Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
[CANCER RESEARCH 41, 3979-3984, 0008-5472/81 /0041-OOOOS02.00 October 1981] Glucocorticoid Receptor Determinations in Leukemia Patients Using Cytosol and Whole-Cell Assays1 Stefano lacobelli,2 Vittoria Natoli, Paola Longo, Franco O. Ranelletti, Giulio De Rossi, Daniela Pasqualetti, Franco Mandelli, and Renato Mastrangelo Laboratory of Molecular Endocrinology [S. I., V. N.. P. L.]. Departments of Histology Department of Hematology [G. D., D. P., F. M.j, University of Rome, 00100 Rome. Italy [F. O. P.] and Pediatrics [R. M.], Sacro Cuore Catholic University; and ABSTRACT al. (13, 14) demonstrated that cells of 22 ALL patients with favorable prognoses contained high GR levels while the cells Parallel determinations of glucocorticoid receptors in the of 6 other patients in relapse who were refractory to treatment cells of patients with various forms of leukemia were made by contained barely detectable GR levels. In a subsequent study, two assay methods, one using cell-free cytosolic extracts and Lippman et al. (15) reported that the cells of AML patients had the other using whole-cell preparations. Both assays revealed appreciable GR levels in only 3 of 16 cases. In independent saturable binding of triamcinolone acetonide in all cases. The studies, Gailani ef al. (4) found that the cells of 4 of 4 lymphomean equilibrium dissociation constant for the interaction of triamcinolone acetonide with the cytoplasmic receptor at 2° sarcoma patients, 3 of 3 ALL patients, 2 of 6 AML patients, 0 of 8 CLL patients, and 0 of 2 CML patients contained appre was 9.45 ±6.33 (S.D.) nM while that for the whole-cell binding ciable GR levels. Furthermore, Terenius ef al. (28) reported at 37°was 6.13 ±3.25 nM, suggesting an increase in receptor that the lymphocytes of only 17 of 27 patients with CLL contained significant levels of GR's, while Homo et al. (8) were affinity at physiological temperatures. Competition experiments with various unlabeled steroids revealed a higher degree of able to demonstrate the presence of specific receptors in all of glucocorticoid specificity at 37°in whole-cell suspensions than the patients with CLL and ALL as well as in normal lymphocytes. at 2°in cytosol. In a comparative analysis of 41 leukemic cell Finally, more recent reports have shown the presence of GR's specimens, it was found that determinations carried out by whole-cell assay, calculated as number of sites per cell, cor related well with those performed by cytosol assay, calculated as fmol/mg protein, independent of the type of leukemia. However, for cells with low receptor content, the two assay methods were more difficult to compare. In agreement with previous reports, the cytosol assay consistently underesti mated the number of receptors with respect to the whole-cell assay, particularly in cases of lymphatic leukemia. Further more, the underestimation decreased for increasing levels of total cellular receptor. These results suggest that, in addition to possible defects in the cytoplasm-to-nucleus translocation process, the acceptor-binding capacity of the nucleus may also represent one of the factors which determines the levels of assayable cytoplasmic receptors. Moreover, they indicate that the two assay methods furnish nonequivalent information. INTRODUCTION The possibility of using steroid receptor analysis to predict the response of a potentially hormone-dependent tumor to endocrine therapy has attracted considerable interest, espe cially in the area of breast cancer (20). On the basis of en couraging results obtained with this disease, many laboratories have attempted analogous studies on GR's3 in human leukemia cells in the hope of finding a correlation with the clinical response to glucocorticoids. So far, however, these studies have produced conflicting results. In early reports, Lippman ef ' Supported by Grant 800156696 of the Finalized Project "Control of Tumor Growth." 2 To whom requests for reprints should be addressed. 3 The abbreviations used are: GR, glucocorticoid receptor; ALL, acute lymphoblastic leukemia; AML. acute myelogenous leukemia; CLL, chronic lympho cyte leukemia; CML, chronic myelogenous leukemia; ANLL, acute nonlymphocytic leukemia; TA, triamcinolone acetonide. Received February 23, 1981 : accepted June 15. 