Investigating USP42 Mutation as Underlying Cause of Familial Non-Medullary Thyroid Carcinoma

In a family with Familial Non-Medullary Thyroid Carcinoma (FNMTC), our investigation using Whole-Exome Sequencing (WES) uncovered a novel germline USP42 mutation [p.(Gly486Arg)]. USP42 is known for regulating p53, cell cycle arrest, and apoptosis, and for being reported as overexpressed in breast and gastric cancer patients. Recently, a USP13 missense mutation was described in FNMTC, suggesting a potential involvement in thyroid cancer. Aiming to explore the USP42 mutation as an underlying cause of FNMTC, our team validated the mutation in blood and tissue samples from the family. Using immunohistochemistry, the expression of USP42, Caspase-3, and p53 was assessed. The USP42 gene was silenced in human thyroid Nthy-Ori 3-1 cells using siRNAs. Subsequently, expression, viability, and morphological assays were conducted. p53, Cyclin D1, p21, and p27 proteins were evaluated by Western blot. USP42 protein was confirmed in all family members and was found to be overexpressed in tumor samples, along with an increased expression of p53 and cleaved Caspase-3. siRNA-mediated USP42 downregulation in Nthy-Ori 3-1 cells resulted in reduced cell viability, morphological changes, and modifications in cell cycle-related proteins. Our results suggest a pivotal role of USP42 mutation in thyroid cell biology, and this finding indicates that USP42 may serve as a new putative target in FNMTC.


Introduction
Thyroid cancer (TC) incidence has increased substantially over the last three decades [1].Up to 10% of TC cases encompass familial forms [2].When arising from thyroid follicular cells, familial cases of thyroid carcinoma are designated as Familial Non-Medullary Thyroid Carcinoma (FNMTC) [3].FNMTCs are further subdivided into syndromic FN-MTC (SFNMTC) and non-syndromic FNMTC (NSFNMTC).In SFNMTC, patients exhibit a predominance of non-thyroid tumors alongside thyroid cancer, and several genetic alterations associated with syndromic forms have been well described (e.g., APC, PTEN, and DICER1) [4,5].NSFNMTC patients present with TC as the main clinical manifestation, comprising a very heterogeneous disorder, both at the phenotypic and genotypic levels, which is usually coupled with various low-to moderate-penetrance susceptibility alleles (e.g., FOXE1 and TITF-1), and incomplete penetrance [6,7].A polygenic pattern was suggested but many questions remain unanswered [5,8,9].
Our team identified a family with FNMTC.The family was composed of a mother and two monozygotic twin daughters with an early-onset diagnosis of Oncocytic Subtype of Papillary Thyroid Carcinoma (OSPTC).The family fulfilled the criteria for hereditary cancer, but no alterations associated with syndromic forms of the disease were found.Using Whole-Exome Sequencing (WES), our team identified a possible disease-causative mutation in a Ubiquitin-Specific Protease gene [USP42 p.(Gly486Arg)].
Many USPs have been put forward as regulators of tumor formation and proliferation, mainly because they are involved in cell cycle progression, regulation of apoptosis-related factors, DNA repair activity, and other important cellular functions [10][11][12].USP42, located on chromosome 7 (7p22.1),was shown to regulate cellular metabolism through p53-histone H2B interaction in cell lines by Hock et al. (2011) and Hock et al. (2014) [13,14].
To our knowledge, no prior studies investigating the role of USP42 gene alterations in NSFNMTC have been performed; therefore, we decided to evaluate USP42 mutation as an underlying cause of NSFNMTC.

Case Report
Our team identified a family presenting with NSFNMTC.The family was composed of a mother (I), diagnosed at the age of 35, and of two monozygotic twin daughters both diagnosed at the age of 19 (II.1 and II.2) (Figure 1A).The histological lesions of the three cases were similar, all presenting the same diagnosis of OSPTC (Figure 1A).The mother, in addition to displaying OSPTC, also presented several lesions of Follicular Nodular Disease (FND) comprising oncocytic cells.FND lesions were also present in the surgical specimens from the twins.

Molecular Studies 2.2.1. USP42 Missense Mutation Validation
In silico analysis of the normal DNA extracted from the three affected individuals was subjected to WES (Figure 1A,B), which revealed the presence of an extremely rare Single-Nucleotide Polymorphisms (SNP) affecting the USP42 gene (Table 1).Although this is a variant of unknown significance according to ClinVar, it shows a predicted CADD score over 20, meaning that this SNP is predicted to be a damaging substitution (Table 1).After confirming the presence of this SNP on IGV (Integrative Genomics Viewer), PCR followed by Sanger sequencing was performed using DNA from blood samples and from FFPE (Formalin-Fixed Paraffin-Embedded) samples of tumoral and non-tumoral tissues (Figure 1C,D; Table 1).The USP42 p.(Gly486Arg) mutation was present in all analyzed samples of the mother (I, proband) and of the monozygotic twin daughters (II.1 and II.2) in heterozygosity (Figure 1D; Figure S1).

