Mechanisms of Filovirus Entry.

Filovirus entry into cells is complex, perhaps as complex as any viral entry mechanism identified to date. However, over the past 10 years, the important events required for filoviruses to enter into the endosomal compartment and fuse with vesicular membranes have been elucidated (Fig. 1). Here, we highlight the important steps that are required for productive entry of filoviruses into mammalian cells.

Female mice haploinsufficient for an inactivated androgen receptor (AR) exhibit age-dependent defects that resemble the AR null phenotype of dysfunctional late follicle development, ovulation, and fertility.

The role of classical genomic androgen receptor (AR) mediated actions in female reproductive physiology remains unclear. Female mice homozygous for an in-frame deletion of exon 3 of the Ar (AR(-/-)) were subfertile, exhibiting delayed production of their first litter (AR(+/+) = 22 d vs. AR(-/-) = 61 d, P < 0.05) and producing 60% fewer pups/litter (AR(+/+): 8.1 +/- 0.4 vs. AR(-/-): 3.2 +/- 0.9, P < 0.01). Heterozygous females (AR(+/-)) exhibited an age-dependent 55% reduction (P < 0.01) in pups per litter, evident from 6 months of age (P < 0.05), compared with AR(+/+), indicating a significant gene dosage effect on female fertility. Ovulation was defective with a significant reduction in corpora lutea numbers (48-79%, P < 0.01) in 10- to 12- and 26-wk-old AR(+/-) and AR(-/-) females and a 57% reduction in oocytes recovered from naturally mated AR(-/-) females (AR(+/+): 9.8 +/- 1.0 vs. AR(-/-): 4.2 +/- 1.2, P < 0.01); however, early embryo development to the two-cell stage was unaltered. The delay in first litter, reduction in natural ovulation rate, and aromatase expression in AR(+/-) and AR(-/-) ovaries, coupled with the restored ovulation rate by gonadotropin hyperstimulation in AR(-/-) females, suggest aberrant gonadotropin regulation. A 2.7-fold increase (AR(+/+): 35.4 +/- 13.4 vs. AR(-/-): 93.9 +/- 6.1, P < 0.01) in morphologically unhealthy antral follicles demonstrated deficiencies in late follicular development, although growing follicle populations and growth rates were unaltered. This novel model reveals that classical genomic AR action is critical for normal ovarian function, although not for follicle depletion and that haploinsufficiency for an inactivated AR may contribute to a premature reduction in female fecundity.

Constitutively active CCK2 receptor splice variant increases Src-dependent HIF-1 alpha expression and tumor growth.

Gastrointestinal (GI) cancers ectopically express multiple splice variants of the cholecystokinin-2 (CCK(2))/gastrin receptor; however, their relative contributions to the cancer phenotype are unknown. The aim of this study was to compare the effects of CCK(2) receptor (CCK(2)R) and CCK(2i4sv)R expression on cell growth both in vitro and in vivo using a human epithelial cell model, HEK239. In vitro, receptor variant expression did not affect cell proliferation either in the absence or presence of agonist. However, in vivo, the expression of CCK(2i4sv)R, but not CCK(2)R, increases HEK293 tumor growth in a constitutive, Src-dependent manner. Enhanced tumorigenicity of CCK(2i4sv)R is associated with an Src-dependent increase in the transcription factor, hypoxia-inducible factor-1alpha, its downstream target, vascular endothelial growth factor and tumor micro-vessel density, suggesting that CCK(2i4sv)R may contribute to the growth and spread of GI cancers through agonist-independent mechanisms that enhance tumor angiogenesis.

Effects of amylin deficiency on trabecular bone in young mice are sex-dependent.

Amylin deficiency in mice results in late-onset osteopenia. Sex differences have been identified in insulin secretion in Amylin-overexpressing transgenic mice, suggesting a possible interaction of sex steroids, growth factors, or cytokines and amylin. The aim of the current study was to compare the effects of amylin deficiency on bone in young and adult male and female mice. The metaphyses of the distal femora from male and female Amylin-deficient mice at 4, 6, and 26 weeks of age were assessed by bone histomorphometry. Femoral length was increased in Amylin-deficient male mice compared to wild-type (WT) mice at 26 weeks of age (P < 0.005) but not in females. This was associated with an increase in growth plate height in Amylin-deficient males at 4 (P < 0.01) and 6 (P < 0.05) weeks of age. Furthermore, young Amylin-deficient males had decreased trabecular number at 4 weeks of age (P < 0.05) and increased trabecular thickness at 4 and 6 weeks of age (P < 0.05) compared to WT mice, with no net change in trabecular bone volume. These effects of amylin deficiency were not observed in female mice. In conclusion, this study demonstrates that amylin deficiency exerts effects on bone during growth that are sex-dependent and suggest a possible interaction between amylin and testosterone, growth factors, or cytokines to regulate bone cell metabolism.

