Deciency in Vitamin B12 and Ferritin Is Associated With Epigenetic Alterations in Cancer Patients

Background: Cancer is the second-leading cause of death worldwide, caused by several mutations in DNA within the cells including epigenetic alteration. The epigenetic changes are external modications to the DNA that switch “on” or “off” gene expression. The present study was conducted to investigate the epigenetic modications and its correlation with the levels of vitamin B12 and ferritin in cancer patients with hepatocellular carcinoma (HCC), breast cancer (BC), lung cancer (LC), or colon cancer (CC). Methods and Results: A total of 200 blood samples were obtained from cancer patients and healthy individuals. The relative expression of DNA methyltransferases (DNMTs), Ten-Eleven translocation (TET), and methionine synthase (MS) was evaluated in patients with the normal level of vitamin B12/ferritin and patients with the decient levels of them. DNA methylation within the promoter regions was investigated of each indicated genes using the methylation-sensitive restriction enzyme HpaII and bisulte PCR. Interestingly, the expression of DNMT1, DNMT3a, and DNMT3b was increased in patients with low levels of vitamin B12 and ferritin, while the expression of MS, TET1, and TET3 was signicantly decreased. DNA methylation analysis in patients with decient levels of vitamin B12/ferritin showed a methylated-cytosine within the location 318/CG and 385/CG in the promoter region of TET1 and TET3, respectively. Moreover, the bisulte PCR assay further conrmed the methylation changes in the promoter region of TET1 and TET3 at the indicated locations. Conclusion: These data indicate that the deciency in vitamin B12 and ferritin in cancer patients plays a key role in the epigenetic exchanges during cancer development. Bisulte-converted DNA revealed a signicant CpG methylated motif in TET1 and TET3 genes associated with vitamin B12/ferritin level in cancer patients


Introduction
between the low levels of vitamin B12/ferritin in patients and hypermethylation of methionine synthase, the essential co-factor of DNMT and ten-eleven translocation (TET) 1 and 2, the methycytosine dioxygenases.

Ethical issues
This work was conducted at the faculty of medicine (Mansora University, Egypt) and Genetic Engineering and Biotechnology Research Institute (University of Sadat City, Egypt) from May 2018 to March 2020. The present study has been approved by the medical ethics and scienti c ethics committees of the Faculty of Medicine and Genetic Engineering and Biotechnology Research Institute, respectively. The aim of this work was explained to patients who were agreed and understood all the rules and regulations. The most important criteria to be stated are that all samples were obtained from patients with a mean age of 40 ± 10 years and before exposure to any treatment. Exclusion criteria were including age (< 50 or > 30 years), patients diagnosed with diabetes mellitus, patients with renal failure, patients treated with drugs affected plasma ferritin, vitamin B 12 , folate, or zinc, and patients who were underweight or suffered from cachexia, or with a history of an eating disorder.

Samples conditions
A total of 160 blood samples were obtained equally from patients with hepatocellular carcinoma (HCC), breast cancer (BC), lung cancer (LC), and colon cancer (CC), in addition to 40 blood samples obtained from healthy individuals investigated as control. All patients were under health care, medical supervision, and accurate diagnosis of the cancer development and they were all in the advanced stage of cancer. Collected blood samples were further analyzed for the signs of cancer including tumor markers, the presence of cancer cells, and proteins or other substances made by cancer.