1981. OCTOBER 1981 in virtually all patients with various forms of hematological cancer (1, 2, 11, 19). Some of these discrepancies could be related to the different methods used to measure the receptors. Many of the studies cited above utilized either a whole-cell or a cytosol assay indifferently. Although cytosol preparations have been used to quantify a variety of steroid receptors, including those for glucocorticoids, several reports claim that this method is un reliable for the measurement of GR content in leukemia (1, 2, 7) since both normal and neoplastic lymphocytes contain a very thin rim of cytoplasm and one must be aware of the efficiency of cell breakage during homogenization. Besides, unbound GR's may be preferentially localized in the nuclei of lymphocytes (23); therefore, cytoplasmic assay may result in underestimation of cellular receptor content. By using a wholecell procedure, some of these problems can be overcome and appropriate measurement of total (cytoplasmic plus nuclear) receptors can be obtained. However, a major technical problem associated with this assay technique is the difficulty of distin guishing the proportion of nonspecifically bound steroid, often quite elevated, from free steroid after cell washes. It appears obvious that a thorough comparison of GR measurement using both the whole-cell assay and the cytosol assay could be useful. This paper presents the results of such a comparison performed on the cells from a group of patients with various types of leukemia. MATERIALS AND METHODS Patients The study was done on 41 patients with various forms of leukemia (14 patients with CLL, 15 patients with ALL, and 12 patients with ANLL). Diagnoses were established by means of conventional morpho logical and cytochemical criteria. None of the ALL and ANLL patients 3979 Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research. S. lacobelli et al. had received antileukemic therapy of any kind at the time of study. None of the CLL patients under treatment had received any glucocor- dissociation ticoid therapy within a week of their inclusion in the study. calculated as the difference between the values obtained with and without the excess of radioinert steroid. Binding capacity was ex pressed as number of sites per cell (whole-cell assay) or fmol/mg Isolation of Leukocytes constant (K :) and the concentration of [3H]TA binding with the aid of an Olivetti P-6060 desk-top computer. Specific binding was Heparinized peripheral blood was diluted 2 to 3 times with 0.9% NaCI solution. Oextran T 500 (Pharmacia, Uppsala, Sweden) was added to a final concentration of 1% (w/v). After gravity sedimentation for 30 min at room temperature, the supernatant containing leukocytes was centrifugea at 400 x g for 6 min. Hypotonie lysis of contaminating erythrocytes was achieved by treating the cells with 0.84% ammonium chloride for 15 min. After 2 washes with phosphate-buffered saline protein (cytosol assay). Repeated determinations of numbers of sites per cell on the same fresh cell suspensions gave an average range of ±7% around the mean value. Repeated determinations on the same group of frozen cells (cytosol assay) showed a similar variability. (0.136 M NaCI: 2.6 mw KCI: 6.4 mw Na2HPO4: 1.4 mM KH2PO4), the cell pellets were resuspended in Roswell Park Memorial Institute Tissue Culture Medium 1640 (Grand Island Biological, Co., Grand Island, N. Y.) supplemented with 25 mM 4-(2-hydroxyethyl)-1-piperazinethane- GR Assays. Since the amountof cells recovered from most of the patients was not sufficient to furnish high enough cytosol protein concentrations to make a complete binding curve, we decided to use the DE 81 ion-exchange filter assay which accurately measures GR levels at very low protein concentra tions (25). Using this procedure, appreciable levels of receptor were measured in all the patients studied. Furthermore, in some patients with a high peripheral white blood count, recep tor determinations were made by using dextran-coated char coal and filter assays simultaneously. The 2 procedures gave comparable results (data not shown). Chart 1 illustrates the time course of [ 'H]TA binding to sulfonic acid buffer (Sigma Chemical Co., St. Louis, Mo.), penicillin (100 units/ml), and streptomycin (100 