USP42 Missense Mutation Validation
In silico analysis of the normal DNA extracted from the three affected individuals was subjected to WES (Figure 1 A,B), which revealed the presence of an extremely rare Single-Nucleotide Polymorphisms (SNP) affecting the USP42 gene (Table 1).Although this is a variant of unknown significance according to ClinVar, it shows a predicted CADD score over 20, meaning that this SNP is predicted to be a damaging substitution (Table 1).After confirming the presence of this SNP on IGV (Integrative Genomics Viewer), PCR followed by Sanger sequencing was performed using DNA from blood samples and from FFPE (Formalin-Fixed Paraffin-Embedded) samples of tumoral and non-tumoral tissues (Figure 1C,D; Table 1).The USP42 p.(Gly486Arg) mutation was present in all analyzed samples of  When looking at mutations affecting the USP42 gene in 514 PTCs subjected to WES that were selected from the TCGA database, two missense mutations were identified, one of which is pathogenic [USP42 p.(Pro669Gln)] (Figure 1E; Figure S1; Table S1).We expanded our analysis and looked at PTC cases harboring pathogenic somatic alterations affecting any genes of the USP family.Of the 17 PTC cases found to have genetic alterations in USPs, 24% and 16% had mutations affecting USP9X and USP6, respectively.In addition, 76% of these 17 PTCs also harbored BRAF hotspot mutations (Figure S1; Table S1).
The tumor sample from the mother subjected to WES had a similar variant-allele frequency of the USP42 p.(Gly486Arg) SNP to the one observed in the matched normal sample (Figure 1F).CNA analysis revealed only a loss of chromosome 22 (Figure 1G).When looking at the somatic mutations, an HRAS p.(Gln61Arg) hotspot mutation was identified (Figure 1H; Table 2).

Immunohistochemistry
FFPE tissues from the affected individuals were assessed to observe tissue expression of USP42.We observed that USP42 protein was overexpressed in all tumoral tissues from all tested individuals (Figure 2B,D,F).Adjacent tissues also presented high expression of the protein, thus suggesting that USP42 mutation may result in protein overexpression (Figure 2A,C,E).As for the other markers analyzed, we observed that Caspase-3 (Figure 3) and p53 (Figure 4) proteins presented tumor overexpression in all tested samples.At least 1000 cells were counted in each field.Quantification was performed in the areas that presented higher expression of p53. 1 ID-Identification; 2 FND-Follicular Nodular Disease.Quantification of p53 showed that the protein was overexpressed in the tumor tissues compared to adjacent tissues and to non-tumoral tissue from the proband (2.71% p53 positive cells) (Table 3; Figure 4).Tumor lesions of the three elements presented a range of 26.33-57.56%p53-positive cells in the available fragments, while adjacent tissue presented 0.75-10.65%p53-positive cells in the available fragments (Table 3; Figure 4).At least 1000 cells were counted in each field.Quantification was performed in the areas that presented higher expression of p53. 1 ID-Identification; 2 FND-Follicular Nodular Disease.

USP42 Silencing and Analysis of mRNA and Protein Expression
We successfully silenced USP42 using two different siRNA treatments, siRNA 2 and siRNA 7, which were evaluated for mRNA expression using RT-qPCR with the 2 −∆∆Ct method (Figure 5A).At the 24 h timepoint, siRNA 2 and siRNA 7 treatments each had a significant silencing effect on the gene (60%) when compared to a negative control (p < 0.0001 for both treatments).At the 48 h timepoint, the silencing effect (60%) was maintained for both treatments (p = 0.0001 and p = 0.0003, respectively).However, at 72 h, a recuperation of USP42 gene expression was observed, but silencing effects of 51% and 45% were still detected for siRNA 2 (p = 0.0002) and siRNA 7 (p = 0.004), respectively (Figure 5A).USP42 protein expression was evaluated by Western blot in these same treatments.We did not observe protein expression alterations at 24 h or 48 h timepoints (Figure 5B).However, at the 72 h timepoint, USP42 protein silencing was significant using the siRNA 7 treatment, with a 32% reduction in protein expression compared to the negative control (p = 0.008).No significant changes were observed for protein expression using siRNA 2 for up to 72 h post-treatment.We successfully silenced USP42 using two different siRNA treatments, siRNA 2 and siRNA 7, which were evaluated for mRNA expression using RT-qPCR with the 2 −ΔΔCt method (Figure 5A).At the 24 h timepoint, siRNA 2 and siRNA 7 treatments each had a significant silencing effect on the gene (60%) when compared to a negative control (p < 0.0001 for both treatments).At the 48 h timepoint, the silencing effect (60%) was maintained for both treatments (p = 0.0001 and p = 0.0003, respectively).However, at 72 h, a recuperation of USP42 gene expression was observed, but silencing effects of 51% and 45% were still detected for siRNA 2 (p = 0.0002) and siRNA 7 (p = 0.004), respectively (Figure 5A).USP42 protein expression was evaluated by Western blot in these same treatments.We did not observe protein expression alterations at 24 h or 48 h timepoints (Figure 5B).However, at the 72 h timepoint, USP42 protein silencing was significant using the siRNA 7 treatment, with a 32% reduction in protein expression compared to the negative control (p = 0.008).No significant changes were observed for protein expression using siRNA 2 for up to 72 h post-treatment.

Cell Viability and Apoptosis Analysis
We performed flow cytometry analysis with Annexin V/PI double staining to measure cell viability/mortality following siRNA treatment (Figure 6; Figure S2).At the 24 h timepoint, the level of viable cells (Annexin V − /PI − ) showed a statistically significant decrease when treated with siRNA 2 compared to the negative control (p = 0.032) (Figure 6A).No statistically significant changes were found between treatments at the 48 h

Cell Viability and Apoptosis Analysis
We performed flow cytometry analysis with Annexin V/PI double staining to measure cell viability/mortality following siRNA treatment (Figure 6; Figure S2).At the 24 h timepoint, the level of viable cells (Annexin V − /PI − ) showed a statistically significant decrease when treated with siRNA 2 compared to the negative control (p = 0.032) (Figure 6A).No statistically significant changes were found between treatments at the 48 h timepoint.However, after 72 h of treatment cell viability levels decreased in cells treated with siRNA 7, compared to negative control (p = 0.010) (Figure 6A).It was also observed that siRNA 7 treatment resulted in lower cell viability compared to siRNA 2 treatment (p = 0.013) (Figure 6A).siRNA 2 treatments (p < 0.0001, for both comparisons) (Figure 6C).The same observation was made at the 72 h timepoint (p = 0.004 and p = 0.005, respectively) (Figure 6C).Late apoptotic cells (Annexin V + /PI + ) were the second most representative population of cells present in the samples (Figure 6D).The frequency of late apoptotic cells was higher following siRNA 2 treatment at the 24 h timepoint than in the negative control (p = 0.027) (Figure 6D).At 48 h, no differences were observed (Figure 6D).At the 72 h timepoint, siRNA 7 treatment resulted in higher percentages of late apoptotic cells compared to negative control and siRNA 2 treatment (p = 0.032 and p = 0.037, respectively) (Figure 6D).

