Minireview: SLCO and ABC Transporters: A Role for Steroid Transport in Prostate Cancer Progression

Abstract

Androgens play a critical role in the development and progression of prostate cancer (PCa), and androgen deprivation therapy via surgical or medical castration is front-line therapy for patients with advanced PCa. However, intratumoral testosterone levels are elevated in metastases from patients with castration-resistant disease, and residual intratumoral androgens have been implicated in mediating ligand-dependent mechanisms of androgen receptor activation. The source of residual tissue androgens present despite castration has not been fully elucidated, but proposed mechanisms include uptake and conversion of adrenal androgens, such as dehdroepiandrosterone to testosterone and dihydrotestosterone, or de novo androgen synthesis from cholesterol or progesterone precursors. In this minireview, we discuss the emerging evidence that suggests a role for specific transporters in mediating transport of steroids into or out of prostate cells, thereby influencing intratumoral androgen levels and PCa development and progression. We focus on the solute carrier and ATP binding cassette gene families, which have the most published data for a role in PCa-related steroid transport, and review the potential impact of genetic variation on steroid transport activity and PCa outcomes. Continued assessment of transport activity in PCa models and human tumor tissue is needed to better delineate the different roles these transporters play in physiologic and neoplastic settings, and in order to determine whether targeting the uptake of steroid substrates by specific transporters may be a clinically feasible therapeutic strategy.

Androgens play a critical role in the development and progression of prostate cancer (PCa) (1), and androgen deprivation therapy (ADT) via surgical or medical castration remains front-line therapy for patients with advanced hormone-sensitive disease (2). However, progression inevitably occurs, resulting in castration-resistant PCa (CRPC) the phenotypically lethal form of the disease.

Persistent signaling via the androgen receptor (AR) axis in CRPC is attributed to AR and ligand-related mechanisms of resistance (3–6). Any mechanism that modulates intracellular androgens may influence the AR signaling axis and thereby PCa progression. There is growing interest in the role of steroid transporters in mediating steroid flux into (or out of) PCa cells and whether these pathways influence PCa development, response to steroid suppression therapy, and progression to CRPC. This concept is particularly salient, because both localized and advanced prostate tumors demonstrate significant levels of residual tissue androgens despite castration. In localized disease, castration reduces prostate androgens by only 75% (7), whereas intratumoral testosterone (T) levels in CRPC metastases are higher than serum T levels in castrate men and 4-fold higher than levels in prostate tissues from eugonadal men (8).

The source of residual tissue androgens despite ADT has not been fully elucidated, but proposed mechanisms include uptake and conversion of adrenal androgens (such as dehydroepiandrosterone-sulfate [DHEA-S]) to T and dihydrotestosterone (DHT), or de novo androgen synthesis from cholesterol or progesterone precursors (8, 9). The specific mechanism by which adrenal androgens or progesterone precursors enter PCa cells, however, is generally attributed to passive diffusion and has been largely ignored as a potential contributing factor in disease progression. Here, we review emerging data that support a role for steroid transporters in modulating intratumoral androgen levels and thereby PCa progression.

Solute carrier organic anion (SLCO)-Encoded Membrane Transporters

SLCO-encoded transporters are involved in hormone transport

The organic anion transporting polypeptides (OATPs) superfamily comprises 11 proteins encoded by the SLCO gene family and mediates the sodium and ATP-independent uptake of compounds such as bile acids, thyroid hormones, steroid conjugates, xenobiotics, and a variety of important pharmaceuticals (10). Although several members can function as bidirectional facilitated diffusion transporters, physiologic conditions under which this occurs are not well understood, and OATPs are typically considered uptake transporters (11). Several family members mediate specific uptake of steroids and steroid precursors, including sulfated forms of DHEA, pregnenolone, and estrone (Table 1) (12), all of which are potentially important substrates in PCa. Among these are SLCO1B1 and SLCO1B3, primarily expressed in the liver, and SLCO1A2 and SLCO2B1, more broadly distributed in liver, kidney, intestine, and brain as well as in steroidogenic tissues, such as testis, ovary, mammary, placenta, and adipose tissue (10, 12–17).

Table 1.

