Glucocorticoid Receptors Are Making a Comeback in Corticotroph Tumorigenesis.

Commentary

Partial resistance to the negative glucocorticoid (GC) feedback is a defining feature of corticotroph tumors and the rationale behind utilizing the dexamethasone suppression test for the differential diagnosis of Cushing’s disease. In the mid-1990s, the discovery of a mutation in the coding region of the NR3C1 gene that encodes for GC receptor (GR) α in a case of progressive corticotroph tumor growth after bilateral adrenalectomy (Nelson’s syndrome) proposed that genetic defects might play a role in the development of GC resistance in corticotroph tumors (1). However, although subsequent studies did report loss of heterozygosity in the coding region, they failed to detect additional NR3C1 mutations, concluding that these must be extremely rare events and therefore unlikely to contribute to corticotroph tumorigenesis [reviewed in (2)].

The advent of next-generation sequencing brought GR back into the spotlight; while some whole exome sequencing efforts did not pick up NR3C1 mutations, other detected them in up to 10% of cases [(3,4); reviewed in (5)]. A meta-analysis calculated the presence of somatic NR3C1 mutations in 6.2% of corticotroph tumors concluding that even if infrequent, they are not the extremely rare sporadic events they were once thought to be (5).

Still, we had no information how these GR variants may affect adrenocorticotropin synthesis and corticotroph tumor growth. In a recent study, Miao et al tackled this issue by investigating NR3C1 gene status in corticotroph tumors and studying in vitro how GR variants affect GC response in corticotroph tumor cells (6). They found mutations in 3 out of 49 cases (6.1%), 1 in the DNA binding domain resulting in a truncated protein (p.R469X) and 2 in the ligand binding domain (p.D590G and p.Y693D). Furthermore, they reported that overexpressing these 3 variants in the immortalized murine corticotroph tumor AtT-20 cells hampered in certain instances their response to dexamethasone treatment in terms of suppression of cell viability, POMC transcription, and adrenocorticotropin secretion (6). Not all variants affected dexamethasone response, and this is equally noteworthy. In this respect, it is of interest that overexpressing the truncated GR variant (p.R469X) compromised the antiproliferative action to dexamethasone and the patient carrying this variant presented with recurrent tumor more than a decade after surgery. In contrast, the impact of the truncated p.R469X variant on dexamethasone suppression of POMC transcription was minimal.

Interestingly, the same p.R469X mutation was found in heterozygous state in a family with GC resistance (or Chrousos syndrome), a syndrome characterized by partial insensitivity to GC caused by NR3C1 mutations [(7); reviewed in (8)]. The p.R469X mutant GR did not translocate to nucleus after dexamethasone treatment and showed no DNA binding and GC response element–dependent transcription, while its selective degradation through nonsense-mediated messenger RNA decay resulted in GR haploinsufficiency in the affected patients (7).

These first data open several questions on the GR regulatory pathways: How do the variants affect GR transcriptional activity? How do they alter GR target promoter occupancy and subsequently gene expression? Moreover, given the lack of outstanding clinical features, why is the effect of these GR mutations so diverse and subtle? Eventually, what does this mean for GC antagonism?

In corticotroph tumors, GC inhibit cell viability, and this is of high importance for corticotroph pathophysiology [example in (9)]. In contrast, the impact of GR on tumor growth is less straightforward in nonhematological cancers, where it is described as antiproliferative/proapoptotic or prosurvival depending on the tumor type [reviewed in (10)]. The reason for this diverging action is not well understood and has been attributed to the complexity of the genomic as well as the less understood nongenomic GR action [reviewed in (11)]. In this context, tumor-associated human GR mutants provide us with invaluable tools to dissect the GR mode of action not only in corticotroph but also in other solid tumors.