Transcriptomic analysis reveals that NIBAN1 overexpression is associated with BRAF^V600E mutation and increases the aggressiveness of thyroid cancer

NIBAN1 overexpression has been reported in a wide range of thyroid carcinoma subtypes, and in other cancers, while it is not expressed in thyroid benign lesions and normal thyroid.1, 2, 3 However, the mechanism associated with its expression and prognostic value remains unclear. In this study, six independent cohorts were enrolled, in which NIBAN1 expression was evaluated. A total of 583 patients with thyroid tumors, 326 patients with skin cancers, 293 patients with colorectal carcinoma, and 189 patients with lung carcinoma were analyzed. Our study demonstrates, for the first time, that NIBAN1 expression varied according to thyroid tumor subtypes, presence of BRAF^V600E mutation, worse clinical features, and aggressive phenotype. Remarkably, the data suggest that BRAF^V600E mutation might influence NIBAN1 expression in an MYC-dependent manner during the thyroid carcinogenic process. Furthermore, NIBAN1 expression was predicted to be associated with stress-induced transcription/translation. In summary, our findings suggested that NIBAN1 expression could be used not only to help preoperative diagnosis of a thyroid nodule but also may have prognostic implications.

NIBAN1 expression levels were assessed in three differentiated thyroid tumors cohorts: (i) 52 thyroid samples, analyzed by RT-qPCR (RT-qPCR_BR) from the Brazilian cohort; (ii) 27 thyroid samples, assessed by RNA-Seq (RNA-Seq_BR) from the Brazilian cohort; and (iii) 504 patients with PTC from TCGA cohort (TCGA-THCA). In the RT-qPCR_BR and RNA-Seq_BR cohorts, NIBAN1 expression was highest in classical PTC (CVPTC), followed by follicular carcinoma (FTC) and follicular subtype of PTC (FVPTC) (Fig. 1A, B). In the RNA-Seq_BR cohort, NIBAN1 expression differed between CVPTC and FVPTC (Fig. 1B). In the TCGA-THCA cohort, its expression rate was prominently different between primary tumors and paired normal thyroid (Fig. S1), being higher in the tall subtype (TVPTC), intermediate in CVPTC, and lower in FVPTC (Fig. S2).

Figure 1.

NIBAN1 expression in relation to histological types, mutational profile, MYC expression, and the NIBAN1 role in thyroid carcinomas. (A) RT-qPCR_BR cohort shows increased NIBAN1 expression in thyroid carcinomas (FTC, FVPTC, and CVPTC) compared with FTAs (P < 0.05). (B) The expression of NIBAN1 in the RNA-Seq_BR cohort shows differences according to the histological subtypes FVPTC and CVPTC (P = 0.02). (C) In the RT-qPCR_BR cohort, an increased expression of NIBAN1 is detected in BRAF^V600E samples in relation to: K-N-HRAS; samples negative for these variants and PAX8-PPRG rearrangements (P < 0.05). (D) Samples were stratified by variants and histology subtypes, showing increased NIBAN1 expression in CVPTC with BRAF^V600E compared with FVPTC, positive for K-N-HRAS variants, and FTAs. In FVPTC samples with BRAF^V600E variants and ETV6-NTKR fusion, NIBAN1 expression is higher compared to FTAs (P < 0.05). (E) The RNA-Seq_BR cohort demonstrates the increase of NIBAN1 expression in BRAF^V600E samples compared with K-N-HRAS (P = 0.03). (F) The RT-qPCR_BR cohort demonstrates the increase of NIBAN1 expression in MYC-High samples compared with MYC-Low (P = 0.00015). (G) Schematic representation of the human chromosome 1 that harbors the NIBAN1 locus and its possible regulation by MYC. The NIBAN1 gene (ENSG00000135842, 184,790,724–184,974,580), at 1q25.3, is transcribed by the reverse strand (antisense). In response to cellular stress, the MYC transcription factor may be activated, thereby increasing NIBAN1 expression. The asterisks indicate significance values; ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001. The data is represented as mean ± standard deviation of RT-qPCR and RNA-Seq values (Log); P represents the statistical analysis from the nonparametric Wilcoxon and Kruskall–Wallis tests.