Summary of Steroid Transporters, Tissue Distribution, Substrates, and Allelic Variants of Functional Significance

Gene	Normal Expression	Neoplastic Dysregulation	Hormone Substrates	Cancer-Related Exogenous Substrates	Significant SNP Variants
SLCO1B3	Liver (76)	Gastric, colon, pancreas, gallbladder, lung, brain, breast (77), prostate (27, 32)	DHEA-S, estradiol-17β-glucuronide, estrone-3-sulfate, T, thyroid hormone (T3, T4) (28)	Methotrexate, taxanes, irinotecan, statins, imatinib, green tea catechins (28)	rs4149117 T>G. impaired T uptake in transfected Cos-7 cell line (32); decreased PSCM (27, 32)
SLCO2B1	Liver, placenta, heart, small intestine, brain (28), spleen, kidney, prostate, ovary (13)	Colon, prostate, breast, thyroid, bladder (33)	Estrone-3-sulfate, DHEA-S (28), pregnenolone sulfate (43)	Statins (59)	rs12422149 A>G, rs1789693 A>T, rs1077858 A>G, shorter time to progression on androgen deprivation (34); rs12422149 A>G increased uptake of DHEA-S in LNCaP and CAPC4 cells (34); rs12422149 A>G decreased risk of PSCM (27)
SLCO1B1	Liver (28)		Estrone-3-sulfate, DHEA-S, thyroid hormone, estradiol-17β-glucuronide (28)	Statins (78)	rs4149056 T>C, increased risk of statin myopathy (78)
SLCO1A2	Brain, liver, kidney, intestine, lymph, breast, testis, prostate (28)	Breast (79, 80), prostate (35)	Estrone-3-sulfate, DHEA-S, thyroid hormone (28)	Statins, green tea catechins (81)	rs11568563 A>C, rs45502302 A>T, decreased transport of estrone-3-sulfate in transfected HeLa cells (38)
Megalin	Kidney, prostate, ovary, breast, uterus (60), brain (82), gallbladder (83)		Vitamin D + binding protein, thyroid hormone + transthyretin, retinol + binding protein, T, and DHT + sex hormone binding globulin (60)		rs830994 A>G, decreased risk of PCa progression (61); rs2300446 A>G, rs830094 A>G, decreased risk of PSCM (61)
ABCC1/MRP1	Liver, kidney, intestine, brain, adrenal, testis, ovaries (84)	Lung, breast, prostate, neuroblastoma (85), bile duct (86)	DHEA-S, estrone-3-sulfate, and 17β-estradiol glucuronide (63)	Methotrexate, doxorubicin, vinca alkaloids (85), flutamide (87), statins (59)
ABCG2/BCRP	Intestine, liver, breast, testis, ovary, kidney, endothelium (88), placenta (89), prostate (rare) (66)	Ovary, breast, colon, stomach, fibrosarcoma (90)	DHEA-S, estrone-3-sulfate, DHT, and 17β-estradiol glucuronide (66)	Methotrexate, imatinib, mitoxantrone, irinotecan, lapatinib, statins, sulfasalazine, topotecan (90)	rs2231142 C->A, worse survival in PCa (68)
ABCC4/MRP4	Prostate, liver, testis, ovary, brain, kidney, adrenal, platelets (91)	Neuroblastoma (92), prostate (70, 71)	DHEA-S, estrone-3-sulfate, and 17β-estradiol glucuronide (63, 69)	6-MP, thioguanine (93)

DHEA, pregnenolone, and estrone, either directly or via downstream metabolites, are capable of activating AR axis signaling (discussed below); thus, their well-recognized place among SLCO substrates is of particular interest. The transport of DHEA-S is perhaps most notable, because DHEA-S circulates at high levels in castrate men, DHEA is a known substrate for the production of T and DHT in the PCa microenvironment (9, 18), and DHEA can serve as a noncanonical ligand capable of promiscuously activating AR molecules mutated in the ligand binding domain (19). Similarly, uptake of pregnenolone provides substrate for the key steroidogenic enzyme, cytochrome p450 17A1 (CYP17A1), which converts pregnenolone to DHEA. This is particularly relevant in men undergoing CYP17A1 inhibition with abiraterone, in whom circulating and tissue levels of pregnenolone are significantly increased (20), such that efficient uptake of pregnenolone into CRPC tumor cells could overcome competitive inhibition of CYP17A1 by intratumoral abiraterone. The uptake of estrone and conversion to estradiol is also of interest, as both estrogen receptor α and β isoforms have been implicated in PCa pathogenesis (reviewed in Refs. 21, 22). Moreover, as with DHEA, both pregnenolone and estradiol can activate mutated AR (23, 24).

