ORIGINAL ARTICLE
Y. Oda . I. Röse . K. Radig . W. Wagemann U. Mittler . A. Roessner
Expression of MDR1/p-glycoprotein and multidrug resistance-associated protein in childhood solid tumours
Received: 6 May 1996 / Accepted: 3 July 1996
Abstract We evaluated the expression of MDR1/p-gly- coprotein in paediatric tumours using reverse transcrip- tase polymerase chain reaction (RT-PCR), RNA dot blot analysis, and immunohistochemistry on formalin fixed paraffin-embedded material with JSB-1 and C-219 monoclonal antibodies, and compared these three tech- niques. The expression of multidrug resistance-associat- ed protein (MRP) gene was examined by RT-PCR assay. We studied MDR1/p-glycoprotein and MRP expression in 13 samples from 10 neuroblastoma patients, 11 sam- ples from 10 nephroblastoma patients, 2 rhabdomyosar- comas, 1 adrenocortical carcinoma and 10 benign tu- mours or tumour-like lesions. Eleven of 13 neuroblasto- mas, 7 of 11 nephroblastomas, 2 rhabdomyosarcomas, 1 adrenocortical carcinoma, and 7 of 10 benign tumours or tumour-like lesions showed MDR1 PCR products. By RNA dot blot analysis, MDR1 transcripts were detect- able in 11 of 34 specimens. Immunohistochemically, we detected positive reaction products for JSB-1 in 26 of 36 samples. There was a significant correlation between the immunoreactivity for JSB-1 and the expression of MDR1 mRNA expression by RT-PCR (P=0.0001). How- ever, the presence of p-glycoprotein immunostaining does not correlate with the MDR1 expression shown by RT-PCR in every case. As for MRP mRNA expression, 9 of 13 neuroblastomas and 10 of 11 nephroblastomas re- vealed PCR products.
Y. Oda · I. Röse · K. Radig · A. Roessner ☒ X Department of Pathology, Otto-von-Guericke University Magdeburg, Leipziger Strasse 44, D-39120 Magdeburg, Germany Tel .: (+49) 391-67-15817, Fax: (+49) 391-67-15818
W. Wagemann Department of Pediatric Surgery, Otto-von-Guericke University Magdeburg, Germany
U. Mittler Department of Pediatric Oncology, Otto-von-Guericke University Magdeburg, Germany
Key words Multidrug resistance · P-glycoprotein · Neuroblastoma · Nephroblastoma · Reverse transcriptase polymerase chain reaction
Introduction
The human MDR1 gene, which encodes the drug efflux pump called p-glycoprotein [26, 36], has been shown to play a major role in the acquisition and maintenance of the multidrug resistance (MDR) phenotype in vitro [16, 31]. Sensitive molecular biological and immunohisto- chemical techniques have been developed to detect MDR1 mRNA/p-glycoprotein expression in clinical samples of paediatric tumours [3, 9, 22, 30, 32]. An in- crease in MDR1 transcript levels has already been re- ported for neuroblastomas by many authors using north- ern blotting or RNA dot blot analysis [4, 5, 7, 14, 23, 28]. To investigate further association of the expression of the MDR1 gene in childhood cancers with drug resis- tance, we studied the expression of MDR1/p-glycopro- tein in paediatric solid tumours using reverse transcrip- tase polymerase chain reaction (RT-PCR), RNA dot blot analysis and immunohistochemistry. To find the most sensitive and reliable tools for the assessment of MDR1/p-glycoprotein expression, these three methods were compared. The recently described multidrug resis- tance-associated protein (MRP) gene, identified by Cole et al. [12], encodes a novel membrane transport protein, the overexpression of which has been associated with a number of non-P-glycoprotein-mediated MDR pheno- types in vitro. Using an RT-PCR assay, the expression of this gene was also examined in this study.