Genomic actions of the androgen receptor are required for normal male sexual differentiation in a mouse model.

Androgens mediate their effects in target cells via the androgen receptor (AR), which acts predominantly as a ligand-dependent transcription factor. In addition, androgens induce rapid activation of second messenger signal transduction cascades, and this is thought to occur via non-genomic mechanisms. We have used the Cre/loxP system to generate an AR knockout (ARKO) mouse targeting exon 3, which encodes the second zinc finger of the DNA-binding domain. To generate universal ARKO mice, floxed AR mice were mated with CMV-Cre mice, which express Cre recombinase ubiquitously. Deletion of the floxed allele in our mice does not disrupt the reading frame, and has been designed so that the mutant AR can bind ligand but not target genes. ARKO males displayed a complete androgen insensitivity phenotype, with female external genitalia and a reduction in body weight compared with wild-type males (P < 0.001). Testes of ARKO males were smaller than control males (P < 0.0001) and were located intra-abdominally. We have demonstrated that genotypically XY mice lacking the second zinc finger of the AR have a female phenotype, and we conclude that the genomic actions of the AR (mediated by DNA binding) are indispensable for normal male sexual differentiation.

Transgenic mice that express Cre recombinase in osteoclasts.

To study the physiological control of osteoclasts, the bone resorbing cells, we generated transgenic mice carrying the Cre recombinase gene driven by either the tartrate-resistant acid phosphatase (TRAP) or cathepsin K (Ctsk) promoters. TRAP-Cre and Ctsk-Cre transgenic mouse lines were characterized by breeding with LacZ ROSA 26 (R26R) reporter mice and immunohistochemistry for Cre recombinase. The Cre transgene was functional in all lines, with Cre-mediated recombination occurring primarily in the long bones, vertebrae, ribs, and calvaria. Histological analyses of the bones demonstrated that functional Cre protein was present in 1) osteoclasts (Ctsk-Cre); 2) osteoclasts, columnar proliferating, and hypertrophic chondrocytes (TRAP-Cre line 4); and 3) round proliferating chondrocytes (TRAP-Cre line 3). In conclusion, we generated transgenic mouse lines that will enable the deletion of floxed target genes in osteoclasts, which will be valuable tools for studying the regulation of osteoclast function.

Local secretion of parathyroid hormone-related protein by an osteoblastic osteosarcoma (UMR 106-01) cell line results in growth inhibition.

Parathyroid hormone-related protein (PTHrP) has been implicated as being important in the growth of tumor cells responsive to the peptide. We utilized a rat osteoblastic osteosarcoma cell line, UMR 106-01, which has PTHrP receptors and a PTHrP-responsive adenylate cyclase/cAMP messenger system, to produce a modified cell line that overexpresses PTHrP. The human PTHrP cDNA sequence was transfected by electroporation into UMR 106-01 cells and the stable cell lines UMR-36 and UMR-34 were established. The modified cell line, UMR-36, had increased levels of PTHrP mRNA compared with control cell lines and secreted PTHrP into the culture medium at levels of 0.01-0.1 pmol/10(7) cells in 12 h. The secreted peptide was biologically active as indicated by its ability to activate adenylate cyclase. The number of UMR-36 cells following 9 days in culture was reduced by up to 80% compared with control lines, which was associated with decreased (3)H-thymidine incorporation into genomic DNA. Addition of 1000-fold excess of the PTHrP antagonist, PTHrP(7-34), to UMR-36 cells resulted in the escape of growth inhibition and increased rate of growth. In vivo, tumors derived from UMR-36 cells were smaller in size compared with tumors derived from control cells. In conclusion, increased autocrine secretion of, and responsiveness to, PTHrP results in inhibited growth kinetics of an osteoblast-like bone tumor cell line in vitro and in vivo.

P-Glycoprotein and transporter MRP1 reduce HIV protease inhibitor uptake in CD4 cells: potential for accelerated viral drug resistance?