Neoplastic tissue expression of SLCO transporters

Altered expression and distribution of various OATP transporters occurs in a number of neoplastic tissues, including gliomas, colon, breast, prostate, and hepatocellular carcinomas (13, 25–28). The association of SLCO transporters with breast cancer is of particular note because it suggests an analogous role for these transporters in mediating steroid uptake into PCa. Specifically, expression of SLCO2B1 has been associated with breast cancer grade (29), and in vitro studies corroborate a functional role of OATP2B1 in breast cell lines, showing transport of estrone-3-sulfate into the hormone-dependent lines MCF-7 and ZR-75-1, but not in the hormone-independent line MDA-MB-231 (25), and showing that inhibition of estrone-sulfate transport suppressed MCF-7 proliferation (29, 30). The impact of SLCO single nucleotide polymorphism (SNPs) on the clinical outcome of established breast cancer patients, however, has not yet been reported (31).

SLCO Genes in PCa

Expression levels in normal vs neoplastic prostate tissues

In prostate tumors, expression of SLCO genes has been demonstrated at the RNA and protein levels in localized and advanced PCa tissues. OATP1B3 was markedly overexpressed by immunofluorescence or mRNA profiling in primary PCa when compared with normal prostate and benign prostate hyperplasia (32, 33), and CRPC metastases showed significantly increased expression of 6 SLCO genes compared with untreated primary prostate tumors, including SLCO1B1, SLCO1B3, SLCO2A1, SLCO2B1, SLCO3A1, and SLCO4A1 (27). This pattern of up-regulation from benign to malignant prostate tissue and from untreated PCa to CRPC suggests that these transporters may play a role in tumor progression, a hypothesis supported by the genetic association and functional studies discussed below.

Impact of SLCO SNPs on steroid transport in PCa cells

The impact of OATP expression and genetic variation on steroid transport in vitro has been evaluated for a number of SLCO genes, of which data for SLCO2B1 and SLCO1A2 exist specifically in PCa cell lines. In LNCaP cells transfected with the SLCO2B1 SNP variant rs1242149 (935G>A; Arg312Gln), the 935G variant exhibited higher maximal DHEA-S uptake compared with 935A or mock-transfected cells (34). DHEA-S treatment of LNCaP and LAPC-4 cells transfected with 935A or 935G also resulted in higher PSA expression and AR mRNA levels in the 935G-transfected cells. Notably (albeit in a non-PCa model), progesterone specifically increased OAT2B1-mediated DHEA-S and E1S transport in HEK1 cells overexpressing OATB2B1 (17). This finding is of relevance to PCa, because CYP17A1 inhibitors, such as abiraterone, significantly increase levels of hormones upstream of CYP17A, such as progesterone (20), which could then promote scavenging of residual DHEA-S by OATP2B1-expressing tumor cells.

SLCO1A2 may also play a role in DHEA-S-induced PCa cell growth under androgen-depleted conditions. SLCO1A2-transfected LNCaP and 22RV1 cells showed increased DHEA-S-stimulated growth (vs growth of SLCO1A2-knockdown cells, which was insensitive to DHEA-S), and SLCO1A2 mRNA expression was higher in PCa cell lines grown under androgen-depleted conditions (and was suppressed by addition of DHT). Together, these data suggest that expression of SLCO1A2 may be up-regulated by androgen deprivation as a potential mechanism of scavenging DHEA-S for conversion to T (35).

