J Cancer Res Clin Oncol (1992) 119:121-126
Cancer Research Clinical 9 9 Springer-Verlag 1992
Adriamycin binding assay: a valuable chemosensitivity test in human osteosarcoma Nicola Baldini 1, Katia Scoflandi 1, Massimo Serra 1, Katsuyuki Kusuzaki 2, Toshiharu Shikita 1,2, Maria C. Manara 1, Daniela Maurici 1, and Mario Campanacci 1 1 Laboratorio di Ricerca Oncologica, Istituti Ortopedici Rizzoli, Bologna, Italy 2 Department of Orthopaedic Surgery, Kyoto Prefectural University of Medicine, Kyoto, Japan Received 14 February 1992/Accepted 23 June 1992
Summary. The reliability of a simple method evaluating the pattern of subcellular binding of Adriamycin (Adriamycin binding assay, A B A ) as an index of sensitivity was demonstrated in different primary cultures and in sensitive and resistant cell lines of human osteosarcoma. After exposure to Adriamycin (10 gg/ml for 30 min at 37 ~ C), living sensitive cells showed selective intranuclear uptake of the drug, whereas in resistant cells no distinct subcellular distribution was observed. The binding pattern of Adriamycin in sensitive and in highly resistant cells was inversely related to the expression of P-glycoprotein. However, low levels of resistance in vitro, not detectable b y increased levels of expression of P-glycoprotein, were revealed by A B A . The use of A B A in combination with the estimate of P-glycoprotein expression is recommended in clinical practice as an accurate means for predicting the sensitivity of osteosarcoma to Adriamycin.
Different patterns of fluorescence have been observed in living sensitive and resistant cells (Willingham et al. 1986) and responsiveness to the drug has been associated with elective binding to the nucleus, the more obvious target of A d r i a m y cin cytotoxic activity (Gigli et al. 1989). A simple method evaluating the pattern of subcellular binding of Adriamycin (Adriamycin binding assay, A B A ) has been recently suggested as a new chemosensitivity test (Kusuzaki et/al. 1989). In this paper we demonstrate the reliability of this method to determine the sensitivity to Adriamycin in different primary cultures and in both sensitive and resistant cell lines of human osteosarcoma, with a view to further application on clinical material.
Materials and methods Cells. Human osteosarcoma cell lines U-20S and Saos-2 were gifts, re-
Key words: A d r i a m y c i n - Chemosensitivity - Osteosarcoma
Introduction Adriamycin is very effective in high-grade osteosarcoma and should be included in any multimodal therapeutic regimen for this tumor (Winkler et al. 1988; Bacci et al. 1990). The heterogeneous response o f osteosarcoma to induction chemotherapy and the occurrence of relapses in 3 0 % - 4 0 % of the patients despite aggressive treatment suggest the need for individualized therapies based on the preliminary assessment of sensitivity. Adriamycin is a natural fluorochrome (Silvestrini et al. 1970), and intracellular drug accumulation m a y be monitored by direct microscopic observation (Egorin et al. 1974). This study was supported by a grant from the Associazione Italiana per la Ricerca sul Cancro (A.I.R.C.). Correspondence to: Nicola Baldini, Laboratorio di Ricerca Oncologica, Istituti Rizzoli, via di Barbiano 1/10, 40136 Bologna, Italy.
spectively, from Dr. GraziellaPratesi and Dr. Maria I. Colnaghi, Istituto Nazionale Tumori, Milano, Italy. Adriamycin-resistantvariants of these sensitive lines were selected by continuous in vitro exposure to the drug, as previously described (Serra et al. 1991). Stepwise increases in Adriamycin concentration produced sublines that were resistant up to 580ng/ml (1 pM) Adriamycin (respectively, U - 2 0 S / D X 3~ Saos2/DX 3~ U-20S/DX ~~176 Saos-2/DX 1~176 U-20S/DX 58~and Saos-2/DX58~ Primary cell cultures (SARG, IOR/MOS, IOPUOSF, IOR/OS7, IOR/OS9, IOR/OS10, IOR/OS14) were obtained from surgical specimens of seven cases of high-grade human osteosarcoma. Of these, six were primary lesions and 1 (IOR/OS9) was a bone metastasis. After mechanical mincing, cells were seeded in Iscove's modified Dulbecco's medium supplemented with penicillin (100 U/ml) and streptomycin (100 gg/ml) (IMDM, Gibco, Paisley, Scotland) and with 10% heat-inactivated fetal calf serum (FCS, Biological Industries, Kibbutz Beth Haemek, Israel). Culture flasks were incubated at 37~ C in a humidified atmosphere of 5% CO2. All the cultures used for this study were before their tenth in vitro passage. Incubations with Adriamycin. Adriamycin was purchased from Farmita-
lia Carlo Erba, Milano, Italy. To identify the optimal conditions for drug incorporation, cell suspensions (200,000 cells/ml IMDM, 10% FCS) from U-20S and Saos-2 were incubated with different Adriamycin concentrations (1 ~tg/ml, 5 p.g/ml, 10 ~tg/ml, and 50 ~tg/ml) at 37~ C for 10, 30, 60, 90, 120, and 180 min. Cells were also exposed to 50 gg/ml Adriamycin at 4 ~ C. After Adriamycin incubation in the above conditions, cells were incubated with 1 mg/ml fluorescein diacetate for
122
Fig. 1. Fluorescein diacetate staining and intracellular distribution of Adriamycin in sensitive (a, b) and resistant (c, d) cells after exposure to 10 gg/ml Adriamycin at 37 ~ C
10 rain at 4 ~ C, washed and resuspended in phosphate-buffered saline prior to microscopic observation and cytofluorometry. Exposure t o 10 btg/ml Adriamycin for 30 min at 37 ~ C was defined as a standard for the assay, in agreement with a previous report (Kusuzaki et al. 1989). To define the relationship between differences in the pattern and the amount of drug incorporation in sensitive and resistant cells, cell suspensions (200000 cells/ml IMDM with 10% FCS) from Saos-2/DX58~were incubated with 10 gg/ml Adriamycin for 30, 60, 90, and 120 rain at 37 ~ C, and then with 1 mg/ml fluorescein diacetate for 10 rain at 4 ~ C prior to observation and cytofluorometry. To evaluate the reliability of the assay on different cultures, cell suspensions (200000 cells/ml IMDM with 10% FCS) from U - 2 0 S / D X 3~ U - 2 0 S / D X 1~176 U - 2 0 S / D X 58~ Saos-2/DX3~ Saos-2/DX I~176 SARG, IOR/MOS, IOR/OS7, IOR/OSF, IOR/OS9, IOR/OS10, and IOR/OS14 were incubated with 10 gg/ml Adriamycin for 30 rain at 37 ~ C and then with 1 mg/ml fluorescein diacetate for 10 min at 4 ~ C prior to microscopic observation.
Microscopic observation and cytofluorometry. Cells were observed with an epi-fluorescence microscope (Nikon FX-A) equipped with a HBO 100-W high-pressure mercury lamp operating on stabilized d.c. power supply, a 40x fluorite lens (0.85 NA), and a high-gain photomultiplier connected to a personal computer for signal recording and analysis. Living cells were first identified with fluorescein diacetate green fluorescence (Rotman and Papermaster 1966). On those cells, the red fluorescence of Adriamycin was examined and two different patterns observed: strong nuclear fluorescence with weak fluorescence in the cytoplasm (Fig. 1 a, b); and diffuse fluorescence without any distinctive intracellular binding (Fig. 1 c, d). As previously suggested, the first pattern was considered sensitive and the second resistant (Kusuzaki et al. 1989). The percentage of sensitive ceils was calculated on 300 cells for each sample.