BACKGROUND: The multidrug transporters P-glycoprotein (P-gp) and MRP1 are functionally expressed in several subclasses of lymphocytes. HIV-1 protease inhibitors interact with both; consequently the transporters could reduce the local concentration of HIV-1 protease inhibitors and, thus, influence the selection of viral mutants. OBJECTIVES: To study the effect of the expression of P-gp and MRP1 on the transport and accumulation of HIV-1 protease inhibitors in human lymphocytes and to study the effects of specific P-gp and MRP1 inhibitors. METHODS: The initial rate and the steady-state intracellular accumulation of radiolabelled ritonavir, indinavir, saquinavir and nelfinavir was measured in three human lymphocyte cell lines: control CEM cells, CEM-MDR cells, which express 30-fold more P-gp than CEM cells, and CEM-MRP cells, which express fivefold more MRP1 protein than CEM cells. The effect of specific inhibitors of P-gp (GF 120918) and MRP1 (MK 571) was also examined. RESULTS: Compared with CEM cells, the initial rates of uptake and the steady-state intracellular concentrations of all protease inhibitors are significantly reduced in CEM-MDR cells. The intracellular concentrations of the protease inhibitors are increased upon co-administration with GF 120918, in some cases to levels approaching those in CEM cells. The intracellular concentrations of the protease inhibitors are also significantly reduced in CEM-MRP cells. Co-administration with MK -571 can partially overcome these effects. CONCLUSIONS: The overexpression of multidrug transporters significantly reduces the accumulation of protease inhibitors at this major site of virus replication, which, potentially, could accelerate the acquisition of viral resistance. Targeted inhibition of P-gp may represent an important strategy by which this problem can be overcome.

Verapamil-stimulated glutathione transport by the multidrug resistance-associated protein (MRP1) in leukaemia cells.

Multidrug resistance mediated by the multidrug resistance-associated protein MRP1 is associated with decreased drug accumulation, which is in turn dependent on cellular glutathione. We have reported that verapamil, an inhibitor of drug transport, caused a decrease in cellular glutathione in CCRF-CEM/E1000 MRP1-overexpressing leukaemia cells (Biochem Pharmacol 55;1283--9, 1998). We now demonstrate that other inhibitors of MRP1-mediated drug transport (e.g. MK571, indomethacin, genistein, and nifedipine) deplete cellular glutathione in these leukaemia cells (>30% decrease; P < 0.01) while having no effect on the parental CCRF-CEM cells. However, treatment with etoposide or vincristine (at similar molar concentrations) caused a 20% decrease in glutathione. Verapamil-stimulated glutathione transport correlated with MRP1 expression in a series of drug-resistant cells, and glutathione was quantitatively recovered in the extracellular media. Further, verapamil-stimulated glutathione transport was rapid (50% decrease in 10 min), dose-dependent, and inhibited by vanadate, an inhibitor of ATPase activity, but not by sulphobromophthalein (BSP) or methionine, inhibitors of hepatic glutathione transporters. Incubation of CCRF-CEM/E1000 cells in 25 mM glutathione not only showed that verapamil-mediated efflux occurred against the concentration gradient, but also demonstrated the MRP1-mediated uptake of glutathione (P < 0.01 compared to the parental CCRF-CEM cells), which was not inhibited by vanadate. These results demonstrate that while MRP1 transports glutathione in the presence of inhibitors of drug transport, there is no convincing evidence for co-transport of glutathione with drug. They further demonstrate that MRP1 mediates the facilitated transport of glutathione into the MRP1-overexpressing CEM/E1000 cells, suggesting that MRP1 may play a major role in cellular glutathione homeostasis.

Modular organization of the Friend murine leukemia virus envelope protein underlies the mechanism of infection.

Retrovirus infection is initiated by receptor-dependent fusion of the envelope to the cell membrane. The modular organization of the envelope protein of C type retroviruses has been exploited to investigate how binding of the surface subunit (SU) to receptor triggers fusion mediated by the transmembrane (TM) subunit. We show that deletion of the receptor-binding domain (RBD) from SU of Friend murine leukemia virus (Fr-MLV) abolishes infection that is restored by supplying RBD as a soluble protein. Infection by this mechanism remains dependent on receptor expression. When membrane attachment of the virus lacking RBD is reestablished by inserting the hormone erythropoietin, infection remains dependent on the RBD/receptor complex. However, infection increases 50-fold to 5 x 10(5) units/ml on cells that also express the erythropoietin receptor. Soluble RBD from Fr-MLV also restores infection by amphotropic and xenotropic MLVs in which RBD is deleted. These experiments demonstrate that RBD has two functions: mediating virus attachment and activating the fusion mechanism. In addition, they indicate that receptor engagement triggers fusion by promoting a subgroup-independent functional interaction between RBD and the remainder of SU and/or TM.