Although steroid transport by the following OATP proteins has yet to be specifically shown in PCa cell lines, genetic variation in SLCO1B3, SLCO1B1, and SLCO1A2 is also known to affect transport of steroids relevant to PCa. In Cos-7 cells transfected with SLCO1B3 rs4149117 (334T>G) and rs7311358 (669G>A) (32), cells transfected with wild-type variants of both SNPs (334T/699G) actively transported T compared with cells carrying the 334G and 699A SNPs. No correlation was observed between genotype and SLCO1B3 mRNA expression in NCI-60 cell lines, suggesting that impact of the variant was on function and not expression. In HEK293 cells transfected with SLCO1B1 SNP rs4149056 (37041T>C or Val174Ala), cells transfected with the variant 37041C allele demonstrated both lower cell-surface expression and lower uptake of estrone-3-sulfate and 17β-D-glucuronide vs wild type (36). In HEK23 cells transfected with SLCO2B1 SNP rs2306168 (1457C>T), intrinsic transport of estrone-3-sulfate by the 1457T allele was less than half the wild-type 1457C allele (37). Finally, 2 SNP variants of SLCO1A2 overexpressed in HeLa cells showed reduced transport of estrone 3-sulfate compared with wild type (rs11568563 516A>C and rs45502302 404A>T) (38). Further studies are needed to reveal whether these transporters are similarly active in PCa cells.

Ethnic variations in SLCO SNP prevalence and PCa risk

Given the ethnic variation in serum androgen levels and PCa risk between Caucasian and African American (AA) men, key SLCO SNPs associated with androgen transport (SLCO1B3 rs4149117 and SLCO2B1 rs1242149) have been examined for an association with ethnicity. The low-T transport SLCO1B3 SNP rs4149117 (334T>G) occurs at a much lower frequency (41%) in AA men compared with European Americans (88%), which might suggest that increased uptake of T plays a role in driving the more aggressive prostate tumors noted in AA men (32, 39). However, Caucasian men are more likely than AA men to have the high DHEA-S transporting G allele of SLCO2B1 rs1242149 (92% vs 87%) (40), which is reverse of what might be expected and suggests that factors other than transport SNPs are likely to be more important drivers of aggressive PCa in AA men.

Clinical analyses of SLCO SNPs in PCa

In conjunction with in vitro studies demonstrating genetic variation in OATP-mediated steroid transport in PCa cell lines, studies in multiple clinical settings have demonstrated an association of SLCO1B3 and SLCO2B1 SNPs with PCa outcomes. Analysis of the rs4149117 (334T>G) and rs7311358 (669G>A) SLCO1B3 variants (associated with T transport) in 180 Caucasian patients with CRPC showed that high transport SLCO1B3 TT/GG and TG/GA haplotypes had an estimated 10-year survival of 23% as compared with 42% for the GG/AA haplotype (P = .023) (32). Genetic variation in SLCO1B3 was also associated with time from initiation of ADT to development of “androgen independence.” In a study of 68 Caucasian patients with advanced PCa, more rapid development of androgen independence was noted in men with 1 or 2 copies of the high transport SLCO1B3 rs4149117 (334T) allele compared with patients with 2 copies of the 334G allele (41), consistent with the hypothesis that increased uptake of androgen levels in the castrate tumor environment may promote tumor progression.

In 538 patients with metastatic hormone-sensitive PCa, the median time to progression (TTP) for patients with each of 3 SLCO2B1 alleles (rs12422149 [935G>A; Arg312Gln], rs1789693, and rs1077858) was 10, 7, and 12 months shorter than the other allele, and the effect was additive (34). A difference in TTP between patients carrying the SLCO1B3 334T and 334G alleles associated with T transport was not observed. However, patients who had multiple “at-risk” SLCO2B1 variants (including the high DHEA-S transport SLCO2B1 allele rs12422149 935G) who also had the high T transport SLCO1B3 SNPs had the shortest time to progression on ADT, with a hazard ratio (HR) of 3.57 (P = .41). (The authors theorized that increased uptake of DHEA-S by SLCO2B1 and intracellular conversion to T might result in increased excretion of T, which, in turn, could be conserved by uptake back into tumor cells via SLCO1B3.) Evaluation of SLCO2B1 in 532 Japanese men similarly showed that 2 copies of theSLCO2B1 rs12422149 935G variant was associated with shorter TTP on ADT (10.0 vs 17.0 mo; P = .004) but no difference in overall survival (42).