USP42 Silencing Effect in Cell Proliferation
To assess if cell proliferation was affected by siRNA treatments, we performed cell counting with Trypan Blue on a Neubauer chamber.No statistically significant changes were found at 24 h and 48 h for both siRNA treatments, although a tendency for decreased proliferation was found when treating cells with siRNA 7 at 48 h (p = 0.065) (Figure 7).However, at the 72 h timepoint, siRNA 7-treated cells were the most affected, showing lower proliferation compared to negative control cells (p = 0.0002) and siRNA 2-treated cells (p = 0.014) (Figure 7).At the 72 h timepoint, compared to the negative control siRNA 2-treated cells proliferated two times less and siRNA 7-treated cells proliferated 7.3 times less (Figure 7).Necrotic cells (Annexin V − /PI + ) represented the smallest population of cells.No statistically significant differences were observed between treatments or timepoints (Figure 6B).For the population of early apoptotic cells (Annexin V + /PI − ), at the 24 h timepoint siRNA 2 treatment seemed to result in a higher frequency of early apoptotic cells compared to siRNA 7 treatment (p = 0.023) (Figure 6C).However, at the 48 h timepoint, siRNA 7 treatment showed to be more effective when compared to both negative control and siRNA 2 treatments (p < 0.0001, for both comparisons) (Figure 6C).The same observation was made at the 72 h timepoint (p = 0.004 and p = 0.005, respectively) (Figure 6C).Late apoptotic cells (Annexin V + /PI + ) were the second most representative population of cells present in the samples (Figure 6D).The frequency of late apoptotic cells was higher following siRNA 2 treatment at the 24 h timepoint than in the negative control (p = 0.027) (Figure 6D).At 48 h, no differences were observed (Figure 6D).At the 72 h timepoint, siRNA 7 treatment resulted in higher percentages of late apoptotic cells compared to negative control and siRNA 2 treatment (p = 0.032 and p = 0.037, respectively) (Figure 6D).

USP42 Silencing Effect in Cell Proliferation
To assess if cell proliferation was affected by siRNA treatments, we performed cell counting with Trypan Blue on a Neubauer chamber.No statistically significant changes were found at 24 h and 48 h for both siRNA treatments, although a tendency for decreased proliferation was found when treating cells with siRNA 7 at 48 h (p = 0.065) (Figure 7).However, at the 72 h timepoint, siRNA 7-treated cells were the most affected, showing lower proliferation compared to negative control cells (p = 0.0002) and siRNA 2-treated cells (p = 0.014) (Figure 7).At the 72 h timepoint, compared to the negative control siRNA 2-treated cells proliferated two times less and siRNA 7-treated cells proliferated 7.3 times less (Figure 7).

Cell Morphology Evaluation by Phalloidin Staining
Cell morphology was evaluated by immunofluorescence following 48 h treatment with siRNAs.Representative images of each condition are shown in Figure 8. Negative control-treated cells presented, as expected, an epithelial-like phenotype, higher cellular confluence, and cell-to-cell adhesion, with actin filaments being evenly distributed within the cells (Figure 8A,B).For both treatments, siRNA 2 and siRNA 7, most of the fields were composed of a significant number of cells with altered morphology.These cells were smaller and their actin filaments appeared to lose shape in the cytoplasm and to be concentrated in the cell periphery (Figure 8C-F).The described effect was more evident with siRNA 2 treatment, which presented a higher number of cells with altered morphology per field at the 48 h timepoint (Figure 8).

Cell Morphology Evaluation by Phalloidin Staining
Cell morphology was evaluated by immunofluorescence following 48 h treatment with siRNAs.Representative images of each condition are shown in Figure 8. Negative control-treated cells presented, as expected, an epithelial-like phenotype, higher cellular confluence, and cell-to-cell adhesion, with actin filaments being evenly distributed within the cells (Figure 8A,B).For both treatments, siRNA 2 and siRNA 7, most of the fields were composed of a significant number of cells with altered morphology.These cells were smaller and their actin filaments appeared to lose shape in the cytoplasm and to be concentrated in the cell periphery (Figure 8C-F).The described effect was more evident with siRNA 2 treatment, which presented a higher number of cells with altered morphology per field at the 48 h timepoint (Figure 8).

β-Actin Expression Analysis by Western Blot
To assess whether β-actin alterations were only morphological or whether siRNA treatment resulted in β-actin protein expression changes, we performed an evaluation using Western blot.At 24 h, no evident changes in β-actin expression were found.However, at 48 h we observed a reduction of 55% in β-actin expression for siRNA 2 treatment compared to the negative control (p = 0.002) (Figure 9).A reduction of 53% in β-actin expression was also verified using siRNA 7 treatment, although this was not statistically significant (p = 0.065) (Figure 9).At the 72 h timepoint, no difference was found between β-actin expression in siRNA 2-treated cells and negative control cells; however, a significant 48% reduction was found in siRNA 7-treated cells (p = 0.010) (Figure 9).We also detected reductions in tubulin and vinculin expression by Western blot (Figure S3).