Drug resistance does not correlate with resistance to Fas-mediated apoptosis.

Recent reports have correlated multidrug resistance (MDR) and P-glycoprotein expression with decreased Fas expression and resistance to Fas-mediated apoptosis. We report the MRP-overexpressing MDR subline CEM/E1000 has the same Fas expression (MFI 74.3 +/- 0.7) as the parental CCRF-CEM T-cell leukaemia cells (MFI 70.0 +/- 3.6; P>0.05), and that the level of apoptosis induced by anti-Fas antibody or drug was similar in both cell lines. Further the P-glycoprotein-expressing CEM/VLB(100) subline of the CCRF-CEM cells showed increased Fas expression (MFI 114.8 +/- 3.6; P<0.001) and no resistance to Fas-mediated apoptosis. This questions the hypothesis that selection of drug resistance results in resistance to Fas-mediated apoptosis, with important implications for the rational use of immunotherapy in the treatment of drug resistant cancer.

Osteoblast gene expression in rat long bones: effects of ovariectomy and dihydrotestosterone on mRNA levels.

The steroid sex hormones exert major effects on bone formation although the molecular events associated with their activity remain unclear. We have investigated the effects of ovariectomy and dihydrotestosterone (DHT) administration to both sham-operated and ovariectomized (ovx) rats on the bone mRNA levels of osteoblast genes. Rats were randomly allocated to either sham or ovariectomy operations and were administered either vehicle or 40 mg/ kg body weight DHT by silastic tube implants at the time of operation for 8 weeks, at which time they were killed and total RNA was extracted from the long bones. Northern blot analysis indicated that the mRNA levels of the bone cell genes alpha1(I) collagen, alkaline phosphatase, osteocalcin, and osteopontin were markedly increased in ovx rats between 6- and 30-fold. DHT administration to ovary-intact, estrogen-sufficient rats increased the mRNA levels of alpha1(I) collagen, alkaline phosphatase, osteopontin, and osteocalcin between 3- and 9-fold. In contrast, DHT did not alter levels of these mRNA species in ovx rats. The data demonstrate that estrogen deficiency increased mRNA levels of genes expressed during osteoblast development and suggest an interplay between estrogen and androgen action in regulating the expression of a number of bone cell genes.

Altered drug sensitivity in response to idarubicin treatment in K562 human leukaemia cells.

Relative to the commonly used anthracyclines, little is known about idarubicin and the development of multidrug resistance. We have previously shown the K562/IDA subline resulting from intermittent treatment of the K562 human leukaemia cell line with 20 ng/ml idarubicin did not develop multidrug resistance but became more sensitive to etoposide. Additional similar treatments of this subline produced the K562/IDA20 subline which partially retained its etoposide sensitivity although these cells expressed P-glycoprotein and were resistant to paclitaxel. Sensitization to etoposide was associated with increased decatenation activity of topoisomerase II, although there were no changes in topoisomerase IIalpha expression or formation of etoposide-dependent cleavable complexes. In comparison, the K562/IDA10 subline produced by intermittent treatment of the K562 cells, firstly with 5 ng/ml then 10 ng/ml idarubicin, showed no detectable expression of P-glycoprotein, decreased topoisomerase IIalpha expression and increased resistance to etoposide and amsacrine, but not to idarubicin or genistein. Even though intermittent treatment with idarubicin caused increased drug resistance in both sublines, they remained sensitive to idarubicin. Therefore the potential of idarubicin as a substitute for other anthracyclines in the treatment of cancer warrants further investigation.

Identification of a receptor-binding pocket on the envelope protein of friend murine leukemia virus.

Based on previous structural and functional studies, a potential receptor-binding site composed of residues that form a pocket at one end of the two long antiparallel helices in the receptor-binding domain of Friend 57 murine leukemia virus envelope protein (RBD) has been proposed. To test this hypothesis, directed substitutions for residues in the pocket were introduced and consequences for infection and for receptor binding were measured. Receptor binding was measured initially by a sensitive assay based on coexpression of receptor and RBD in Xenopus oocytes, and the findings were confirmed by using purified proteins. Three residues that are critical for both binding and infection (S84, D86, and W102), with side chains that extend into the pocket, were identified. Moreover, when mCAT-1 was overexpressed, the infectivity of Fr57-MLV carrying pocket substitutions was partially restored. Substitutions for 18 adjacent residues and 11 other previously unexamined surface-exposed residues outside of the RBD pocket had no detectable effect on function. Taken together, these findings support a model in which the RBD pocket interacts directly with mCAT-1 (likely residues, Y235 and E237) and multiple receptor-envelope complexes are required to form the fusion pore.