Biochemical analysis
The levels of serum ferritin, vitamin B12, and folate were measured using Elecsys Ferritin (REF 04491785190), Elecsys B12 II (REF 07212771190), and Elecsys Folate III (REF 07559992190), respectively via the electrochemiluminescence immunoassay (ECLIA) (Roche) [17]. For serum ferritin measurement, the blood samples collected in a trace-element free 8 mL tube (Becton Dickinson) were allowed to clot and centrifuged. 10 µL of each sample was incubated with a biotinylated monoclonal ferritinspeci c antibody and a monoclonal ferritinspeci c antibody labeled with a ruthenium complex for 9 min. Then the streptavidincoated microparticles were added to the mixture which incubated for 9 min. The reaction was then transferred into the measuring cell and the chemiluminescent emission was measured by a photomultiplier [18].
To measure vitamin B12, 15 µL of the collected serum was incubated with vitamin B12 pretreatment 1 and pretreatment 2 reagents for 9 min. Then the pretreated samples were incubated with the ruthenium-labeled vitamin B12 binding factor for 9 min followed by incubation with streptavidin-coated microparticles and vitamin B12 labeled with biotin for 9 min. The reaction was then transferred into measuring cells and the induced chemiluminescent emission was measured by a photomultiplier. The nal values were determined using the master curve provided via reagents barcode. Similarly, the level of folate in prepared serum was measured using 25 µL serum incubated with folate pretreated 1 and 2 reagents, ruthenium-labeled folate binding protein, and folate labeled with biotin.

DNA isolation and epigenetic analysis
The genomic DNA was isolated from obtained samples by using a DNA puri cation kit (Qiagen, USA) according to the manufacturer's protocol. Genomic DNA was digested with the methylation-dependent restriction enzyme HpaII (Thermo Scienti c, USA) that unable to cleave the 5-methyle cytosine in its own restriction site 5-CCGG-3 by using ve units of the enzyme to digest 200 ng of DNA during the incubation period of 4 hrs at 37 º C. The digested products were loaded in 1% agarose gel and electrophoresed using 1X-TBE buffer. The electrophoresed fragments were monitored using the UV transilluminator with a long-wave (320 nm) UV and gel documentation system [19].

DNA elution and sequences
The HpaII-cleaved genomic DNA of cancer patients were eluted form the agarose gel using GeneJET gel extraction kit (Thermo Scienti c, USA). Brie y, the gel slice was excised using a clean scalper and was weighted in 1.5ml-eppendorf tube. The gel slices (100mg) was then incubated with 100µL of binding buffer for 15min at 60 º C. The dissolved gel was mixed very well by inversion and vortex, and then 100µL isopropanol was added to the mixture before loaded on the puri cation columns. The columns were centrifuged at 12000 rpm for 1 min and nally the DNA fragment was dissolved and collected from the columns using 50µL elusion buffer. Sequencing analysis of eluted fragments we performed using Illumina NovaSeq 6000 sequencing system [20]. The DNA alignment between sequences data and the 3`-UTR of Atg7 and LC3B was done by using online tool [21].
Bisul te PCR assay Bisulphite analysis of TET1 and TET3 promoter methylation was performed using the spin columns of EpiTect Bisulphite kit (Qiagen, USA). According to the manufacturer protocol, the puri ed genomic DNA was thermal denaturized and treated with sodium bisul te using DNA protect buffer and incubated for 7 hrs for complete conversion. The converted DNA was applied to an EpiTect spin column using optimized buffers, then washed to remove all traces of sodium bisul te and eluted from spin columns. Firstly designed oligoniclutids have been used to detect the methylated or unmethylated fragments into the promoter region of TET1 and TET3 at the location 318/CG and 385/CG, respectively. These speci c oligonucleotides contain the forward primer speci c for each gene location and two reverse primers. The rst reverse primer contains the original or wild type sequences indicating the ampli ed unmethylated fragments. While the second reverse primer or the methylation-insensitive primer (MIP) has degenerated bases (G/A) to cope with the uncertain C/U conversion indicates the ampli ed methylated fragments. Therefore, the oligonucleotides contain triple primers speci c for TET1; TET1-F-C-5'-GGCTCGGGCCTTGACTGTGCTG-3, TET1-R-W-5'-AGGTTTTGGTCGCTGGCCGGGT-3' and TET1-R-M-5'-CACTGGCCAGGTCACATTCCCA − 3'. The following triple primers have been used for ampli cation of TET3 fragment, TET3-F-C-5'-GAGGCGGCGGGCAGGCAGCACT-3', TET3-F-W-5'-AAAGGGATCATCTTGGCCCGGT-3', and TET3-R-M-5'-GCCCAGTGGGTTCTTCACCCTC-3' [22]. The conventional PCR was used to amplify the methylated and unmethylated fragments using the following parameters; 95 º C for 5 min, and 40 cycles (95 º C for 30 sec, 62 º C for 30sec, and 72 º C for 30 sec) [23,24]. The PCR products were loaded in 1% agarose gel and electrophoresed using 1X-TBE buffer and the ampli ed fragments were monitored using the UV trans-illuminator with a long-wave (320 nm) UV and gel documentation system.