Materials and methods
KB-C1, a drug resistant cell line, and the KB-3-1 parental epider- moid carcinoma cell line (kindly provided by Prof. M. Kuwano, First Department of Biochemistry, Kyushu University, Fukuoka, Japan) were used as standard material for quantification of MDR1 transcript level and p-glycoprotein expression by immunohisto-
| Gene | Length | Primer sequence | Thermal cycling |
|---|---|---|---|
| MDR1 | 167 bp | 5'-CCC ATC ATT GCA ATA GCA GG-3' 5'-GTT CAA ACT TCT CGT CCT GA-3' | 94°C (30 s) 55° C (1 min) 72° C (2 min) 33 cycles |
| MRP | 614 bp | 5'-CTG AGA AGG AGG CGC CCT G-3' 5'-GTG TCC GGA TGG TGG ACT G-3' | 94° C (1 min) 57° C (1 min) 72°C (2 min) 30 cycles |
| B2MG | 120 bp | 5'-ACC CCC ACT GAA AAA GAT GA-3' 5'-ATC TTC AAA CCT CCA TGA TG-3' | 94° C (30 s) 55° C (1 min) 72°C (2 min) 35 cycles |
chemistry. A multidrug resistant mutant, KB-C1, was selected with increasing concentration of colchicine, and was maintained in the presence of 1 ug/ml colchicine [1]. The KB-C1 cell line expressed 270-fold the MDR1 mRNA compared with KB-8, which expressed a minimal level of MDR1 as seen by slot-blot analysis [34].
We examined 13 samples from 10 neuroblastoma patients, 11 specimens from 10 nephroblastoma patients, 2 rhabdomyosarco- mas and 1 adrenocortical carcinoma. Furthermore, 10 benign tu- mours or tumour-like lesions were also investigated. For neuro- blastoma, histological subclassification was made according to Hughes et al. [24]. The clinical stages of the disease were deter- mined according to the system of Evans et al. [18]. For nephro- blastomas, histological grade and clinical stage were determined according to “Gesellschaft für Pädiatrische Onkologie und Hämatologie” [33] and “Societe Internationale d’Oncologie Pediatrique” [25], respectively. Tissues with necrosis, haemor- rhage, or calcification were excluded. Representative tissue was frozen in liquid nitrogen and stored at -80° C for RNA analysis. The remainder of the tissue was fixed in 4% neutral buffered for- malin for pathological analysis and immunostaining.
Total RNA was isolated from the tissue specimens and cell lines by the single extraction method as described by Chomczyn- ski and Sacchi [11]. Total RNA was also obtained from peripheral blood mononuclear cells (PBMC) isolated from freshly drawn heparinized normal human blood by centrifugation through Hist- opaque 1077 and washed with phosphate-buffered saline before lysis. Cellular RNA from PBMC of a healthy volunteer was used for positive control of MRP mRNA [2, 12]. Purity was confirmed on 1% agarose-formaldehyde gel, and quantitation was performed by spectrophotometry analysis.
For reverse transcriptase-polymerase chain reaction (RT-PCR), 1 µg purified cellular RNA was converted to single strand cDNA using a random primer and the Promega (Madison, Wis.) RT system according to the manufacturer’s instructions. PCR was per- formed in a programmable thermal cycler (PTC-100TM: MJ Re- search, Watertown, Mass.). The oligonucleotides used as ampli- mers in this study were synthesized by Pharmacia and conditions of thermal cycling were listed in Table 1. cDNA derived from 50 ng of RNA was incubated with 2.5U of Taq DNA polymerase (GI- BCO) in 50 ul containing 1.5 mM magnesium chloride (0.5 mM for MRP primers), 0.2 mM of each dNTPs and 1xPCR buffer (GI- BCO) and 37.5 pmol (25 pmol for MRP) gene specific amplimers. After amplification, 5 ul of each reaction was analysed by electro- phoresis on a 4% agarose gel (1% agarose + 3% NuSieve) stained with ethidium bromide, and quantified by laser densitometry. The intensity of each mRNA was standardized with beta 2-micro- globulin (B2MG) expression. The level of MDR1 was determined by comparison of the ratio of the intensities of the MDR1 and B2MG PCR products for each samples with the positive control KB-C1 cells (100 units). MDR1 values were estimated as follows: 0 unit, grade 0; < 10 units, grade 1; >10 units and <40 units, grade 2; >40 units, grade 3.
The level of MRP mRNA was classified as follows: grade 0, not observed; grade 1, the intensity of the amplified band was less than that of PBMC; grade 2, the intensity was stronger than PBMC.
