2.1. Why Silent?
There are several hypotheses on the silencing mechanism of SCAs. Kovacs et al.
[14] observed that SCA cells have a high number of lysosomes in the cytoplasm and show fusion of these lysosomes with secretory granules as evidenced by electron microscopy, leading to the hypothesis that ACTH is destroyed before it can be released. The possibility that a defective
POMC gene was responsible for the silent nature of SCAs was denied by ribonuclease mapping analysis as no mutation was detected of
POMC [15]. There is also speculation that SCA cells may not derive from anterior pituitary lobe ACTH-expressing cells, which give rise to CD cells, and rather they may arise from the pars intermedia POMC-expressing cells. Therefore, SCAs may have distinct characteristics from CD adenomas
[16].
Recent advances have focused on the transcription, post-transcriptional regulation of POMC, and post-translational regulation of ACTH to explain the silencing mechanism of SCAs.
Araki et al.
[17] observed that an additional regulatory region near the
POMC gene, functioning as a second promoter of
POMC, is highly methylated in SCAs. In contrast, it is highly demethylated in pituitary ACTH-secreting tumors, suggesting that methylation of the second promoter might play a role in transcription of the
POMC gene, possibly silencing
POMC in SCAs.
Given the fact that SCAs and CDs in most studies expressed similar
POMC [11][18][19], a most likely and feasible hypothesis has been established that the function or expression of prohormone convertase (PC) 1/3, a POMC-processing enzyme, may be disturbed in SCAs. PC1/3 has been extensively studied in the literature
[11][18][19][20][21][22][23]. PC1/3, encoded by the
PCSK1 gene, is involved in the processing of
POMC into mature and biologically active ACTH
[23]. Concerning
POMC post-transcriptional regulation, Tateno et al.
[20] and Jahangiri et al.
[11] observed that the expression of PC1/3 was present at 15-fold and 30-fold higher levels in CDs than SCAs, respectively, indicating defects in the conversion of
POMC to ACTH in SCAs. ISABIAL’s lab
[18] also observed lower
PCSK1 and PC1/3 expression in SCAs compared with CDs, reinforcing the hypothesis that defective PC1/3 is implicated in the impaired
POMC post-transcriptional regulation.
MicroRNAs(miRNAs) have also been proposed to play a potential role in the silencing mechanism of SCAs. miRNome analysis revealed differences between SCAs and CDs as they were arranged in two clusters with a different miRNAs signature, suggesting a potential part played by the miRNAs in the pathophysiology of different corticotroph adenomas
[24]. Comparing 23 SCAs and 24 CDs, García-Martínez
[1] observed significantly higher levels of miR-200a and miR-103 in SCAs than in CDs (indeed, both miRNAs appeared normo-regulated in SCAs while downregulated in CDs), and suggested these miRNAs may serve as potential diagnostic markers to distinguish SCAs from CDs. In SCAs, miR-200a and miR-103 could participate in the regulation of
POMC expression.
POMC transcription is associated with the
PKA-MAP3K8-MEK-ERK1/2-NGFIB pathway
[25], while miR-200a and miR-103 negatively regulated this pathway by inhibiting the
MAP3K8 expression. Additionally, miR-103 correlated negatively with the expression of
MEK. Therefore, miR-200a and miR-103 are involved in the post-transcriptional regulation of
POMC expression in SCAs by inhibiting its upstream pathway
[1]. Moreover, the authors demonstrated other miRNAs, such as miR-375, miR-488, and miR-383, could participate in the regulation of
POMC transcription, probably through their association with different pathways and relevant transcription factors.
Post-translational regulation of ACTH has also been studied
[11][18][20][26]. The proconvertase PC2 is proposed to contribute to the silencing mechanism of SCAs. Encoded by
PCSK2, PC2 is an endoproteolytic enzyme responsible for the processing of biologically active ACTH into α-melanocyte-stimulating hormone and corticotropin-like intermediate lobe peptide in the intermediate lobe
[27]. Most studies reported a decreased degradation of ACTH in corticotroph adenomas (SCAs and CDs) compared with other subtypes of NFPAs because PC2 mRNA levels in SCAs and CDs were less than those in other subtypes of NFPAs, without differences between SCAs and CDs
[11][20]. Further study by García-Martínez et al., however, found a stronger correlation between
PCSK2 and PC2 expression in SCAs than in CDs. Moreover, the authors found that SCAs exhibited higher
PCSK2 expression than CDs (though not statistically significant). These results are consistent with a more efficient and accelerated degradation of ACTH in SCAs than in CDs, thereby partly accounting for silent manifestations of Cushing’s syndrome in SCAs
[18]. In addition, more peptidylglycine α-amidating monooxygenase (PAM) and carboxypeptidase E (CPE) expression in SCAs than in CDs, both involved in the post-translational regulation of ACTH, might lead to the inability to secrete ACTH in these SCAs.