The relationship between modulation of MDR and glutathione in MRP-overexpressing human leukemia cells.

Multidrug resistance-associated protein (MRP) causes multidrug resistance (MDR) involving the anthracyclines and epipodophyllotoxins. Many studies show modulation of anthracycline levels and cytotoxicity in MRP-overexpressing cells, but there is limited data on the modulation of etoposide levels and cytotoxicity in MRP-overexpressing or in P-glycoprotein-expressing cells. Etoposide accumulation was 50% reduced in both the CEM/E1000 MRP-overexpressing subline and the CEM/VLB100 P-glycoprotein-expressing subline compared to the parental CEM cells, correlating with similar resistance to etoposide (200-fold) of the two sublines. For the CEM/VLB100 subline, the P-glycoprotein inhibitor SDZ PSC 833, but not verapamil, was able to increase etoposide accumulation and cytotoxicity. For the CEM/E1000 subline, neither SDZ PSC 833 nor verapamil had any effect on etoposide accumulation. However, verapamil caused a 4-fold sensitization to etoposide in this subline, along with an 80% decrease in cellular glutathione (P < 0.05). Buthionine sulfoximine (BSO), which depletes glutathione, also caused a 2.5-fold sensitization to etoposide with no effect on accumulation in the CEM/E1000 subline. In contrast, SDZ PSC 833 was able to increase daunorubicin accumulation in the CEM/E1000 subline (P < 0.05), but had no effect on daunorubicin cytotoxicity, or cellular glutathione. These results show that modulation of etoposide cytotoxicity in MRP-overexpressing cells may be through changes in glutathione metabolism rather than changes in accumulation and confirm that changes in drug accumulation are not related to drug resistance in MRP-overexpressing cells.

ARA, a novel ABC transporter, is located at 16p13.1, is deleted in inv(16) leukemias, and is shown to be expressed in primitive hematopoietic precursors.

ATP-binding cassette (ABC), ATP-dependent transporters are a large superfamily of proteins that include the multidrug resistance proteins, P-glycoprotein and MRP (multidrug resistance protein). The ARA (anthracycline resistance-associated) gene that codes for a putative member of the ABC transporters has recently been cloned and shown to have high sequence homology to the gene for MRP. We have previously shown MRP to be deleted in a subset of inv(16) leukemic patients. The deletion of MRP was associated with an improved patient survival compared with inv(16) patients who did not have such a deletion. In this study, the ARA gene is mapped to 16p13.1, in the same physical interval as the inv(16) short-arm breakpoint. It is shown to be situated proximal to both MYH11, the gene involved in the primary breakpoint on the short arm of the inv(16), and MRP. A YAC clone has been isolated containing both MRP and ARA. FISH analysis of metaphase chromosomes from inv(16) patients has established the gene order as telomere-MYH11-MRP-ARA-centromere and demonstrated that both ARA and MRP are deleted in a subgroup of the inv(16) leukemias. ARA and MRP are both shown to be expressed in normal hematopoietic precursors including CD34(+) cells. The mapping of ARA to this region and its homology to MRP raises questions about its potential role in the biology of the inv(16) leukemias.

Amplification and expression of the ABC transporters ARA and MRP in a series of multidrug-resistant leukaemia cell sublines.

E1000, the most drug-resistant subline from the E-series (CCRF-CEM/E16 to E1000), has been previously shown to express high mRNA levels from two ABC transporter genes associated with multidrug resistance, ARA and MRP. The expression and amplification of both genes has now been characterized for each member of the E-series of drug-resistant sublines and is reported here. Both ARA [detected by reverse transcriptase polymerase chain reaction (RT-PCR)] and MRP (detected by Northern blot analysis) were expressed at low levels in the sensitive parental CEM cell line. An equivalent level of MRP mRNA expression was detected throughout the CEM, E16, E25 and E50 sublines, and there was increasing expression in the E100, E200 and E1000 sublines. ARA expression was not detected in the E16, E25, E50 and E100 sublines but was detected by both RT-PCR and Northern blot analysis in the E200 and E1000 sublines. Southern blot analysis indicated the increased levels of MRP and ARA expression resulted from gene amplification and that MRP was first amplified in the E100 subline and ARA in the E200 subline, suggesting that the two genes were not initially co-amplified. Cytogenetic analysis of E1000 cells demonstrated a large addition to chromosome 16p, around the region where the ARA and MRP genes are located. Increased expression of ARA is associated with increased colchicine resistance in the E-series of sublines and combined with MRP may account for their resistance phenotype.

