3. Discussion
The present study demonstrated that MTHFR knockdown in liver-origin cells may assist the defense against folate depletion-induced chromosome segregation and uracil misincorporation in the DNA (
Figure 4).
To better understand the interactions between MTHFR function and chromosome stability under different folate supplies, we successfully established in vitro cell models of liver origin with stabilized inhibition of MTHFR using small hairpin RNA. The cell model was then characterized with respect to cell cycle, SAM homeostasis, folate-dependent de novo nucleotide biosynthesis, and DNA stability under folate depletion.
In our cell model, reduced MTHFR function is associated with decreased cell populations in the G1 phase and increased cell populations in the G2/M phase. The proportion of cell population in the G2/M phase was highly inversely correlated with MTHFR gene level, protein abundance, and enzyme activity, suggesting that MTHFR function is closely related to the cellular events during G2/M.
The impacts of MTHFR function have been reported to be closely related to folate status. We previously demonstrated that human lymphoblasts with MTHFR TT genotype had significantly reduced folate-dependent remethylation and SAM contents when folate supply was restricted. In folate restriction, the reduction of SAM supply in TT genotype was threefold (decreased by 27%) of that in the CC genotype (decreased by 9%) [
26]. We also demonstrated an increase in purine synthesis in the TT lymphoblasts compared to that of the CC lymphoblasts when folate was abundant, presumably due to the increased formylated folate pool [
26]. We suggested that the advantage of de novo purine synthesis found in the MTHFR TT genotype may account for its protective effect in hematological malignancies. Methyl donor and/or folate deficiency is also associated with genomic damage and cell death in human lymphocytes in vitro [
10]. Folate deficiency may deplete cellular SAM supply and perturb methylation reactions of DNA, RNA, and histones, leading to altered gene expression and genome stability. By catalyzing the irreversible conversion from 5,10-methyleneTHF to 5-methylTHF, MTHFR may regulate the competition between folate-dependent dTMP and SAM biosynthesis. However, genetic predisposition, including variations in MTHFR function, could impact folate mediated transmethylation and nucleotide synthesis in a tissue-specific manner [
27]. MTHFR C677T mutation induces cell-specific changes in genomic DNA methylation and uracil misincorporation that could in part account for the molecular basis for the site-specific risk modification in tumors from different tissue origins [
27]. Unlike the lymphoblast models with TT genotype that were more sensitive to folate depletion in SAM synthesis, shMTHFR did not intensify folate depletion induced G2/M cell cycle arrest or SAM reduction in our HepG2 cell model. These findings suggest that, unlike extrahepatic cells that have enhanced nucleotide biosynthesis at the cost of reductions in the methyl group supply during folate depletion, lower MTHFR in cells of liver origin may take advantage of the nucleotide biosynthesis without severe impacts on SAM supply, possibly due to other methyl sources such as choline (via the betaine homocysteine
S-methyltransferase pathway).
shMTHFR appeared to facilitate HepG2 cells entering the S phase and prolong G2/M in the cell cycle. Knockdown of MTHFR has been shown to decrease gastric cancer cell survival and result in cell cycle arrest at the G2 phase [
55]. The shMTHFR induced cell cycle arrest raised our interest as cells in G2/M undergo multiple processes involved in DNA damage and repair. DNA instability (strand breakage, uracil misincorporation, and defective repair) is increased by folic acid depletion in human lymphocytes in vitro [
9].
The nuclear localization of SHMT/TS/DHFR multienzyme complex has been determined as a function of cell cycle that is required for of de novo thymidylate biosynthesis [
43]. The nuclear translocation of SHMT1 is cell cycle-dependent and occurs during the S and G
2/M phases [
56,
57]. DHFR and TYMS also localize to the nucleus during S and G
2/M phases but not in the G
1 phase. In the present study, folate depletion resulted in a significant reduction in SHMT/DHFR/TYMS, and shMTHFR-promoted nuclei SHMT1/DHFR/TYMS protein expression and translocation, in particular, under folate deficiency.
