| Background Changes of epigenetic modification play an important role in tumorigenesis.5-hydroxymethylcytosine (5hmC), related to DNA demethylation, is a kind of epigenetic modification of DNA. Both amount and modification of 5hmC are greatly changed in many tumors, likes kidney cancer, breast cancer and glioma. Ten-eleven translocation 2 (TET2) is a kind of dioxygenases responsible for catalyzing 5-methylcytosine (5mC) into 5hmC,5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) step by step. Mutations of TET2 occur frequently in myeloid diseases and influence 5hmC levels. Besides, TET2 expression and distribution are directly related with 5hmC amount. Epigenetic factors have been proven to contribute to pituitary adenomas formation, however it is unclear yet if 5hmC is involved in it. There are few studies about 5hmC dysregulation in nonfunctioning pituitary adenomas (NFPAs). We should pay attention to mechanisms underlying NFPAs due to their high prevalence, tremendous harm and difficult treatment.Purpose Our study aimed to detect cytosine modifications in NFPAs, and to explore the mechanism underlying the changes of 5hmC. We detected the gene variations, protein expression and subcellular localization of TET2 in NFPA patients, and explored their possible correlation with 5hmC abundance. Our study included 3 parts. Firstly, we collected samples and clinical information. Samples were separated into 3 groups. We measured 5hmC amount in NFPAs and normal pituitary glands. Secondly, TET2 sequencing was performed in all NFPAs for TET2 mutations. Thirdly, we conducted immunohistochemical analysis using samples of NFPAs and a normal pituitary gland.Methods We studied 57 NFPAs (all for ultrahigh performance liquid chromatography-electrospray ionization-tandem mass spectrometry (UPLC-ESI-MS/MS) and TET2 sequencing,26 of them for immunohistochemical analysis) and 6 normal pituitary glands (5 of them for UPLC-ESI-MS/MS and 1 of them for immunohistochemical analysis) and collected clinical informations. NFPAs were separated into invasive NFPAs and noninvasive NFPAs. This work was respectively conducted by two excellent neurosurgeons in Peking Union Medical College Hospital judging by imaging results and surgery field.5hmC amount was detected by mass spectrum.TET2 alteration was searched by sequencing after PCR. TET2 expression and distribution was shown by immunohistochemistry. Expression levels were shown through H-score calculated by software.Results Compared with those in normal pituitary glands, genomic 5hmC levels in NFPAs were significantly decreased (0.38%o (0.13-1.23%o) vs 2.47%0 (0.82-2.86%o), P<0.0001), while genomic 5caC levels were significantly increased (0.20‰(0.04-0.24‰) vs0.16‰ (0.14-0.19‰), P=0.005). Patients with lower 5hmC levels had bigger tumors (34.6±12.7mm vs 27.4±8.5mm, P=0.023) and higher Ki-67 indexes (9/5 vs 16/0, P=0.014). There was no difference of age, gender, course of disease, invasion, P53 and Knosp grade, between low 5hmC patients and high 5hmC patients. Genomic 5hmC levels in noninvasive NFPAs were similar to those in invasive NFPAs (0.44%o (0.15-1.04‰) vs 0.35‰ (0.13-1.23%o), P=0.30). There were 3 single nucleotide polymorphisms (SNP) in TET2. They were TET2c.86C>G, c.5162T>G and c.5284A>G in plus strand, leading to TET2 p.P29R, p.L1721W and p.11762V. Compared with those in TET2 P29 group, genomic 5hmC levels in TET2 R29 group were greatly decreased (0.25‰(0.13-0.80‰) vs 0.46‰(0.16-1.23‰), P=0.013). We did not find correlations of TET2 p.L1721V, p.11762V with 5hmC amount. There was medium positive correlation of 5hmC amount with TET2 expression in NFPAs (r=0.461, P=0.018; r=0.458, P=0.019). According to genomic 5hmC levels detected by UPLC-ESI-MS/MS, samples performing immunohistochemistry were separated into two groups. TET2 expression of high 5hmC group was higher than that of low 5hmC group (192.78±79.87 vs 129.58±60.18, P=0.032<0.05; 121.49±49.21 vs 80.07±36.68, P=0.023). There was no significant difference of TET2 expression and distribution between noninvasive NFPAs and invasive NFPAs.Conclusions 1. Genomic 5hmC loss is present in NFPAs. Lower 5hmC levels imply bigger tumor sizes and higher Ki-67 indexes. The decrease of 5hmC amount may be responsible for NFPAs tumorigenesis, and related to tumor size and Ki-67 index.2. Samples with TET2 p.P29R show lower 5hmC. It is unclear how TET2 p.P29R alteration affects 5hmC levels. With increases in TET2 expression and nuclear localization, genomic 5hmC levels increase. Genomic 5hmC levels may be affected by TET2 expression and subcellular localization. |
PDF Full Text Request