β-Actin Expression Analysis by Western Blot
To assess whether β-actin alterations were only morphological or whether siRNA treatment resulted in β-actin protein expression changes, we performed an evaluation using Western blot.At 24 h, no evident changes in β-actin expression were found.However, at 48 h we observed a reduction of 55% in β-actin expression for siRNA 2 treatment compared to the negative control (p = 0.002) (Figure 9).A reduction of 53% in β-actin expression was also verified using siRNA 7 treatment, although this was not statistically significant (p = 0.065) (Figure 9).At the 72 h timepoint, no difference was found between β-actin expression in siRNA 2-treated cells and negative control cells; however, a significant 48% reduction was found in siRNA 7-treated cells (p = 0.010) (Figure 9).We also detected reductions in tubulin and vinculin expression by Western blot (Figure S3).

Cell Cycle-Related Proteins Analyzed by Western Blot
As USP42 has been described to regulate cell cycle and apoptosis [13], protein expression of cell cycle-related proteins (p53, Cyclin D1, p27, and p21) was evaluated (Figure 10).

Cell Cycle-Related Proteins Analyzed by Western Blot
As USP42 has been described to regulate cell cycle and apoptosis [13], protein expression of cell cycle-related proteins (p53, Cyclin D1, p27, and p21) was evaluated (Figure 10).
siRNAs had not altered USP42 protein expression at 48 h post-treatment; however, p53 expression was decreased at this timepoint (Figure 10A).Compared to the negative control, siRNA 2 treatment led to a 40% reduction in p53 expression (p = 0.004) and siRNA 7 treatment lead to a 53% reduction (p = 0.037) (Figure 10A).After 72 h, siRNA 2 treatment did not give a statistically significant change compared to the negative control, but p53 expression in siRNA 2-treated cells presented a large discrepancy compared to siRNA 7treated cells (p = 0.002) (Figure 10A).At the 72 h timepoint, we observed the most significant reduction in p53 expression (66%) after treating cells with siRNA 7 (p < 0001) (Figure 10A).
Cyclin D1 expression was only found to be significantly altered at the 24 h timepoint, with a 38% reduction in its expression following siRNA 7 treatment compared to negative control (Figure 10B).No significant changes were observed at the remaining timepoints (Figure 10B).p21 protein expression was reduced following siRNA 2 treatment.At 24 h, we observed a 27% decrease of expression compared to negative control (p = 0.033) and at the 48 h timepoint, a 58% reduction was found (p = 0.0004) (Figure 10C).However, at the 72 h timepoint with siRNA 2 treatment, cells seemed to recuperate p21 expression to the same level as in negative control cells (Figure 10C).Using siRNA 7, only at the 48 h timepoint a statistically significant association was found, interestingly with a marked increase of 140% (2.4-fold) of p21 expression compared to negative control (p = 0.045) (Figure 10C).When comparing p21 expression between the two treatments, a 2-fold difference was found (p = 0.011) (Figure 10C).

Discussion
We report here the identification of a USP42 germline mutation that segregated the disease in a family compatible with the diagnosis of NSFNMTC.The family was posed of a mother and two monozygotic twin daughters with an early-onset diagno OSPTC.Despite filling the criteria for hereditary cancer, no alterations associated siRNAs had not altered USP42 protein expression at 48 h post-treatment; however, p53 expression was decreased at this timepoint (Figure 10A).Compared to the negative control, siRNA 2 treatment led to a 40% reduction in p53 expression (p = 0.004) and siRNA 7 treatment lead to a 53% reduction (p = 0.037) (Figure 10A).After 72 h, siRNA 2 treatment did not give a statistically significant change compared to the negative control, but p53 expression in siRNA 2-treated cells presented a large discrepancy compared to siRNA 7treated cells (p = 0.002) (Figure 10A).At the 72 h timepoint, we observed the most significant reduction in p53 expression (66%) after treating cells with siRNA 7 (p < 0001) (Figure 10A).
Cyclin D1 expression was only found to be significantly altered at the 24 h timepoint, with a 38% reduction in its expression following siRNA 7 treatment compared to negative control (Figure 10B).No significant changes were observed at the remaining timepoints (Figure 10B).p21 protein expression was reduced following siRNA 2 treatment.At 24 h, we observed a 27% decrease of expression compared to negative control (p = 0.033) and at the 48 h timepoint, a 58% reduction was found (p = 0.0004) (Figure 10C).However, at the 72 h timepoint with siRNA 2 treatment, cells seemed to recuperate p21 expression to the same level as in negative control cells (Figure 10C).Using siRNA 7, only at the 48 h timepoint a statistically significant association was found, interestingly with a marked increase of 140% (2.4-fold) of p21 expression compared to negative control (p = 0.045) (Figure 10C).When comparing p21 expression between the two treatments, a 2-fold difference was found (p = 0.011) (Figure 10C).

