not-yet-known not-yet-known not-yet-known unknown 4. DISCUSSION The incidence of lung cancer is increasing worldwide in women and nonsmokers across all ages and races/ethnicities.35 It has been shown that among nonsmokers, women with a history of HPV infection have an increased risk of lung cancer.36 Additionally, regional differences in HPV incidence in lung cancer patients are greater in Asia than in North America and Europe.9 MCPyV, which has been shown to be strongly involved in the development of MCC, has also been investigated for its association with several human tumors, including skin cancer.12,37 In a study by Hashida et al., viral integration and mutation in MCC were confirmed in NSCLC, a specific cancer other than MCC, and 17.9% (20 of 112) of Japanese NSCLC patients were positive for MCPyV DNA.16 Another study by Xu et al. showed that early-stage I NSCLC patients were more likely to be infected with MCPyV than patients with cancers in other stages, with an MCPyV infection rate of 16.9% among NSCLC samples.17 Reports by Gheit and Behdarvand revealed MCPyV DNA in 4.7% and 38.5% of NSCLC patients, respectively,15,18 but Shikova et al. did not detect MCPyV DNA in any specimens of patients with nonmalignant chronic lung disease or lung cancer.38 Overall, the prevalence of MCPyV in adjacent non-NSCLC patients has been reported to be significantly low or nonexistent, at 1.04% (1 of 96) or 6.67% (1 of 15), although the sample sizes of these studies were small.13,17-19,39 Such differences in detection rates may be the result of different experimental setups in laboratories (including differences in DNA quality) and geographic epidemiological variations in patients. In the present study, we detected MCPyV DNA in 22.7% (34 of 150) of NSCLC tissues and 8.0% (12 of 150) of adjacent non-NSCLC tissues from Korean patients (Table 1 and Figure 1 ). Interestingly, the 12 patients with MCPyV-positive adjacent non-NSCLC tissues included 8 female AD patients without a history of smoking (Table 2 ). In general, the incidence of lung cancer in nonsmoking women is increasing worldwide;40 it has been suggested that lung cancer in nonsmoking women should be considered different from smoking-related lung cancer, and HPV has been proposed as a risk factor.41 Moreover, previous studies have reported that coinfection with different oncogenic DNA viruses can play an important role in long-term immunosuppression/immunocompromised status and further activation/promotion of proto-oncogenes.42 Therefore, we investigated patients with NSCLC to determine whether a direct association between MCPyV and HPV infection and lung cancer exists using PCR and DNA sequence analysis. Among the 34 MCPyV-positive patients with NSCLC, 9 (26.5%) were HPV-positive; among the 116 non-MCPyV patients, 15 (12.9%) were HPV-positive (Table 2 , Table 3 and Supplementary Table S2 ). HPV-16/-18 are common high-risk genotypes found in lung cancer. Regarding the HPV-type distribution in the present study, 9 HPV-positive NSCLCs with MCPyV were detected, including 8 HPV-16 and 1 HPV-33 sample (Table 2 ). Among 15 HPV-positive NSCLCs without MCPyV, 12 had HPV-16, and 3 had HPV-18 (Supplementary Table S2 ). Conversely, adjacent normal lung tissues showed no cases of HPV. Overall, HPV-16 was predominant regardless of MCPyV-infection status. Based on these data, we showed that the prevalence of MCPyV-positive/HPV-positive NSCLC was greater than that of MCPyV-negative/HPV-positive NSCLC, but the associations between the two variables were not statistically significant. Thus, further research with more statistical power is needed to determine the impact of MCPyV/HPV infections (i.e., mixed infections) on NSCLC. Additionally, we investigated sequence data obtained from amplified full-length MCPyV genomes and two target regions of MCPyV the LT and VP1 genes for in-depth comparisons of the sequences and specific genotypes of Korean patients with NSCLC. The sequences obtained from 34 MCPyV-positive samples were compared with those of six reported reference isolates from North America (MCC350; MCC339), Japan (TKS), China (HB039C), Sweden, France and Italy (MKL-1), and Korea (KIB). For both samples that were MCPyV-positive in the LT region, four single-nucleotide substitutions of specific Korean origin, which differed from a previously reported Korean isolate (KIB), were found (JL93 and JL95) (Figure 2 and Supplementary Figure S1 ). These four mutations were specifically detected only in Koreans and were not detected in the six reference strains. The VP1 region was also compared with the reported six reference strains, and two specific Korean mutations not yet reported were discovered (at positions 4,152 and 4,153 in three samples, BL1, BL4, and BL28) (Figure 3 and Supplementary Figure S2 ). At positions 4,362 and 4,368, the point mutations in North American/European strains were G > A mutations, but those in Asian strains of Chinese/Korean origin were A > G mutations (Japanese strain, TKS at location 4,368 is the same as that of North American/European strains). Furthermore, among the detected MCPyV-positive samples, two mutations, at positions 4,362 and 4,368, revealed trends of a mixed type with North American/European strains and Chinese/Japan/Korean strains. Rearrangements in the NCCR, including mutations, deletions, and duplications, of human polyomaviruses such as BKV and JCV impact the transcriptional activity of the promoter and the functions of enhancer elements.43 Variants with rearrangements are known to show high variability within the NCCR, and increasing gene expression and replication ability are closely related to disease development.24-25,43 In a study by Hashida and colleagues, an analysis of the genetic variability of the MCPyV NCCRs recovered from skin swab specimens of healthy individuals of distinct ethnicities and geographical origins was performed, and