Extra-pituitary ACTH secretion is a well-known phenomenon that represents the biological basis for the so-called ACTH-dependent ectopic Cushing syndrome (CS) and is associated with a variety of neoplastic conditions [6]. The molecular and cellular pathways leading to ACTH expression outside the pituitary gland are matter of ongoing debate, considering that their clarification would allow to identify potential therapeutic targets, useful to treat a complex and severe disease. In the last years, the identification of transcription factors driving cell lineage differentiation in the pituitary gland has elucidated the mechanisms underlying hormone secretion in normal and neoplastic cells, besides providing robust tools for the pathological diagnosis of pituitary diseases [4]. In pituitary corticotroph cells and tumors, the transcription factor TPIT seems to be essential in inducing POMC expression and, consequently, ACTH production [1]. This mechanism, however, does not explain why a subset of corticotroph PitNETs, which are indeed ACTH and TPIT-positive by immunohistochemistry, do not cause an excess of circulating ACTH and, ultimately, are not related to CS [11]. In turn, to our knowledge, no data is currently available on the presence and role of TPIT in extra-pituitary ACTH-producing neoplasms.
This study was designed with the aim to explore TPIT expression, at the protein and the mRNA level, in a series of extra-pituitary ACTH-producing tumors, either CS-associated or not, mainly represented by pulmonary and pancreatic NETs. Our results show complete absence of TPIT expression in PanNETs, which were all CS-associated, as well as in the pheochromocytomas. In contrast, half of the lung ACTH-producing tumors showed some TPIT expression, although, except for one case, no definite relationship was found either between immunohistochemistry and molecular analysis or between TPIT expression and the presence or absence of Cushing syndrome. To exclude the possible, albeit remote, event of a non-specific immunostaining for TPIT in ACTH-secreting Lu-NETs, we decided to perform the immunohistochemical study also on a comparable number of ACTH-negative LuNETs and PanNETs and we did not find any positive case, supporting the reliability of our data.
Our results suggest, as a whole, that ACTH-secreting PanNETs are TPIT-negative. In contrast, LuNETs seem to show, at least partially, some TPIT expression, suggesting that a “corticotroph-like” transcriptional program may be active in such tumors. The lack of complete overlap between protein and mRNA expression may be explained in several different ways, including technical and biological reasons, as we have already discussed elsewhere [12]. However, the analysis of all other tumor types that we have explored (ACTH-positive and ACTH-negative PanNETs and ACTH-negative LuNETs) supports that TPIT expression may be really limited to LuNETs and that ACTH secretion in these tumors may be, at least in part, related to TPIT activity. In PanNETs, other mechanisms of ACTH production may be active. In fact, TPIT function seems not to be the only driver of ACTH production and secretion even in corticotroph PitNETs, where TPIT is always expressed but the amount and functionality of ACTH secretion is variable. Indeed, the clinical spectrum of these PitNETs is wide, not only in terms of endocrinological function, but also of oncological aggressiveness [4, 11]. In this regard, Araki and co-workers have demonstrated the existence of two distinct POMC gene promoter regions, one of which is bound by TPIT and the other by E2F1 [9, 13, 14]. Interestingly, the first promoter is demethylated in all POMC-expressing cells and is highly demethylated only in pituitary ACTH-secreting tumors harboring the ubiquitin-specific protease 8 (USP8) mutation, whereas the second promoter region is highly methylated in non-functioning corticotroph PitNETs, partially demethylated in normal corticotrophs, and highly demethylated in pituitary and ectopic ACTH-secreting tumors [13]. In addition, it has been shown that the demethylation of the second promoter was related to markers of aggressive phenotype (large size, invasion, Crooke’s changes) in corticotroph PitNETs [13]. Moreover, the same authors were able to show, in vitro, that POMC transcription in human extra-pituitary ACTH-producing cell lines is independent of TPIT/PITX1 function and is E2F1-mediated [9]. These findings are particularly interesting in the context of our study, if we consider that ACTH-producing PanNETs (which are TPIT-negative and, conceivably, should show high demethylation of the second POMC promoter) are, in general, oncologically aggressive neoplasms and are always related to signs and symptoms of CS [10]. In contrast, ACTH-producing LuNETs (which are TPIT-positive, at least in part, and in which POMC expression may be related to the demethylation of the first promoter) do not show a different course compared with non-ACTH-producing grade- and stage-matched LuNETs and may be non-functioning [8], whether or not they express TPIT.
In conclusion, this study further highlights the clinical and biological heterogeneity of extra-pituitary ACTH secretion and ectopic CS and supports the idea that the differences between ACTH-secreting PanNETs and LuNETs mirror distinct molecular mechanisms underlying POMC expression. In fact, our results point towards the recognition of a real corticotroph-like phenotype of ACTH-producing LuNETs, that is not a feature of ACTH-producing PanNETs.