Discussion
We report here the identification of a USP42 germline mutation that segregated with the disease in a family compatible with the diagnosis of NSFNMTC.The family was composed of a mother and two monozygotic twin daughters with an early-onset diagnosis of OSPTC.Despite filling the criteria for hereditary cancer, no alterations associated with syndromic forms of FNMTC were found.Using WES, we identified a possible causative mutation for the disease in a Ubiquitin-Specific Protease gene-[USP42 p.(Gly486Arg)].
USP42 is a thiol-dependent deubiquitinating enzyme (DUB) predicted to be active in cytosol and nucleus [15,16].The ubiquitination pathway is a dynamic and complex posttranscriptional process, regulated by E3 enzymes and reversible through DUBs [11,17,18].DUBs are enzymes capable of removing ubiquitin from ubiquitinated proteins, which may regulate their stability and activity, protecting them from proteasomal degradation [19][20][21].
USP42 gene alterations have been previously described as an underlying cause of cancer.Wang et al. (2022) suggested that USP42 might be a target gene for breast cancer [22].They have shown that USP42 was overexpressed in breast cancer tissues and further demonstrated, using in vitro studies, that USP42 overexpression promoted cell invasion and migration, as well as that USP42 silencing led not only to a reduction in proliferation, but also to increased apoptosis, suggesting that this gene may act as an oncogene [22].USP42 regulation of cellular metabolism through p53-histone H2B interaction has been demonstrated [13,14].
Other USPs have been put forward as regulators of tumor formation and proliferation in TC.Jianing et al. ( 2022) showed an interaction of USP26 protein with TAZ protein, leading to its stabilization and the consequent proliferation of Anaplastic Thyroid Carcinoma (ATC) [23].Liang et al. (2021) reported that TC patients with low expression of USP18 had worse survival rates than those with high USP18 expression [24].An et al. (2015) demonstrated that USP39 underexpression was associated with decreased cell proliferation due to G2/M cell cycle arrest in Medullary Thyroid Carcinoma (MTC) cells [25].More recently, using WES, Maria et al. (2022) described a USP13 p.(Val495Met) mutation as an underlying cause of PTC in a family with NSFNMTC.In this family, the USP13 mutation that was shown to result in USP13 overexpression, which played a role in the tumorigenesis of PTC [26].
These results suggest that USPs may play a role in both sporadic and familial thyroid carcinoma.We examined the data available in the TCGA database and found that 17 PTC cases had somatic genetic alterations in USP genes.USP9X and USP6 were the most frequently affected USP genes.Two cases presented variants in USP42 (different from the one identified in the family examined in this study).BRAF hotspot mutations were present in 76% of these 17 PTCs, suggesting that USP alterations can cooperate with well-known TC genetic alterations during the transformation process.Recently, a polygenic mode of action in NSFNMTC tumorigenesis has been proposed [8,9].NSFNMTC is described as presenting low-to moderate-penetrance susceptibility alleles with incomplete penetrance, which leads to TC susceptibility [5][6][7].It is hypothesized that the combined influences of other modified genes along with these susceptibility alleles triggers the onset of cancer [6][7][8][9].Of note, in the studied family we detected a somatic HRAS mutation in the proband tumor samples, but that alteration was not present in the other elements.
The USP42 p.(Gly486Arg) mutation was present in all of the tumor samples from the family members in heterozygosity.Using immunohistochemistry (IHC) we noted that, although USP42 was expressed in the normal tissue, an increased expression was observed in the tumor specimens, indicating overexpression.The tumor samples also presented higher expression of p53 and Caspase-3 than adjacent tissues.
To our knowledge, no prior studies investigating the role of USP42 gene alterations in NSFNMTC have conducted, and we therefore performed several in vitro experiments to validate USP42 mutation as an underlying cause of NSFNMTC.
We undertook downregulation of USP42 using two siRNAs, both of which efficiently silenced USP42 at 24 h and 48 h, also showing some recuperation of the silencing effect at 72 h.A marked effect on apoptosis was shown, with a significant increase in apoptosis following silencing of the USP42 gene.These results are consistent with those of Wang et al. (2022), who silenced USP42 and observed increased apoptosis in breast cancer cells [22].Other USPs, such as USP15, have been demonstrated to impact the stability and activity of Caspase-3 [27].
The effect of USP42 siRNA silencing on proliferation was delayed and not so marked, although a significant reduction in the proliferation of USP42-silenced cells at 72 h posttreatment was noted.
The ubiquitination process is a dynamic process that has already been described to regulate actin filaments in the cells [28].In our study, we observed that USP42 silencing led to cytoskeletal alterations in the cells, as shown by phalloidin staining of actin filaments, and also a reduction in cytoskeletal proteins as shown by Western blot for actin, vinculin, and tubulin.Whether USP42 silencing has a direct or indirect effect on cytoskeleton formation is still to be uncovered, but other USPs, such as USP8 and USP14, have been shown to regulate cytoskeletal proteins [28,29].
It has been described that USP42 stabilizes p53 by removing ubiquitin and preventing its proteasomal degradation [13,14,30].Using IHC, increased expression of p53 was found in FFPE tissues for all family members.Usually, increased p53 expression results from mutations [31], but no p53 mutations were found in the samples from the proband (I).Therefore, we hypothesize that increased p53 expression in the tissues is related to USP42 overexpression, which was also noted in the FFPE; supporting this hypothesis, after silencing USP42 we observed a significant reduction in p53 expression at 48 h post-treatment for both siRNAs.However, at 72 h post-transfection, while siRNA 7 treatment maintained a significant reduction in p53, this effect was not verified for siRNA 2. This might be due to the transient effects of the siRNAs, which presented a recovery at 72 h, as referred to above.As the USP42 gene can present different isoforms, we can also hypothesize that different siRNAs may target different isoforms, thus resulting in variable effects in the cell [13].Hou et al. (2016) has shown that USP42 was overexpressed in gastric cancer patients and that overall survival of these patients was reduced compared to patients whose tumors presented lower expression of USP42 protein [32].They further demonstrated, using siRNAs against the USP42 gene, that cell cycle-related proteins, namely Cyclin D1, Cyclin E1, and the Proliferating Cell Nuclear Antigen (PCNA), were downregulated following USP42 silencing [32].Furthermore, Maria et al. (2022) showed that USP13 silencing by siRNAs decreased expression of Cyclin D1, leading to cell cycle arrest [26].In our study, in concordance with the aforementioned studies, we demonstrated that, following USP42 silencing using siRNA 2, Cyclin D1 is significantly downregulated at 24 h post-treatment.
We also showed that p21 is downregulated at 24 h and 48 h after treating cells with siRNA 2, although the same effect is not evident when treating cells with siRNA 7 for 48 h.These discrepancies may be associated with the specific binding site of the siRNA that might lead to variable effects in the cells or they may be related to the cytotoxicity associated with siRNA 7. Hock et al. (2014) reported that USP42 silencing resulted in lower expression of p21 mRNA and hypothesized that USP42 is recruited to the transcriptional site of p21, but further investigation on this interaction is still required [14].A similar effect was obtained for p27 expression following USP42-silencing treatments.siRNA 2 treatment resulted in p27 downregulation between 24 h and 48 h of treatment, and siRNA 7 treatment resulted in the upregulation of p27 at the 72 h timepoint.
Our results, along with the available information in the literature, leads us to hypothesize that the USP42 p.(Gly486Arg) missense mutation is the underlying cause of NSFNMTC in this family.The mutation seems to be activating, leading to overexpression of the mutated protein, thus possibly acting as an oncogene.
As for the limitations of this study, it is important to emphasize that our sample is small, being composed of only three affected elements.Furthermore, in the in vitro siRNA studies we cannot rule out the possibility that some of the observed changes might be due to the transient nature of siRNA and/or that some effects may be due to transfection cytotoxicity.Future assays using CRISPR-Cas9 or a similar methodology, in which the Nthy-Ori 3-1 cell line will be transformed to harbor the specific SNP that was found in this family by using WES [USP42 p.(Gly486Arg)], will be necessary.
The genetics behind NSFNMTC is rather heterogeneous and poorly understood [2,4].The causative genes in many affected families remain to be identified and the establishment of molecular markers would lead to better genetic counselling and to the development of new therapies.Disease prevention and active surveillance are key factors for reducing aggressive treatments, high comorbidities, and invasive surgeries, in addition to avoiding high burdens for the health care system.