In order to allow us to perform RNA dot blotting pHDR5A was a kind gift from Dr. M. Gottesman. It consists of a 1.4 kb in-
Rhabdomyosa. 1
Marker
KB-C1
KB-3-1
NB 1
NPB 7
300 bp -
200 bp
-167 bp (MDR 1)
100 bp -
-120 bp ( B2MG )
sert of MDR1 in transcription vector pGEM4 [37]. The plasmid was digested with PuvII which cuts midway in the MDR1 cDNA. Digoxigenin (dig)-labelled antisense RNA probe was transcribed using SP6 RNA polymerase with a dig-RNA-labelling kit (Boeh- ringer-Mannheim, Biochemica, Mannheim, Germany) according to the manufacturer’s instructions. Serial dilutions of 10, 5 and 1 ug of each sample of denatured total cellular RNA were spotted on the nylon membrane (Hybond N, Amersham) using a BRL Hy- bridot Manifold apparatus. Samples from both MDR1 negative KB-3-1 and positive KB-C1 cell lines were included on each blot. After fixation by UV-irradiation (UV Stratalinker 2400, Strata- gene), the filters were prehybridized for 4 h at 55° C in 50% form- amide, 5x standard saline citrate (SSC), 2% blocking reagent (Boehringer-Mannheim), 0.1% N-lauroylsarcosine and 0.02% so- dium dodecyl sulphate (SDS). The hybridization was carried out overnight at 55° C in the same solution containing dig-labelled MDR1 antisense RNA probe. After hybridization, the filters were washed with 2x SSC, 0.1% SDS twice for 10 min each at room temperature and then three times with 0.1x SSC, 0.1% SDS for 15 min each at 68° C. The detection of hybridization signals were carried out with DIG nucleic acid detection kit (Boehringer-Mann- heim) and chemiluminescent substrates: 1,2-dioxetane chemi- luminescent enzyme subtrate (CSPD) (Serva, Heidelberg, Germa- ny). Finally, the membrane was exposed to X-ray film and the hy- bridization signals were quantified by densitometer. The specifici- ty of the generated cRNA probe was confirmed by northern blot- ting. Values for MDR1 expression were determined by substrating
| Case number | Age/ Sex | Histological diagnosis | Prior CT | Gradeª | Stage | MDR1 RT-PCR | DB | PGP JSB-1 | IHC C-219 | MRP RT-PCR | Prognosis (m) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 9m/F | NB | No | 3 | IV-S | 3 | 3 | 1 | 0 | 2 | Alive (84) |
| 2 | 72m/F | NB | No | 3 | IV | 3 | 2 | 2 | 1 | 0 | Dead (12) |
| 2' | Recurrence | Yes | 3 | 1 | 2 | 1 | 0 | ||||
| 3 | 5m/F | NB | No | 3 | IV-S | 0 | 0 | NE | NE | 1 | Dead (11) |
| 3' | Recurrence | Yes | 2 | 0 | 0 | 0 | 1 | ||||
| 4 | 72m/F | NB | Yes | 2 | IV | 3 | 0 | 1 | 0 | 0 | Dead (7) |
| 5 | 12m/F | NB | Yes | 3 | III | 3 | 0 | 2 | 1 | 1 | Aive (72) |
| 6 | 36m/M | NB | No | 3 | III | 3 | 1 | 1 | 0 | 2 | Alive (16) |
| 7 | 10m/F | NB | No | 3 | II | 3 | 3 | 0 | NE | 1 | Alive (12) |
| 8 | 24m/F | NB | Yes | 3 | IV | 0 | 0 | 0 | 0 | 0 | Alive (5) |
| 9 | 5m/M | NB | No | 3 | 1 | 1 | 0 | 1 | 0 | 1 | Alive (2) |
| 10 | 5m/F | NB | Yes | 3 | IV | 1 | 0 | 2 | 1 | 1 | Dead (2) |
| 10' | Liver metastasis | Yes | 3 | 3 | 2 | 2 | 2 | ||||
| 11 | 96m/M | NPB | No | S | II | 2 | 0 | 2 | 1 | 2 | Alive (84) |
| 12 | 72m/M | NPB | No | S | IV | 2 | 0 | 1 | 1 | 2 | Alive (70) |
| 13 | 48m/M | NPB | No | S | II | 0 | 0 | 0 | 0 | 1 | Alive (63) |
| 14 | 47m/M | NPB | No | S | II | 0 | 0 | 0 | 0 | 1 | Dead (6) |
| 15 | 15m/M | NPB | Yes | H | III | 2 | 0 | 1 | 0 | 1 | Dead (18) |
| 15' | Recurrence | Yes | 0 | 0 | 0 | 0 | 1 | ||||
| 16 | 60m/F | NPB | Yes | S | I | 0 | 0 | 0 | 0 | 1 | Alive (54) |
| 17 | 42m/M | NPB | Yes | H | III | 2 | NE | 1 | 1 | 0 | Dead (8) |
| 18 | 48m/M | NPB | Yes | S | IV | 3 | 2 | 1 | 0 | 1 | Alive (24) |
