Abstract
The large majority of clinical prostate cancers remain dependent on androgen receptor (AR) activity for proliferation even as they lose their responsiveness to androgen deprivation or antagonism. AR activity can be maintained in these circumstances by increased AR synthesis—often reflecting increased NF-κB activation; upregulation of signaling pathways that promote AR activity in the absence of androgens; and by emergence of AR mutations or splice variants lacking the ligand-binding domain, which render the AR constitutively active. Drugs targeting the N-terminal transactivating domain of the AR, some of which are now in preclinical development, can be expected to inhibit the activity not only of unmutated ARs but also of the mutant forms and splice variants selected for by androgen deprivation. Concurrent measures that suppress AR synthesis or boost AR turnover could be expected to complement the efficacy of such drugs. A number of nutraceuticals that show efficacy in prostate cancer xenograft models—including polyphenols from pomegranate, grape seed, and green tea, the crucifera metabolite diindolylmethane, and the hormone melatonin—have the potential to suppress AR synthesis via downregulation of NF-κB activity; clinical doses of salicylate may have analogous efficacy. The proteasomal turnover of the AR is abetted by diets with a high ratio of long-chain omega-3 to omega-6 fatty acids, which are beneficial in prostate cancer xenograft models; berberine and sulforaphane, by inhibiting AR’s interaction with its chaperone Hsp90, likewise promote AR proteasomal degradation and retard growth of human prostate cancer in nude mice. Hinge region acetylation of the AR is required for optimal transactivational activity, and low micromolar concentrations of the catechin epigallocatechin-3-gallate (EGCG) can inhibit such acetylation—possibly explaining the ability of EGCG administration to suppress androgenic activity and cell proliferation in prostate cancer xenografts. Hence, it is proposed that regimens featuring an N-terminal domain-targeting drug, various nutraceuticals/drugs that downregulate NF-κB activity, and/or supplemental intakes of fish oil, berberine, sulforaphane, and EGCG have potential for blocking proliferation of prostate cancer by targeting its characteristic addiction to androgen receptor activity.
Metastatic Prostate Cancer: A Massive and Intransigent Problem
Prostate cancer is the second leading cause of cancer deaths in US males, accounting for about 30 000 cancer deaths annually. 1 Its incidence and mortality are markedly increasing in Asia, as Westernized diets and lifestyles become more prevalent. 2 Although curable by local control measures while still confined to the prostate, it is considered an incurable disease once it metastasizes—albeit in many cases it progresses so indolently that other causes of death intervene. Since the 1940s, androgen suppression therapy has been the standard initial therapy for metastatic prostate cancer. Initially, 70% to 80% of metastatic prostate cancers regress in response to treatments that either block testosterone production or that inhibit the activating interaction of testosterone with the androgen receptor (AR). The classical strategy for androgen suppression, orchidectomy, has now been largely replaced by the use of injectable gonadotropin-releasing hormone agonists (eg, leuprolide, triptorelin) or antagonists (eg, cytorelix, abarelix) that block pituitary secretion of luteinizing hormone and follicle-stimulating hormone, and thereby suppress testicular testosterone synthesis. 3 Since many prostate cancers evolve the ability to synthesize their own testosterone from adrenal precursors, or upregulate their responsiveness to weak androgens of adrenal origin (such as dehydroepiandrosterone), the irreversible CYP17A/C17,20 inhibitor abiraterone acetate has recently been approved for post–docetaxal management of castrate-resistant prostate cancer, used in conjunction with prednisone. 4 This drug blocks adrenal production of weak androgens at a proximal stage, leading to a profound reduction in androgen availability; prednisone is coadministered since CYP17 also is required for corticosteroid synthesis. Drugs that bind to the AR, such that its activation by testosterone is inhibited, are also employed, as an alternative or adjunct to suppression of androgen synthesis. These include flutamide, bicalutamide, and nilutamide. The recently approved enzalutamide has a much higher affinity for the native AR; unlike some other AR antagonists, it does not promote translocation of the AR to the nucleus and is effective in some prostate cancers that no longer respond to earlier AR antagonists. 5
Nonetheless, despite this impressive armory of androgen antagonists, in almost all cases the cancer eventually recurs in a form that is unresponsive to androgen deprivation and AR blockade. Chemotherapy with docetaxel or the newer cabazitaxel (which is effective in many docetaxel-resistant cancers), and immunotherapy with the recently approved sipuleucel-T, can extend survival by an average of 2 to 4 months, but lack curative efficacy. 5
This essay explores various novel drug and nutraceutical strategies that may have potential for suppressing AR signaling and hence prolonging remission and extending survival in clinical prostate cancer.
