FGFR4 Polymorphism as Molecular Determinant of the Efficacy of mTOR Inhibitors In GH-Secreting Pituitary Adenomas

Pituitary adenomas are intracranial benign tumors with high prevalence, reaching up to 20% in the general population, and are among the most studied human tumors. However, several puzzling issues concerning their pathogenesis and the determinants of response to therapy still need to be unveiled. For example, although a powerful pharmacological arsenal is available for specific adenoma histotypes, including somatostatin and dopamine agonists, too many unresponsive or incompletely controlled cases are world-wide reported (1). Moreover, due to the absence of predictive markers, the selection of a given therapy is still based on investigator's choice. In fact, the simple somatostatin receptor expression pattern, the most common molecular evaluation used, cannot encompass all the pituitary adenoma molecular variants that influence the positive clinical outcome in response to the available drugs (2). Furthermore, although pituitary adenomas are classically considered of monoclonal origin, recent studies questioned this assumption. Starting from the identification of pituitary stem cells and the relative niche in adult pituitary (3, 4), cellular heterogeneity was proved to persist also within human pituitary adenomas, and the possibility that stem-like tumor cells drive pituitary tumorigenesis has been postulated (5).

Several genetic alterations responsible for pituitary adenoma development were identified (6). These include, among others, activating mutations (GNAS or USP8) or overexpression (pituitary tumor transforming gene or NeuroD1) of oncogenes, hypofunction of oncosuppressors (growth arrest and DNA damage inducible gene-45γ, zinc finger regulator of apoptosis and cell cycle arrest-1), or alterations in the expression of cell cycle regulators (p16, p21, p27).

Moreover, the activity of growth factor receptors and down-stream kinase cascades is increased in different pituitary adenoma subtypes, further contributing to tumor cell gain of function, leading to proliferative and survival advantages or increased hormone secretion. In particular, fibroblast growth factors (FGFs) and their receptors (7) play a relevant role in pituitary tumor development. FGFs are a large family of single chain polypeptides classified in 3 major groups: 1) autocrine/paracrine FGFs (15 molecules, including FGF1–FGF10, FGF16–FGF18, FGF20, and FGF22), acting locally on the tissue where they are produced; 2) endocrine FGFs (FGF19, FGF21, and FGF23), reaching target cells through the blood stream; and 3) noncanonical FGFs (FGF11–FGF14) that do not bind their cognate receptors but act intracellularly on signal transduction pathways. Canonical FGFs elicit their biological effects via 4 tyrosine kinase receptors (FGF receptor [FGFR]1–FGFR4), showing high redundancy of binding: one receptor may bind to several FGFs with similar affinity, and each FGF may bind to several distinct receptors. FGF/FGFR interaction exhibits diverse, yet relevant, roles in pituitary function. Repression of FGF10 or FGFR2 expression prevents normal pituitary development, whereas FGF2 overexpression is frequently observed in pituitary adenomas. However, several studies focused on FGFR4, because a truncated, constitutively active isoform was detected in about 40% of pituitary adenomas, but not in normal pituitary, and its expression was correlated with increased tumor aggressiveness. Moreover, FGFR4 single nucleotide polymorphism at codon 388, in which a glycine is substituted by an arginine within the transmembrane region, was identified in several human tumors, including pituitary adenomas (8). FGFR4(R388) overexpression increases GH expression in GH4C1 cells, and, in mice, the knock-in of this receptor variant induces the development of GH-secreting tumors. Different signal transduction pathways mediated by FGFR4 were proposed to affect pituitary tumorigenesis, because signal transducer and activator of trasnscription 3 (STAT3) activation increases cell proliferation after serine phosphorylation, whereas tyrosine phosphorylation potentiates GH secretion. In this context, in the current issue of Endocrinology, Jalali et al (9) analyzed the influence of different FGFR4 genotypes on the response to mammalian target of rapamycin (mTOR) inhibitors in GH4C1 tumors. mTOR is a key regulator of the growth of several solid neoplasia, including neuroendocrine tumors, allowing the development of molecularly targeted drugs, some of them already available for clinical use (ie, everolimus or temsirolimus). In pituitary adenomas, mTOR signaling is up-regulated and is considered a molecular marker for these tumors. Preclinical studies showed efficacy of mTOR inhibitors in human pituitary adenoma cells (10, 11), and, in few cases, these drugs were tested in patients with temozolomide-resistant pituitary tumors (12). However, how the genetic cell context affects everolimus efficacy in pituitary adenomas is still unknown. In the Jalali study, the authors analyzed the in vivo efficacy of everolimus (RAD001) in GH4C1 cells expressing either G388 or R388 FGFR4 variants. Interestingly, the presence of this single nucleotide polymorphism did not affect tumor development and growth rate, or responsiveness to everolimus. These effects were correlated with a biased inhibition of mTOR activity by everolimus, resulting in the abolishment of S6 but not 4EBP1 phosphorylation. On the other hand, everolimus affected GH synthesis and secretion only in cells expressing the polymorphic FGFR4(R388). Thus, this amino acid substitution within FGFR4 selectively controls a specific mTOR-dependent pathway leading to GH production. This dissociation of effects well matches the differential role of STAT3 in pituitary cell proliferation and GH secretion, regulated by serine or tyrosine phosphorylation, respectively. Thus, it was hypothesized that wild-type FGFR4(G388) induces a biased mTOR-dependent signaling toward proliferation-induced pathways, whereas GH secretion is differently regulated. In this perspective, only in the presence of the G388R polymorphism, FGFR4 controls hormone secretion. If confirmed in human tumors, this observation could support the clinical use of mTOR inhibitors in unresponsive patients bearing GH-secreting adenomas which express FGFR4(R388).

Finally, another relevant information is coming out from the analysis of tumor formation driven by FGFR4 isoforms. In particular, G388R polymorphism, which does not affect tumorigenicity, seems to dictate the adenoma subtype that will be generated, with the R388 variant favoring GH-secreting tumors and the wild-type G388 isoform mainly prolactin-releasing adenomas. This alternative fate of tumor cell differentiation could shed light on the possible interference of signal transduction mechanisms in the pituitary adenoma development. Therefore, we can hypothesize that, as shown in several solid tumors, putative undifferentiated (or in some cases dedifferentiated) cells, acquiring proliferative gain of function, can develop different tumor phenotypes according to the activated signaling pathways, supporting the hypothesis of the presence of stem-like adenoma cells also at the origin of pituitary tumorigenesis. In fact, if this will be confirmed to occur also in human tumors, adenoma formation might not merely reflect the expansion of single cells with uncontrolled proliferation, but it should require the presence of undifferentiated cells able to acquire, according to genotype or functional alterations, multiple differentiated adenoma subtypes.

For article see page 3577

Acknowledgments

Disclosure Summary: The author has nothing to disclose.

Abbreviations

 
• FGF

  fibroblast growth factor

 
• FGFR

  FGF receptor.