| 19 | 12m/M | NPB | Yes | S | II | 3 | 2 | 1 | 1 | 2 | Alive (22) |
| 20 | 96m/M | NPB | Yes | S | III | 3 | 0 | 2 | 1 | 1 | Alive (24) |
| 21 | 3y/M | Rhabdomyosarcoma | 2 | NE | 2 | 1 | 1 | Dead (12) | |||
| 22 | 17y/F | Rhabdomyosarcoma | 1 | 0 | 1 | 0 | 1 | Alive (48) | |||
| 23 | 8y/F | Adrenocortical carcinoma | 3 | 1 | 2 | 1 | 2 | Dead (6) | |||
| 24 | 3y/F | Lipoblastoma | 0 | 0 | 0 | 0 | 1 | ||||
| 25 | 8m/M | Lipoblastoma | 0 | 0 | 0 | 0 | 2 | ||||
| 26 | 10d/M | Haemangioendothelioma | 1 | 0 | 0 | 0 | 1 | ||||
| 27 | 15y/M | Haemangioendothelioma | 2 | 0 | 1 | 1 | 1 | ||||
| 28 | 14y/M | Paraganglioma | 3 | 3 | 1 | 1 | 2 | ||||
| 29 | 14y/M | Ganglioneuroma | 2 | 1 | 2 | 2 | 1 | ||||
| 30 | 7y/F | Neurofibroma | 1 | NE | 2 | 1 | 0 | ||||
| 31 | 3y/F | Plasma cell granuloma | 1 | 0 | 1 | 0 | 2 | ||||
| 32 | 5y/M | Plasma cell granuloma | 1 | 0 | 1 | 0 | 1 | ||||
| 33 | 5y/F | Eosinophilic granuloma | 0 | 0 | 1 | 0 | 0 |
a For neuroblastomas the grade is given according to Hughes [24]. For nephroblastomas the grade in that according to Gesellschaft für Pädiatrische Onkologie und Hämatologie [33] with S indicating standard risk and H indicating high risk
the value obtained in the negative control line KB-3-1, and then normalizing all samples to a value of 100 units for 10 µg total RNA of the positive control line KB-C1. The negative signals were scored as grade 0. The value less than 3 units was considered to be a low expression, grade 1; 3-10 units intermediate, grade 2; more than 10 units high, grade 3.
For immunohistochemistry formalin-fixed, paraffin-embedded pellets of KB-3-1 and KB-C1 cell lines were used as controls. Sec- tions were cut at 4 um from paraffin-embedded samples and pellets of cell lines and were dewaxed. After inhibition of endogenous per- oxidase, tissue sections were exposed to JSB-1 monoclonal anti- body (Crawly, England; dilution 1:20) and C-219 murine monoclo- nal antibody (Centocor, Malvern, Pa .; dilution 1:5) for 16 h at 4º C. A secondary horse anti-mouse antibody conjugated with biotin was followed by streptavidin. The colour reaction was developed in di- aminobenzidine solution for 10 min and sections were counter- stained with haematoxylin. A section without the primary antibody served as a negative control. The slides were reviewed by three pa- thologists independent of other experimental results and scored as negative, grade 0; < 50% of cells showing staining, grade 1; >50% of cells showing immunoreactivity, grade 2.
The correlation between RT-PCR and immunohistochemistry in MDR1 expression was evaluated by chi-square test. For testing the effect of MDR1 and MRP expression on the prognosis, Ka- plan-Meier plots were constructed and Wilcoxon-test was per- formed. A P value less than 0.05 was considered significant.
Results
RT-PCR for MDR1
The feasibility of RT-PCR for MDR1 was confirmed with the cell lines. The drug sensitive KB-3-1 showed no MDR1 product, whereas a strong positive signal was ob- tained with the drug resistant cell line KB-C1 (Fig. 1). Out of the 37 specimens, 28 (76%) had positive PCR products. For neuroblastomas, 11 of 13 specimens showed MDR1 expression (Table 2 and Fig. 2A). Four of five chemotherapy free samples revealed MDR1 expres-
NB 2 ( Post - Chemo Tx )
NB 2 ( Pre - Chemo Tx )
NB 10 Liver meta.
KB-C1
KB-3-1
NB 6
NB 4
NB 5
NB 1
NB 8
NB 10 Ad.
NB 7
NB 9
GN
A
B
MDR1
B2 MG
NPB 9
NPB 10
NPB 8
NPB 5 Primary
NPB 5 Recurrence
NPB 6
Rhabdomyosa. 1
Rhabdomyosa. 2
Paraganglioma
Hemangioendothelioma Hemangiopericytoma
Eosinophilic granuloma
Plasma cell granuloma
Adrenocartical ca.