How Prostate Cancer Becomes Androgen Independent
The large majority of prostate cancers—including those that are no longer responsive to androgen deprivation measures, and thus are characterized as androgen-independent or androgen-refractory—are dependent on the transactivating activity of the AR.6-8 In prostate epithelium and most prostate cancers, AR-mediated transcription is required for G1-S cell cycle transition; absence of AR activity is associated with suppression of G1 cyclin expression and increased levels of cyclin-dependent kinase inhibitors.9,10 A small minority of prostate cancers—which fail to express prostate-specific antigen (PSA)—somehow manage to maintain proliferation despite methylation-mediated silencing of the AR gene. 11 In prostate cancers that relapse during androgen deprivation, a marked upregulation of AR mRNA is usually observed; amplification of the AR gene occasionally accounts for this, but more commonly this stems from accelerated transcription of the AR gene.7,12-15 It is presumed that a sufficient increase in the number of ARs can compensate for a reduction in androgen level. In part, this reflects the fact that, even when testicular testosterone synthesis is markedly suppressed or eliminated, weak androgens of adrenal origin such as DHEA may interact with the AR. As noted above, this problem is now being addressed with abiraterone acetate, which inhibits androgen production within the adrenal gland and prostate cancer itself. Nonetheless, even this strategy ultimately often fails owing to the cancer’s acquisition of additional genetic or epigenetic changes that further alter AR function.
For example, various tyrosine and serine/threonine kinases, via phosphorylations of the N-terminal domain of the AR or of its coactivators, can promote the intranuclear uptake and transactivational activity of the AR, independent of androgen binding; these include Src (activated via EGFR), Ack1 (activated via HER2/HER3 receptors), and MAP kinase.16-22 Intranuclear Stat3 can boost AR transactivational activity via direct binding to the N-terminal domain. 22 Akt can increase AR activity by recruiting beta-catenin to the nucleus as a coactivator and has the potential to disinhibit AR activity by excluding FOXO1 from the nucleus.23-29 Increased expression of the coactivators required for AR activity is also commonly encountered in castration-resistant cancers.30,31 It seems that achievement of “androgen independence” often reflects both an increase in expression of AR—and possibly of its coactivators—as well as an upregulation of kinase activities that can directly activate or potentiate AR activity.
Alterations of AR structure can also contribute to androgen independence. Mutations of the ligand-binding region of the AR may arise, which make it more responsive to weak androgens or other steroid hormones, turn androgen antagonists such as bicalutamide into agonists, or which render the AR constitutively active.31-34 Moreover, recent research has established that advanced androgen-insensitive prostate cancers commonly express splice variants of the AR that lack the ligand-binding domain but nevertheless possess constitutive transactivating activity.35-38 Evidently, no degree of androgen deprivation will abolish AR activity in prostate cancers expressing AR variants that are constitutively active in the total absence of androgens.