the MCPyV strains, with the two major subtypes I and II, were identified based on the presence/absence of 25 bp tandem repeats at nucleotide positions 5177-5178. In addition, based on sequence variability as well as two insertions and deletions, MCPyV strains are subdivided into five genotypes.26 Interestingly, subtype I, with 25 bp tandem repeats, is the dominant type in East Asia (Chinese), including Japan. Therefore, we investigated patients with NSCLC to confirm the MCPyV NCCR genotype in Korean individuals residing in the same geographical region of East Asia, and only two genotypes, subtypes I and IIc, were identified. The overall prevalence of the MCPyV NCCR was 27.3%, and among the genotypes, the prevalence of subtype I was 21.3%, and that of subtype IIc was 6% (Table 4 and Figure 4 ). For subtype I, Hashida et al.26 reported a prevalence of 93.5% for skin swab specimens from asymptomatic healthy individuals and a prevalence of 85% for MCC patient tumors, which were higher than those found in tissue samples from Korean patients with NSCLC (21.3%). The detection rate of 21.3% determined in this research may be an underestimate due to the very low number of MCPyV DNA copies recovered from tissue specimens. Indeed, a previous study44 reported that very low copy numbers of DNA can lead to low prevalence results based on DNA quality and experimental protocols using buffy coat samples. Within the globally distributed subtype II, only subgroup IIc (6%) was detected in our study. Alignments of the MCPyV NCCR obtained by sequencing, including one full-length MCPyV genome, revealed only point mutations, excluding tandem repeats, insertions, and deletions, which are characteristics of each subtype previously reported by a group from Japan.26 We performed binding site analysis of various nucleotide positions detected using previously reported45 putative binding sites (NF1, NFkB, Tst-1, OCT1, AP-1, and TATA) of NCCR sequences. The MCPyV NCCR sequence (nucleotide positions 5153-5211) obtained from patients with NSCLC was compared to that of the North American strain MCC350 as a reference,12 and changes in the nucleotide sequence were confirmed at the three binding sites (TATA, AP-1, and Tst-1). In sample BL13, the putative TATA element (nucleotide positions 5146-5160 and 5153) also showed a T to C change, as observed in AP-1 at positions 5164-5172; at position 5168, an A to G change was observed; and the Tst-1 binding site at positions 5210-5211 showed T to C and G to A changes at positions 5238 and 5239, respectively. Surprisingly, a change at position 5239 was observed in several samples, namely, BL78, JL138, and JL150, which share common clinical characteristics with AD and often metastasize to lymph nodes (Figure 5 and Supplementary Figure S3 ). Previous studies have reported that Tst-1 in the POU-domain stimulates the expression of small and large tumor antigens and alters the cellular transcription pattern, and Tst-1 also stimulates the transcription of both early and late viral genes, affecting viral genomes in infectious particles.46 It is possible that these point mutations affect MCPyV virulence by increasing the Tst-1 binding capacity and viral DNA replication and may occur more frequently as the cancer progresses. However, the effect of these mutations on viral virulence has not yet been determined. The NCCRs of MCPyV and other human polyomaviruses, including KIPyV and WUPyV, HPyV6 and HPyV7, and TSPyV, HPyV9, HPyV10, and STLPyV, contain many putative transcription factor binding motifs,47 and mutations at these sites affect binding and are closely related to DNA replication. For instance, some nucleotide mutations (G143, C145A, A173, and C176) within the putative LT-binding motif have been shown to be involved in the replication of tumor-derived viral strains (MCV339 and MCV350),47,48 and it has been suggested that the common V392G mutation in JCV TAg, which inhibits viral replication, may be involved in progressive multifocal leukoencephalopathy (PML) lesions depending on its frequency.49 In light of these findings, mutations in the HPyV NCCR might stimulate transcriptional activity of the promoter and affect viral replication and virulence.45-47,50-51 However, NCCR variants (mutations, insertions, or deletions) may create putative binding motifs for transcription factors, but information about their function in regulating promoter activity is lacking. Future studies of NCCR rearrangement will require large-scale, diverse, in-depth investigations of the mechanisms regulating the binding of common transcription factors with promoter activity to determine their effects on viral replication and specific diseases. In conclusion, our data revealed several key findings: (a) MCPyV DNA was detected in both NSCLC and adjacent non-small cell lung cancer (NSCLC) tissues. (b) However, we did not find an association between coinfection/infection with HPV/MCPyV and NSCLC. (c) We identified G-to-A point mutations in the MCPyV NCCR Tst-1 binding site in several NSCLC patients with adenocarcinoma (AD) and lymph node metastases. (d) Additionally, we revealed Korean variant genotypes and sequences of the MCPyV NCCR in patients with NSCLC. Although this study did not find evidence of coinfection with other oncogenic viruses in lung cancer patients, it is noteworthy that viral coinfection has been reported in many metastatic cancers.42,52-53 The cooperative and pathogenic effects of coinfecting viruses warrant further investigation. To our knowledge, this is the first report to identify a Korean variant genotype of the MCPyV NCCR and to evaluate the NCCR sequence. The findings provide information for ethnic genotyping and suggest that future studies are needed to gain a comprehensive understanding of viral reactivation caused by NCCR rearrangements and their potential to affect the immune system, leading to specific diseases.