MDR1
B2 MG
100
80
Survival rate (%)
60
11
40
MDR1+ (n=15)
20
MDR1- (n=5)
0
0
1
2
3
4
5
6
7
Years after diagnosis
NB 10 Liver meta.
KB-C1
KB-3-1
Paraganglioma
NPB 9
28 S
- 4.5 kb ( MDR1)
28 S -
18 S -
A
KB-3-1
Paraganglioma
NB 10 Liver meta.
KB-C1
NB 7
NPB 9
10 μg
☒
☒
☒
☒
☒
5 µg
☒
☒
☒
☒
☒
1 µg
☒
☒
☒
B
☒
☒
sion initially. Two cases had matched tumours before and after chemotherapy. In case 3, MDR1 expression was recognized only after chemotherapy but in case 2, both pre and post chemotherapy specimens expressed MDR1 at high level.
In the nephroblastomas, 7 of 11 samples revealed MDR1 expression at intermediate or high levels (Table 2 and Fig. 2B). Two of four untreated tumours showed MDR1 expression. There was no significant relationship between MDR1 expression. There was no significant re- lationship between MDR1 expression and prognosis in
A
B
our neuroblastoma and nephroblastoma series (Fig. 3: P=0.131). In other tumours, one adrenocortical carcino- ma and one paraganglioma showed high MDR1 expres- sion, compared with other types of tumour or tumour- like lesions (Table 2 and Fig. 2B).
Dot blot analysis of MDR1 expression
The specificity of cRNA probe and dot blot analysis was verified by northern blotting (Fig. 4A). We examined 34 specimens and MDR1 transcripts were detectable in 11 samples (32%). Among 11 positive specimens, 6 were from 13 neuroblastoma specimens, 2 from 10 nephro- blastomas and 3 from 11 other tumours or tumour-like lesions (Table 2 and Fig. 4B). Out of these 11 dot blot positive cases, 10 showed a high level of MDR1 expres- sion by RT-PCR and 1 showed an intermediate level. There were no dot blot positive cases in which RT-PCR showed no MDR1 gene product.
Immunohistochemical detection of p-glycoprotein
Paraffin-embedded cell pellets were used as control. Drug resistant KB-C1 shows positive reaction products in the plasma membrane, while parental KB-3-1 revealed no immunoreactivity for JSB-1 or C-219 (Fig. 5A). Of 36 specimens, 26 (72%; 6/13 neuroblastomas, 7/11 nephroblastomas, 2/2 rhabdomyosarcomas, 1/1 adreno- cortical carcinoma and 7/10 benign tumours or tumour- like lesions) showed positive reaction for JSB-1 antibody (Table 2 and Fig. 5B). However, 16 of 35 (46%; 5/11 neuroblastomas, 5/11 nephroblastomas, 1/2 rhabdomyo- sarcomas, 1/1 adrenocortical carcinoma and 4/10 benign
| Number of samples | RT-PCR | DB | IHC |
|---|---|---|---|
| 10 | + | + | + |
| 1 | + | + | - |
| 12 | + | - | + |
| 2 | + | - | - |
| 1 | - | - | + |
| 7 | - | - | - |
tumours or tumour-like lesions) revealed immunoreactiv- ity for C-219 (Table 2). Concordance between RT-PCR, dot blot and immunohistochemistry with JSB-1 was found in 17 samples (47; Table 3). Nevertheless, concor- dance between the positivity of MDR1 expression by RT-PCR and p-glycoprotein expression by immunohisto- chemistry with JSB-1 was recognized in 33 (92%) of 36 samples (Table 2). Two neuroblastomas and one haeman- gioendothelioma showed negative reaction for JSB-1 de- spite the positivity of RT-PCR. There was a significant correlation between the immunoreactivity for JSB-1 and the expression of MDR1 mRNA by RT-PCR (chi-square test: P=0.0001).
RT-PCR for MRP
Of 37 samples, 31 (84%) showed MRP positive prod- ucts. In the neuroblastomas, 9 of 13 samples revealed MRP expression (Fig. 6). In two matched neuroblasto- mas, no change was recognized at the levels of MRP be- fore and after chemotherapy. In the nephroblastomas, 10 of 11 specimens had positive products. There was no sig-
NB 2 ( Post-Chemo Tx )
NB 2 ( Pre-Chemo Tx )
KB-C1
PBMC
NB 6
NB 4
NB 5
NB 8
NB 10 Ad.
NB 10 Liver meta.
NB 7
NB 9
GN
MRP
B2 MG
nificant relationship between MRP expression and prog- nosis in neuroblastomas or nephroblastomas (P=0.473).