Induction of broad drug resistance in small cell lung cancer cells and its reversal by paclitaxel.

The H82 "variant" and the H69 "classic" small cell lung cancer (SCLC) cell lines were treated with low levels of epirubicin (69 and 14 nM) which caused little cell death but produced the H82/E8 and H69/E8 extended-multidrug resistant sublines. Both were resistant to drugs associated with multidrug resistance (MDR), and to chlorambucil (9.5- and 5.6-fold, respectively) and cisplatin (2.3- and 8.5-fold, respectively). There was increased expression of the multidrug resistance-associated protein (MRP1) in the H82/E8 subline while P-glycoprotein expression was not detected in any cells or sublines. Treatment of the H82 cells for 1 hr with 69 nM epirubicin increased MRP1-mRNA expression within 4 hr and this was associated with an increase in the resistance to epirubicin, chlorambucil, cisplatin and paclitaxel. Further, a 1 hr treatment with non-cytotoxic doses of chlorambucil (2.5 microM), cisplatin (1.3 microM) or paclitaxel (13 nM), drugs not normally associated with MRP1-mediated MDR, also increased MRP1-mRNA expression in the H82 cells with paclitaxel causing the highest increase (4.5-fold). For chlorambucil treatment, this increased MRPI-mRNA expression was accompanied by increased drug resistance while paclitaxel treatment had no effect on drug resistance in the H82 cells. For the drug resistant H82/E8 subline, these drug treatments had no effect on the MRP1-mRNA expression and little effect on increasing the subline drug resistance. However, pretreatment with paclitaxel sensitised the H82/E8 subline to chlorambucil and cisplatin returning the subline to the sensitivity of the H82 cell line. We conclude that treatment with low levels of MDR and non-MDR drugs can induce extended-multidrug resistance in SCLC cells, a process that probably involves the co-ordinate upregulation of MRP1 and other resistance mechanisms. The results also suggest paclitaxel may have a role as a response modifier in the treatment of refractory SCLC.

In vitro binding of purified murine ecotropic retrovirus envelope surface protein to its receptor, MCAT-1.

An amino-terminal portion of the Friend murine leukemia virus (MLV) envelope surface protein [SU, residues 1 to 236 [SU:(1-236)]] and its receptor, MCAT-1, were each purified from insect cells after expression by using recombinant baculoviruses. Friend SU:(1-236) bound specifically to Xenopus oocytes that expressed MCAT-1 with an affinity (Kd, 55 nM) similar to that of viral SU binding to permissive cells. Direct binding of Friend SU:(1-236) to purified MCAT-1 was observed in detergent and after reconstitution into liposomes. Analysis of binding demonstrated that MCAT-1 and Friend SU:(1-236) interact with a stoichiometry of near 1:1. These findings demonstrate that the amino-terminal domain from the SU of ecotropic murine retroviruses contains an MCAT-1 binding domain.

Increased MRP expression is associated with resistance to radiation, anthracyclines and etoposide in cells treated with fractionated gamma-radiation.

The failure of chemotherapy is often associated with the failure of radiotherapy in the treatment of cancer. To investigate this relationship, the CCRF-CEM (CEM) human T-cell leukaemia cell line was treated with fractionated gamma-radiation totalling 75 Gy (10 cycles of 1.5 Gy daily for 5 days). This produced the CEMRR subline which was 1.5-fold resistant to radiation compared with the parental CEM cells. The CEMRR subline was also resistant to daunorbicin, idarubicin and etoposide but not to paclitaxel, cis-platinum or chlorambucil. Treatment with 50 microM buthionine sulphoximine, an inhibitor of glutathione synthesis, reversed the daunorubicin resistance in the CEMRR subline. Multidrug resistance-associated protein (MRP) mRNA was 6-fold higher in the CEMRR subline than in the CEM cells, and there was no detectable expression of P-glycoprotein in either the CEM cells or the CEMRR subline. Treatment of the CEM cells with 2 Gy of gamma-radiation caused an increase in MRP-mRNA within 4 hr which, by 24 hr, was greater than 5-fold that of the untreated CEM cells. No change in MRP mRNA was observed in the CEMRR subline with similar treatment. We conclude that MRP is involved in the immediate response to radiation and it may account for the drug resistance that often develops following radiation treatment.