Patients
The surgical procedures and sample collection were performed at Cluj-Napoca Municipal Clinical Hospital, Romania.Peripheral blood and FFPE tumor tissues were collected from the three family members.We obtained a total of 26 FFPE blocks from the three elements: 10 blocks from the mother (I), 12 from the first twin (II.1), and 4 from the second twin (II.2) (Figure 1A).DNA was extracted from all samples and preserved at −20 • C until further processing was performed.All subjects gave their informed consent for inclusion before they participated in the study.This study was conducted in accordance with the Declaration of Helsinki.

DNA Extraction from Patient Samples
DNA was extracted from peripheral blood samples using the GRS Total DNA Kit-Blood & Cultured Cells (GRiSP Research Solutions, Porto, Portugal).DNA extraction from FFPE tissues was performed using the GRS Genomic DNA Kit BroadRange (GRiSP Research Solutions, Porto, Portugal), following the manufacturer's instructions.Quantitative and qualitative sample analysis was performed using spectrophotometry with a Nanodrop N-1000 Spectrophotometer for microvolume UV-Visible measurements (Thermo Scientific, Waltham, MA, USA).

Variant Calling and Annotation
Normal DNA was extracted from the blood of the three affected individuals, and the tumor DNA from the mother was microdissected from FFPE tissue blocks as previously described.All DNA samples were subjected to WES using the Illumina DNA Prep.

Variant Prioritization
A step-by-step approach was used to identify potential pathogenic germline variants, using the following criteria: (i) having an allele fraction ≤1% in the overall population, as given by gnomAD (version 2.1.1)[36,37], (ii) being present in all three affected individuals, (iii) having a combined annotation dependent depletion (CADD (version 1.7, University of Washington, Hudson-Alpha Institute for Biotechnology, WA, USA and Berlin Institute of Health at Charité -Universitätsmedizin Berlin, Germany) [38] PHRED score 10), and (iv) being present on IGV(version 2.16.2 Broad Institute, Cambridge and Harvard, MA, USA) [39] when visualizing the sequenced reads.Shared candidate variants were subjected to confirmation by Sanger sequencing.

Analysis of Somatic Variants in the Tumor DNA
The presence of somatic SNVs and indels in the tumor DNA of the mother were obtained by filtering all germline variants present in the matched normal DNA and by visualizing the sequenced reads of each somatic variant on IGV (version 2.16.2, Broad Institute, Cambridge and Harvard, MA, USA) [39].Mutational hotspots were annotated according to Chang et al. (2016) [40].FACETS (version 0.5.6,Memorial Sloan-Kettering Cancer Center, New York, NY, USA) [41] was used to determine Copy Number Alterations (CNAs) and whether genes harboring somatic or germline mutations were targeted by Loss Of Heterozygosity (LOH), as previously described [42].The cancer cell fractions of all somatic mutations were computed using ABSOLUTE (version 1.0.6,Broad Institute, Cambridge and Harvard, MA, USA) [43].A mutation was classified as clonal if its probability of being clonal was >50% or if the lower bound of the 95% confidence interval of its Cancer Cell Fraction (CCF) was >90%, as previously described [42].