fig/ml). Viability was always greater than 90%, as determined by the trypan blue exclusion proce dure (24). A portion of the cell suspension was used immediately for GR determination by whole-cell assay while the remaining part was centrifuged again and the cell pellet stored frozen at -85° for subse quent analysis of GR content by cytosol assay. Under these storage conditions, no loss of receptor activity was noted for up to 3 months. Binding Studies Whole-Cell Assay. Duplicate 0.1-ml aliquots of the leukocyte sus pensions (10 to 20 x 10'v ml) were pipeted into 10 x 85 plastic tubes which already contained 0.15 ml of Roswell Park Memorial Institute Tissue Culture Medium 1640 with [3H]TA (25 Ci/mmol; the Radiochemical Centre, Amersham, England) at 10 different concentrations varying from 0.8 to 25 RM, with or without a 100-fold excess of radioinert TA. Tubes were incubated under 5% CO.. in air at 37° for 60 min except for experiments examining the time course of binding. After incubation, 1.5 ml of phosphate-buffered saline were added to each tube; the tubes were kept in ice for 5 min and then centrifuged at 200 x g for 6 min. The washing process was repeated twice. The extent of nonspe cific bound counts was reduced to a minimum, and less than 7% of specifically bound counts were lost in this washing procedure. Cell pellets were suspended in 1 ml of 80% ethanol, and the suspensions were transferred to counting vials. In order to determine the exact number of cells recovered at the end of the washing procedure, 4 additional tubes containing the same amounts of cells were prepared and processed in the same manner as the tubes used for binding determinations. Cytosol Assay. Cell pellets were rapidly thawed and resuspended in approximately 4 volumes (v/v) of 10 mM Tris:1.5 mw EDTA:12 mM monothioglycerol:20% glycerol, pH 7.4 at 2°. After homogenization with 20 strokes in an all-glass homogenizer, the cytosol fraction was obtained by centrifugation at 145,000 x g for 45 min at 2°.Duplicate 0.06-ml aliquots were transferred into 10 x 75 glass tubes which already contained 0.06 ml of 10 mM Tris:1.5 mM EDTA:12 mM monothioglycerol:20% glycerol, pH 7.4 at 2°, with [3H]TA at 7 different RESULTS receptor in whole cells as well as in cytosol preparations. In the cells, the highest level of specific binding is achieved between 20 and 30 min of incubation at 37°and remains stable for at least 60 min thereafter. At 2°,specific steroid binding in cytosol reaches a maximum after 120 min of incu bation and remains constant between 120 and 240 min. The affinity of various steroids for GR sites in the cells of some patients was evaluated by both assay methods. A rep resentative experiment carried out in one case of CLL is illus trated in Table 1 where the results are expressed as relative binding affinity. In both assays, the most efficient competitor was TA itself (relative binding affinity defined as 100) followed by dexamethasone, prednisolone, and cortisol. Corticosterone and aldosterone displayed an intermediate level of competition which was not dissimilar in both intact cells and cytosol prep arations. 17/¿-Estradici,testosterone, and biologically inactive glucocorticoids such as 21-deoxycortisone and prednisone did not compete. However, there are definite differences in the 100 o 5 m concentrations varying from 1 to 20 nM, with or without a 100-fold excess of radioinert TA. Tubes were incubated at 2°for 3 hr except for experiments examining the time course of binding. After incubation, duplicate 0.05-ml aliquots were deposited onto DEAE-cellulose filter paper discs, DE 81 (2.5-cm diameter; Whatman Co., W. & R. Baiston, Ltd., Maidstone, England); filters were washed as described (25) and transferred to scintillation vials. The protein content of cytosols was determined by the method of Lowry ef al. (17) using bovine serum albumin as standard. Calculations Radioactivity was measured with a 38% mean efficiency. Binding data were analyzed by Scatchard plot (26) to determine the equilibrium 3980 9« 60 120 MINUTES 180 240 Chart 1. Time course of specific |'H |TA binding to GR in intact cells ( ) or in cytosol (•) in one case of ALL. Aliquots of cell suspensions (0.1 ml; 18 x 10* cells/ml) or cytosol (0.06 ml) were incubated at 37°and 2°,respectively, for the times indicated with 20 HM[ 'H]TA alone (total binding) or in the presence of 2 /IM radioinert TA (nonspecific binding). Specific binding was calculated as the difference between total binding and nonspecific binding. The maximum binding attainable during the time of the kinetic experiment was taken as 100% and the other values were expressed as a percentage thereof. The maximum binding for whole cells and cytosol was 4200 sites/cell and 318 fmol/mg protein, respec tively. CANCtR RESEARCH Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research. VOL. 41 Gf? 