TERT Promoter Mutated Follicular Thyroid Carcinomas Exhibit a Distinct microRNA Expressional Profile with Potential Implications for Tumor Progression

The role of microRNAs (miRNAs) as regulators of various important biological processes in human tissues, including cancer development, is established. In thyroid cancer, subsets of miRNAs exhibit oncogenic or tumor suppressive functions, influencing diverse neoplastic processes such as tumor growth, apoptosis, invasive behavior, and survival [1]. Moreover, recurrent somatic mutations in the miRNA processor genes DICER1 and DGCR8 have been reported in subsets of well-differentiated thyroid carcinomas (WDTCs) and/or poorly differentiated thyroid carcinomas (PDTCs) [2–5], and altered levels of DICER1 and DGCR8 messenger RNA (mRNA) promote aberrant miRNA expression patterns in vitro and in vivo [2, 3, 5]. In all, these advances reinforce the notion that dysregulation of the miRNA landscape should be an important mechanism underlying thyroid tumor development and progression.

In terms of genetic events coupled to worse clinical prognosis in WDTC, recurrent somatic mutations of the telomerase reverse transcriptase (TERT) promoter region have been established as one of the most confident molecular markers of poor patient outcome [6, 7]. The mutations activate TERT gene output, enhancing immortalization through elongation of telomere repeats. TERT promoter mutations increase in frequency with more aggressive thyroid cancer subtypes and advanced stages of disease [6]. Apart from the well-established anti-senescence function of TERT, little is known regarding non-canonical effects on cellular function in thyroid cancer. We recently observed a specific global mRNA pattern in TERT promoter mutated follicular thyroid carcinomas (FTCs) compared to FTCs with TERT promoter wild-type sequences, and the former group was enriched in mRNAs associated to various metabolic pathways [3]. Intrigued by the notion that TERT promoter mutated tumors may exhibit a distinct molecular phenotype, we therefore sought to investigate the global miRNA expressional landscape in FTCs with known TERT promoter mutational status.

We extrapolated data from a previously performed Nanostring nCounter platform analysis (Nanostring Technologies, Seattle, WA, USA) in which expression of 827 human miRNAs was interrogated in 11 FTCs with known TERT promoter genotypes [3]. Ethical approval was granted by the Swedish Ethical Review Authority (approval number 2015-959-31). The data was analyzed using Rosalind (Rosalind Inc. https://rosalind.onramp.bio/) as previously described [3]. Data normalization was performed in two steps, a positive control normalization and codeset normalization. During both steps, the geometric mean of each probeset was used to create a normalization factor. ROSALIND then calculates fold changes and p values for comparison using the t-test method and Benjamini-Hochberg method for estimating false discovery rates, assuming normal distribution. A differential expression analysis between TERT promoter mutated FTCs (n=4) and TERT promoter wild-type FTCs (n=7) was performed and visualized in a heatmap. Using a cutoff of ≥ 1.5 or ≤ −1.5 expressional fold change difference for up-regulated and down-regulated miRNAs respectively, and a p value set to <0.05, a total of 111 miRNA genes were found to be differentially expressed in TERT promoter mutated FTCs as compared to wild-type cases (Fig. 1A). The TERT promoter mutated cases displayed a high number of down-regulated miRNAs compared to the wild-type cases, and the top 5 most up- and down-regulated miRNAs in the former group are detailed in Table 1. The top down-regulated miR among TERT promoter mutated FTCs was hsa-miR-195-5p (fold change −5.73), and the top up-regulated miRNA was hsa-miR-200b-3p (fold change 8.25) (Table 1; Fig. 1B-C). Interestingly, hsa-miR-195-5p has been endowed with putative tumor suppressor functions in thyroid cancer, as it inhibits cell proliferation and invasion in thyroid cancer cell lines, while promoting apoptosis [8]. The most up-regulated miRNA in the TERT promoter mutated FTCs, hsa-miR-200b-3p, has been shown to be down-regulated in anaplastic thyroid carcinoma (ATC)—but little is known regarding the roles of this miRNA in WDTCs [9].

Fig. 1.