In addition, cytofluorometric measurement of intracellular Adriamycin content was made on Saos-2, and Saos-2/DXsS~under the abovementioned conditions. Drug fluorescence intensity was determined on 100 living cells for each sample and expressed in arbitrary units after background value had been set at zero.
In vitro sensitivity to Adriamycin. Cells (13x103/cm 2) were seeded in IMDM with 10% FCS. After 24 h, the medium was changed with IMDM with 10% FCS plus 580 ng/ml, 58 ng/ml, and 5.8 ng/ml Adriamycin, or with 1MDM 10% with FCS alone (control). After 48 h, viability was screened with the trypan blue dye exclusion method. Cell growth inhibition was determined as the ratio between the numbers of treated and untreated cells.
Effects on cell cycle. The effects of Adriamycin on cell kinetics were evaluated by bromodeoxyuridine (BrdUrd) incorporation (Dolbeare et al. 1983). Cytospin preparations were obtained from cell suspensions incubated with 50 btM BrdUrd (Sigma, St. Louis, Mo.) at 37~ for 10 rain. After fixation with 70% ethanol for 10 min and treatment with 4 M HCI at room temperature for 30 min, the specimens were incubated with the anti-BrdUrd monoclonal antibody MBU (Eurodiagnostic, Apeldoorn, The Netherlands) at a 1 : 10 dilution at 37 ~ C for 30 rain. A fluorescein-isothiocyanate-conjugatedsecondary antibody was used for labelling. The BrdUrd labelling index was estimated on 300 cells for each slide. Changes of cell cycle were further investigated by determining the DNA content on cytospin preparations obtained before and after exposure to the drug. After fixation with 70% ethanol at 4 ~ C for 30 rain, cells were treated with 0.5 mg/ml ribonuclease A (Sigma, St. Louis, Mo.) at 37 ~ C for 1 h and stained with 25 btg/ml propidium iodide (Sigma, St. Louis, Mo.). Nuclear DNA content was determined by cytofluorometry (Ashihara 1985). The effects of drug exposure on the cell cycle were
123 oo,
U-20S
.~ ...................= ...................~..., .., ~=,,.g~.............
100
/,~.._.;;;:::,,
90
KI/
80 W t..) I.U E 09 z ILl 09
IJ. O
80
.........
tt ....
70
o
6o
t= '~
40
0
40
0
30
/
100 --i
1
a
n
n
i
60
90
120
150
180
TIME
(minutes)
.~............................... ..........::::::--~....
001 / U 7
il/
7o
/
/,
,'I/
'ot'k" / /
40 1 ! / i
/i
~o
lo
30
60 TIME
Saos-2
O3 U. O o~
580
50
. . . . ...-~ 0 .... ~ . ! 0 30
E
Saos-2
1
60
10
09 z ILl
-] Saos-2/DX
20
_J [iJ O [JJ
lo0
":..,
/
/
,-'*~ .........
_-
..-"
,.
/
/
o
0
30
60
90
120
TIME
(minutes)
150
180
Fig. 2. Percentage of cells with selective nuclear binding of Adriamycin in sensitive cell lines U - 2 0 S and Saos-2 after exposure to 50 gg/ml ( . . . i . . . ) , 10 gg/ml ( - - . - - ) , 5 gg/ml (... m.--) and 1 gg/ml ( . . . . . . . ) at 37 ~ C, andto 50 gg/ml ( - - = - - ) at4 ~ C
evaluated and expressed as a ratio between the percentage of untreated and treated cells in phases S/GJM.
P-glycoprotein expression. Cytospin preparations obtained from cul-
120
Fig. 3. Cytofluorometric analysis of intracellular Adriamycin content in Saos-2 ( - - . - - ) and Saos-2/DX58~ ( - - o - - ) after exposure to 10 gg/ml Adriamycin at 37 ~ C
served at different times to detect the pattern of intracellular uptake of the drug. As shown in Fig. 2, selective nuclear binding, typical of sensitive cells, was observed in over 90% of the cells after 30 min of incubation with both 50 gg/ml and 10 gg/ml Adriamycin, after 60 rain incubation with 5 gg/ml Adriamycin, and after 120 min with 1 pg/ml Adriamycin. No uptake was detected even after 120 min exposure to the highest dose of 50 pg/ml Adriamycin when incubation was at 4 ~ C. Cell viability after Adriamycin exposure was always over 90%. On the basis of our results and as previously suggested in mouse leukemia cells (Kusuzaki et al. 1989), optimum conditions to demonstrate drug incorporation in human osteosarcoma were fixed at 10 gg/ml for 30 rain at 37 ~ C. Cytofluorometric measuremend of intracellular Adriamycin did not show any significant difference between Saos-2 and Saos-2/DX 58~ even after 120min of exposure to 10 gg/ml Adriamycin at 37 ~ C (Fig. 3). However, under the same conditions, direct observation revealed a striking difference in the binding pattern between sensitive and resistant cells (Fig. 4). More than 90% of the Saos-2 cells were shown to be sensitive after 30 min, and less than 5% of Saos2/DX 58~even after 120 rain of exposure, confirming that the percentage of sensitive cells, as shown by the selective nuclear binding of Adriamycin, is a valuable index of the in vitro sensitivity to the drug. The reliability of ABA as a means to assess the sensitivity to Adriamycin was further investigated on seven primary osteosarcoma cell cultures obtained in our laboratory. All the
tures were fixed in acetone at room temperature for 10 rain. The expression of P-glycoprotein was evaluated by indirect immunofluorescence with the monoclonal antibody JSB-1 (Sanbyo, Uden, The Netherlands) at a 1 : 10 dilution (Scheper et al. 1988). The percentage of positive cells was determined on 300 cells for each slide.
..i
Results
O ua I--
lO0-
co
z I.U
The two human osteosarcoma cell lines U - 2 0 S and Saos-2 sensitive to Adriamycin, and their resistant variants U-2 OS/DX 58~ and Saos-2/DX 58~ (Serra et al. 1991) were a suitable in vitro model with which to verify the reliability of ABA, a new chemosensitivity test based on the differential binding of Adriamycin in sensitive and resistant cells. Cell suspensions from U - 2 0 S and Saos-2, incubated at 37 ~ C with different Adriamycin concentrations, were ob-
90
(minutes)
~.