Polymerase Chain Reaction and Sanger Sequencing Analysis
The candidate germline variant was validated by Polymerase Chain Reaction (PCR) using the QIAGEN multiplex PCR kit (QIAGEN, Hilden, Germany) and by Sanger sequencing analysis, as previously described [44].Primer design was performed (Forward: 5 ′ -CAGCCAAAATGGGTTCTGTT-3 ′ ; Reverse: 5 ′ -TTTGGACAGAAGCTGAAGCA-3 ′ ) (Integrated DNA Technologies, Coralville, IA, USA).The annealing temperature of 52 • C was established after protocol optimization for the USP42 gene.All alterations were confirmed by performing a new and independent analysis.

siRNAs-USP42 Silencing
Nthy-Ori 3-1 cells were plated in 6-well plates (1 × 10 5 cells/well).After 24 h, cells were transfected with pre-designed siRNAs (Hs_USP42_2-SI00758618 and Hs_USP42_7-SI03068583; QIAGEN, Aarhus, Denmark) or with a non-coding siRNA (Allstars Negative Control siRNA-SI03650318, QIAGEN, Aarhus, Denmark) using Lipofectamine 2000 transfection reagent (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) diluted in Opti-MEM™ I Reduced Serum Medium (Gibco, Thermo Fisher, Waltham, MA, USA).siRNA complexes were formed for 20 min in the dark and were added to the cells at the concentration of 25 nM.Medium change was performed 4h post-transfection (t = 0 h).Each experiment was performed with two replicates and at least three independent experiments were conducted.Cells were collected at 24 h, 48 h, and 72 h timepoints and counted with Trypan Blue (T8154, Merck KGaA, Darmstadt, Germany) on a Neubauer chamber for proliferation assessment.

RNA Extraction and cDNA Conversion
RNA was extracted using the GRS Total RNA Kit-Blood & Cultured Cells (GrisP, Porto, Portugal), following the manufactures' instructions.RNA concentration and sample purity and quality were analyzed using spectrophotometry, as previously described.1000 ng of RNA was converted to cDNA using Reverse Transcriptase following DNAse I and EDTA treatments.Samples were amplified and converted using the SimplyAmp Thermal cells were used as autofluorescence controls.Cells were filtered (70 µm filter) and read in the BD Accuri C6 cytometer (BD Biosciences, Qume Drive San Jose, CA, USA).At least 20,000 events per sample were evaluated using FlowJo (FlowJo™ Software version 10.9, BD Biosciences, Qume Drive San Jose, CA, USA).4.4.9.Statistical Analysis Data analyses were performed using GraphPad Prism (version 8.0.2,La Jolla, CA, USA).Two-way ANOVA or mixed-effect model analysis (non-parametric two-way ANOVA) was performed when applicable.Results are presented as the respective mean and standard error.Samples were considered significantly different when p < 0.05 (* p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001; **** p ≤ 0.0001) and confidence interval ̸ = 1.

Conclusions
We described a germline USP42 p.(Gly486Arg) mutation as an underlying cause of FNMTC in a family presenting OSPTC in all the elements under study.Our in silico results support that USPs, other than USP42, are altered in sporadic thyroid cancer, and that USPs can be in line with other mutations associated with thyroid cancer development, such as BRAF p.(Val600Glu).Our in vitro results demonstrate that silencing of the USP42 gene leads to changes in thyroid cell biology.This is shown by differences in cell proliferation, apoptotic rate, and cell cycle-related protein expression, which correlated with the altered protein expression observed in the patients' tumor samples.We hereby present evidence that USP42 gene alterations could represent a new susceptibility factor for NSFNMTC.

Figure 1 .
Figure 1.Case description and in silico analysis of non-syndromic FNMTC patients subjected to Whole-Exome Sequencing.(A) Family composed of a mother and two monozygotic twin daughters diagnosed with Oncocytic Subtype of Papillary Thyroid Carcinoma (OSPTC).Diagnosis at young age in the three elements of the family and in the absence of syndromic forms of the disease, falling into the category of Non-Syndromic Familial Non-Medullary Thyroid Carcinoma (NSFNMTC).(B) Sequencing analysis strategy for the detection of rare, potentially pathogenic Single-Nucleotide Polymorphisms (SNPs) affecting the three individuals included in this study.(C) Sequencing reads in Integrative Genomics Viewer showing the USP42 p.(Gly486Arg) variant detected in the three individuals.(D) Representative Sanger electropherograms showing the G > A alteration in the normal DNA taken from the mother (top) and the twin daughters (middle and bottom).(E) Lollipop plots depicting the frequencies of the identified somatic mutations affecting USP42 in 514 Papillary Thyroid Carcinomas subjected to Whole-Exome Sequencing that were selected from the TCGA database.UCH-ubiquitin carboxyl-hydrolase.(F) Sequencing reads in Integrative Genomics Viewer showing the USP42 p.(Gly486Arg) variant detected in the tumor DNA of the mother.(G) Copy number plot of the tumor DNA of the mother.The log-ratios are plotted on the y-axis according to genomic positions (x-axis).Chromosomes are depicted by alternating blue and red bands.(H) Heatmaps depicting non-synonymous somatic mutations (top) and cancer cell fractions (bottom) in the tumor DNA of the mother.Mutations and cancer cell fractions are color coded according to the legend.

Figure 1 .
Figure 1.Case description and in silico analysis of non-syndromic FNMTC patients subjected to Whole-Exome Sequencing.(A) Family composed of a mother and two monozygotic twin daughters diagnosed with Oncocytic Subtype of Papillary Thyroid Carcinoma (OSPTC).Diagnosis at young age in the three elements of the family and in the absence of syndromic forms of the disease, falling into the category of Non-Syndromic Familial Non-Medullary Thyroid Carcinoma (NSFNMTC).(B) Sequencing analysis strategy for the detection of rare, potentially pathogenic Single-Nucleotide Polymorphisms (SNPs) affecting the three individuals included in this study.(C) Sequencing reads in Integrative Genomics Viewer showing the USP42 p.(Gly486Arg) variant detected in the three individuals.(D) Representative Sanger electropherograms showing the G > A alteration in the normal DNA taken from the mother (top) and the twin daughters (middle and bottom).(E) Lollipop plots depicting the frequencies of the identified somatic mutations affecting USP42 in 514 Papillary Thyroid Carcinomas subjected to Whole-Exome Sequencing that were selected from the TCGA database.UCH-ubiquitin carboxyl-hydrolase.(F) Sequencing reads in Integrative Genomics Viewer showing the USP42 p.(Gly486Arg) variant detected in the tumor DNA of the mother.(G) Copy number plot of the tumor DNA of the mother.The log-ratios are plotted on the y-axis according to genomic positions (x-axis).Chromosomes are depicted by alternating blue and red bands.(H) Heatmaps depicting non-synonymous somatic mutations (top) and cancer cell fractions (bottom) in the tumor DNA of the mother.Mutations and cancer cell fractions are color coded according to the legend.