RNA isolation and qRT-PCR
Total RNA were isolated from obtained samples by using TriZol (Invitrogen, USA), chloroform and isopropanol. Then the total RNA was puri ed by using RNA puri cation kit (Invitrogen, USA). The complementary DNA (cDNA) was synthesized from total RNA using QuantiTech Reverse Transcriptase Kit (Qiagen, USA) according to the manufacturer's protocol. The relative gene expression of DNMT1, DNMT2, DNMT3a, DNMT3b, MS, TET1, TET2 and TET3 was assessed in samples derived from patients compared with their expression in samples obtained from healthy individuals using speci c oligonucleotides listed in Table 1. The steady state mRNA of the indicated genes was quanti ed using the QuantiTect SYBR Green PCR Kit (Qiagen, USA). Expression level of the housekeeping gene, GAPDH, was used for normalization. The following PCR parameters were used to detect fold changes in genes expression, 94°C for 5 min, 40 cycles (94°C for 30 sec, 60°C for 30 sec, 72°C for 45 sec). Delta-delta Ct analysis has been used to determine the relative gene expression indicated by fold changes in steady state mRNA [14,25].

Statistical analysis and prediction tools
The data obtained by q-RT-PCR that reveal the cycle threshold values (Ct) were analyzed using the previously described ΔΔCt equations. Thus, the relative gene expression of targeted genes at RNA level was indicated by fold changes that equaled to 2 −ΔΔct [26]. Student's two tails t-test was used to determine the differences in analyzed data. *P < 0.05 was considered statistically signi cant and **P < 0.01 as highly signi cant.