's in Leukemia whole-cell assay are given in Chart 3. The values of receptors are: mean, 570 ±540 fmol/mg protein; median, 325 fmol/mg Standard competition curves were obtained by incubating replicate aliquota of protein (N = 14) in CLL; mean, 1492 ±1521 fmol/mg protein; cytosol or cell suspensions with 20 nM |H|TA in the presence of varying median, 706 fmol/mg protein (N = 15) in ALL; mean, 1239 concentrations (1 nM to 1 /IM) of different radioinert competitors. Incubation was done at 2°for 3 hr (for cytosol) or 37°for 60 min (for cells). Bound radioactivity ±1156 fmol/mg protein; median, 720 fmol/mg protein (N = was measured as described in "Materials and Methods." The ratio of the 12) in ANLL. The Kd values are: mean, 12.8 ±8.2 nM; median, concentration of ligand to the concentration of competitor required to displace 9.18 nM (N = 14) in CLL; mean, 8.34 ±4.65 nM; median, 7.2 the radioligand by 50% (relative binding affinity) was determined. The relative binding affinity of TA was taken to be 100. OM (A/ = 15) in ALL; mean, 9.1 ±7.44 nM; median, 7.3 nM (N Relative binding affinity = 12) in ANLL. The values of receptor sites are significantly different among the various types of leukemia (P < 0.05, CompetitorTADexamethasonePrednisoloneCortisolCorticosteroneAldosteroneProgesteroneCortexolonePrednisone21 Kruskal-Wallis test). Analogous to previous findings (18), the Kd value measured by cytosol assay in all types of leukemia (9.45 ±6.33 nM) is significantly higher (p < 0.01, Student's Table 1 for [3HJTA binding in one case of CLL Competition f test) compared to that found in intact cells (6.13 ±3.25 nM). Comparison of GR Measurement: Whole-Cell versus Cy tosol Assay. In Chart 4, the numbers of GR sites per cell measured by whole-cell assay are plotted against the amounts -Deoxycortisone1 of fmol per mg protein determined by cytosol assay. There 7/3-EstradiolTestosteroneCytosol100"92858028235349<1<1<1<1Cells1008886772229141B5<1<1<1 appears to be a good linear correlation between the 2 series of measurements (r = 0.91 ) independent of the type of leukemia Mean of 2 determinations. (CLL, r = 0.78; ALL, r = 0.92; ANLL, r = 0.90). However, for low binding site concentrations (less than 10,000 sites/cell or 1000 fmol/mg protein), the points are less aligned along the 50.00040000y ia° regression line. In some of the previous reports on the levels of GR sites in o0 leukemic cells, the binding studies were done on cytosol prep 30.000UÕS arations at 0-4° (4, 13). It has been shown that this procedure ;o *8 ÃŒ *^^ 20.00010,000i o Õ— *^ST »r- ~*~ :H *» •¿ CLL 0 -K S i! J 8 cfla2010 ALL ANLL Chart 2. QR levels O and steroid-binding affinities (•)in leukocytes isolated from the blood of 41 patients with various types of leukemia. Receptor determi nations were made using whole-cell assay as described in "Materials and Methods." specificity of the binding sites in the cytosol as compared to those in whole cells. In particular, cortexolone and progester one have a lower affinity for receptor sites in whole cells at 37° than in cytosol at 2° which is in agreement with previous observations made on rat splenic lymphocytes (18). Chart 2 shows the values of GR sites determined by wholecell assay in the cells of 41 patients with various forms of leukemia. The number of sites varies widely from one type of leukemia to another and from one patient to another within each type as well. The values of receptors are: mean, 8061 ±4491 sites/cell; median, 6501 sites/cell (N = 14) in CLL; mean, 17,106 ± 13,737 (S.D.) sites/cell; median, 15,070 sites/cell (N = 15) in ALL; mean, 14,954 ±10,083 sites/cell; median, 11,287 sites/cell (N = 12) in ANLL. The Kd values obtained by the slopes of Scatchard plots are: mean, 3.97 ± 2.28 