Moreover, the NIBAN1 expression combined with the mutational profile revealed the presence of distinct subgroups. In the RT-qPCR_BR cohort, NIBAN1 expression was higher in PTC with BRAF^V600E compared with those with RAS mutation, followed by those negative for known driver mutations and PAX8-PPRG fusions, respectively (Fig. 1C). The stratification of these samples according to histological subtypes and mutational profiles showed the highest expression of NIBAN1 in CVPTC and FVPTC with BRAF^V600E or ETV6-NTRK3 fusion compared with FTC and FVPTC with different mutational profile. A noticeable difference was observed between FVPTC BRAF^V600E and FVPTC RAS (Fig. 1D). Similar findings were observed in the RNA-Seq_BR cohort (Fig. 1E) and in the TCGA-THCA cohort (Fig. S3). When all PTC were grouped, NIBAN1 expression was consistently higher in all PTC BRAF^V600E variants than those with K-N-HRAS (Fig. S4). To investigate the influence of BRAF and RAS mutations on NIBAN1 expression, three TCGA cohorts, with high prevalence of BRAF mutations, were evaluated (Fig. S5). NIBAN1 expression differed in colon (TCGA-COAD) and lung adenocarcinomas (TCGA-LUAD) cohorts (Fig. S6).

Since NIBAN1 expression varies according to the mutational profile, its expression in TCGA-THCA, classified as either BRAF-Like or RAS-Like, was explored in all PTC samples with other mutations. This classification was built because the BRAF^V600E and RAS variants were mutually exclusive, allowing the discrimination of PTC in BRAF-Like or RAS-Like expressions, according to the 71-gene expression signature.⁴ Thus, it was observed that NIBAN1 expression was higher in BRAF-Like samples compared with RAS-Like, regardless of the PTC histological subtype (Fig. S7).

To verify if the clinicopathological features could be related to NIBAN1 expression, the TCGA-THCA samples were stratified according to the TNM stage. The results showed that NIBAN1 expression is higher in more aggressive tumors (Fig. S8).

To create molecularly and biologically homogeneous subsets, the TCGA-THCA cohort was also stratified according to NIBAN1 expression in two subsets: high (NIBAN1-High) and low (NIBAN1-Low) profiles (Fig. S9), and their molecular and clinical features were also evaluated (Fig. S10). The incidence of the BRAF mutation is higher in NIBAN1-High than in NIBAN1-Low, while RAS mutations are more prevalent in NIBAN1-Low than in NIBAN1-High (Fig. S11A). Remarkably, the HRAS mutation was detected exclusively in 13% of NIBAN1-Low (Fig. S11A). Both NIBAN1 subsets contain PTC carrying other mutations (Other), corresponding to 56% in NIBAN1-Low and 31% in NIBAN1-High (Fig. S11B). Samples with Other mutations are composed of 11% of BRAF-Like profile and 89% RAS-Like, in NIBAN1-Low (Fig. S11C) and 45% of BRAF-Like profile and 55% RAS-Like, in NIBAN1-High (Fig. S11C). Knowing the importance of the mutational profile for the clinical phenotypes of thyroid tumors, the next step was to analyze the subsets according to histopathological subtypes and clinicopathological features. The results show that the NIBAN1 subsets differed across all these clinical parameters (Fig. S12).

Considering that the tumor microenvironment can recruit different types of immune cells, the fractions of infiltrating immune cells between NIBAN1-Low and NIBAN1-high were predicted using 22 immune cell types. The NIBAN1-High subset showed higher levels of all immune cell infiltration types compared with NIBAN1-Low (Fig. S13). These results suggested that it is possible that NIBAN1 overexpression may be clinically useful for inferring responsiveness to immunotherapy.