9
Saos-2
806040
09
2oo~
0
=
~ 30 TIME
~t
~
60
90
580
~
Saos-2/DX
120
(minutes)
Fig. 4. Percentage of cells with selective nuclear binding of Adriamycin in Saos-2 ( - - . - - ) and Saos-2/DX58~ ( - - o - - ) after exposure to 10 gg/ml Adriamycin at 37 ~ C
124 Table 1. Cell growth inhibition after 48 h exposure to 580 ng/ml,
58 ng/ml, and 5.8 ng/ml Adriamycin (ADR) Cell lines
Table 3. Changes in the cell-cycle distribution before and after 48 h exposure to Adriamycin (ADR) (580 ng/ml); data were obtained from the mean of two experiments
Growth inhibition (%) after ADR at Cell lines 580 ng/ml
58 ng/ml
5.8 ng/ml
U-20S Saos-2 SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS10 IOR/OS14
94.0+- 5.9 67.3+_15.0 96.5+_ 2.3 65.0+12.1 68.4_+ 9.3 57:8+-15,1 84.4+- 2.5 95.3+_ 2.5 57.4+_ 1.1
79.4_+4.2 60.0+_8.3 53.8+_7.4 21.3+-9.2 ND ~ 64.7_+5.6 71.0+_3.1 69.5+_2.8 17.5+_3.7
51.7_+13.5 11.5+_ 5.7 17.8+_ 9.8 10.6+- 4.5 ND ~ 20.2_+ 8.2 50.5+_ 7.6 51.2+ 1.7 2.1+_ 1.1
U20S/DX 58~ Saos-2/DX5s~
15.7_+ 2.2 16.4+_ 3.6
2.0+_1.3 1.1+_0.5
0 0
S/G2/M(%)
S/G2/M (%) after ADR
Increase in
control U-20S Saos-2 SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS 10 IOR/OS 14
39 45 50 31 17 15 43 51 52
72 58 71 66 29 33 68 80 76
45.8 22.4 29.6 53 41.4 54.5 36.8 36.2 31.6
U-20S/DX58~ Saos-2/DX5s~
70 85
71 80
-
S/2/M(%)
1.4
a ND, not determined
Table 2. Cell kinetics before and after 48 h in vitro exposure to Adria-
Table 4. Comparison of sensitivity by Adriamycin binding assay (ABA)
mycin (ADR) (580 ng/ml); data were obtained from the means of two experiments
and resistance evaluated by P-glycoprotein (PGP) expression; data were obtained from the means of three experiments
Cell lines
Cell lines
Sensitivity by ABA (%)
SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS 10 IOR/OS 14
99 88 90 91 83 94 98
0 0 0 0 0 0 0
U-20S U-20S/DXs~ U-20S/DX I~176 U-20S/DX5.~
97 44 6 2
0 12 59 81
Saos-2 Saos-2/DX3~ Saos-2/DX1~176 Saos-2/DX580
99 24 14
0 27 28 78
BrdUrd labelling index Control
Resistance by PGP (%)
After ADR
U-20S Saos-2 SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS 10 IOR/OS 14
33.5 15.8 2.8 5.5 1.6 13.3 8.5 7.1 2.0
0 0 0 0 0 0 0 0 0
U-20S/DX5s~ Saos-2/DX5s~
17.0 18.9
16.9 21.9
primary cell cultures (SARG, IOR/MOS, IOR/OS7, IOR/ OSF, IOR/OS9, IOR/OS10, and IOR/OS14) showed a marked inhibition of the in vitro proliferative activity after 48 h of exposure to Adriamycin. The inhibition of cell growth in vitro (Table 1), the decrease of BrdUrd uptake (Table 2) and the slowing down or the arrest in G J M (Table 3) showed that these cells were sensitive to the in vitro treatment with Adriamycin. The same response to the drug was observed in the sensitive cell lines U - 2 0 S and Saos-2, whereas the Adriamycin-resistant sublines U - 2 0 S / D X 58~ and Saos/DX 58~ did not show significant changes after exposure to the same amount of drug. After exposure to 10 gg/ml Adriamycin for 30 m i n at 37 ~ C, nuclear binding was evident in over 80% of the cells in all the primary cultures (SARG, IOR/MOS, IOR/OSK IOR/OS7; IOR/OS9, IOR/OS 10, and IOR/OS 14). A n inverse relation was found between the percentage of sensitive ceils by A B A and the expression of P-glycoprotein (Table 4). Likewise, no expression of P-glycoprotein was evident in the sensitive cell lines U 2 0 S and Saos-2. In the resistant variants, the expression of P-glycoprotein corresponded to the levels of in vitro resistance to Adriamycin (Serra et al. 1991), but not necessarily to the percentage sen-
1
sitivity by ABA, In fact, the prevalence of the resistant pattern of subcellular Adriamycin distribution was detected even with low levels of in vitro resistance and of P-glycoprotein overexpression.
Discussion
Drug resistance may be present at clinical onset, as an intrinsic feature of the tumor, or develop during or after chemotherapy tllrough the prevalence of unresponsive subpopulations. In all'three instances, even if other mechanisms, such as cell kinetics and drug diffusion, may contribute to drug treatment failure, the basis of drug resistance resides mainly at the cellular level (Goldie and Coldman 1984). Multidrug resistance (MDR) is characterized by loss of sensitivity to a variety of structurally unrelated drugs, including Adriamycin (Riordan and Ling 1985). The M D R pheno-
125 type is usually characterized by increased expression of Pglycoprotein, a membrane protein associated with intracellular drag levels (Juliano and Ling 1976; Kartner et al. 1983). The relationship between drug accumulation and in vitro responsiveness to Adriamycin has been frequently reported (Ganapathi et al. 1982; Fojo et al. 1985; Luk and Tannock 1989), although this finding has not been confirmed by other investigators (Kessel and Corbett 1985; Marsh et al. 1986). In our experience, quantitative analysis of Adriamycin fluorescence within the cell was unable to detect a significant difference between sensitive and resistant cells. Although multiple mechanisms may contribute to the cytotoxic activity of Adriamycin, intercalation with nucleic acids is probably the most relevant cause (Myers 1981). On the basis of the natural fluorescence of anthracyclifies (Bachur and Cradock 1970) it is possible to observe Adriamycin binding to the nucleus (Silvestrini et al. 1970; Egorin et al. 1974) and detect changes in subcellular concentration of the drug at its main target. Sensitive and resistant cells show differences in the intracellular distribution of anthracyclines (Chauffert et al. 1984; Willingham et al. 1986; Gigli et al. 1989; Hindenburg et al. 1989; Gervasoni et al. 1991; Weaver et al. 1991) and may be easily distinguished by direct microscopic observation. A B A was developed as a simple chemosensitivity assay based on the assessment of different patterns of intracellular Adriamycin distribution (Kusuzaki et al. 1989). The assay was originally tested on the mouse leukemia cell lines P388 and P388/DX, respectively sensitive and resistant to Adriamycin. We have investigated the reliability of this assay on different primary cultures and on sensitive and resistant cell lines of human osteosarcoma. In all the samples, after incubation with 10 gg/ml Adriamycin under the standard conditions of 30 rain exposure at 37 ~ C, two different binding patterns might be observed in living cells: sensitivity was identified by a bright fluorescence in the nucleus and a weak fluorescence in the cytoplasm; cytoplasmic fluorescence with little if any fluorescence in the nucleus defined resistance. In all the cultures, the prevalence of these patterns corresponded to the in vitro response to Adriamycin. An inverse relation was found between the percentage of sensitive cells as detected by A B A and the expression of Pglycoprotein in all the primary cultures, in the sensitive cell lines U - 2 0 S and Saos-2, and in the highly (over 300-fold) resistant variants U - 2 0 S / D X 5s~and Saos-2/DX 5s~ However, in U - 2 0 S / D X 3~ Saos-2/DX 3~ U - 2 0 S / D X 1~176 and Saos2/DX 1~176 all showing lower degrees of resistance in vitro (14to 100-fold higher than the parental line), the majority of cells showed resistance by A B A but only a few had increased expression of P-glycoprotein. Several reports suggest that alternative mechanisms of drug resistance should be considered in the absence of P-glycoprotein overexpression (Baas et al. 1990; Samuels et al. 1991; Toffoli et al. 1991). P-glycoprotein is more commonly involved in conditions with high levels of resistance, occurring late in the in Vitro selection of resistant variants and, in clinical situations, more commonly after treatment failure (Baas et al. 1990). In the early phases of development of resistance in vitro, and in inherent resistance, which may be present at clinical onset, detection of P glycoprotein activity may not be able to show resistance. On the other hand, as shown by our data, the assessment of intra-
cellular drug distribution by A B A can define sensitivity to Adriamycin regardless of the level of expression of P-glycoprotein. The possibility of detecting low levels of resistance to Adriamycin may be of great importance in the treatment of osteosarcoma. This tumor generally shows sensitivity to Adriamycin and other drugs, but resistance may occur either at the onset, as an intrinsic feature of the tumor, or late, after several courses of chemotherapy. In both cases, the assessment of chemosensitivity by A B A in conjunction with the analysis of P-glycoprotein expression may provide useful indications of the individual response to therapy and possibly help the selection of patients in which alternative treatments can be considered.