Using a very different study population, Wright et al (27) evaluated SLCO genotype vs recurrence/progression and PCa-specific mortality (PSCM) in 1309 Caucasian patients from a population-based cohort of incident PCa cases (78% with localized disease). Consistent with findings in advanced PCa, the high-T import SLCO1B3 rs4149117 (334T) was associated with a 76% increased risk of PCSM (HR = 1.76; 95% confidence interval, 1.00–3.08). A statistically significant difference between recurrence/progression events was not associated with this SNP.

However, although Yang et al (34) observed a more rapid time to progression on ADT in castrate men bearing the high-DHEA-S import SLCO2B1 rs12422149 (935G, 312Arg), Wright et al (27) observed an increased risk of PCSM in men bearing the low-import allele, with 1 or 2 copies of the variant 935A allele (312Gln) associated with a 2-fold increased risk of PCSM. Notably, allelic variation in DHEA-S uptake by SLCO2B1 (and accordingly, any impact of cellular DHEA-S uptake on PCa outcome) may differ in the eugonadal and castrate settings, because T can inhibit SLCO2B1-mediated transport of DHEA (43). Thus, in eugonadal men, the genetic variation in OATP2B-mediated transport of DHEA-S occurring in castrate men may be inhibited by the presence of circulating T and have no role in primary prostate tumors.

Together, these data suggest that genetic variations in SLCO-mediated androgen transport may impact PCa progression and response to therapy by modulating the ability of cancer cells to scavenge residual androgens from the prostate tumor microenvironment and thereby facilitate continued AR activity despite castration. Importantly, the role of OATP proteins over the entire life of patients may vary from that observed in patients with advanced disease on ADT. In particular, the uptake of steroids such as DHEA-S by OATP2B1 is likely most important in men with reduced circulating androgens, and the association of this gene with outcome in men with earlier disease may reflect transport of an alternative substrate.

Other SLCO substrates of relevance to PCa

Steroids represent a small subset of the wide range of endogenous and exogenous substrates transported by OATP proteins, including agents such as statins, cardiac glycosides, glitazones, metformin, and taxanes (44, 45), all of which have known or postulated impacts on prostate carcinogenesis and/or progression (46–52). Statins are a common OATP substrate postulated to influence PCa risk and progression. Statins inhibit proliferation of LNCaP and PC-3 PCa cell lines in vitro and prostate tumors in C3(1)/Tag transgenic mice in vivo (53, 54), and several large population-based studies have found a reduced incidence of advanced PCa (55) or a decreased risk of PCa mortality in statin users (HR = 0.76) (56). Notably, transporters implicated in PCa progression, such as OATP1B3 and OATP2B1, mediate the uptake of rosuvastatin, atorvastatin, and fluvastatin (46, 57–59), and transport of steroids of potential importance in PCa progression, such as uptake estrone-3-sulfate by OATP2B1, is inhibited by atorvastatin, simvastatin, and gemfibrozil (43). The influence of statins on OATP-mediated transport of other steroids has not been reported but is certainly possible. In addition, these data raise the possibility that alterations in statin transport due to SLCO genotype, or alterations in OATP-mediated steroid transport due to competing statin uptake, may influence the association of statins with PCa.

Other Transporters of Potential Relevance to PCa

Megalin

Megalin, a member of the low density lipoprotein receptor gene family, is expressed by prostate epithelial cells and can internalize androgens bound to sex hormone binding globulin (60). Immunohistochemical assays demonstrated increased megalin gene expression in malignant vs benign prostate cells (Mostaghel, E. A., unpublished data), and several polymorphisms within the megalin gene have been associated with PCa recurrence/progression and mortality (61). Functional studies directly demonstrating a role for megalin in androgen uptake in PCa tissue, however, have not been reported to date.

