Evidencing the presence of merkel cell polyomavirus in papillary thyroid cancer

Merkel cell polyomavirus (MCPyV) infects most people asymptomatically, but recent reports indicate that the virus may be related to carcinogenesis. This study aimed to evaluate the impact of MCPyV on the development of papillary thyroid cancer (PTC). Totally, 1057 samples, including 412 fresh biopsy samples (FBS) and 645 paraffin-embedded PTC biopsy samples (PEBS), and 1057 adjacent non-cancerous samples were assessed for the presence of MCPyV DNA and RNA. MCPyV DNA was positive in 215 (20.3%) of samples, including 126 (30.6%) in FBS and 89 (13.8%) in PEBS. In MCPyV-positive samples, the mean MCPyV copy number was higher in the patients with FBS (2.3 × 10–1 ± 0.5 × 10–1 copies/cell) compared to PEBS (0.7 × 10–4 ± 0.1 × 10–4 copies/cell) and adjacent non-PTC normal samples (0.3 × 10–5 ± 0.02 × 10–5 copies/cell), indicating a statistically significant difference (P < 0.001). The LT-Ag RNA expression was higher in FBS compared to PEBS, while VP1 gene transcript was not detected in any samples. Although our findings showed the presence of MCPyV in a subset of PTC Iranian patients, further research is required to confirm these findings.


Determination MCPyV DNA viral load. The MCPyV DNA viral load was evaluated by LightCycler® 96
Real-Time PCR System (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) with the PCR program and primer sequences, which were previously described 7 . The MCPyV DNA viral load was evaluated by dividing the viral DNA copy number by half of the RNase P gene copy number; each diploid cell had two copies of the RNase P gene. Plasmids for MCPyV LT-Ag and human RNase P gene (standard for real-time PCR) were prepared on the basis of previous research 8 . To draw a standard curve, real-time PCR was performed on a series of tenfold dilutions of the purified plasmids of MCPyV LT-Ag and RNase P, ranging from 2 × 10 1 to 2 × 10 6 copies/μl. In order to eliminate the possibility of contamination and false positive results, adjacent normal non-PTC samples were tested.

Determination of MCPyV RNA expression. Total RNA from the PEBS and FBS sections was extracted
with the RNeasy Kits (QIAGEN, CA, USA), according to the manufacturer's instructions. The LT-Ag transcripts were amplified by using qualitative real-time reverse transcription PCR (real-time RT-PCR) to indicate the presence of MCPyV by using primers and PCR program, which were previously explained 9 . RNA extracted from MCC patients was used as a positive control. To confirm the specificity of the amplification, a melting curve analysis was performed on the PCR products. Statistical analysis. Data analysis was performed using IBM SPSS version 22.0 software (SPSS. Inc., Chicago, IL, USA). The Shapiro-Wilk test was utilized to assess the data normality of continuous data. Pearson's chisquare and Mann-Whitney U tests were also applied to assess quantitative variables and continuous variables, respectively. Two-tailed P-value less than 0.05 was considered statistically significant.
The mean MCPyV copy number in adjacent non-PTC normal samples was 0.3 × 10 -5 ± 0.02 × 10 -5 copies/cell that was very lower than the cancerous samples.   8,[11][12][13] . The findings of our study indicated that the rate of cigarette smoking in women was more than men. Evidences are available that cigarette smoking may be correlated with thyroid cancer risk. Cho et al. in a cohort study demonstrated current smoking was correlated with a decreased risk of incident PTC in men but not in women 14 . The findings of pooled analysis of five prospective reports suggest that smoking is related to a 30-40% decreased risk of PTC 15 . Cigarette smoking could potentially affect the risk of thyroid cancer by changing sex steroid hormone level, serum thyroid antibodies, and thyroid stimulating hormone. Because of self-administered questionnaires have been used in all studies, selection and recall bias may have contributed to the null correlation 16 .