The distinct microRNA (miRNA) expressional profile in TERT promoter mutated follicular thyroid carcinoma (FTC). A Differential expression analysis of 827 miRNAs in FTC yielded 111 miRNAs with a significant expressional fold difference (≥ 1.5 or ≤ −1.5) with a p value < 0.05 in TERT promoter mutated FTCs marked in blue (n=4; cases 102, 201, 203, and 206) compared to wild-type cases marked in orange (n=7; cases 101, 104, 105, 106, 204, 205, and 207). In the heatmap, down-regulated miRNAs are represented in blue and up-regulated miRNAs in orange. B Volcano plot (top row) highlighting hsa-miR-200b-3p (largest purple dot), graph is shown with inverted logarithmic p values on the Y-axis and logarithmic fold change on the X-axis. Significantly up- and down-regulated miRNAs in the TERT promoter mutated FTCs are shown in the top left and right corner respectively. Below is the logarithmic normalized expression of hsa-miR-200b-3p in TERT promoter mutated FTCs (blue bars) compared to that of TERT promoter wild-type cases (orange bars). C Similar data shown for hsa-miR-195-5p (largest green dot in the volcano plot), the top down-regulated miRNA in the TERT promoter mutated group compared to TERT promoter wild-type tumors

Table 1.

Top 5 down- and up-regulated microRNAs in TERT promoter mutated follicular thyroid carcinomas compared to wild-type cases

A.	Down-regulated miRNAs
	miRNA	Fold change	p value*
	hsa-miR-195-5p	−5.73173	0.00078
	hsa-miR-145-5p	−5.14378	0.01635
	hsa-miR-143-3p	−3.46785	0.04721
	hsa-miR-497-5p	−3.39323	0.00724
	hsa-miR-519e-3p	−3.17536	0.02091
B.	Up-regulated miRNAs
	miRNA	Fold change	p value*
	hsa-miR-200b-3p	8.25061	0.00685
	hsa-miR-200a-3p	5.85009	0.00907
	hsa-miR-222-3p	5.43847	0.00640
	hsa-miR-135a-5p	5.14878	0.01027
	hsa-miR-3151-5p	3.57794	0.02489
C.	Selected gene products with associations to cancer targeted by the differentially expressed miRNA pool
	Gene	Function	#miRNAs targeting
	MDM2	P53 regulator	34
	CDK6	Cell cycle regulator	31
	DICER1	miRNA regulator	30
	CCND1	Cell cycle regulator	29
	CDKN1A	Cell cycle regulator	29

^*p values < 0.05 were considered statistically significant

In terms of gene products most frequently targeted by the differentially expressed miRNAs, the top 30 most affected validated genes included MDM2, CDK6, DICER1, CCND1 (encoding cyclin D1), and CDKN1A (encoding p21), all exhibiting some association to the development or progression of thyroid cancer (Table 1). Using the Reactome Pathway Database (https://reactome.org/) for the top 30 most affected genes, the abovementioned observations were reinforced, with significant associations between these genes and the pathway terms “cell cycle,” “transcriptional regulation by TP53,” and “oncogene induced senescence.”

Two TERT promoter wild-type FTCs (cases 204 and 205) clustered together with the four TERT promoter mutated cases (Fig. 1), suggesting that other molecular mechanisms besides TERT promoter mutations establish the miRNA landscape of FTCs. A closer look at these two tumors in terms of TERT mRNA expression and alternative TERT gene aberrancies revealed neither TERT gene copy number gain, TERT promoter hypermethylation, or detectable TERT mRNA expression, all mechanisms previously described in FTCs [10]. Thus, other contributing mechanisms to the observed miRNA pattern are likely to occur in these two cases.

In this pilot series, we demonstrate that TERT promoter mutated FTCs carry a unique miRNA signature and highlight a number of interesting candidates for future studies in terms of disease progression and prognosis. The observed down-regulation of hsa-miR-195-5p should be of particular interest for future studies, as this miRNA could constitute a potent tumor suppressor in thyroid cancer cell lines by targeting TERT mRNA [8]. Moreover, cancer-related mRNA targets of the differentially expressed miRNAs were amassed in cell cycle and P53 signaling, providing a possible link between TERT promoter mutations and dysregulation of these pathways in FTCs. Interestingly, DICER1 was also found among the top targeted mRNAs in this dataset, which may suggest a complex loop in which the mRNA translation of this miRNA processor itself is affected by differentially expressed miRNAs in TERT promoter mutated FTCs. However, the results obtained from this study are based on a small sample series in need of experimental validation. For example, our findings of overexpressed hsa-miR-200b-3p in TERT promoter mutated FTCs are in conflict with the previously demonstrated down-regulation in ATCs, thereby serving as a good example for a miRNA candidate reliant on further analyses using larger tumor cohorts as well as the inclusion of normal controls [9].

In conclusion, TERT promoter mutated FTCs may exhibit a specific miRNA pattern that in part regulates key cancer pathways, an observation that may support non-canonical TERT functions in thyroid cancer. Given the dismal outcome for many TERT promoter mutated FTCs, increased knowledge of the molecular underpinnings driving these tumors should be of clinical interest.

Funding

Open access funding provided by Karolinska Institute.