References Ashihara T (1985) Quantitative cytochemistry: microscope-based cytofluorometry and its application to the analysis of human cancer cells. Acta Histochem Cytochem 17:681-682 Baas F, Jongsma APM, Broxterman HJ, Arceci RJ, Housman GL, Scheffer GL, Riethorst A, van Groenigen M, Nieuwint AWM, Joenje H (1990) Non-p-glycoprotein mediated mechanism for multidrug resistance precedes p-glycoprotein expression during in vitro selection for doxorubicin resistance in a human lung cancer cell line. Cancer Res 50:5392-5398 Bacci G, Picci P, Ruggieri P, Mercuri M, Avella M, Capanna R, Brach Del Preyer A, Mancini A, Gherlinzoni F, Padovani G, Leonessa C, Biagini R, Ferraro A, Fem~zzi A, Cazzola A, Manfrini M, Campanacci M (1990) Primary chemotherapy and delayed surgery (neoadjuvant chemotherapy) for osteosarcoma of the extremities. Cancer 65:2539-2553 Bachur NR, Cradock JC (1970) Daunomycin metabolism in rat tissue slices. J Pharmacol Exp Ther 175:331-337 Chauffert B, Martin F, Caignard A, Jeannin JF, Leclerc A (1984) Cytofluorescence localization of Adriamycin in resistant colon cancer cells. Cancer Chemother Pharmacol 13:14-18 Dolbeare F, Gratzner H, Pallavicini MG, Gray JW (1983) Flow cytometric measurement of total DNA content and incorporated bromodeoxyuridine. Proc Natl Acad Sci USA 80:5573-5577 Egorin MJ, Hildebrand RC, Cimino EF, Bachur NR (1974) Cytofluorescence localization of Adriamycin and daunombicin. Cancer Res 34:2243-2245 Fojo A, Akiyama SI, Gottesman MM, Pastan I (1985) Reduced drug accumulation in multiple drug resistant human KB carcinoma cell lines. Cancer Res 45:3002-3007 Ganapathi R, Reiter W, Krishan A (1982) Intracellular Adriamycin levels and cytotoxicity in Adriamycin-sensitive and Adriamycin-resistant P388 mouse leukemia cells. J Natl Cancer Inst 68:1027-1032 Gervasoni JE, Fields SZ, Krishna S, Baker MA, Rosado M, Thuraisamy K, Hindenburg AA, Taub RN (1991) Subcellular distribution of daunorubicin in p-glycoprotein-positive and -negative drug-resistant cell lines using laser-assisted confocal microscopy. Cancer Res 51:4955-4963 Gigli M, Rasoanaivo TWD, Millot J-M, Jeannesson P, Rizzo V, Jardillier J-C, Arcamone F, Manfalt M (1989) Correlation between growth inhibition and intranuclear doxorubicin and 4-deoxy-4-iododoxorubicin quantitated in living K562 cells by microspectrofluorometry. Cancer Res 49:560-564 Goldie JH, Coldman AJ (1984) The genetic origin of drug resistance in neoplasms: implications for systemic therapy. Cancer Res 44:3643-3653 Hindenburg AA, Gervasoni JE, Krishna S, Stewart VJ, Rosado M, Lutzky J, Bhalla K, Baker MA, Taub RN (1989) Intracellular distribution and pharmacokinetics of daunorubicin in anthracycline-sensitive and -resistant HL-60 ceils. Cancer Res 49:4607-4614
126 Juliano RL, Ling V (1976) A surface glycoprotein modulating chug permeability in chinese hamster ovary cell mutants. Biochim Biophys Acta 455:152--162 Karmer N, Riordan JR, Ling V (1983) Cell surface p-glycoprotein is associated with multidrng resistance in mammalian cell lines. Science 221:1285-1288 Kessel D, Corbett T (1985) Correlations between anthracycline resistance, drug accumulation and membrane glycoprotein patterns in solid minors of mice. Cancer Lett 28:187-193 Krishan A, Ganapathi R (1980) Laser flow cytometric studies on the intracellular fluorescence of anthracyclines. Cancer Res 40: 3895-3900 Kusuzaki K, Litwak GI, Gebhardt MC, Springfield DS, Mankin HJ (1989) Assessment of intracellular Adriamycin (Adr) binding in bone and soft-tissue tumor cells by fluorescence miroscopy. Trans Orthop Res Soc 14:549 Luk CK, Tannock IF (1989) Flow cytometric analysis of doxornbicin accumulation in cells from human and rodent cell lines. J Natl Cancer Inst 81:55-59 Marsh W, Sickeri D, Center MS (1986) Isolation and characterization of Adriamycin-resistant HL-60 ceils which are not defective in the initial intracellular accumulation of drug. Cancer Res 46:187-193 Myers CE (1981) Antitumor antibiotics: I. Anthracycline. In: Pinedo HM (ed) Cancer chemotherapy 1980. Exeerpta Medica, Amsterdam, pp 66-93 Riordan JR, Ling V (1985) Genetic and biochemical characterization of multi-drng resistance. Pharmacol Ther 28:51-75 Rotman B, Papermaster BW (1966) Membrane properties of living cells as studied by enzymatic hydrolysis of fluorigenic esters. Proc Natl Acad Sci USA 55:134-141
Samuels BL, Murray JL, Cohen MB, Safa AR, Sinha BK, Townsend A J, Beckett MA, Weichelsbaum RR (1991) Increased glutathione peroxidase activity in a human sarcoma cell line with inherent doxorubicin resistance. Cancer Res 51:521-527 Scheper RJ, Bulte JWM, Brakkee JGP, Quak JJ, Schoot E, Balm AJM, Meijer C, Broxterman H, Kuiper CM, Lankelma J, Pinedo HM (1988) Monoclonal antibody (JSB-I) detects a highly conserved epitope on the p-glycoprotein associated with multidrng-resistance. Int J Cancer 42:389-394 Serra M, Scotlandi K, Olivari S, Manara MC, Baldini N (1991) Characterization of Adriamycin-resistant variants of two human osteosa> coma cell lines. Eur J Cancer 27 [Suppl 3]:$44 Silvestrini R, Gambarucci C, Dasdia T (1970) In vitro biological activity of Adriamycin. Tumori 56:13%148 Toffoli G, Viel A, Tumiotto L, Biscontin G, Rossi C, Boiocchi M (199I) Pleiotropic-resistant phenotype is a multifactorial phenomenon in human colon carcinoma cell lines. Br ~ Cancer 63:51-56 Weaver JL, Pine PS, Aszalos A, Schoenlein PV, Currier SJ, Padmanabhan R, Gottesman MM (1991) Laser scanning and confocal microscopy of daunorubicin, doxorubicin, and rhodamine 123 in multidrug-resistance cells. Exp Cell Res 196:323-329 Willingham MC, Cornwell MM, Cardarelli CO, Gottesmann MM, Pastan I (1986) Single cell analysis of daunomycin uptake and efflux in multidrng-resistant and -sensitive KB cells: effects of verapamil and other drugs. Cancer Res 46:5941-5946 Winkler K, Beron G, Delling G, Heise U, Kabisch H, Purfiist C, Berger J, Ritter J, Jfirgens H, Gerein V, Graf N, Russe W, Gruemayer ER, Ertelt W, Kotz R, Preusser P, Prindull G, Brandeis W, Landbeck G (1988) Neoadjuvant chemotherapy of osteosarcoma: results of a randomized cooperative trial (COSS-82) with salvage chemotherapy based on histological tumor response. J Clin Oncol 6:329-337
Cancer Research Clinical 9 9 Springer-Verlag 1992
Adriamycin binding assay: a valuable chemosensitivity test in human osteosarcoma Nicola Baldini 1, Katia Scoflandi 1, Massimo Serra 1, Katsuyuki Kusuzaki 2, Toshiharu Shikita 1,2, Maria C. Manara 1, Daniela Maurici 1, and Mario Campanacci 1 1 Laboratorio di Ricerca Oncologica, Istituti Ortopedici Rizzoli, Bologna, Italy 2 Department of Orthopaedic Surgery, Kyoto Prefectural University of Medicine, Kyoto, Japan Received 14 February 1992/Accepted 23 June 1992
Summary. The reliability of a simple method evaluating the pattern of subcellular binding of Adriamycin (Adriamycin binding assay, A B A ) as an index of sensitivity was demonstrated in different primary cultures and in sensitive and resistant cell lines of human osteosarcoma. After exposure to Adriamycin (10 gg/ml for 30 min at 37 ~ C), living sensitive cells showed selective intranuclear uptake of the drug, whereas in resistant cells no distinct subcellular distribution was observed. The binding pattern of Adriamycin in sensitive and in highly resistant cells was inversely related to the expression of P-glycoprotein. However, low levels of resistance in vitro, not detectable b y increased levels of expression of P-glycoprotein, were revealed by A B A . The use of A B A in combination with the estimate of P-glycoprotein expression is recommended in clinical practice as an accurate means for predicting the sensitivity of osteosarcoma to Adriamycin.
