Pathogenesis of canine cortisol-secreting adrenocortical tumors

Abstract

In dogs, hypercortisolism is one of the most frequently observed endocrine disorders, with an estimated incidence of about 1-2 cases per 1000 dogs per year. Approximately 15% of these cases is due to a cortisol-secreting adrenocortical tumor (AT). Cortisol-secreting ATs are characterized by uncontrolled growth and excessive ACTH-independent cortisol secretion, the molecular mechanisms of which are largely unknown. Treatment options for canine ATs are limited and, thus far, no reliable prognostic markers have been identified. The higher incidence of cortisol-secreting ATs in dogs when compared to humans, raises the question whether the dog might be a good spontaneous animal model for human adrenocortical carcinoma (ACC). In this thesis the molecular mechanisms behind ACTH-independent cortisol secretion are addressed. The ACTH/cyclic AMP (cAMP) signaling pathway regulates cortisol secretion in the normal adrenal cortex. Activating mutations in the gene encoding G-protein stimulatory subunit alpha (GNAS) were found in one third of the ATs. This could well explain the ACTH-independent cortisol secretion in the affected subset of ATs and may also play a role in adrenocortical tumorigenesis. Activation of cAMP signaling may also be due to altered hormone receptor expression. In recurrent carcinomas, a remarkably higher relative expression of dopamine receptor DRD1 was detected. As DRD1 activates adenylate cyclase through G-protein stimulatory subunit α, its overexpression may represent an alternative mechanism for cAMP activation. The expression of one of the key effectors of ACTH/cAMP signaling, steroidogenic factor 1 (SF-1), was higher in recurrent carcinomas. Recurrence was mostly due to metastasis, therefore a role for SF-1 expression in autonomous cell growth and metastasis is likely. Our data provide a preclinical rationale for evaluating the potential of SF-1 inverse agonists to reduce AT cell proliferation and steroidogenesis. Angiopoietin 2 (ANGPT2) is a downstream target of SF-1. In canine ATs, the mRNA expression of ANGPT2 and the ANGPT2/ANGPT1 ratio were higher. ANGPT2 protein expression was detected in tumor cells of all ATs, while ANGPT receptor Tie2 was expressed in the tumor vascular endothelial lining. Taken together, our results indicate an ANGPT2-induced angiogenic state which is likely to contribute to AT growth and metastasis. The use of ANGPT2-specific anti-angiogenic drugs may thus hold therapeutic benefits. Another downstream target of cAMP/PKA activation is the canonical Wnt pathway. In approximately one third of the ATs, activation of this pathway was detected. The cause of activation was unclear, only 2 ATs harbored an activating β-catenin mutation. However, the activation of the Wnt pathway in a substantial portion of ATs may well contribute to uncontrolled growth and metastasis, and thus represent an attractive therapeutic target. The phosphatidylinositol 3 kinase (PI3K) pathway is induces tumor cell proliferation. Target gene expression indicated PI3K activation in carcinomas, which may partly be caused by the tendency toward higher ERBB2 expression. In carcinomas with recurrence, IGF binding protein 5 (IGFBP5) and inhibitor of differentiation 1 and 2 (ID1 and 2) were expressed at higher levels, and might serve as prognostic markers. ERBB2, ID1 and ID2 may represent attractive new therapeutic targets.