Another interesting phenomenon is that SCAs can transform into active CDs after years of silence. The phenotypic transition has been related to three mechanisms: first, an increase in
PCSK1 and PC1/3 expression has been reported in three cases of SCAs that evolved during follow-up to functioning ones, thus underscoring the role of PC1/3 as one of the most feasible mechanism for silencing corticotroph adenomas
[28]. Second, though decreased in SCAs, the accumulative effect of PC1/3 could still process POMC into biologically active ACTH if time allows. Little by little, ACTH could gradually and eventually rise to a higher level to cause clinical manifestations of Cushing’s syndrome. Indeed, the transformation took a median of 30 months, with the longest interval being 10 years summarized by Zheng et al.
[29]. Additionally, a most recent mechanism proposed by Araki et al.
[17] suggested that the sequential hypomethylation of the second promoter might activate
POMC and promote ACTH secretion in the phenotypic transition.
2.2. SCAs Could Represent Silent Cortico-Gonadotroph Adenomas?
Cooper et al.
[30] proposed a pathologic and clinically distinct typification of SCAs as silent cortico-gonadotroph adenomas as they exhibit features consistent with both the corticotroph and the gonadotroph cell lines.
In this study, SCAs have been shown to be immunopositive for corticotroph markers, ACTH, NeuroD1, as well the as gonadotroph markers, SF-1 and DAX-1. Some cells co-expressed ACTH with either SF-1 or LH. In contrast, their secreting counterparts are immunonegative for DAX-1. It is noteworthy that only 44% of SCAs are immunoreactive for T-PIT, which could be attributed to T-PIT antibodies that were not readily available and reliable when the research was conducted. Under electron microscopy, SCAs demonstrated ultrastructural features consistent with both corticotroph and gonadotroph cells. It is thus hypothesized that SCAs derive from an intermediary cell that shares both gonadotroph and corticotroph characteristics.
Intriguingly, further investigation reported that 20.45% (9/44) of the corticotroph adenomas expressed
GATA2, a gene of the transcription factor involved in the development of gonadotrophs
[31], also considered a marker of gonadotroph adenomas
[32]. Eight of the nine cases that expressed
GATA2 are SCAs. The expression of
GATA2 in SCAs suggests that this subtype could have a component of gonadotrophic features
[31].
Coincidentally, by using a canonical transcriptome signature for each pituitary cell type, Neou et al. observed that 88% (7/8) of SCAs displayed both corticotroph and gonadotroph signatures
[24]. It was later confirmed by co-expression of GATA3 (a genomic paralogy of GATA2) to be characteristic of pituitary gonadotroph adenomas
[33], ACTH, and T-PIT IHC in these tumors. The result highlights that SCAs manifest both corticotroph and gonadotroph cytologic features, supporting the arguments by some authors
[31][30]. Further research by Ricklefs et al.
[34] analyzing genome-wide DNA methylation profiles also demonstrated that SCAs show a spatial relationship with gonadotroph adenomas. These findings suggest that SCAs might arise from a specific cell lineage distinct from other corticotroph adenomas and that they could represent a new pituitary tumor subtype named cortico-gonadotroph adenomas, but it still requires further investigations in a larger cohort of SCAs.
2.3. Progression and Growth
SCAs are among the five types of “high-risk” tumors proposed by the 2017 WHO classification system that demonstrate aggressive behavior. The concept of aggressiveness in SCAs is a composite of high invasion, proliferation, progression, and frequent recurrences.
Over the past years, advanced techniques have allowed the exploration of either mono-omic or multi-omics, such as genomic, transcriptomic, epigenomic, metabolomic, proteomic, and lipidomic, to study the characteristics of NFPAs
[35][36][37][38][39][40][41]. Though attempts to explore the underlying mechanism of invasiveness and aggressiveness of SCAs are still limited and inconclusive, researchers have already witnessed some advances, especially those using various -omics approaches to provide novel insights into the invasive and aggressive behavior of SCAs.
Various factors have been implicated in the development of SCA progression and growth including cyclin-dependent kinases inhibitor 2A, galectin-3, kallikrein 10, osteopontin, and O6-methylguanine-DNA methyltransferase. They have been reviewed by Ben-Shlomo et al.