Different patterns of fluorescence have been observed in living sensitive and resistant cells (Willingham et al. 1986) and responsiveness to the drug has been associated with elective binding to the nucleus, the more obvious target of A d r i a m y cin cytotoxic activity (Gigli et al. 1989). A simple method evaluating the pattern of subcellular binding of Adriamycin (Adriamycin binding assay, A B A ) has been recently suggested as a new chemosensitivity test (Kusuzaki et/al. 1989). In this paper we demonstrate the reliability of this method to determine the sensitivity to Adriamycin in different primary cultures and in both sensitive and resistant cell lines of human osteosarcoma, with a view to further application on clinical material.
Materials and methods Cells. Human osteosarcoma cell lines U-20S and Saos-2 were gifts, re-
Key words: A d r i a m y c i n - Chemosensitivity - Osteosarcoma
Introduction Adriamycin is very effective in high-grade osteosarcoma and should be included in any multimodal therapeutic regimen for this tumor (Winkler et al. 1988; Bacci et al. 1990). The heterogeneous response o f osteosarcoma to induction chemotherapy and the occurrence of relapses in 3 0 % - 4 0 % of the patients despite aggressive treatment suggest the need for individualized therapies based on the preliminary assessment of sensitivity. Adriamycin is a natural fluorochrome (Silvestrini et al. 1970), and intracellular drug accumulation m a y be monitored by direct microscopic observation (Egorin et al. 1974). This study was supported by a grant from the Associazione Italiana per la Ricerca sul Cancro (A.I.R.C.). Correspondence to: Nicola Baldini, Laboratorio di Ricerca Oncologica, Istituti Rizzoli, via di Barbiano 1/10, 40136 Bologna, Italy.
spectively, from Dr. GraziellaPratesi and Dr. Maria I. Colnaghi, Istituto Nazionale Tumori, Milano, Italy. Adriamycin-resistantvariants of these sensitive lines were selected by continuous in vitro exposure to the drug, as previously described (Serra et al. 1991). Stepwise increases in Adriamycin concentration produced sublines that were resistant up to 580ng/ml (1 pM) Adriamycin (respectively, U - 2 0 S / D X 3~ Saos2/DX 3~ U-20S/DX ~~176 Saos-2/DX 1~176 U-20S/DX 58~and Saos-2/DX58~ Primary cell cultures (SARG, IOR/MOS, IOPUOSF, IOR/OS7, IOR/OS9, IOR/OS10, IOR/OS14) were obtained from surgical specimens of seven cases of high-grade human osteosarcoma. Of these, six were primary lesions and 1 (IOR/OS9) was a bone metastasis. After mechanical mincing, cells were seeded in Iscove's modified Dulbecco's medium supplemented with penicillin (100 U/ml) and streptomycin (100 gg/ml) (IMDM, Gibco, Paisley, Scotland) and with 10% heat-inactivated fetal calf serum (FCS, Biological Industries, Kibbutz Beth Haemek, Israel). Culture flasks were incubated at 37~ C in a humidified atmosphere of 5% CO2. All the cultures used for this study were before their tenth in vitro passage. Incubations with Adriamycin. Adriamycin was purchased from Farmita-
lia Carlo Erba, Milano, Italy. To identify the optimal conditions for drug incorporation, cell suspensions (200,000 cells/ml IMDM, 10% FCS) from U-20S and Saos-2 were incubated with different Adriamycin concentrations (1 ~tg/ml, 5 p.g/ml, 10 ~tg/ml, and 50 ~tg/ml) at 37~ C for 10, 30, 60, 90, 120, and 180 min. Cells were also exposed to 50 gg/ml Adriamycin at 4 ~ C. After Adriamycin incubation in the above conditions, cells were incubated with 1 mg/ml fluorescein diacetate for
122
Fig. 1. Fluorescein diacetate staining and intracellular distribution of Adriamycin in sensitive (a, b) and resistant (c, d) cells after exposure to 10 gg/ml Adriamycin at 37 ~ C
10 rain at 4 ~ C, washed and resuspended in phosphate-buffered saline prior to microscopic observation and cytofluorometry. Exposure t o 10 btg/ml Adriamycin for 30 min at 37 ~ C was defined as a standard for the assay, in agreement with a previous report (Kusuzaki et al. 1989). To define the relationship between differences in the pattern and the amount of drug incorporation in sensitive and resistant cells, cell suspensions (200000 cells/ml IMDM with 10% FCS) from Saos-2/DX58~were incubated with 10 gg/ml Adriamycin for 30, 60, 90, and 120 rain at 37 ~ C, and then with 1 mg/ml fluorescein diacetate for 10 rain at 4 ~ C prior to observation and cytofluorometry. To evaluate the reliability of the assay on different cultures, cell suspensions (200000 cells/ml IMDM with 10% FCS) from U - 2 0 S / D X 3~ U - 2 0 S / D X 1~176 U - 2 0 S / D X 58~ Saos-2/DX3~ Saos-2/DX I~176 SARG, IOR/MOS, IOR/OS7, IOR/OSF, IOR/OS9, IOR/OS10, and IOR/OS14 were incubated with 10 gg/ml Adriamycin for 30 rain at 37 ~ C and then with 1 mg/ml fluorescein diacetate for 10 min at 4 ~ C prior to microscopic observation.
Microscopic observation and cytofluorometry. Cells were observed with an epi-fluorescence microscope (Nikon FX-A) equipped with a HBO 100-W high-pressure mercury lamp operating on stabilized d.c. power supply, a 40x fluorite lens (0.85 NA), and a high-gain photomultiplier connected to a personal computer for signal recording and analysis. Living cells were first identified with fluorescein diacetate green fluorescence (Rotman and Papermaster 1966). On those cells, the red fluorescence of Adriamycin was examined and two different patterns observed: strong nuclear fluorescence with weak fluorescence in the cytoplasm (Fig. 1 a, b); and diffuse fluorescence without any distinctive intracellular binding (Fig. 1 c, d). As previously suggested, the first pattern was considered sensitive and the second resistant (Kusuzaki et al. 1989). The percentage of sensitive ceils was calculated on 300 cells for each sample.
