (10 ng/ml) of IGF-I and IGF-II led to a significant (P<0.01) increase in IGFBP-2 secretion. We conclude that post-transcriptional steps rather than cellular mechanisms affecting IGFBP-2 mrna accumulation regulate IGFBP-2 gene expression. Insulin used in a dose (<100 ng/ml) below that affecting IGF-I-R (20), was completely ineffective with respect to a specific stimulation of IGFBP-2 synthesis. However, as known from previous studies (21), insulin enhanced leukemic cell growth (data not shown) by way of the insulin receptor. These results confirm that IGF-II is, as was hypothesized earlier (14), the most effective known regulator of IGFBP-2 expression. The second major finding is that IGF-I-R is present in both the leukemic B- and T-cells, as described earlier (20) and that it is involved in the mediation of the autocrine/paracrine effects of IGF-II. Inhibition of IGF-I-R with JB1, a peptide which blocks the IGF signal transduction (17), clearly demonstrated that IGF-I-R mediates the autocrine/paracrine actions of IGFs. In particular, cell growth, IGFBP-2 expression, IGFBP-4 mrna and IGF-I-R mrna accumulation were affected. However, as IGFBP-2 and -4 mrna levels were not associated with IGF-I-R mrna levels, the IGF-I-R mrna accumulation may not represent the mechanism which triggers the differential expression of IGFBP-2 and -4 and growth of leukemic T- and B-cells. IGF-I-R mrna levels were also unaffected by exogenous IGF-I or IGF-II. This conflicts with other studies (13, 21), where altered IGF-I-R expression was found to be associated with tumor growth. However, the IGF-I-R mrna level was significantly increased by the use of the IGF-II antibody (Table 1). Clearly, IGF-I-R gene expression (13) is modulated by the autocrine/paracrine action of IGF-II, which is mediated through the IGF-I-R itself. Although the IGF type-ii/mannose-6-phosphate receptor had been detected in non-malignant lymphocytes (7, 18, 19), it was found to be mainly regulated by developmental factors. Hence, the IGF-I-R appears to be a mediator of the IGF actions in leukemic cells. In an earlier study (9), it was demonstrated that leukemic cells secrete IGFBP-2 and -4. Here we found evidence that the mrna expression of both factors is affected in the same way by locally secreted IGF-II via the IGF-I-R. As both IGFBP-2 and -4 are known to modulate the mitogenic action of IGFs (4, 7, J L Föll, L Dannecker, C Zehrer, S Hettmer, M W Elmlinger, D Niethammer, M B Ranke & G E Dannecker, unpublished observations), it is likely that the expression of both factors is increased to modulate the autocrine/ paracrine effects of elevated IGF-II. Acknowledgements This work was supported by the Growth Research Center, Tuebingen, Germany. We would like to thank Karin Schmitt and Daniela Köndgen for their excellent technical assistance. References 1 Rechler MM. Insulin-like growth factor binding proteins. Vitamins and Hormones Shimasaki S & Ling N Identification and molecular characterization of insulin-like growth factor binding proteins (IGFBP-1, -2, -3, -4, -5 and -6). Progress in Growth Factor Research Oh Y, Nagalla SR, Yamanaka Y, Kim HS, Wilson E & Rosenfeld RG. Synthesis and characterization of insulin-like growth factor binding-protein (IGFBP)-7: recombinant human mac25 protein specifically binds IGF-I and IGF-II. Journal of Biological Chemistry Ross M, Francis GL, Szabo L, Wallace JC & Ballard FJ. Insulinlike growth factor (IGF)-binding proteins inhibit the biological activities of IGF-I and IGF-II but not des-(1 3)-IGF-I. Biochemical and Biophysical