Discussion
Recent studies have revealed high levels of MDR1/p-gly- coprotein in some normal tissue including kidney, colon, adrenal gland and liver [19, 29]. The tumour arising from these tissues also shows high levels of MDR1 [22]. Ex- pression of MDR1/p-glycoprotein has also been docu- mented in neuroblastoma [3, 32]. Two authors [7, 23] us- ing RNA dot blot analysis on neuroblastoma samples showed that a significantly greater number of treated ver- sus untreated cases expressed higher levels of MDR1 mRNA, suggesting that the acquisition of a multidrug re- sistant phenotype is related to chemotherapy. In our study, there was no significant correlation between MDR1 mRNA expression and therapy status or progno- sis, possibly due to the small number of cases.
For the evaluation of MDR1 mRNA expression, Bro- phy et al. [8] recommended RT-PCR in comparison with RNA dot blot analysis, in situ hybridization and immunohistochemistry, because of its relative simplici- ty and specificity. Apart from Bordow et al. [6], no in- vestigator has studied the MDR1 expression in paedi- atric tumours using RT-PCR. In this study, 28 of 37 specimens (76%) showed MDR1 expression by RT- PCR, while 11 of 34 specimens (32%) revealed MDR1 mRNA expression by dot blot analysis even when em- ploying a cRNA probe which is considered to be more sensitive than a cDNA probe. We could not detect the MDR1 transcript by dot blot in 14 samples which showed positive PCR product. Therefore, we consider that RT-PCR is more sensitive than dot blot analysis to detect MDR1 mRNA.
Generally, nephroblastoma is a highly chemorespon- sive tumour [15], and infrequent MDR1 expression was previously demonstrated by RNA dot blot analysis [22]. However, 7 of 11 specimens revealed MDR1 expression by RT-PCR in the current study, whereas the MDR1 tran- script was demonstrated in only 2 specimens by RNA dot blot analysis. Our results suggest that nephroblasto-
mas of childhood are not necessarily all chemorespon- sive tumours and the clinical significance of a low level of MDR1 expression should be investigated within large systematically-treated patient groups by employing sen- sitive techniques such as RT-PCR.
In this study, formalin-fixed paraffin-embedded clini- cal materials were examined for the presence of p-glyco- protein by immunohistochemical staining using two monoclonal antibodies, JSB-1 and C-219. We tried to use C-219 for paraffin-embedded sections, but the immu- noreactivity was weaker than JSB-1. Kandel et al. [27] described that C-219 antibody from a different lot num- ber decreased the number of cases which showed posi- tive immunostaining. Furthermore, Cordon-Cardo et al. [13] recommended using the C-219 antibody only for frozen sections because the determinants recognized by C-219 are irreversibly masked or denatured through for- malin fixation and processing. Several investigators have pointed out that applying immunohistochemistry alone is insufficient and must be supplemented by additional as- says such as northern blotting and quantitative RT-PCR [17]. Therefore, careful interpretation of the immunohis- tochemistry of p-glycoprotein should be done.
MRP, a gene recently isolated from a non-p-glycopro- tein-mediated multidrug-resistant small cell lung cancer, is a candidate multidrug-resistance gene [12]. It is con- sidered to encode a 190 kDa transporter membrane pro- tein [2, 12]. It is still unclear whether MRP mRNA ex- pression is associated with drug resistance in vitro. Some authors have found no correlation between MRP expres- sion and multidrug resistance in vitro [10, 20]. In the in vivo study, however, some authors have found a contri- bution of MRP mRNA expression to the multidrug resis- tance of haematopoietic neoplasms [21] and thyroid can- cer [35]. Recently, Bordow et al. [6] demonstrated the MRP mRNA expression in 5 human neuroblastoma cell lines and in 25 primary neuroblastomas. In our study, 9 of 13 neuroblastomas (69%) and 10 of 11 nephroblasto- mas (90%) showed MRP expression.
This study confirms the usefulness of RT-PCR to de- tect the MDR1 mRNA in paediatric tumours. The clini- cal correlation with MDR1 expression shown by RT- PCR should be demonstrated in a large number of sys- tematically-treated patient groups.
Acknowledgements The authors are grateful to Dr. Michael Got- tesman, the Laboratory of Molecular Biology, National Institute of Health, Bethesda, Md., for providing the probe pHDR5A. Also, we thank Prof. Michihiko Kuwano, First Department of Biochem- istry, Kyushu University, Fukuoka, Japan for the donation of KB- 3-1 and KB-C1 cell lines. We thank Mr. B. Wusthoff for editing the manuscript. Yoshinao Oda is granted an Alexander-von-Hum- boldt Fellowship. This study was supported by Deutsche For- schungsgemeinschaft.