De ciency of serum vitamin B12 and ferritin is recognized in cancer patients
Out of 200 patients screened and offered baseline hemoglobin, serum ferritin, serum folate, and vitamin B12 estimation, 40 patients in each type of cancer were nally enrolled in the study. As shown in Table 2, the prevalence of anemia was almost 80% in patients with HCC, since the persons who showed de cient levels of hemoglobin were 32 out of 40. These patients also showed de cient levels of ferritin, vitamin B12, transferrin saturation, and increasing levels of homocysteine. In patients with BC, 75% of patients have severe anemia with de cient levels of ferritin, vitamin B12, transferrin saturation, and increasing levels of homocysteine. The prevalence of anemia in patients with LC was 70% of patients accompanied by de cient levels of ferritin, vitamin B12, and transferrin saturation, and high levels of homocysteine. Likewise, 85% of patients with CC were anemic with de cient levels of ferritin, vitamin B12, and transferrin saturation, and high levels of homocysteine. Unlikely, more than 80% of cancer patients showed normal levels of serum folate, while a maximum of 20% of cancer patients showed low levels of serum folate. These data indicate that the cancer patients with HCC, BC, LC, or CC represented de cient levels of hemoglobin, serum ferritin, vitamin B12, and subsequently low saturation levels and accumulated levels of homocysteine.
The relative expression of DNMTs is associated with de ciency of vitamin B12/ferritin in cancer patients To assess the correlation between DNA methylation and vitamin B12/ferritin de ciency in cancer, the relative gene expression of DNMT1, DNMT2, DNMT3a, and DNMT3b in obtained blood samples from patients with HCC, BC, LC, and CC were compared to blood samples derived from healthy individuals using qRT-PCR. Our results revealed a signi cantly increased expression level of DNMT1, DNMT3a, and DNMT3b, but not DNMT2 that connected with the low level of vitamin B12 and ferritin, while their expression signi cantly reduced in patients represented normal levels of vitamin B12 and ferritin (Fig. 1A, B, C, and D). Notably, the calculated P values for the expression level of DNMT1 and DNMT3b were lower than 0.01 in samples derived from patients with HCC, BC, LC, and CC. Whereas the calculated P values for the expression level of DNMT3b were about 0.05 (Tables 3-6).
These ndings indicate the possible connection between DNMTs expression, particularly DNMT1, DNMT3a, and DNMT3b, and levels of vitamin B12/ferritin in patients with the indicated cancer diseases.   To achieve the correlation between low levels of vitamin B12/ferritin and gene expression of other essential cofactors involved in DNA methylation in cancer patients, the relative gene expression of MS, TET1, TET2, and TET3 was evaluated by qRT-PCR in obtained blood samples. Interestingly, the results showed a signi cant reduction in the expression level of indicated genes; expect TET2 expression, associated with the de ciency of vitamin B12/ferritin in patients with HCC, BC, LC, and CC ( Fig. 2A, B, C, and D). Importantly, the calculated P values for MS, TET1, and TET3 expression level were lower than 0.01 in obtained samples from patients with low level of vitamin B12/ferritin compared with samples derived from patients with normal level of vitamin B12/ferritin and healthy individuals (Tables 7, 8, and 10). Whereas the calculated P values for TET2 expression level showed insigni cant differences between patients and healthy individuals (Table 10). These data reveal that the de ciency of serum vitamin B12 and ferritin in patients with cancer may contribute in the regulation of MS, TET1, and TET3 genes expression and subsequently controlling the DNA methylation process.   Patients with de cient vitamin B12 and ferritin revealed a methylation activity particularly in TET1 and TET3 promoter region DNA methylation at CpG dinucleotides is a major epigenetic marker leading to gene silencing when located in the promoter region of a certain gene [27].