Research Communications Adamo ML, Shao Z-M, Lanau F, Chen JC, Clemmons DR, Roberts CT, LeRoith D & Fontana JA. Insulin-like growth factor-i (IGF-I) and retinoic acid modulation of IGF-binding proteins (IGFBPs): IGFBP-2, -3, -4 gene expression and protein secretion in a breast cancer cell line. Endocrinology Elmlinger MW, Wimmer K, Biemer E, Dannecker L, Blum WF, Ranke MB & Dannecker GE. Insulin-like growth factor binding protein-2 (IGFBP-2) is differentially expressed in leukemic T- and B-cell lines. Growth Regulation Daughaday WH. Editorial: the possible autocrine/paracrine and endocrine roles of insulin-like growth factors of human tumors. Endocrinology Reeve JG, Kirby LB, Brinkman A, Hughes SA, Schwander J & Bleehen NM. Insulin-like growth factor-binding protein gene expression and protein production by human tumor cell lines. International Journal of Cancer Neely EK, Smith SD & Rosenfeld RG. Human leukemic T and B lymphoblasts produce insulin-like growth factor binding proteins- 2 and -4. Acta Endocrinologica Minniti CP & Hellman LJ. IGF-II in the pathogenesis of rhabdomyosarcoma: a prototype of IGFs involvement in human tumorigenesis. Advances in Experimental Medicine and Biology Ogawa O, Eccles MR, Szeto J, Menoe LA, Yun K, Maw MA, Smith PJ & Reeve AE. Relaxation of insulin-like growth factor-ii gene imprinting implicated in Wilms tumor. Nature Werner H, Rauscher FJ III, Sukhatme VP, Drummond IA, Roberts CT Jr & LeRoith D. Transcriptional repression of the insulin-like growth factor-i receptor (IGF-I-R) gene by the tumor suppressor WT1 involves binding to sequences both upstream and downstream of the IGF-I-R gene transcription start site. Journal of Biological Chemistry LeRoith D, Baserga R, Helman L & Roberts CT. The regulation of IGF-I receptor gene expression. Biochemical and Cellular Biology Blum WF, Horn N, Kratzsch J, Jorgensen JOL, Juul A, Teale D, Mohnike K & Ranke MB. Clinical studies of IGFBP-2 by radioimmunoassay. Growth Regulation Mohnike K, Kluba U, Mittler U, Aumann V, Vorwerk P & Blum WF. Serum levels of insulin-like growth factor-i, -II and insulin-like growth factor binding proteins-2 and -3 in children with acute lymphoblastic leukaemia. European Journal of Pediatrics Lee PD, Rosenfeld RG, Hintz RL & Smith SD. Characterization of insulin, insulin-like growth factors-i and -II, and growth hormone

7 EUROPEAN JOURNAL OF ENDOCRINOLOGY (1998) 138 IGFBP-2 and IGFBP-4 regulation by IGF-II in leukemic cells 343 receptors on human leukemic lymphoblasts. Journal of Clinical Endocrinology and Metabolism Pietrzkowski Z, Wernicke D, Porcu P, Jameson BA & Baserga R. Inhibition of cellular proliferation by peptide analogues of insulinlike growth factors. Cancer Research Nyman T & Pekonen F. The expression of insulin-like growth factors and their binding proteins in normal human lymphocytes. Acta Endocrinologica Yang Y, Hoeflich A, Butenandt O & Kiess W. Opposite regulation of IGF-I and IGF-I receptor mrna and concomitant changes of GH receptor and IGF-II/M6P receptor mrna in human IM-9 lymphoblasts. Biochimica et Biophysica Acta Kellerer M, Obermaier-Kusser B, Ermel B, Häring H-U & Petrides PE. An altered IGF-I receptor is present in leukemic cells. Journal of Biological Chemistry Clemmons DR, Cascieri MA, Camacho-Hübner C, McCusker RH & Bayne ML. Discrete alterations of the insulin-like growth factor-i molecule which alter its affinity for insulin-like growth factor binding proteins result in changes in bioactivity. Journal of Biological Chemistry Received 16 June 1997 Accepted 31 October 1997