References
1. Akiyama S, Fojo A, Hanover JA, Pastan I, Gottesman MM (1985) Isolation and genetic characterization of human KB cell lines resistant to multiple drugs. Somatic Cell Genet 11:117-126
2. Barrand MA, Heppel-Parton AC, Wright KA, Rabbitts PH, Twentyman PR (1994) A 190-kilodalton protein overexpres- sion in non-p-glycoprotein-containing multidrug-resistant cells and its relationship to the MRP gene. J Natl Cancer Int 86: 110-117
3. Bates SE, Shieh CY, Tsokos M (1991) Expression of mdr1/P- glycoprotein in human neuroblastoma. Am J Pathol 139:305- 315
4. Benard J, Bourhis J, Riou G (1991) Activation of N-myc on- cogene and associated genes to chemoresistance, prognostic value for neuroblastoma. Bull Cancer (Paris) 78:91-97
5. Benard J, Bourhis J, de-Vathaire F, Ferrandis E, Terrier-La- combe MJ, Lemerle J, Riou G, Hartmann O (1994) Prognostic value of MDR1 gene expression in neuroblastoma: results of a multivariante analysis. Prog Clin Biol Res 385:111-16
6. Bordow SB, Haber M, Madafiglio J, Cheung B, Marshall GM, Norris MD (1994) Expression of the multidrug resistance-as- sociated protein (MRP) gene correlates with amplification and overexpression of the N-myc oncogene in childhood neuro- blastoma. Cancer Res 54:5036-5040
7. Bourhis J, Benard J, Hartmann O, Boccon-Gibod L, Lemerle J, Riou G (1989) Correlation of MDR1 gene expression with chemotherapy in neuroblastoma. J Natl Cancer Inst 81:1401- 1405
8. Brophy NA, Marie JP, Rojas VA, Warnke RA, McFall PJ, Smith SD, Sikic BI (1994) Mdr1 gene expression in childhood acute lymphoblastic leukemias and lymphomas: A critical evaluation by four techniques. Leukemia 8:327-335
9. Chan HSL, Haddad G, Thorner PS, DeBoer G, Lin YP, Ond- rusek N, Yeger H, Ling V (1990) P-glycoprotein expression as a predictor of the outcome of therapy for neuroblastoma. N Engl J Med 325:1608-1614
10. Chen G, Jaffrezou J-P, Fleming WH, Duran GE, Sikic BI (1994) Prevalence of multidrug resistance related to activation of the mdr1 gene in human sarcoma mutants derived by sin- gle-step doxorubicin selection. Cancer Res 54:4980-4987
11. Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159
12. Cole SPC, Bhardwaj G, Gerlach JH, Mackie JE, Grant CE, AI- mquist KC, Stewärt AJ, Kurz EU, Duncan AMV, Deeley RG (1992) Overexpression of a transporter gene in a multidrug-re- sistant human lung cancer cell line. Science 258:1650-1654
13. Cordon-Cardo C, O’Brien JP, Boccia J, Casals D, Bertino Jr, Melamed MR (1990) Expression of the multidrug resistance gene product (p-glycoprotein) in human normal and tumor tis- sues. J Histochem Cytochem 38:1277-1287
14. Corrias MV, Cornaglia-Ferraris P, Di-Martino D, Stenger AM, Lanino E, Boni L, Tonini GP (1990) Expression of multiple drug resistance gene, MDR1, and N-myc oncogene in Italian population of human neuroblastoma patients. Anticancer Res 10:897-902
15. D’Angio GJ, Breslow N, Beckwith B, Evans A, Baum E, De- Limier A, Fernbach D, Hrabovsky E, Jones B, Kelalis P, Oth- ersen B, Tefft M, Thomas PRM (1989) Treatment of Wilms’ tumor: Results of the third national Wilms’ tumor study. Can- cer 64:349-360
16. Dietel M (1991) What’s new in cytostatic drug resistance and pathology. Pathol Res Pract 187:892-905
17. Dietel M (1993) Meeting report. Second international sympo- sium on cytostatic drug resistance. Cancer Res 53:2683-2688
18. Evans AE, D’Andio GJ, Randolph JA (1971) A proposed stag- ing for children with neuroblastoma. Cancer 27:374-378
19. Fojo AT, Ueda K, Slamon DJ, Poplack DG, Gottesman MM, Pastan I (1987) Expression of a multidrug-resistance gene in human tumors and tissues. Proc Natl Acad Sci USA 84:265- 269