Drug Targeting of the Androgen Receptor N-Terminal Domain
For that reason, drugs that target the N-terminal domain (NTD) of the AR receptor, responsible for its transactivating activity, would appear to have greater potential for durable inhibition of AR activity than measures which target the C-terminal ligand-binding region, which demonstrably is not essential for AR activity.39,40 Such agents also have the potential to block the interaction of the AR with activating kinases targeting the NTD. (Theoretically, drugs targeting the DNA-binding domain could also be durably effective, but this domain is so homologous to that of various other steroid receptors that it would be difficult to find such a drug that inhibited only the AR. 41 ) Sadar and colleagues, following this line of reasoning, have characterized several agents with high affinity for the NTD of the AR that impair its interaction with essential coactivators and hence block its transactivational activity. EPI-001, a naturally occurring chlorinated derivative of the industrial chemical bisphenol A diglycidyl ether (BADGE), is one of these agents; this agent inhibits AR activity in all prostate cancer cell lines tested, and moreover induces shrinkage of human androgen-independent prostate cancers implanted in nude mice, in doses that are not overtly toxic. 41 The initial interaction of EPI-001 with the AR NTD catalyzes formation of an epoxide within EPI-001; subsequent nucleophilic attack of this epoxide results in formation of a covalent linkage between this drug and the AR, such that the AR is permanently incapacitated. 42 Chlorinated peptides, known as sinkotamides, derived from a marine sea sponge, likewise target the NTD of the AR and inhibit its activity. 43
It remains to be seen whether the AR NTD can mutate in ways that markedly lessen the binding affinity of these agents while preserving its ability to interact appropriately with key coactivators; if such mutations are rare or impossible, these agents may possess durable anti-androgenic activity. However, the fact that EPI-001’s activity may require attack by a nucleophile in the NTD raises the prospect that mutagenic loss of this nucleophilic amino acid could impede this drug’s efficacy, possibly without impairing the NTD’s interaction with its key coactivators. Whether or not this proves to be true, efforts to develop drugs (or discover phytochemicals) that can target the NTD should be given high priority, as it is clear that targeting the ligand-binding domain, as current AR-directed drugs do, can provide only temporary benefit.
It is quite conceivable that feasible and tolerable concentrations of drugs targeting the NTD of the AR may fail to achieve a total elimination of AR activity. And, not unlikely, some prostate cancers may increase their expression of membrane transport proteins that can evict these drugs from the cell. In these cases, increased AR expression would represent an escape mechanism for the cancer, as it does when prostate cancers are treated with anti-androgen drugs. Therefore, the concurrent use of adjuvant measures that suppress the accelerated transcription of the AR gene would seem likely to aid therapeutic outcomes.
Downregulating NF-κB Activity to Suppress Androgen Receptor Synthesis
Undoubtedly, there are a variety of mechanisms that are responsible for the increase in AR expression observed in androgen-insensitive prostate cancers. However, there is good reason to suspect that a constitutive increase in NF-κB activity, also commonly observed in androgen-insensitive prostate cancers, and a negative prognostic sign, is the chief reason for the increase in AR expression.44-50 The promoter region of the AR gene possesses NF-κB response elements, and several studies show that stimulation of NF-κB activity increases AR levels in prostate cancer cells and enhances transcription of reporter genes containing the AR gene promoter.49,51-54 This phenomenon appears to reflect direct binding of NF-κB to the AR promoter. Hence, feasible measures that suppress NF-κB activation in androgen-insensitive prostate cancer could be expected to decrease AR expression.
Further studies are required to define why NF-κB is typically constitutively active in advanced prostate cancer. Several recent investigations have shown that “atypical” forms of protein kinase C (eg, PKC-zeta, PKC-iota/kappa) are activated in various prostate cancer cell lines, and stimulate NF-κB activation, presumably via phosphorylation of IKK-β.55-57 These atypical PKCs appear to mediate the activation of NF-κB by TNF-α in prostate cancer. Increased production of oxidative stress via NADPH oxidase activity would also seem likely to upregulate NF-κB activation in many prostate cancers.53,54,58-65 Akt activation—typical of the many prostate cancers that are PTEN deficient—might also participate in NF-κB activation.66-69
Intriguingly, certain nutraceuticals that have shown potential for prostate cancer control in cell cultures or xenograft models—such as polyphenols derived from pomegranate, grape seed, or green tea, the crucifera-derived compound diindolylmethane, and the hormone melatonin—have been found to blunt NF-κB activation in prostate cancer cell lines, either in vitro or in vivo.70-82 It would be of interest to define the mechanisms responsible for such inhibition. In addition, feasible clinical concentrations of the drug salicylate are known to suppress NF-κB activation, reflecting a direct inhibitory interaction with IKK-β.82-86 The phycocyanobilin in spirulina has the potential to blunt oxidative stress-mediated upregulation of NF-κB activity via inhibition of NADPH oxidase complexes.87,88 Correction of poor selenium status with supplemental selenium, by optimizing glutathione peroxide activity and hence aiding control of hydrogen peroxide, could be expected to provide a similar benefit. Relatively high concentrations of selenium can inhibit NF-κB activation in prostate cancer cell lines, and administration of methylselenocysteine and selenite has been reported to decrease AR levels, suppress PSA production, and slow tumor growth in mice xenografted with human prostate cancer.89-93 Hence, a regimen composed of appropriate polyphenols, melatonin, spirulina, selenium, and salicylate (best administered as well-tolerated salsalate 94 ) may have potential for lessening the NF-κB activity of prostate cancers and thereby diminishing their AR expression.