nM; median, 3.7 nw (N = 14) in CLL; mean, 6.57 ±2.98 nM; median, 2.7 OM (N = 15) in ALL; mean, 8.1 ±3.25 nM; median, 7.2 nw (N = 12) in ANLL. The K«values appear similar to those determined for the receptor-dexamethasone interac tion in the leukemic cells of previous studies (1, 6). However, they are significantly different among the various types of leukemia (p < 0.01, Kruskal-Wallis test). The results of GR determinations by cytosol assay using cytosol obtained from the same leukocyte specimens used for OCTOBER 1981 leads to an underestimation of the total number of binding sites present in the cell. To verify this, the number of GR sites per cell determined by whole-cell assay has been compared with the number of sites per cell obtained by cytosol assay. This latter parameter was calculated by dividing the fmol bound/mg of cytosol protein by the number of cells per mg protein [based on cell count and protein determination by the method of Lowry ef a/. (17)] and multiplying by Avogadro's number. The results of this comparison indicate that the relationship between these 2 estimates is not linear and may be well represented by a parabolic equation (Chart 5). Specifically, in cells with low GR content (less than 10,000 sites/cell), the amount of receptor measured by the cytosol assay does not vary greatly as a function of the variations in the number of sites per cell, whereas it increases rapidly in cells with higher receptor levels. Furthermore, it is interesting to note that, even though all 3 types of leukemia follow the same behavior, ANLL is predomi nantly distributed above the theoretical equation curve with respect to the others. This behavior pattern is further illustrated 5.0004.0003.0002,0001.000• e•• ¡8^115 0o :I: ars i g>3jl ^frT'- s _^_o "T- r"* *a" «,8 CLL ° ^.. t* ALL r30 ANLL Chart 3. GR levels O and steroid-binding affinities (•)in leukocytes isolated from the blood of 41 patients with various types of leukemia. Receptor determi nations were made using cytosol assay as described in "Materials and Methods." 3981 Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research. S. lacobelli et al. Table 2 5.000 Underestimation of GR content (cytosol assay versus whole-cell assay) in leukocytes from patients with various types of leukemia Sites/cellLeukemia 4.000 typeCLL cell8,061 ± 4,491" (%)78.7 ±18.4° ±1,201 73.1 ±14.4o ALL (15) 17,106 ±13,737 5,434 ±6,81 7 57.8 ±22.3e ANLL (1 2)Whole 14,964 ±10,083Cytosol1,551 6.536 ±6,822Underestimation (14)" 1000 Numbers in parentheses, number of cases. 6 Mean ±S.D. c' ' * Mean difference (Student's f test): c versus e, p < 0.05; d versus e, p 2.000 < 0.02; c versus d, not significant. Table 3 1.000. Stability of GR in the cytosol of leukocytes from patients with various types of leukemia 10.000 30.000 50,000 SITES/CELL Chart 4. Correlation between QR levels measured by whole-cell assay and cytosol assay in leukocytes from patients with CLL (•). ALL (A), and ANLL, (O). Receptor determinations were carried out as described in "Materials and Meth ods." Cytosols were prepared and maintainedat 2°,and at hourly intervals thereafter up to 6 hr, aliquots were incubated with 20 HM| H|TA alone (total binding) or in the presence of 2 «M radioinert TA (nonspecific binding) for 3 hr at 2°.The loss of binding capacity referred to specific binding which was calculated as the difference between total binding and nonspecific binding. One-half of the maxi mum binding capacity was calculated as 50% of the initial (zero time) amount of specific binding. Loss of one-half of the maximum binding capacity Leukemia typeCLL ALL ANLLRapid hr)7A«3 (87.5)" (>3 hr)'A % (67) %(50)Slow % % Numbers in parentheses, percentage. were arbitrarily established as less or more than 3 hr as shown in Table 3. A rapid loss (less than 3 hr) was observed in 87% (7 of 8) OfCLL, 67% (4 of 6) of ALL, and 50% (3 of 6) of ANLL. DISCUSSION 10000 SITES/CELL 30,000 [whole cell assay] 50,000 Chart 5. Correlation between the number of QR sites measured by whole-cell assay and cytosol assay in leukocytes from patients with CLL (•), ALL (A), and ANLL (O). The number of sites per cell relative to the cytosol assay were calculated from the binding data as described in "Results." The best fit for the regression curve was represented by a parabolic