In addition, cytofluorometric measurement of intracellular Adriamycin content was made on Saos-2, and Saos-2/DXsS~under the abovementioned conditions. Drug fluorescence intensity was determined on 100 living cells for each sample and expressed in arbitrary units after background value had been set at zero.
In vitro sensitivity to Adriamycin. Cells (13x103/cm 2) were seeded in IMDM with 10% FCS. After 24 h, the medium was changed with IMDM with 10% FCS plus 580 ng/ml, 58 ng/ml, and 5.8 ng/ml Adriamycin, or with 1MDM 10% with FCS alone (control). After 48 h, viability was screened with the trypan blue dye exclusion method. Cell growth inhibition was determined as the ratio between the numbers of treated and untreated cells.
Effects on cell cycle. The effects of Adriamycin on cell kinetics were evaluated by bromodeoxyuridine (BrdUrd) incorporation (Dolbeare et al. 1983). Cytospin preparations were obtained from cell suspensions incubated with 50 btM BrdUrd (Sigma, St. Louis, Mo.) at 37~ for 10 rain. After fixation with 70% ethanol for 10 min and treatment with 4 M HCI at room temperature for 30 min, the specimens were incubated with the anti-BrdUrd monoclonal antibody MBU (Eurodiagnostic, Apeldoorn, The Netherlands) at a 1 : 10 dilution at 37 ~ C for 30 rain. A fluorescein-isothiocyanate-conjugatedsecondary antibody was used for labelling. The BrdUrd labelling index was estimated on 300 cells for each slide. Changes of cell cycle were further investigated by determining the DNA content on cytospin preparations obtained before and after exposure to the drug. After fixation with 70% ethanol at 4 ~ C for 30 rain, cells were treated with 0.5 mg/ml ribonuclease A (Sigma, St. Louis, Mo.) at 37 ~ C for 1 h and stained with 25 btg/ml propidium iodide (Sigma, St. Louis, Mo.). Nuclear DNA content was determined by cytofluorometry (Ashihara 1985). The effects of drug exposure on the cell cycle were
123 oo,
U-20S
.~ ...................= ...................~..., .., ~=,,.g~.............
100
/,~.._.;;;:::,,
90
KI/
80 W t..) I.U E 09 z ILl 09
IJ. O
80
.........
tt ....
70
o
6o
t= '~
40
0
40
0
30
/
100 --i
1
a
n
n
i
60
90
120
150
180
TIME
(minutes)
.~............................... ..........::::::--~....
001 / U 7
il/
7o
/
/,
,'I/
'ot'k" / /
40 1 ! / i
/i
~o
lo
30
60 TIME
Saos-2
O3 U. O o~
580
50
. . . . ...-~ 0 .... ~ . ! 0 30
E
Saos-2
1
60
10
09 z ILl
-] Saos-2/DX
20
_J [iJ O [JJ
lo0
":..,
/
/
,-'*~ .........
_-
..-"
,.
/
/
o
0
30
60
90
120
TIME
(minutes)
150
180
Fig. 2. Percentage of cells with selective nuclear binding of Adriamycin in sensitive cell lines U - 2 0 S and Saos-2 after exposure to 50 gg/ml ( . . . i . . . ) , 10 gg/ml ( - - . - - ) , 5 gg/ml (... m.--) and 1 gg/ml ( . . . . . . . ) at 37 ~ C, andto 50 gg/ml ( - - = - - ) at4 ~ C
evaluated and expressed as a ratio between the percentage of untreated and treated cells in phases S/GJM.
P-glycoprotein expression. Cytospin preparations obtained from cul-
120
Fig. 3. Cytofluorometric analysis of intracellular Adriamycin content in Saos-2 ( - - . - - ) and Saos-2/DX58~ ( - - o - - ) after exposure to 10 gg/ml Adriamycin at 37 ~ C
served at different times to detect the pattern of intracellular uptake of the drug. As shown in Fig. 2, selective nuclear binding, typical of sensitive cells, was observed in over 90% of the cells after 30 min of incubation with both 50 gg/ml and 10 gg/ml Adriamycin, after 60 rain incubation with 5 gg/ml Adriamycin, and after 120 min with 1 pg/ml Adriamycin. No uptake was detected even after 120 min exposure to the highest dose of 50 pg/ml Adriamycin when incubation was at 4 ~ C. Cell viability after Adriamycin exposure was always over 90%. On the basis of our results and as previously suggested in mouse leukemia cells (Kusuzaki et al. 1989), optimum conditions to demonstrate drug incorporation in human osteosarcoma were fixed at 10 gg/ml for 30 rain at 37 ~ C. Cytofluorometric measuremend of intracellular Adriamycin did not show any significant difference between Saos-2 and Saos-2/DX 58~ even after 120min of exposure to 10 gg/ml Adriamycin at 37 ~ C (Fig. 3). However, under the same conditions, direct observation revealed a striking difference in the binding pattern between sensitive and resistant cells (Fig. 4). More than 90% of the Saos-2 cells were shown to be sensitive after 30 min, and less than 5% of Saos2/DX 58~even after 120 rain of exposure, confirming that the percentage of sensitive cells, as shown by the selective nuclear binding of Adriamycin, is a valuable index of the in vitro sensitivity to the drug. The reliability of ABA as a means to assess the sensitivity to Adriamycin was further investigated on seven primary osteosarcoma cell cultures obtained in our laboratory. All the
tures were fixed in acetone at room temperature for 10 rain. The expression of P-glycoprotein was evaluated by indirect immunofluorescence with the monoclonal antibody JSB-1 (Sanbyo, Uden, The Netherlands) at a 1 : 10 dilution (Scheper et al. 1988). The percentage of positive cells was determined on 300 cells for each slide.
..i
Results
O ua I--
lO0-
co
z I.U
The two human osteosarcoma cell lines U - 2 0 S and Saos-2 sensitive to Adriamycin, and their resistant variants U-2 OS/DX 58~ and Saos-2/DX 58~ (Serra et al. 1991) were a suitable in vitro model with which to verify the reliability of ABA, a new chemosensitivity test based on the differential binding of Adriamycin in sensitive and resistant cells. Cell suspensions from U - 2 0 S and Saos-2, incubated at 37 ~ C with different Adriamycin concentrations, were ob-
90
(minutes)
~.