Comparison of RNA expression of MCPyV LT-Ag in FBS and
To the best of our knowledge, this is the first study that evaluated the relationship between MCPyV infection and PTC patients according to pathologic features. To date, few studies have evaluated the association of polyomavirus with thyroid cancer. In those studies, three of them detected sequences of simian vacuolating virus 40 (SV40) and the other one, BK in post-operative thyroid [17][18][19][20] .
However, realization of Koch's four postulates has demonstrated difficult for any of the carcinogenic viruses discovered to date, the oncogenicity of polyomaviruses in thyroid cancer is still controversial. Genome integration, instead of the mere viral genome sequences or proteins identification, has been proposed as a means of elucidating the link between viruses and cancer, which could potentially provide answers 21 . More convincing evidence is definitely needed. The current study is just the beginning of a long research journey to clarify whether these viruses are the cause of the development of thyroid cancer or just represent innocent bystanders. However, this study indicated the relationship between MCPyV positivity and PTC.
In this study, we compared the MCPyV positivity and DNA viral load between FBS and PEBS samples. The findings of this study indicated that MCPyV DNA in 30.6% of FBS was positive, while the frequency of MCPyV DNA in PEBS was 13.8%. Also, the mean MCPyV DNA viral load was higher in the patients with FBS compared to PEBS, indicating a statistically significant difference. It seems that the quality of the samples was the reason for the differences in the number of positive samples by PCR with LT3 primers and the genome copy number per cell. As a result, the FBS samples were suitable for detection of viral infection in PTC patients.
However, FBS are advantageous in the ways where the process is much faster compared to the PEBS process. While PEBS are not adequate for molecular analysis, FBS is very suitable for this issue. This is because of the PEBS preparation that affects the molecular data. FBS are also preferred in analyses such as next generation sequencing, mass spectrometry, and quantitative real-time PCR. Therefore, this type of samples is the gold standard for DNA and RNA analysis 22,23 .
The inability to detect MCPyV LT-Ag DNA and RNA in most positive PEBS, and the low number of DNA copies per cell, may be due to many reasons. First, it is well-known that DNA and RNA degradation occurs in PEBS, resulting in poor quality of DNA and RNA yields for routine and real-time PCR analysis 11,24 . Second, the low copy number of viral RNA and DNA in PEBS may indicate that MCPyV in these samples exists as a passenger virus without certain pathological findings. Third, MCPyV might play a role in cancer pathogenesis through the "hit-and-run" mechanism 8,12 . Viral hit-and-run oncogenesis scenarios indicate that transient acquisition of the genomes of oncoviruses can induce a permanent change in the gene expression pattern of the host cell, leading to malignant transformation 11 . In this case, viral genomes or small viral fragments may be found in malignant tumors or in the precursor stages of the tumor. The T-Ag expression of is necessary for growth of MCC cell lines in vitro, and in vivo has been established MCPyV as a causative factor in the MCC oncogenesis. Ironically, both the transformational functions and the mechanisms of hit and run in non-transforming viruses that appear to be unrelated to human cancer have been explored 11,25,26 .
To define MCPyV as another infectious agent correlated with PTC, DNA positivity of MCPyV alone is not enough to determine its etiology 9 . Several reports have indicated that the MCPyV LT-Ag expression is required for the oncogenesis of MCPyV-positive MCC 27,28 . In this case, we evaluated the LT-Ag gene expression at the RNA level. In our study, 75.4% of FBS and 26.2% of PEBS samples expressed LT-Ag RNA; while VP1 gene transcript was not detected in any samples. Polyomaviruses replication, like MCPyV, shows an ordered gene expression cascade in which the LT gene transcript is expressed first as an early gene transcription, followed by the VP1 gene expression as late gene transcription 29 . However, the loss of the viral replication ability is a common feature of virus-related tumors 30 . In most MCCs infected with MCPyV in which viral replication is disrupted, only the truncated LT gene is constitutively expressed, but not the VP1 gene that cannot support replication (DNA binding domain is lacking) and the viral genome is integrated 6,31 . According to these findings, it is rational that we only found the LT RNA expression in PTC samples, but not the RNA expression of the VP1 gene that reflects virus replication. To prove it, Hashida et al. used immunohistochemistry to evaluate the MCPyV LT-Ag expression and localization. The localization of powerful immunoreactivity in the tumor cell nuclei revealed the MCPyV LT-Ag expression in the lung cancer cells and also the genome integration of MCPyV into the host genome is considered as a key element in carcinogenesis 9 . The lack of performing immunohistochemistry especially for small T-Ag, truncated LT-Ag expression, and viral integration were the limitations of this study.
In this study, the MCPyV LT-Ag DNA loads were higher in cancerous cells than in non-cancerous samples. Several studies in different types of cancer showed the frequency of MCPyV DNA was significantly higher in the tumor cells than non-cancerous tissues 11,32 . This finding implies a transitory role for MCPyV in cell transformation, since its genome can be silenced or lost during cancer progression, which refers to the hit-and-run mechanism 33 . www.nature.com/scientificreports/ In conclusion, the present study, for the first time, not only identified MCPyV DNA, but also the LT RNA transcripts expressions in PTC. Also, we indicated the viral DNA load in FBS was higher than in PEBS. However, more epidemiological and virological studies are required to determine the relationship between the pathogenicity of MCPyV with PTC.