Noteworthy, the identi ed restriction site for HapII enzyme is (CCGG) with an independent sensitivity to cleave the methylated CpG [28,29]. Based on this, we incubated the puri ed genomic DNA with HapII to compare and analyze the differences in cleaved fragments in cancer patients that are associated with the level of vitamin B12 and ferritin. Interestingly, agarose gel electrophoresis of HapII-cleaved fragments showed an alteration in the number of cleaved fragments in patients with a normal level of vitamin B12/ferritin and healthy on one side and patients with a low level of vitamin B12/ferritin on the other side. As shown in Figs. 3A and B, almost 7 cleaved fragments were found in patients with the normal level of vitamin B12/ferritin, while only 2 cleaved fragments were detected in patients with the low level of vitamin B12/ferritin. These ndings clearly indicate that the HapII resection enzyme has failed to digest the genomic DNA isolated from cancer patients with a de cient level of vitamin B12/ferritin due to the potential DNA methylation activity on its binding site. Furthermore, to identify the potential methylated region, the low-intensity band cleaved from genomic DNA puri ed from patients with the normal level of vitamin B12/ferritin (approximately 100bp) was sequenced and aligned with deferent sequences of various related genes. Fortunately, the sequence and alignment analysis showed a consensus sequence of about 80 nucleotides within the promoter region of TET1 and TET3, started at the location 318/CG and 385/CG, respectively ( Fig. 3C and D). These ndings strongly suggest that TET1 and TET3 were downregulated in cancer patients in a vitamin B12/ferritin dependent manner due to the methylation potential in their promoter regions.
Bisul te-converted DNA revealed a signi cant CpG methylated motif in TET1 and TET3 genes associated with vitamin B12/ferritin level in cancer patients Bisul te DNA sequencing was recognized as an accurate method to detect DNA methylation-based on the conversion of genomic DNA using sodium bisul te [30]. In this way, the methylated cytosine is converted into uracil and recognized in PCR ampli cation as thymine [31]. Therefore, we designed speci c oligoniclutids for both TET1 and TET3 methylated regions including common forward primers and two reverse primers for each TET gene. The rst reverse primer contains the complementary sequences of the normal motif CG named wild-type reverse primer (R-W) at the 3 − end of the primer.
While the second reverse primer contains the complementary sequences of converted cytosine in CpG motif and named methylated reverse primer (R-M) at the 5 − end of the primer, to make the ampli ed fragment shorter than the ampli ed wild-type fragment, and contained adenine nucleotide instead of guanidine nucleotide. These triple primers were used to amplify the speci c fragments from bisulfate-converted DNA isolated from cancer patients with a vary level of vitamin B12 and ferritin. Interestingly, the agarose gel electrophoresis of PCR products showed only one speci c band with a molecular size of about 175bp when used the triple speci c primers for TET1 and genomic DNA isolated from cancer patients with the normal level of vitamin B12/ferritin (Fig. 4A). Whereas, two speci c bands with a molecular size of 175bp and 165bp were detected using the genomic DNA isolated from cancer patient with a low level of vitamin B12 and ferritin (Fig. 4B). Likewise, the speci c triple primers for TET3 ampli ed only the wild-type fragment when using the genomic DNA isolated from cancer patients with a low level of vitamin B12/ferritin. Meanwhile, the same primers ampli ed both of the fragments with the molecular size of 175bp and 165bp using the obtained DNA from cancer patients with a low level of vitamin B12 and ferritin ( Fig. 4C and D). These ndings reveal the methylation potential in the promoter region of TET1 and TET3 associated with de ciency of vitamin B12 and ferritin in cancer patients like HCC, BC, LC, and CC.