ATP binding cassette (ABC) transporters

The ABC transporter superfamily encodes a large number of ATP-dependent unidirectional transporters (important in the efflux of many drugs and endogenous metabolites), of which several transport steroid metabolites of potential relevance to PCa, including DHEA-S, estrone-3-sulfate, and 17β-estradiol glucuronide (62–65). Functional data supporting a role for ABC transporters in PCa cells are not as well developed as for the SLCO family. However, altered expression of several family members in advanced vs primary prostate tumors suggests that a decrease in the steroid efflux capacity of these transporters may potentially facilitate disease progression by conserving tumor androgen levels.

ABCG2 (ATP-binding cassette subfamily G2, also known as breast cancer resistance protein [BCRP]) is expressed in prostate basal epithelial cells and can mediate efflux of DHT in in vitro studies (66, 67). Huss et al (66) demonstrated expression of ABCG2/BCRP in putative prostate tumor stem cells and showed in rat prostate progenitor cells that constitutive efflux of DHT by this transporter prevents stabilization and nuclear transport of AR. A clinical association of ABCG2 with ABC in PCa has also been reported, with men who were homozygous for wild-type ABCG2 SNP at rs2231142 (421C>A) demonstrating improved overall survival when compared with men who carried 1 or 2 A alleles (68). Whether this SNP affects DHT transport has not been reported but would be consistent with the hypothesis that efflux of DHT by ABCG2 might underlie a difference in clinical outcomes.

ABCC4 (also known as multidrug resistance protein 4 [MRP4]) is also capable of transporting steroid hormones, including the key precursor of T, DHEA-S (63, 69). Interestingly, MRP4 expression is reduced in primary prostate tumors after neoadjuvant androgen ablation and is significantly lower in CRPC tissues compared with primary prostate tissues (70, 71). These observations suggest that MRP4-mediated efflux of DHEA-S may be specifically suppressed after androgen suppression and/or in CRPC tumors as a mechanism of conserving intracellular androgen precursors (Figure 1).

Figure 1.

Steroid hormones, such as estrone-sulfate (circles labeled E), dihydroepiandrosterone-sulfate (DHEA-S, circles labeled D), pregnenolone (circles labeled P), and T (circles labeled T), can be transported into cancer cells by transporters, such as the SLCO/OATP family proteins and/or Megalin. Presence of the appropriate intracellular enzymes may allow conversion of these hormones into estradiol (E2), DHEA, T, and DHT. E2 and DHT can bind to estrogen receptor (ER) and AR, respectively, which homo-dimerize and enter the nucleus to influence transcription of genes that enhance growth and survival. DHEA can activate AR bearing certain mutations in the ligand binding domain. Steroid hormones can also be metabolized by enzymes, and the hormones and metabolites can be actively effluxed through the ABC-transporter family. Polymorphisms or differential expression of these transporters can impact the influx and efflux of steroid hormones, which may in turn affect the balance of intratumoral steroids and overall tumor growth.

Conclusion

A variety of mechanisms exist by which PCa grows despite castrate serum T levels. Emerging evidence supports a potential role for SLCO- or ABC-mediated androgen transporters in facilitating continued AR signaling in CRPC via the increased uptake or decreased efflux of androgen precursors. However, the reported associations of steroid transporters with PCa outcomes may also be influenced by the transport of nonsteroid substrates, such as statins, metformin, and glitazones, which are known to influence prostate tumor development and progression. Moreover, many commonly ingested substances, such as indomethacin, salicylates, soy, citrus juices, and nutriceuticals, such as green tea, ginkgo, and others, are known to inhibit the activity of certain transporters (72, 73). Thus, detailed attention to medication and supplement histories is required to identify factors that may easily confound small association studies. Nevertheless, available data suggest several consistent disease associations, particularly for SLCO1B3 and SLCO2B1, in PCa.

Whether or not cancer-associated expression of transporters will be feasible targets for inhibition remains a significant question, given the critical function these proteins play in normal clearance of endogenous metabolites and in excretion of drugs and toxins (74, 75). However, tumoral expression of transporters may still serve as relevant biomarkers for cancers with efficient steroid uptake for which more aggressive targeting of the androgen axis may be warranted. Continued assessment of transport activity in PCa models and in human tumor tissue will better delineate the role of transporters in the neoplastic settings and whether targeting the uptake of steroid substrates by specific transporters may be a clinically feasible strategy in CRPC therapy.