9
Saos-2
806040
09
2oo~
0
=
~ 30 TIME
~t
~
60
90
580
~
Saos-2/DX
120
(minutes)
Fig. 4. Percentage of cells with selective nuclear binding of Adriamycin in Saos-2 ( - - . - - ) and Saos-2/DX58~ ( - - o - - ) after exposure to 10 gg/ml Adriamycin at 37 ~ C
124 Table 1. Cell growth inhibition after 48 h exposure to 580 ng/ml,
58 ng/ml, and 5.8 ng/ml Adriamycin (ADR) Cell lines
Table 3. Changes in the cell-cycle distribution before and after 48 h exposure to Adriamycin (ADR) (580 ng/ml); data were obtained from the mean of two experiments
Growth inhibition (%) after ADR at Cell lines 580 ng/ml
58 ng/ml
5.8 ng/ml
U-20S Saos-2 SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS10 IOR/OS14
94.0+- 5.9 67.3+_15.0 96.5+_ 2.3 65.0+12.1 68.4_+ 9.3 57:8+-15,1 84.4+- 2.5 95.3+_ 2.5 57.4+_ 1.1
79.4_+4.2 60.0+_8.3 53.8+_7.4 21.3+-9.2 ND ~ 64.7_+5.6 71.0+_3.1 69.5+_2.8 17.5+_3.7
51.7_+13.5 11.5+_ 5.7 17.8+_ 9.8 10.6+- 4.5 ND ~ 20.2_+ 8.2 50.5+_ 7.6 51.2+ 1.7 2.1+_ 1.1
U20S/DX 58~ Saos-2/DX5s~
15.7_+ 2.2 16.4+_ 3.6
2.0+_1.3 1.1+_0.5
0 0
S/G2/M(%)
S/G2/M (%) after ADR
Increase in
control U-20S Saos-2 SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS 10 IOR/OS 14
39 45 50 31 17 15 43 51 52
72 58 71 66 29 33 68 80 76
45.8 22.4 29.6 53 41.4 54.5 36.8 36.2 31.6
U-20S/DX58~ Saos-2/DX5s~
70 85
71 80
-
S/2/M(%)
1.4
a ND, not determined
Table 2. Cell kinetics before and after 48 h in vitro exposure to Adria-
Table 4. Comparison of sensitivity by Adriamycin binding assay (ABA)
mycin (ADR) (580 ng/ml); data were obtained from the means of two experiments
and resistance evaluated by P-glycoprotein (PGP) expression; data were obtained from the means of three experiments
Cell lines
Cell lines
Sensitivity by ABA (%)
SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS 10 IOR/OS 14
99 88 90 91 83 94 98
0 0 0 0 0 0 0
U-20S U-20S/DXs~ U-20S/DX I~176 U-20S/DX5.~
97 44 6 2
0 12 59 81
Saos-2 Saos-2/DX3~ Saos-2/DX1~176 Saos-2/DX580
99 24 14
0 27 28 78
BrdUrd labelling index Control
Resistance by PGP (%)
After ADR
U-20S Saos-2 SARG IOR/MOS IOR/OSF IOR/OS7 IOR/OS9 IOR/OS 10 IOR/OS 14
33.5 15.8 2.8 5.5 1.6 13.3 8.5 7.1 2.0
0 0 0 0 0 0 0 0 0
U-20S/DX5s~ Saos-2/DX5s~
17.0 18.9
16.9 21.9
primary cell cultures (SARG, IOR/MOS, IOR/OS7, IOR/ OSF, IOR/OS9, IOR/OS10, and IOR/OS14) showed a marked inhibition of the in vitro proliferative activity after 48 h of exposure to Adriamycin. The inhibition of cell growth in vitro (Table 1), the decrease of BrdUrd uptake (Table 2) and the slowing down or the arrest in G J M (Table 3) showed that these cells were sensitive to the in vitro treatment with Adriamycin. The same response to the drug was observed in the sensitive cell lines U - 2 0 S and Saos-2, whereas the Adriamycin-resistant sublines U - 2 0 S / D X 58~ and Saos/DX 58~ did not show significant changes after exposure to the same amount of drug. After exposure to 10 gg/ml Adriamycin for 30 m i n at 37 ~ C, nuclear binding was evident in over 80% of the cells in all the primary cultures (SARG, IOR/MOS, IOR/OSK IOR/OS7; IOR/OS9, IOR/OS 10, and IOR/OS 14). A n inverse relation was found between the percentage of sensitive ceils by A B A and the expression of P-glycoprotein (Table 4). Likewise, no expression of P-glycoprotein was evident in the sensitive cell lines U 2 0 S and Saos-2. In the resistant variants, the expression of P-glycoprotein corresponded to the levels of in vitro resistance to Adriamycin (Serra et al. 1991), but not necessarily to the percentage sen-
1
sitivity by ABA, In fact, the prevalence of the resistant pattern of subcellular Adriamycin distribution was detected even with low levels of in vitro resistance and of P-glycoprotein overexpression.
Discussion
Drug resistance may be present at clinical onset, as an intrinsic feature of the tumor, or develop during or after chemotherapy tllrough the prevalence of unresponsive subpopulations. In all'three instances, even if other mechanisms, such as cell kinetics and drug diffusion, may contribute to drug treatment failure, the basis of drug resistance resides mainly at the cellular level (Goldie and Coldman 1984). Multidrug resistance (MDR) is characterized by loss of sensitivity to a variety of structurally unrelated drugs, including Adriamycin (Riordan and Ling 1985). The M D R pheno-
125 type is usually characterized by increased expression of Pglycoprotein, a membrane protein associated with intracellular drag levels (Juliano and Ling 1976; Kartner et al. 1983). The relationship between drug accumulation and in vitro responsiveness to Adriamycin has been frequently reported (Ganapathi et al. 1982; Fojo et al. 1985; Luk and Tannock 1989), although this finding has not been confirmed by other investigators (Kessel and Corbett 1985; Marsh et al. 1986). In our experience, quantitative analysis of Adriamycin fluorescence within the cell was unable to detect a significant difference between sensitive and resistant cells. Although multiple mechanisms may contribute to the cytotoxic activity of Adriamycin, intercalation with nucleic acids is probably the most relevant cause (Myers 1981). On the basis of the natural fluorescence of anthracyclifies (Bachur and Cradock 1970) it is possible to observe Adriamycin binding to the nucleus (Silvestrini et al. 1970; Egorin et al. 1974) and detect changes in subcellular concentration of the drug at its main target. Sensitive and resistant cells show differences in the intracellular distribution of anthracyclines (Chauffert et al. 1984; Willingham et al. 1986; Gigli et al. 1989; Hindenburg et al. 1989; Gervasoni et al. 1991; Weaver et al. 1991) and may be easily distinguished by direct microscopic observation. A B A was developed as a simple chemosensitivity assay based on the assessment of different patterns of intracellular Adriamycin distribution (Kusuzaki et al. 1989). The assay was originally tested on the mouse leukemia cell lines P388 and P388/DX, respectively sensitive and resistant to Adriamycin. We have investigated the reliability of this assay on different primary cultures and on sensitive and resistant cell lines of human osteosarcoma. In all the samples, after incubation with 10 gg/ml Adriamycin under the standard conditions of 30 rain exposure at 37 ~ C, two different binding patterns might be observed in living cells: sensitivity was identified by a bright fluorescence in the nucleus and a weak fluorescence in the cytoplasm; cytoplasmic fluorescence with little if any fluorescence in the nucleus defined resistance. In all the cultures, the prevalence of these patterns corresponded to the in vitro response to Adriamycin. An inverse relation was found between the percentage of sensitive cells as detected by A B A and the expression of Pglycoprotein in all the primary cultures, in the sensitive cell lines U - 2 0 S and Saos-2, and in the highly (over 300-fold) resistant variants U - 2 0 S / D X 5s~and Saos-2/DX 5s~ However, in U - 2 0 S / D X 3~ Saos-2/DX 3~ U - 2 0 S / D X 1~176 and Saos2/DX 1~176 all showing lower degrees of resistance in vitro (14to 100-fold higher than the parental line), the majority of cells showed resistance by A B A but only a few had increased expression of P-glycoprotein. Several reports suggest that alternative mechanisms of drug resistance should be considered in the absence of P-glycoprotein overexpression (Baas et al. 1990; Samuels et al. 1991; Toffoli et al. 1991). P-glycoprotein is more commonly involved in conditions with high levels of resistance, occurring late in the in Vitro selection of resistant variants and, in clinical situations, more commonly after treatment failure (Baas et al. 1990). In the early phases of development of resistance in vitro, and in inherent resistance, which may be present at clinical onset, detection of P glycoprotein activity may not be able to show resistance. On the other hand, as shown by our data, the assessment of intra-
cellular drug distribution by A B A can define sensitivity to Adriamycin regardless of the level of expression of P-glycoprotein. The possibility of detecting low levels of resistance to Adriamycin may be of great importance in the treatment of osteosarcoma. This tumor generally shows sensitivity to Adriamycin and other drugs, but resistance may occur either at the onset, as an intrinsic feature of the tumor, or late, after several courses of chemotherapy. In both cases, the assessment of chemosensitivity by A B A in conjunction with the analysis of P-glycoprotein expression may provide useful indications of the individual response to therapy and possibly help the selection of patients in which alternative treatments can be considered.