[4], but they are merely associative and remain to be further elucidated. Of note, as is pointed out in the 2017 WHO edition
[7], galectin-3 has been reported to be weak or absent in SCAs compared with CDs and thus can be used as an important marker to distinguish between the two forms
[42]. However, the direct effect of the absence of galectin-3 on SCAs’ behavior remains further elucidated. Since NFPAs are best studied in the emerging multi-omics studies, here, researchers summarize insights from the recent developments in SCA progression and growth.
Recently, Wang et al. investigated the invasiveness-associated lipid alterations in SCAs through ultra-performance liquid chromatography-mass spectrometry (UPLC-MS)-based lipidomic analysis
[43]. Twenty-eight differential lipids were identified in 54 SCAs (34 invasive/20 noninvasive) and found to be associated with active proliferation of SCAs. Transcriptome analysis was also conducted in another cohort of 42 NFPAs (23 invasive/19 noninvasive). Then, a multiomic functionally connected network was established with two lipids, 17 differentially expressed genes, and four molecular pathways to explore their potential molecular mechanism in the invasiveness of SCAs. These lipids and genes were found enriched in the energy metabolism-related pathway, leading to the invasiveness of SCAs since rapid proliferation and invasion of tumors require the metabolism of a large amount of lipids. Finally, by using multiple machine learning methods, the authors identified the four most critical lipids contributing to the invasive phenotype of SCAs, including three membrane-constituting lipids and one signal-transducing lipid, and thereby constructed a four-lipid risk model, which demonstrated excellent predictive ability to discriminate the invasive SCAs from non-invasive ones.
Neou et al.
[24] generated a pangenomic classification of PAs. In their comprehensive research, transcriptomic analysis revealed low epithelial-mesenchymal transition (EMT) in SCAs along with their functioning
USP8-mutated counterparts, while increased EMT was seen in their functioning
USP8 wild-type counterparts showing increased invasiveness. As EMT is associated pathologically with tumor invasion and progression, the result might suggest that MET is not the main process that drives for the invasiveness of SCAs, and factors or processes other than MET may play a more crucial role in the proliferation and invasiveness in SCAs.
Epigenetic mechanisms, such as DNA methylation, have been shown to well distinguish between SCAs and CDs through DNA methylomic analysis
[24], which has also been suggested by Ricklefs et al.
[34]. DNA methylation of specific genes is related to the growth and invasiveness of PAs
[44][45][46]. However, further research is needed to elucidate which key genes are involved in invasiveness-related aberrant epigenetic deregulation that lead to the invasiveness and aggressiveness of SCAs.
By high-throughput sequencing, weighted gene co-expression network analysis, and functional annotation, researchers observed gene signatures correlated with the invasiveness of SCAs
[47]. Twenty differentially expressed genes were identified enriched in the “Pathway in cancers” and “MAPK pathway” and contributed to alterations of related factors. For instance, several receptor tyrosine kinases (EGFR and NTRK1) and their downstream signal transducers (RRAS and JAK1) showed enhanced expression in invasive SCAs. These components jointly regulate the tumorigenesis of SCAs. researchers next analyzed three invasion-related molecular markers (INSM1, HSPA2, and CDK6) in three subtypes of NFPAs (SCAs, SGAs, and null-cell adenomas) and found that INSM1 and HSPA2 expressions were higher in SCAs than other subtypes and that CDK6 showed a tendency of positive correlation with invasive SCAs. These results further supported the strong invasiveness of SCAs
[47].
Uraki et al. studied the mismatch repair genes
mutS homologs 6/2 (
MSH6/2) and
programmed cell death 1 ligand 1 (
PD-L1), which are involved in tumor growth and tumor immunity, respectively, in a cohort of 73 NFPAs, 23 of them SCAs
[48]. Results showed that reduced expressions of
MSH6/2 and
PD-L1 mRNA in SCAs partially account for the molecular mechanism causing proliferative and invasive characteristics as the reduction of
MSH6/2 could decrease the rate of apoptosis and promote cell-cycle progression, and the reduction of
PD-L1 could provide a relatively functional tumor immunity. However, their detailed mechanism warrants further in-depth studies.
In corticotroph adenomas, pituitary-specific hormone gene expression of arginine vasopressin receptor 1B (
AVPR1B) is observed to express higher in SCAs than in CDs. Since arginine vasopressin is capable of activating pathways associated with cell proliferation, higher
AVPR1B mRNA expression could be related to the growth and aggressiveness of SCAs
[18]. However, another study by ISABIAL’s lab found comparable
AVPR1B mRNA between SCAs and CDs, inconsistent with their previous finding
[49].