equation (y - a + ox + ex2); r, the correlation coefficient for this curve. The results of this study demonstrate that the 2 GR assay methods furnish correlated values, independent of the type of leukemia studied. The correlation is less evident for cells with low receptor content in which the 2 series of measurements are more difficult to compare. Moreover, our data indicate that the cytosol assay consistently leads to an underestimation of the number of binding sites as compared to the values obtained by whole-cell assay. There may be several factors accounting for this. First, the whole-cell assay procedure might ensure the de termination of the total (cytoplasmic and nuclear) receptor content. It has long been reported that in rat thymocytes exposed to glucocorticoids at 37°the majority (approximately 80%) of the receptor is found to be associated with the nucleus in the form of receptor-hormone complex (22). Similar obser vations have recently been made on human thymocytes (10). Interestingly enough, a similar percentage of receptor under estimation was seen in the cytosol of leukemic cells of lym phatic origin (Table 2). Thus, a simple explanation which could account for the low number of binding sites measured by the Loss of Binding Capacity. For these experiments, the cy cytosol assay is that in intact cells at physiological tempera tosol was prepared and maintained at 2°,and at hourly intervals tures the majority of receptor is localized in the nucleus. On aliquots were incubated with saturating concentrations (20 HM) the other hand, the higher number of binding sites detected in of [3H]TA in the absence or presence of radioinert TA to the cytoplasm of ANLL could reflect an altered distribution of examine the binding during the subsequent 3-hr incubation receptor between the cytoplasmic and nuclear compartments period. The rate of loss of one-half of the binding capacity in possibly due to a variable rate of receptor translocation as 20 different cytosols fell into 2 basic time categories which reported elsewhere (1). by the data in Table 2 which compares the means of values for receptor sites obtained by the 2 assay methods. The percent age of underestimation (cytosol assay/whole-cell assay) is significantly lower in ANLL than in CLL and ALL (p < 0.05 and 0.02, respectively). 3982 CANCER RESEARCH Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research. VOL. 41 GR 's in Leukemia Second, possible breakage of lysosomes and release of proteolytic enzymes that partially inactivate the cytoplasmic receptor must also be considered. Our data show that the receptor-binding capacity in cytosol from certain types of leukemic cells, notably those of CLL, decreases rather rapidly, suggesting that substances capable of inactivating the receptor molecule (which could account for the generally lower binding capacity of these particular cells) are active in these cells. Although these substances (possibly enzymes) have not yet been identified, it is interesting that Sloman and Bell (27) have shown that the cytosol of CML contains receptor-inactivating factor(s) which could be activated during the homogenization process. The presence of similar factors in leukemic myeloblasts of AML has also been proposed by other authors (1). Third, it is possible that some cells are particularly resistant to the breaking process during homogenization. Although this possibility has been postulated to explain the negative results obtained by the cytosol assay (1 ), it seems to be unlikely since, at least in our hands, complete cell breakage was observed in all cases. The concentrations of GR sites in the different types of leukemic cells used in our study are analogous to those ob served elsewhere (1, 10). The receptor content of these leu kemic cells greatly exceeds that found in normal leukocytes of both lymphatic and myelomonocytic origin (12, 23). This ca pacity of the leukemic cell to express a greater quantity of receptor protein with respect to its normal counterpart seems to be a more general characteristic of the neoplastic phenotype. Thus, for example, the normal mammary cell does not seem to possess appreciable amounts of estrogen receptor while as a result of the neoplastic transformation process the quantity of receptor molecules can reach