Discussion
In the current work, we sought to investigate the DNA methylation process and the expression pro le of its related cofactors in cancer patients to address the correlation between epigenetic changes and the levels of vitamin B12 and ferritin in patients' blood. Interestingly, the biochemical analysis of the patients and their collected samples demonstrated that more than 75% of the cancer patients showed low levels of vitamin B12 and ferritin, while more than 90% of the patients showed normal levels of serum folate. Therefore, the blood samples obtained from patients with HCC, BC, LC, or CC were divided into two main groups according to the levels of vitamin B12 and ferritin, patients with normal levels and patients with de cient levels. Our ndings showed that the expression pro le of DNMT1, DNMT3a, and DNMT3b was increased in patients with low levels of vitamin B12 and ferritin, while the expression of MS, TET1, and TET3 was signi cantly decreased. The methylation analysis of puri ed DNA using the HapII restriction enzyme and sequences analysis of cleaved fragments showed a potential methylated-cytosine at the location 318/CG and 385/CG in the promoter region of TET1 and TET3, respectively. These methylation changes were also achieved in patients with a low level of vitamin B12 and ferritin. Moreover, the bisul te PCR using triple speci c primers further con rmed the methylation changes in the promoter region of TET1 and TET3 at same indicated sites. Accordingly, the current data demonstrated that the low levels of vitamin B12 and ferritin in cancer patients plays a crucial role in hypermethylation and gene silencing of TET1 and TET3 and subsequently the balance between hyper and hypo-methylation processes in cancer cells.
The methylation activity of DNA is maintained by DNMTs that contain three major types, DNMT1, DNMT2, and DNMT3. DNMT2 mainly regulates the methylation activity of transfer RNA molecules (tRNA), although very recent evidence indicating the ability of DNMT2 to make methylation changes in the genomic DNA of aged macrophages [14,32]. DNMT3 includes two different subtypes, DNMT3a and DNMT3b, which can make methylation changes in the genomic DNA [33]. Several studies demonstrated the variation of mRNA levels of DNMT1 and DNMT3 in association with human diseases such as atopic dermatitis and systemic lupus erythematosus in which their mRNA levels were decreased [34,35]. While the overexpression of DNMT1 and DNMT3 has been connected with cancer development and autoimmune diseases indicated by the hypermethylation of certain genes expression [36]. In this way, our ndings revealed the overexpression of DNMT1, DNMT3a, and DNMT3b in blood samples obtained from cancer patients with low levels of vitamin B12 and ferritin when compared with the blood samples obtained from cancer patients with normal levels of vitamin B12 and ferritin. This conclusion supports the hypothesis suggesting the role of vitamin B12 and ferritin in hypermethylation activity of genomic DNA during cancer development such as HCC, BC, LC, and CC. The current data also indicated the depletion of MS expression in patients with a de cient level of vitamin B12/ferritin. Noteworthy, MS is a crucial cofactor of DNMTs which responsible for the metabolism of SAMe, the donor of methyl groups required for DNA methylation, nucleic acids metabolism, and membrane structure and function [37].
Consequently, we addressed how DNA hypermethylation of certain genes can occur during cancer development; however, the methyl donor is interrupted due to de cient levels of vitamin B12 and ferritin. We hypothesize that the downregulation of TET1 and TET3 gene expression is indirectly responsible for the DNA hypermethylation of cancerous cells since TET1 and TET3 are oxidize 5-methylcytosines (5mCs) and promote locus-speci c reversal of DNA methylation [38]. Most likely, the downregulation of TET1 and TET3 is accumulated due to epigenetic exchanges in their promoter regions, as suggested by the current data, by which the expression pro le and the molecular function of these effectors are prevented in cancer patients. Numerous studies reported that TET proteins contribute to preventing the onset and transformation of malignancies; however, the exact mechanism is still unclear [38,39]. Other evidences indicated that TET genes are often mutated in several cancers including myeloid malignancies and solid tumors such as colorectal cancer, breast cancer, lung and liver cancer [40][41][42]. In normal cells, the hemimethylated DNA strands are normally converted to symmetrical methylation condition by the contribution of DNMT1 and ubiquitin-like, containing PHD and RING nger domains 1 (UHRF1) protein complex which recognizes hemimethylated CpGs and remains DNA methylation progress [43].
Recently, the TET proteins family were identi ed as dioxygenases that can converts 5-methylcytosine in DNA to 5-hydroxymethylcytosine, which plays a critical role in restoring and decreasing the symmetrical methylation progress [44]. Thereby, we provide evidence for the mechanism by which TET genes are mutated in cancer patients with de cient levels of vitamin B12 and ferritin via methylation of CpG islands in the promoter region of TET1 and TET3 genes. Interestingly, we developed an accurate method to e ciently determine single-base pair DNA methylation patterns on TET1 and TET 3 promoter regions via exploiting triple speci c primers for each region that contain a forward primer and two reverse primers recognized the methylated CpG island with either wild-type sequences or complementary sequences for the methylated islands. Based on the conversion of 5 − methyl cytosine at CpG islands to uracil in bisul te-converted DNA, the triple primers were designed to recognize a certain fragment with CpG frequencies. In this method, the forward primer recognizes the upper sequences of the targeted fragment, while the wild-type reverse primer recognizes the end of targeted fragment that contains unmethylated CpG islands. The methylated reverse primer recognized the methylated CpG islands and ampli es a shorter fragment in comparison with the ampli ed one using the wild-type reverse primer. Accordingly, the agarose gel electrophoresis of bisul te-PCR products using the indicated tripe primers can be used to distinguish between the two ampli ed fragments and con rm the potential methylation activity in the targeted region.

Declarations Ethics statement
The current study has been approved by the medical ethics committee and scienti c ethics committee of the Faculty of Medicine, Mansora University and Genetic Engineering and Biotechnology Research Institute, University of Sadat City, respectively.
Author's contributions HK, SF, and SA established the study design, provided the research strategy and supervised overall the research plan. HK and KE performed the experiments, interpreted the data and provided the scienti c and statistical analysis. HK prepared and wrote the manuscript. All authors read and approved the manuscript.