Boosting Androgen Receptor Turnover
Theoretically, measures that boost the proteasomal degradation of the AR would likewise have utility in management of prostate cancer. In that regard, there is a recent report that, when mice bearing PTEN-null castrate-resistant prostate cancer were fed a diet with a 1:1 ratio of omega-6 to omega-3 (the omega-3 provided by equal amounts of EPA and DHA), as contrasted with a diet with an omega-6/omega-3 ratio of 40, the expression of AR protein in the cancers was notably diminished, despite the fact that AR mRNA was not decreased. 95 Cell culture studies revealed that long-chain omega-3 fats promoted proteasomal degradation of the AR in these cancer cells. Likewise, a previous study had reported lower levels of AR in prostate cancer cells exposed to EPA and DHA. 96 These findings may help rationalize other studies in which diets with high omega-3/omega-6 ratio have retarded growth of prostate cancers in mice, as well as epidemiology correlating increased oily fish intake with decreased risk for prostate cancer mortality.97-106
Interaction with Hsp90 and co-chaperones protects AR from proteasomal degradation. 107 There is recent evidence that low micromolar concentrations of the antidiabetic phytochemical berberine reduce the expression of AR within prostate cancer cell lines by promoting its proteasomal degradation, while impeding its interaction with Hsp90. 108 Intriguingly, this effect was even more notable in androgen-independent cell lines expressing AR splice variants lacking the ligand-binding domain; whether this reflects interaction with Hsp90 is questionable, as some of these AR variants are not dependent on Hsp90.109,110 Although the concentrations of berberine employed in these cell culture studies were above those achievable clinically, administration of berberine to nude mice xenografted with LnCaP human prostate cancer, in a dose thought to be clinically relevant and well tolerated (5 mg/kg ip), markedly slowed growth of the cancer and suppressed AR expression within the cancer. 108 The direct molecular target of berberine in these effects is not clear; although berberine, like metformin, is known to activate AMPK in clinical concentrations, there is no evidence that AMPK modifies the interaction of the AR with Hsp90.
Hsp90’s chaperone activity is impeded by acetylations that are reversed by HDAC6 111 ; inhibition of the latter has been shown to decrease AR expression and activity in prostate cancer cells.112,113 Intriguingly, the crucifera-derived phase 2 inducer sulforaphane has been reported to inhibit HDAC6 and thereby destabilize AR in prostate cancer cells. Administration of sulforaphane can slow the growth of prostate cancer xenografts in nude mice, and ingestion of broccoli sprouts rich in sulforaphane inhibits monocyte histone deacetylase activity in humans. 114
Blocking Androgen Receptor Acetylation
A range of enzymes with acetyltransferase activity can promote acetylation of 3 lysine groups in the hinge region of the AR—including p300, PCAF, TIP60, and N-acetyltransferase arrest-defect 1 protein.115-117 Such acetylation aids the nuclear translocation of the AR, as well as its interaction with its key co-activators. Conversely, the AR’s association with corepressors is enhanced when AR is underacetylated. There is recent evidence that epigallocatechin-3-gallate (EGCG), the chief catechin in green tea, can act as an inhibitor of acetyltransferase activity in prostate cancer cells; at 25 µM, this activity was suppressed by about 80%. 118 In particular, EGCG inhibits AR and histone acetylation, reduces the ability of androgens to promote nuclear uptake of the AR, and suppresses AR-mediated gene transcription. 118 This novel finding may rationalize, at least in part, previous studies reporting that administration of EGCG to nude mice with prostate cancer xenografts (1 mg ip daily or several days a week) inhibits nuclear translocation of the AR and suppresses expression of AR-dependent genes such as PSA.119,120 Whether tolerable oral doses of EGCG can suppress AR acetylation in prostate cancer patients is not yet known, but when 800 mg EGCG daily was administered to such patients for about 6 weeks prior to radical prostatectomy, PSA levels declined by an average of 10%. 121 It should be noted that EGCG also has been reported to reduce the protein expression of the AR in prostate cancer xenografts 120 —possibly reflecting a decrease in NF-κB activity owing to inhibition of p65 acetylation.122,123 It is unclear whether EGCG might suppress inhibitory acetylations of Hsp90, and hence antagonize the utility of sulforaphane; in any case, its net impact appears to decrease the AR activity of prostate cancers.