References Ashihara T (1985) Quantitative cytochemistry: microscope-based cytofluorometry and its application to the analysis of human cancer cells. Acta Histochem Cytochem 17:681-682 Baas F, Jongsma APM, Broxterman HJ, Arceci RJ, Housman GL, Scheffer GL, Riethorst A, van Groenigen M, Nieuwint AWM, Joenje H (1990) Non-p-glycoprotein mediated mechanism for multidrug resistance precedes p-glycoprotein expression during in vitro selection for doxorubicin resistance in a human lung cancer cell line. Cancer Res 50:5392-5398 Bacci G, Picci P, Ruggieri P, Mercuri M, Avella M, Capanna R, Brach Del Preyer A, Mancini A, Gherlinzoni F, Padovani G, Leonessa C, Biagini R, Ferraro A, Fem~zzi A, Cazzola A, Manfrini M, Campanacci M (1990) Primary chemotherapy and delayed surgery (neoadjuvant chemotherapy) for osteosarcoma of the extremities. Cancer 65:2539-2553 Bachur NR, Cradock JC (1970) Daunomycin metabolism in rat tissue slices. J Pharmacol Exp Ther 175:331-337 Chauffert B, Martin F, Caignard A, Jeannin JF, Leclerc A (1984) Cytofluorescence localization of Adriamycin in resistant colon cancer cells. Cancer Chemother Pharmacol 13:14-18 Dolbeare F, Gratzner H, Pallavicini MG, Gray JW (1983) Flow cytometric measurement of total DNA content and incorporated bromodeoxyuridine. Proc Natl Acad Sci USA 80:5573-5577 Egorin MJ, Hildebrand RC, Cimino EF, Bachur NR (1974) Cytofluorescence localization of Adriamycin and daunombicin. Cancer Res 34:2243-2245 Fojo A, Akiyama SI, Gottesman MM, Pastan I (1985) Reduced drug accumulation in multiple drug resistant human KB carcinoma cell lines. Cancer Res 45:3002-3007 Ganapathi R, Reiter W, Krishan A (1982) Intracellular Adriamycin levels and cytotoxicity in Adriamycin-sensitive and Adriamycin-resistant P388 mouse leukemia cells. J Natl Cancer Inst 68:1027-1032 Gervasoni JE, Fields SZ, Krishna S, Baker MA, Rosado M, Thuraisamy K, Hindenburg AA, Taub RN (1991) Subcellular distribution of daunorubicin in p-glycoprotein-positive and -negative drug-resistant cell lines using laser-assisted confocal microscopy. Cancer Res 51:4955-4963 Gigli M, Rasoanaivo TWD, Millot J-M, Jeannesson P, Rizzo V, Jardillier J-C, Arcamone F, Manfalt M (1989) Correlation between growth inhibition and intranuclear doxorubicin and 4-deoxy-4-iododoxorubicin quantitated in living K562 cells by microspectrofluorometry. Cancer Res 49:560-564 Goldie JH, Coldman AJ (1984) The genetic origin of drug resistance in neoplasms: implications for systemic therapy. Cancer Res 44:3643-3653 Hindenburg AA, Gervasoni JE, Krishna S, Stewart VJ, Rosado M, Lutzky J, Bhalla K, Baker MA, Taub RN (1989) Intracellular distribution and pharmacokinetics of daunorubicin in anthracycline-sensitive and -resistant HL-60 ceils. Cancer Res 49:4607-4614
126 Juliano RL, Ling V (1976) A surface glycoprotein modulating chug permeability in chinese hamster ovary cell mutants. Biochim Biophys Acta 455:152--162 Karmer N, Riordan JR, Ling V (1983) Cell surface p-glycoprotein is associated with multidrng resistance in mammalian cell lines. Science 221:1285-1288 Kessel D, Corbett T (1985) Correlations between anthracycline resistance, drug accumulation and membrane glycoprotein patterns in solid minors of mice. Cancer Lett 28:187-193 Krishan A, Ganapathi R (1980) Laser flow cytometric studies on the intracellular fluorescence of anthracyclines. Cancer Res 40: 3895-3900 Kusuzaki K, Litwak GI, Gebhardt MC, Springfield DS, Mankin HJ (1989) Assessment of intracellular Adriamycin (Adr) binding in bone and soft-tissue tumor cells by fluorescence miroscopy. Trans Orthop Res Soc 14:549 Luk CK, Tannock IF (1989) Flow cytometric analysis of doxornbicin accumulation in cells from human and rodent cell lines. J Natl Cancer Inst 81:55-59 Marsh W, Sickeri D, Center MS (1986) Isolation and characterization of Adriamycin-resistant HL-60 ceils which are not defective in the initial intracellular accumulation of drug. Cancer Res 46:187-193 Myers CE (1981) Antitumor antibiotics: I. Anthracycline. In: Pinedo HM (ed) Cancer chemotherapy 1980. Exeerpta Medica, Amsterdam, pp 66-93 Riordan JR, Ling V (1985) Genetic and biochemical characterization of multi-drng resistance. Pharmacol Ther 28:51-75 Rotman B, Papermaster BW (1966) Membrane properties of living cells as studied by enzymatic hydrolysis of fluorigenic esters. Proc Natl Acad Sci USA 55:134-141
Samuels BL, Murray JL, Cohen MB, Safa AR, Sinha BK, Townsend A J, Beckett MA, Weichelsbaum RR (1991) Increased glutathione peroxidase activity in a human sarcoma cell line with inherent doxorubicin resistance. Cancer Res 51:521-527 Scheper RJ, Bulte JWM, Brakkee JGP, Quak JJ, Schoot E, Balm AJM, Meijer C, Broxterman H, Kuiper CM, Lankelma J, Pinedo HM (1988) Monoclonal antibody (JSB-I) detects a highly conserved epitope on the p-glycoprotein associated with multidrng-resistance. Int J Cancer 42:389-394 Serra M, Scotlandi K, Olivari S, Manara MC, Baldini N (1991) Characterization of Adriamycin-resistant variants of two human osteosa> coma cell lines. Eur J Cancer 27 [Suppl 3]:$44 Silvestrini R, Gambarucci C, Dasdia T (1970) In vitro biological activity of Adriamycin. Tumori 56:13%148 Toffoli G, Viel A, Tumiotto L, Biscontin G, Rossi C, Boiocchi M (199I) Pleiotropic-resistant phenotype is a multifactorial phenomenon in human colon carcinoma cell lines. Br ~ Cancer 63:51-56 Weaver JL, Pine PS, Aszalos A, Schoenlein PV, Currier SJ, Padmanabhan R, Gottesman MM (1991) Laser scanning and confocal microscopy of daunorubicin, doxorubicin, and rhodamine 123 in multidrug-resistance cells. Exp Cell Res 196:323-329 Willingham MC, Cornwell MM, Cardarelli CO, Gottesmann MM, Pastan I (1986) Single cell analysis of daunomycin uptake and efflux in multidrng-resistant and -sensitive KB cells: effects of verapamil and other drugs. Cancer Res 46:5941-5946 Winkler K, Beron G, Delling G, Heise U, Kabisch H, Purfiist C, Berger J, Ritter J, Jfirgens H, Gerein V, Graf N, Russe W, Gruemayer ER, Ertelt W, Kotz R, Preusser P, Prindull G, Brandeis W, Landbeck G (1988) Neoadjuvant chemotherapy of osteosarcoma: results of a randomized cooperative trial (COSS-82) with salvage chemotherapy based on histological tumor response. J Clin Oncol 6:329-337