20. Futscher BW, Abbaszadegan Mr, Domann F, Dalton WS (1994) Analysis of MRP mRNA in mitoxantrone-selected, multidrug-resistant human tumor cells. Biochem Pharmacol 47:1601-1606
21. Gekeler V, Beck J, Noller A, Wilisch A, Frese G, Neumann M, Handgretinger R, Ehninger G, Probst H, Niethammer D (1994) Drug-induced changes in the expression of MDR-asso- ciated genes: investigations on cultured cell lines and chemo- therapeutically treated leukemias. Ann Hematol [Suppl] 69: 519-524
22. Goldstein LJ, Galski H, Fojo A, Willingham M, Lai S-L, Gaz- dar A, Priker R, Green A, Crist W, Brodeur GM, Lieber M, Cossman J, Gottesman MM, Pastan I (1989) Expression of a multidrug resistance gene in human cancers. J Natl Cancer Inst 81:116-124
23. Goldstein LJ, Fojo AT, Ueda K, Crist W, Green A, Brodeur G, Pastan I, Gottesman MM (1990) Expression of the multidrug resistance, MDR1, gene in neuroblastoma. J Clin Oncol 8: 128-136
24. Hughes M, Marsden HB, Palmer MK (1974) Histologic pat- terns of neuroblastoma related to prognosis and clinical stag- ing. Cancer 34:1706-1711
25. Jereb B, Tournade MF, Jemerle J, Voute PA, Delemarre JF, Ahstrom L, Flamant R, Gerard-Marchant R, Sandstedt B (1980) Lymph node invasion and prognosis in nephroblasto- ma. Cancer 45:1632-1636
26. Juliano R, Ling A (1976) A surface glycoprotein modulation drug permeability in Chinese hamster ovarian cell mutants. Biochim Biophys Acta 455:152-162
27. Kandel RA, Campbell S, Noble-Topham S, Bell R, Andrulis IL (1995) Correlation of p-glycoprotein by immunohistochem- istry with mdr-1 mRNA levels in osteosarcomas. Pilot study. Diagn Mol Pathol 4:59-65
28. Nakagawara A, Kadomatsu K, Sato S, Kohno K, Takano H, Akazawa K, Nose Y, Kuwano M (1990) Inverse correlation between expression of multidrug resistance gene and N-myc oncogene in human neuroblastoma. Cancer Res 50:3043- 3047
29. Noonan K, Beck C, Holtzmayer T, Chin JE, Wunder JS, Andr- ulis IL, Gazdar AF, Willman CL, Griffith B, Von Hoff DD, Roninson IB (1990) Quantitative analysis of MDR1 (multi- drug resistance) gene expression in human tumors by polymer- ase chain reaction. Proc Natl Acad Sci USA 87:7160-7164
30. McDowell H, Peuchmaur M, Dominici C, Flamant F (1993) Multidrug resistance gene transcript level, and p-glycoprotein expression in paediatric malignant mesenchymal tumours. An- ticancer Res 13:1863-1866
31. Pastan I, Gottesman M (1987) Multidrug resistance in human cancer. N Engl J Med 316:1388-1393
32. Ramani P, Dewchand H (1995) Expression of mdr1/P-glyco- protein and p110 in neuroblastoma. J Pathol 175:13-22
33. Schmidt D, Harms D (1983) Histology and prognosis of neph- roblastoma - with particular reference to special variants. Klin Padiatr 195:214-221
34. Shen DW, Fojo A, Chin JE, Roninson IB, Richert N, Pastan I, Gottesman MM (1986) Human multidrug-resistant cell lines: Increased mdr1 expression can precede gene amplification. Science 232:643-645
35. Sugawara I, Arai T, Yamashita T, Yoshida A, Masunaga A, Itoyama S (1994) Expression of multidrug resistance-associat- ed protein (MRP) in anaplastic carcinoma of the thyroid. Can- cer Lett 82:185-188
36. Ueda K, Cornwell MM, Gottesman MM, Pastan I, Roninson IB, Ling V, Riordan JR (1986) The mdr1 gene, responsible for multidrug-resistance, codes for P-glycoprotein. Biochem Bio- phys Res Commun 141:956-962
37. Ueda K, Clark DP, Chen C, Roninson IB, Gottesman MM, Pa- stan I (1987) The human multidrug resistance (mdr1) gene: cDNA cloning and tanscription initiation. J Biol Chem 262:505-508