abnormally high levels (3). Although the mechanisms which regulate receptor levels in the cell are only now beginning to be clarified, it is possible to hypothesize that the redundance of GR's in leu kemic cells, especially in ALL and ANLL as indicated in this study, represents a class of heterogeneous receptor molecules with functional defects such as wide ranges of steroid-binding affinity (this paper and Ref. 1) as well as variable capacities of cytoplasm-to-nucleus translocation and of interaction with ac ceptor sites in the chromatin (5). In the presence of this redundance, the nuclear acceptor-binding capacity could be come the limiting step in the process, independent of any translocation defect. Therefore, in cells with low receptor con tent (i.e., closer to normal cell values), the quantity of binding sites measured in the cytoplasm could represent a minimal part of the total cellular receptor sites which in the presence of endogenous hormone would be preferentially localized in the nucleus. This could explain the high frequency of negative results in GR determinations made by cytosol assay in leukemic cells with low receptor levels, particularly in CLL (4, 28). However, when the cellular receptor levels exceed the nuclear acceptor capacity, as in the case of cells with extremely high receptor content, high proportions of binding sites are found in the cytoplasmic compartment. This hypothesis could explain the nonlinear correlation between the quantity of receptor sites measured by the 2 assay methods. It has been suggested that steroid receptors are not re stricted to the cytoplasm alone and that in some cases they can be localized exclusively in the nucleus (9). As a conse quence, the measurement of cytoplasmic binding sites alone OCTOBER 1981 may lead to falsely negative results if nuclear receptors are not considered as well. On the other hand, demonstration of total cellular receptor sites in whole-cell preparations does not give information about their localization, i.e., if in the cytoplasm and/or in the nucleus. Moreover, binding to receptor in intact cells may depend on such factors as the metabolic state of the cells (21) and the rate of steroid transport through the cell membrane. While the conclusions drawn from this study may be used to partly explain the contrasting data on GR determinations in leukemic cells so far reported, at least in cases of low receptor content, they indicate that receptor assessments by whole-cell and cytosol assays provide nonequivalent but possibly comple mentary information. Interestingly, each of the only 2 studies correlating GR levels in leukemic cells with clinical aspects furnished entirely different prognostic indications and each was based on one of the 2 assay procedures herein discussed (13, 16). Therefore, it seems reasonable not to depend exclusively on only one of these procedures, at least until their precise prognostic significance is more fully clarified by further correl ative clinical studies. REFERENCES 1. Crabtree, G. R . Smith, K. A., and Munck. A. Glucocorticoid receptors and sensitivity of isolated human leukemia and lymphoma cells. Cancer Res.. 38. 4268-4272. 1978. 2. Duval, D., and Homo, F. Prognostic value of steroid receptor determination in leukemia. Cancer Res.. 38 4263-4267. 1978. 3. Feherty, P., Ferrer-Brown, G.. and Kellie, A. E. Oestriol receptors in carci noma and benign disease of the breast: an in vitro assay. Br. J. Cancer, 25. 697-710, 1972. 4. Gailani, S . Minowada. J . Silvermail, P.. Nussbaum. A., Kaiser, N . Rosen, F., and Shimaoka, K. Specific glucocorticoid binding in human hemopoietic cell lines and neoplastic tissue. Cancer Res., 33: 2653-2657. 1975. 5. Gehring. U. Genetic and biochemical studies on glucocorticoid receptors. In: S. lacobelli, R. J. B. King, H. R. Lindner, and M. E. Lippman (eds.). Hormones and Cancer, p. 79. New York: Raven Press, 1980. 6. Homo, F.. Duval. 0., Harousseau. J. L, Marie. J. P.. and Zittoun, R. Heterogeneity of the in vitro responses to glucocorticoids in acute leukemia. Cancer Res., 40: 2601-2608, 1980. 7. Homo, F., Duval, D.. Harousseau. J. L, and Thierry. C. Glucocorticoid receptors in normal and neoplastic human lymphoid cells, lit S. lacobelli, R. J. B. King, H. R. Lindner, and M. E. Lippman (eds.). Hormones and Cancer, p. 387. New York: Raven Press, 1980. 