Knowing Your ABCCs: Novel Functions of ABCC Transporters

Neuroblastoma, a tumor of sympathetic neuronal precursors, is the most common and deadly solid tumor of childhood. Interestingly, this tumor has the highest rate of spontaneous differentiation or regression of any human cancer (1,2). However, the majority of neuroblastoma patients have unresectable or metastatic tumors and a poor outcome despite aggressive multimodality therapy (3). Several acquired genetic changes have been identified that are associated with a worse outcome, such as MYCN amplification, deletion of loci on 1p36 or 11q23, unbalanced gain of distal 17q, and diploidy (4–15). Nevertheless, at the present time, the clinical and biological variables currently used for risk stratification and therapy selection (16) are imperfect, and treatment failure occurs for high-risk patients as well as some lower-risk patients. The most common reason for treatment failure of high-risk neuroblastoma is the development of drug resistance. A variety of mechanisms have been proposed to account for acquired drug resistance, including increased DNA repair, inhibited apoptosis, or altered drug handling, such as decreased uptake, increased metabolism, or increased excretion (17).

One of the most important causes of acquired drug resistance to chemotherapeutic agents is increased expression of multidrug transporter genes (17). The most important of these multidrug resistance genes encode the ATP-binding cassette (ABC) superfamily, including the B and C subfamilies (ABCB and ABCC). ABCB1 (formerly known as MDR1) and ABCC1 (formerly known as MRP1) are among the best studied. Both are presumably responsible for more rapid efflux and clearing of selected chemotherapeutic agents. High expression of ABCC1/MRP1 in neuroblastomas is associated with a poor outcome (18,19), but the prognostic significance of ABCB1/MDR1 expression in neuroblastomas is less clear (20–22).

In this issue of the Journal, Henderson et al. (23) explore the impact of ABCC family gene expression on neuroblastoma cell behavior and patient outcome independent of their effect on chemotherapy efflux. Indeed, such a role has been suggested by observations that high ABCC gene expression had an adverse effect on outcome even in patients who had not received chemotherapy [reviewed in (24)]. In this report, the authors used a variety of in vitro and in vivo approaches to address this issue. Their data provide compelling evidence for a role of ABCC transporters in modulating neuroblastoma behavior that is independent of their effect on chemotherapy efflux but also hint at even greater complexity.

For in vitro studies, Henderson et al. (23) depleted ABCC1 and ABCC4 using short interfering RNA, inhibited ABCC1 using a small molecule called Reversan, and increased ABCC1 expression using stable transfection. They studied a MYCN-amplified neuroblastoma line SK-N-BE(2)C, as well as SH-SY5Y and SH-EP, two subclones (neuronal and epithelial subclones, respectively) of the non-amplified line SK-N-SH (25). They examined cell proliferation, morphological differentiation, cell motility, and colony formation. Depletion or inhibition of ABCC1 led to impaired cell motility, decreased proliferation, and decreased colony formation but had only a modest effect on differentiation. Depletion of ABCC4 caused a more dramatic increase in morphological differentiation, with markedly reduced proliferation and colony formation. In contrast, increased expression of ABCC3 caused reduced cell migration and colony formation. In general, similar effects were seen in the three cell lines tested, although effects on morphology and proliferation were more variable than migration and colony formation.

For in vivo studies, Henderson et al. (23) used a mouse model in which the mice carrying one or two copies of a TH-MYCN transgene develop neuroblastomas that mimic the human disease (26). They treated the TH-MYCN mice with the Abcc1 inhibitor Reversan, and this treatment delayed the development of tumors in treated mice compared with untreated controls. TH-MYCN transgenic mice were crossed with mice that had the Abcc1 gene knocked out. Abcc1 +/− mice (carrying either one or two copies of the TH-MYCN transgene) had decreased prevalence or delayed development of tumors compared with wild type Abcc1 +/+. Interestingly, this effect was lost when both copies of Abcc1 were knocked out (Abcc −/−) (23).

For primary tumor studies, Henderson et al. (23) examined ABCC transporter expression (ABCC1–12) in primary tumors using quantitative, real-time, reverse-transcriptase polymerase chain reaction, and they determined the effect on outcome in a prospective cohort of 209 neuroblastoma patients. High expression of ABCC1 and ABCC4, as well as low expression of ABCC3, were all associated with a worse outcome. Henderson et al. (23) tested the prognostic impact of these three ABCC genes in a multivariable model. Only stage (not age or MYCN amplification) retained prognostic significance after adjusting for ABCC expression (23). Furthermore, patients could be stratified into groups with excellent, intermediate, and poor outcomes, based on expression of 0, 1, or 2–3 of the unfavorable ABCC profiles, respectively, and this result was validated in an independent cohort of patients.

This work expands our understanding of the complexity of ABCC transporter genes and proteins and provides compelling biological evidence for a role beyond the efflux of chemotherapy drugs and multidrug resistance, because no drugs were used in these experiments. Furthermore, none of the known chemotherapy substrates of ABCC4 are currently used to treat neuroblastoma patients, and low expression of ABCC3 was associated with a worse outcome, even though it is responsible for efflux of etoposide, an important agent for treating this disease. Clearly, the evolutionary role of ABCC transporters is not to facilitate the efflux of chemotherapy agents. So to fully understand their functions, we need to understand better the full range of endogenous cellular substrates that each transporter regulates (27) and whether or not there are non-transporter functions of these membrane-spanning proteins.

However, the Henderson et al. (23) data also suggest even greater complexity in the role of ABCC transporters in neuroblastoma behavior. Hemizygous deletion of Abcc1 in the TH-MYCN transgenic mouse background delays development of tumors or reduces the incidence of tumors, but homozygous inactivation of Abcc1 abrogates this effect. The reasons for this are unclear at present but suggest either a dosage effect (haploinsufficiency) or that there is a compensatory cellular response to complete absence of ABCC1. Second, high expression of ABCC1 and ABCC4, but low expression of ABCC3, are associated with an adverse outcome. Thus, compared with the other two genes, increased expression of ABCC3, and the types of substrates it regulates, has the opposite effect on tumor development or aggressiveness, despite its effect on etoposide efflux. The authors propose a number of logical next steps to examine endogenous substrates, as well as their effect on other cellular functions, such as inflammatory mediators or immune response. They also are planning crosses of TH-MYCN transgenic mice with Abcc3 and Abcc4 knockout mice, which will further elucidate the effect of these genes on tumor development in this model. We will need a greater understanding of the functions of ABCC transporters and the consequences of their inhibition to consider therapeutic interventions with small-molecule inhibitors such as Reversan or others under development (27).

Finally, Henderson et al. (23) show that the pattern of ABCC expression is strongly predictive of outcome (high ABCC1, high ABCC4 and low ABCC3 are bad). These three genes are correspondingly regulated by MYCN expression (whereas other ABCC family genes are not), and so at least some of the consequences of MYCN amplification may be mediated by its effect on their expression. It is unlikely that looking at the expression of any three genes will provide a complete picture of the biology and behavior of every tumor. However, their data argue strongly for the inclusion of these genes in any predictive algorithm for neuroblastomas based on gene expression, and they strengthen the argument for targeted therapy directed at these genes, proteins, and pathways. Clearly, getting to know your ABCCs will be important for fully understanding their role in the pathogenesis of neuroblastoma and perhaps many other human neoplasms.

Funding

This work was supported in part by grants CA039771, CA094194, and CA097323 from the National Cancer Institute, a V Foundation Translational Research Grant, and the Audrey E. Evans Endowed Chair.