Figure 2 .
Figure 2. USP42 analysis by immunohistochemistry for FFPE tissues from the affected individuals.Adjacent (A) and tumor tissue (B) from the proband (I); adjacent (C) and tumor tissue (D) from twin II.1; and adjacent (E) and tumor tissue (F) from twin II.2.USP42 is overexpressed in the analyzed tumor tissues compared to adjacent tissue in the three individuals.Staining intensity was evaluated by semi-quantitative analysis.Bar: 50 µm.

Figure 2 .
Figure 2. USP42 analysis by immunohistochemistry for FFPE tissues from the affected individuals.Adjacent (A) and tumor tissue (B) from the proband (I); adjacent (C) and tumor tissue (D) from twin II.1; and adjacent (E) and tumor tissue (F) from twin II.2.USP42 is overexpressed in the analyzed tumor tissues compared to adjacent tissue in the three individuals.Staining intensity was evaluated by semi-quantitative analysis.Bar: 50 µm.

22 Figure 3 .
Figure 3. Caspase-3 analysis by immunohistochemistry for FFPE tissues from the affected individuals.Adjacent (A) and tumor tissue (B) from the proband (I); adjacent (C) and tumor tissue (D) from twin II.1; and adjacent (E) and tumor tissue (F) from twin II.2.Caspase-3 is overexpressed in the analyzed tumor tissues compared to adjacent tissue in the three individuals.Black arrows indicate positive cells.Staining intensity was evaluated by semi-quantitative analysis.Bar: 50 µm.

Figure 3 .
Figure 3. Caspase-3 analysis by immunohistochemistry for FFPE tissues from the affected individuals.Adjacent (A) and tumor tissue (B) from the proband (I); adjacent (C) and tumor tissue (D) from twin II.1;

Figure 4 .
Figure 4. p53 analysis by immunohistochemistry for FFPE tissues from the affected individuals.Adjacent (A) and tumor tissue (B) from the proband (I); adjacent (C) and tumor tissue (D) from twin II.1; and adjacent (E) and tumor tissue (F) from twin II.2.p53 is overexpressed in the analyzed tumor tissues compared to adjacent tissue in the three individuals.Black arrows indicate positive cells.Bar: 50 µm.

Figure 4 .
Figure 4. p53 analysis by immunohistochemistry for FFPE tissues from the affected individuals.Adjacent (A) and tumor tissue (B) from the proband (I); adjacent (C) and tumor tissue (D) from twin II.1; and adjacent (E) and tumor tissue (F) from twin II.2.p53 is overexpressed in the analyzed tumor tissues compared to adjacent tissue in the three individuals.Black arrows indicate positive cells.Bar: 50 µm.

22 Figure 7 .
Figure 7. Nthy-Ori 3-1 cell proliferation analysis following USP42 silencing treatment.Two to three cell counts were performed per well to establish cell proliferation rates.Fold-change was calculated from the initial number of plated cells [cell count/100,000)-1].This evaluation was conducted using two replicates and two independent experiments were conducted (* p ≤ 0.05; *** p ≤ 0.001).

Figure 7 .
Figure 7. Nthy-Ori 3-1 cell proliferation analysis following USP42 silencing treatment.Two to three cell counts were performed per well to establish cell proliferation rates.Fold-change was calculated from the initial number of plated cells [cell count/100,000)-1].This evaluation was conducted using two replicates and two independent experiments were conducted (* p ≤ 0.05; *** p ≤ 0.001).

Figure 8 .
Figure 8. Immunofluorescence analysis of cell morphology by phalloidin staining of actin filaments (green) contrasted with DAPI staining of cell nuclei (blue).(A,B) Negative control cells under 200× and 400× magnification, respectively.(C,D) siRNA 2-treated cells under 200× and 400× magnification, respectively.(E,F) siRNA 7-treated cells under 200× and 400× magnification, respectively.White arrows indicate cells with altered morphology, presenting loss of shape and actin concentrated in cell periphery.Images are representative of three experiments conducted independently (each with two replicates).Bar: 50 µm at 200× magnification and 20 µm at 400× magnification.

Figure 8 .
Figure 8. Immunofluorescence analysis of cell morphology by phalloidin staining of actin filaments (green) contrasted with DAPI staining of cell nuclei (blue).(A,B) Negative control cells under 200× and 400× magnification, respectively.(C,D) siRNA 2-treated cells under 200× and 400× magnification, respectively.(E,F) siRNA 7-treated cells under 200× and 400× magnification, respectively.White arrows indicate cells with altered morphology, presenting loss of shape and actin concentrated in cell periphery.Images are representative of three experiments conducted independently (each with two replicates).Bar: 50 µm at 200× magnification and 20 µm at 400× magnification.

Table 1 .
Candidate SNP identified in the three affected individuals included in this study.

Table 2 .
Somatic mutations detected in the tumor DNA subjected to Whole-Exome Sequencing.

Table 3 .
p53 quantification in different lesions from the three family elements.