8. Homo, F., Duval, D., and Meyer, P. Etude de la liaison de la dexamethasone tntiiu- dans les lymphocytes de sujets normaux et leucémiques. C. R. Hedb. Seances Acad. Sci., 280. 1923-1926, 1975. 9. Horwitz, K. B., Zava. D. T., Thilagar, A. K., Jensen, E. M., and McGuire, W. Steroid receptor analyses of nine human breast cancer cell lines. Cancer Res., 38. 2434-2437. 1968. 10. lacobelli, S.. Longo, P.. Mastrangelo, R.. Malandrino. R., and Ranelletti. F. O. Glucocorticoid receptors and steroid sensitivity of acute lymphoblastic leukemia and thymoma. in S. lacobelli, R. J. B. King. H. R. Lindner, and M. E. Lippman (eds.). Hormones and Cancer, p. 371. New York: Raven Press, 1980. 11. Kontula, K.. Andersson, L. C., Paavonen, T.. Myllyla, G.. Teerenhovi, L., and Vuopio. P. Glucocorticoid receptors and glucocorticoid sensitivity of human leukemic cells. Int. J. Cancer, 26. 177-183, 1980. 12. Lippman, M. E.. and Barr. R. Glucocorticoid receptors in purified subpopu lations of human peripheral blood lymphocytes. J. Immunol., 118: 19771981. 1977. 13. Lippman, M. E., Halterman. R., Leventhal, B. G. Perry, S., and Thompson. E.B. Glucocorticoid binding proteins in human acute lymphoblastic leukemic blast cells. J. Clin. Invest., 52. 1715-1725, 1973. 14. Lippman, M. E., Halterman, R.. Perry, S., Leventhal. B., and Thompson, E. B. Glucocorticoid binding proteins in human leukaemic lymphoblasts. Nat. New Biol.. 242. 157-158, 1973. 15. Lippman, M. E., Perry, S., and Thompson, E. B. Glucocorticoid binding proteins in myeloblasts of acute myelogenous leukemia. Am. J. Med.. 59. 224-227, 1975. 16. Lippman, M. E.. Yarbro. G. K.. and Leventhal. B. G. Clinical implications of glucocorticoid receptors in human leukemia. Cancer Res., 38. 4251-4256, 1978. 3983 Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research. S. lacobelli et al. 17. Lowry, O. H., Rosebrough, N. J., Farr, A. L . and Randall, R. J. Protein measurement with the Polin phenol reagent. J. Biol. Chem., 793: 265-275, 1951. 18. MacDonald, R , and Cidlowski, J. A. Alterations in specificity of the glucocorticoid receptor with temperature in rat splenic lymphocytes. J. Steroid Biochem., to. 21-29, 1979. 19. Mastrangelo, R., Malandrino, R., Riccardi, R., Longo, P., Rane-netti.F. O., and lacobelli, S. Clinical implications of glucocorticoid receptor studies in childhood acute lymphoblastic leukemia. Blood, 56: 1036-1040, 1980. 20. McGuire, W. L. (ed.). Hormones, Receptors, and Breast Cancer. New York: Raven Press, 1978. 21. Munck, A., and Brinh-Johnsens. T. Specific and nonspecific physicochemical interactions of glucocorticoids and related steroids with rat thymus cells. J. Biol. Chem., 243: 5556-5565, 1968. 22. Munck, A., and Wira. C. Glucocorticoids receptors in rat thymus cells, in G. L. Raspé,(ed.), Advances in the Biosciences, p. 301. Elmsford, N. Y.: 3984 Pergamon Press, Inc., 1971. 23. Neifeld, J. P., Lippman, M E., and Tormey, D. C. Steroid hormone receptors in normal human lymphocytes. J. Biol. Chem., 252.-2972-2977, 1977. 24. Raff, M. C. Surface antigenic markers for distinguishing T and B lymphocytes in mice. Transplant. Rev., 6: 52-80,1971. 25. Santi. D. V., Sibley. C. H., Perriard. E. R.. Tomkins, G. M., and Baxter, J. D. A filter assay for steroid hormone receptors. Biochemistry, 72:2412-2416, 1973. 26. Scatchard, G. The attraction of proteins for small molecules and ions. Ann. N. Y. Acad. Sci.. 57: 660-672, 1949. 27. Sloman, J. C., and Bell, P. Glucocorticoids and myeloid leukemia, in. Glucocorticoid Action and Leukemia, p. 161. Cardiff, Wales: Alpha Omega Alpha Publishing Ltd., 1979. 28. Terenius, L., Simonssons, B., and Nilssons, K. Glucocorticoid receptors, DNA synthesis, membrane antigens and their relation to disease activity In chronic lymphatic leukemia. J. Steroid Biochem., 7:905-909, 1976. CANCER RESEARCH Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research. VOL. 41 Glucocorticoid Receptor Determinations in Leukemia Patients Using Cytosol and Whole-Cell Assays Stefano Iacobelli, Vittoria Natoli, Paola Longo, et al. Cancer Res 1981;41:3979-3984. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/41/10/3979 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on August 3, 2017. © 1981 American Association for Cancer Research.