Curcumin can function as an inhibitor of p300 acetyltransferase activity and has been found to diminish AR transactivational activity both in prostate cancer cell cultures and in prostate cancer xenografts.124,125 The clinical pertinence of these findings can be doubted, however, owing to the poor pharmacokinetics of oral curcumin in humans; even when its inefficient absorption is enhanced, little native curcumin reaches the blood owing to rapid reductive metabolism in the intestinal mucosa. 126
Targeting Growth Factor Pathways
As has been noted, signaling via various growth factor-stimulated pathways can phosphorylate the native AR in ways that render it active in the absence of androgens. Therefore, agents that could intervene in some of these pathways may have potential for controlling androgen-independent prostate cancer. EGFR and the related family of receptors (HER2/HER3) appear to be logical targets. Unfortunately, erlotinib has not been shown to be effective in androgen-independent prostate cancer to date, 127 likely because such prostate cancers upregulate their expression of HER2/HER3 receptors, or lose PTEN expression.128,129 However, it is hoped that novel agents with broader activity for the range of HER receptors, alone or in conjunction with erlotinib, may eventually demonstrate some clinical utility in this regard.130,131
Overview
The large majority of prostate cancers, including those considered “androgen-independent,” are addicted to AR activity; this may constitute an “Achilles heel” for this cancer. Strategies that depend on inhibition of androgen synthesis, or that block AR interaction with androgens, ultimately fail owing to genetic or epigenetic shifts within the cancer that give rise to AR variants not dependent on androgens for activity. Development of a drug that blocks the transactivational activity of AR—including the splice-variants that lack a ligand-binding domain—may provide a powerful tool for blocking prostate cancer growth; its activity hopefully will prove more durable than that of other AR antagonists, but this remains to be seen. A range of feasible measures that suppress synthesis of AR by downregulating NF-κB activity (eg, polyphenols, melatonin, spirulina, salsalate), that promote proteasomal degradation of AR (eg, omega-3, berberine, sulforaphane), or that inhibit AR acetylation (EGCG) may also have potential for controlling prostate cancer by targeting AR activity.
It should of course be noted that AR activity is a necessary but not sufficient condition for the emergence and survival of prostate cancer; therapeutic strategies that downregulate AR activity can be complemented by measures that target other pathways promoting cancer spread. Quite aside from its impact on AR expression, NF-κB activity can work in a number of ways to promote survival and aggressive behavior in prostate cancer; hence, inhibitors of NF-κB may not only slow proliferation but also upregulate apoptosis and lessen invasiveness. Moreover, some of the agents mentioned above have potential for intervening in other signaling pathways—downstream from cytokine or growth factor receptors—that promote survival and aggressiveness in many prostate cancers. (And certain of these signaling pathways increase androgen-independent AR activity via interaction with the N-terminal domain, or via other mechanisms, as discussed above.) Additional agents may have some utility in this regard; of particular interest is the HIV protease inhibitor nelfinavir, which, in a concentration (10 µM) approximately twice the standard clinical plasma level, has been found to decrease activation of both Akt and Stat3 and suppress AR transcription activity in LNCaP human prostate cancer cells. 132 Other feasible adjuvant measures with potential for aiding control of prostate cancer include lycopene, vitamin D, soy isoflavones, metformin, low-dose aspirin, and a low-fat plant-based diet.133-152
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