According to the results demonstrated herein, NIBAN1 expression seems to be associated with histopathological characteristics and aggressiveness of thyroid tumors. Thus, the next step was to verify in silico the predicted biological functions of NIBAN1 gene. To predict NIBAN1 functions and signaling pathways associated with its expression, the differentially expressed genes (DEGs) in the NIBAN1-High subset compared with NIBAN1-Low were analyzed. This analysis found 14,602 down-regulated and 468 up-regulated DEGs (Fig. S14). Functional enrichment analysis showed that the overexpressed genes seem to be closely related to the regulation of transcription involved in the G1/S transition of the mitotic cell cycle, cytoplasmic translational initiation, and other functions (Fig. S15). These DEGs were enriched in the ER to Golgi transport vesicles and nuclear matrix. On the other hand, the down-regulated genes seem to be closely related to the cellular response to copper and other metal ions, and the regulation of G1/S transition of mitotic cell cycle (Fig. S16). These results, analyzed in NIBAN1-High samples with BRAF^V600E mutation and BRAF-Like samples, corroborate the NIBAN1 involvement in the G1/S transition of the mitotic cell cycle, as well as transcription and translation during the cell cycle described before in functional experiments from our group.³

As demonstrated herein, the most prevalent genotype profile in the samples with high NIBAN1 expression was the BRAF^V600E mutation. Thus, it was evaluated in silico whether the transcription factors, activated downstream on the BRAF pathway in MAPK signaling, may regulate the NIBAN1 expression. The results showed that the transcription factors JUNB, JUND, and MYC are predicted to bind in the promoter region of NIBAN1, while FOS appears to act as a proximal enhancer. Although JUNB, JUND, and FOS are annotated as NIBAN1 regulators, only MYC has increased expression in the NIBAN1-High samples (Table S1). Importantly, MYC is deregulated across many human cancers, and it is a main regulator of gene transcription. Our group has previously shown that MYC is overexpressed in more aggressive thyroid carcinoma subtypes.⁵ To investigate the influence of MYC on NIBAN1 expression, samples from the RT-qPCR_BR cohort were classified based on MYC expression levels (Log values). Samples were classified into MYC-Low (<0.7) and MYC-High (>0.7). The results showed, for the first time, that samples with MYC-High present higher levels of NIBAN1 compared with samples with MYC-Low (Fig. 1F). Therefore, it is possible that MYC could be activated in NIBAN1-High/BRAF^V600E samples, increasing NIBAN1 levels in response to cellular stress (Fig. 1G).

Overall, these results, together with the functional experiments from our group,³ provide a global genetics picture of NIBAN1 expression in thyroid carcinogenesis and demonstrate a strong association between NIBAN1 overexpression and poor clinical features and more aggressive phenotype. Furthermore, NIBAN1 expression also appears to be associated with the mutational profile of the tumor. These results suggested that BRAF^V600E mutation may influence NIBAN1 expression, corroborating that its high expression level may lead to more aggressive conditions. All these data corroborate the insights, described by Diana and Carvalheira,³ which point out that NIBAN1 could be a strong biomarker candidate in a preoperative FNA diagnosis in thyroid lesions. Despite the promising results, this retrospective analysis must be further validated in a greater number of clinical cases and multicentric studies.

Ethics declaration

The study followed the Declaration of Helsinki and was analyzed (ID: CEP/UNIFESP: 0008/2016, CEP/UNIFESP: 0781/2016) by the Clinical Research Ethics Committee of the Universidade Federal de São Paulo. Signed informed consent was obtained from each patient.

Author contributions

Paula Diana and Gianna Maria Griz Carvalheira wrote the draft of the paper. Paula Diana, Reginaldo Massanobu Kuroshu, Thaise Nayane Ribeiro Carneiro, and Gianna Maria Griz Carvalheira designed the experiments. Paula Diana, Thaise Nayane Ribeiro Carneiro, Reginaldo Massanobu Kuroshu, Gianna Maria Griz Carvalheira, and Janete Cerutti analyzed the data. Paula Diana, Reginaldo Massanobu Kuroshu, and Gianna Maria Griz Carvalheira prepared the figures and table. Paula Diana, Thaise Nayane Ribeiro Carneiro, Janete Maria Cerutti, Reginaldo Massanobu Kuroshu, and Gianna Maria Griz Carvalheira reviewed and edited the manuscript.

Conflict of interests

The authors declare no conflict of interests.

Funding

This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) (N°. 88882.430337/2019-01); Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) (N°. 2014/06570-6 and 2015/04164-3) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (N°. 470441/2013-5).

Appendix A. Supplementary data

The following are the Supplementary data to this article.