Does the autonomic nervous system contribute to the initiation and progression of prostate cancer?

In the July 12 issue of Science magazine, researchers from the Albert Einstein College of Medicine, the Mount Sinai School of Medicine, the Durham VA Medical Centre and Duke University published an elegant study demonstrating that the sympathetic nervous system, acting through β₂ and β₃-adrenoceptors in the prostate, plays an important role in the initiation of prostate cancer, while the parasympathetic nervous system plays a role in the dissemination of tumour metastases via M₁ muscarinic receptors. These findings are significant because they indicate that receptors associated with the autonomic nervous system may be viable targets for prostate cancer therapy.

The influence of the autonomic nervous system on physiological systems changes with age. In particular, the sympathetic nervous system is implicated in disorders of the aging male, including hypertension and benign prostatic hyperplasia. In both disorders, the sympathetic nervous system shows signs of overactivity and symptoms can be controlled by the use of therapeutic drugs which block the effects mediated by adrenoceptors (e.g., α₁-adrenoceptor antagonists, β-blockers). It has been hypothesized previously that there may be a similar association between increased sympathetic activity and prostate cancer.^1,2

Cells innervated by the sympathetic nervous system respond to the neurotransmitter noradrenaline via nine different adrenoceptor (AR) subtypes, the α_1A-AR, α_1B-AR and α_1D-AR, three α₂-AR subtypes, and β₁-AR, β₂-AR and β₃-ARs. These receptors are also activated by adrenaline released from the adrenal gland in response to behavioural stress. There is robust evidence linking activation of the β₂-AR with progression and/or metastasis of breast cancer,³ ovarian cancer^4,5 and acute lymphoblastic leukaemia,⁶ but these studies have focused on animals with high circulating levels of adrenaline. A recent study has also shown that in mouse models of prostate cancer, immobilisation stress and systemic adrenaline promote prostatic intraepithelial neoplasia (PIN) and resistance of tumours to apoptosis normally induced by the PI3K inhibitor ZSTK474.⁷

Unlike studies based on tumour xenografts in the flank of immunocompromised mice, the work described by Magnon et al.⁸ employed orthotopic injection of human PC-3 prostate cancer cells directly into the ventral prostate gland, mirroring the tumour microenvironment found in human disease. The PC-3 cells were labelled with luciferase so that they could be visualized and quantified in vivo or ex vivo using bioluminescence imaging. Tumours were readily detected after 5 weeks. At 11 weeks, tumour-infiltrating sympathetic fibres arising from the surrounding normal tissue were apparent (characterized by positive staining for the noradrenaline-synthesizing enzyme tyrosine hydroxylase), as were intratumour parasympathetic fibres identified by positive staining for vesicular acetylcholine transporter.

Both chemically-induced sympathectomy using 6-hydroxydopamine, or surgical sympathetic denervation, prevented the development of tumours in the prostate. The target cells for sympathetic stimulation were shown to be stromal cells rather than the PC-3 tumour cells, as recipient mice with genetic deletion of β₂-/β₃-ARs also displayed a marked decrease in tumour development. This is interesting, as the previous study on prostate cancer⁷ indicated that β₂-ARs expressed by tumour cells mediate the effects of behavioural stress. There are key differences between these studies,^7,8 namely, the sites of tumour injection (orthotopic injection vs. flank xenograft), the stimulus investigated (increased sympathetic innervation of the prostate vs. acute behavioural stress leading to increased circulating adrenaline), and the time course of the measured responses (resistance to apoptosis over a period of 6–72 h vs. tumour development over the course of 11 weeks).

Magnon et al.⁸ also investigated the functional significance of parasympathetic nerve infiltration in prostate tumours. Stimulation of cholinergic pathways using the acetylcholine receptor agonist carbachol promoted metastasis rather than affecting primary tumour development. This effect on metastasis was blocked by the nonselective muscarinic receptor antagonist scopolamine as well as the M₁ muscarinic receptor antagonist pirenzepine. In immunocompromised mice with genetic deletion of the M₁ receptor, carbachol no longer promoted metastasis. Again, cells within the recipient prostate rather than the injected tumour cells mediated the effects of parasympathetic stimulation.

The findings above were confirmed in Hi-Myc mice that spontaneously develop prostate cancer.⁹ Hi-Myc mice displayed PIN by 2 weeks after birth, progressing to invasive adenocarcinomas within 6 months. Chemical sympathectomy of Hi-Myc mice prior to one month of age led to greatly reduced tumour development, whereas surgical or chemical sympathectomy at 2 or 5 months had no effect on PIN or invasive cancer, indicating that the sympathetic nervous system is important primarily in the initiation and early stages of prostate tumour development. In contrast, Hi-Myc mice treated with carbachol even at 3 months of age displayed substantial increases in PIN and progression to invasive lesions. These effects of carbachol were abolished in M₁ knockout mice or in wild-type mice treated with the antagonist pirenzepine. Thus in both Hi-Myc animals and immunocompromised mice injected with PC-3 cells, sympathetic and parasympathetic activity within the prostate promote disease progression via complementary but distinct mechanisms. This characteristic of malignant disease is not observed in normal physiology, as the sympathetic and parasympathetic nervous systems generally produce opposing responses. In normal prostate gland, for example, parasympathectomy leads to increased prostate weight, while sympathectomy inhibits growth.¹⁰

Magnon et al.⁸ addressed the clinical relevance of their mouse data by examining prostatectomy samples taken from 43 treatment-naive patients with prostate cancer. Patients were classified as having low (30) or high (13) risk based on prostate-specific antigen levels, Gleason score and disease stage. Prostate tissue surrounding the cancers in high-risk patients was found to have increased noradrenergic fibre density, while cholinergic fibres were restricted to the tumours. A high density of both noradrenergic and cholinergic fibres was associated with a higher proliferative index. Interestingly, correlations were also found between nerve fibre density and preoperative prostate-specific antigen levels, time until recurrence and tumour spread. The authors suggest that assessment of autonomic nerve density may provide a useful diagnostic tool for prediction of tumour aggressiveness. This would be a major prognostic advance, as it may be feasible to use more conservative treatment regimes in patients with low nerve density combined with low risk based on Gleason score and prostate-specific antigen levels. Such assessment would also highlight high-risk patients most likely to benefit from treatment with β-blockers or muscarinic antagonists.

Recent epidemiological evidence is consistent with the findings of this study, as treatment with β-blockers is associated with improved survival of prostate cancer patients.^11,12 However, both clinical studies show a β-blocker effect on high-grade, aggressive and metastatic tumours, in contrast to the tumour initiating effect of β-adrenoceptors in the mouse study.⁸ Furthermore, it is noteworthy that most clinically used β-blockers are somewhat selective for the β₁-AR subtype, in contrast to the β₂- and β₃-AR subtypes found to initiate tumours in the mouse study. Note that all three adrenoceptor and muscarinic receptor subtypes investigated in this study were involved in tumour initiation and/or progression. There is also epidemiological evidence to suggest that α₁-AR antagonists may be of benefit in prostate cancer,^13,14 so their involvement in disease development also requires investigation.

Given the findings of this study,⁸ the targeting of G protein-coupled receptors associated with the autonomic nervous system for the treatment of prostate and other cancers may be a viable proposition. This is exciting, as potent and selective antagonists for these receptors are already available and generally well tolerated in patients, in contrast to the severe side effects associated with commonly used chemotherapeutic agents. The authors of the present study have not ruled out contributions by other G protein-coupled receptor subtypes; thus, future studies should investigate additional receptors as well as the signalling pathways that they activate, as these may offer even more novel therapeutic targets for the treatment of cancer.