The reductive methylation of uridylate monophosphate (dUMP) to thymidylate monophosphate (dTMP) involves the transfer and, simultaneously, the reduction of the one-carbon moiety from 5,10-methyleneTHF. The 5,10-methyleneTHF is produced either by the activity of SHMT or by MTHFD1 [
43]. Metabolic studies in MCF-7 cells demonstrated that SHMT contributes approximately 30% whereas MTHFD1 contributes 70% of one-carbon groups used in the reductive methylation of dUMP catalyzed by TYMS [
23]. The THF is regenerated by the activity of DHFR [
44]. In addition to the increased abundance of SHMT1/DHFR/TYMS proteins, shMTHFR also induced MTHFD1 expression both in the cytosol and nucleus in our HepG2 cell model.
Folate depletion impairs nucleotide biosynthesis that promotes DNA strand breakage, uracil misincorporation, and defective repair in human lymphocytes [
9]. Reduced dTMP synthesis results in uracil incorporation into DNA. The DNA repair machinery can remove the uracil, but in the presence of a high deoxyuridine triphosphate (dUTP) to deoxythymidylate triphosphate (dTTP) ratio, it can be incorporated into DNA again. This futile cycle of uracil incorporation and deprived repair ultimately results in DNA double strand breaks and genome instability [
9]. Unlike all other nucleotide synthesis that occurs in the cytoplasm, the de novo thymidylate biosynthesis pathway localizes to the nucleus and it is required to prevent uracil accumulation in DNA [
50]. The nuclear compartmentation of de novo thymidylate biosynthesis at the replication fork enables the regulation of dUTP incorporation into DNA, as opposed to its misincorporation into DNA, for the regulation of transcription [
43]. The present study provided direct evidence that by promoting the formation of the nuclear thymidylate biosynthesis multienzyme complex, shMTHFR effectively suppresses folate deficiency induced uracil accumulation thus helps protect DNA.
The rate of DNA synthesis is dependent on de novo purine synthesis. During the G1/S phase, the rates of purine synthesis via the
de novo and the salvage pathway increased 5-fold and 3.3-fold, respectively, in human colon carcinoma cell HCT116 [
58], indicating that when cells progress from mid-G1-phase to early S-phase, they must substantially increase the synthesis of purine nucleotides/deoxynucleotides. Defected purine synthesis due to folate depletion leads to chromosome instability, and reduced purine synthesis causes cytostasis and cytotoxicity, as well as aberrant DNA synthesis, repair, and mutagenesis [
59,
60]. Our study suggests that shMTHFR may assist cells passing the G1 and S phase and entering the G2/M stage more rapidly by shuffling more nucleotides. This may partially account for the observation that shMTHFR had more cell populations in the G2/M phase. Furthermore, the incorporations of serine derived one-carbon moiety into the purine ring were significantly increased by shMTHFR during folate depletion and such impacts were sustained after folinate supplementation followed by the initial folate depletion. These findings suggest that by preserving more one-carbon moieties in nucleotide biosynthesis, shMTHFR can help maintain cell cycle progression during starvation of metabolic intermediates, such as purines and pyrimidines, in this cell model. These data demonstrated an important role of MTHFR in the regulation of cell cycle progression, as well as the formation and translocation of the nuclear multi-enzyme complex during S and G2/M phases. Our study also suggested that MTHFR can assist nuclear de novo thymidylate synthesis during DNA replication and repair in folate deficiency.
In the present study, we discovered that shMTHFR promoted nuclear p53 (and MLH1) expression, especially under folate deficiency. p53 is the most commonly mutated gene in human cancer that acts as a major cell cycle checkpoint regulator. p53 is involved in various DNA-repair systems and takes on multiple mechanisms to prevent cancer development by maintaining genome stability [
61]. p53 has been identified as a component of a spindle checkpoint that ensures the maintenance of diploidy [
62].
Certain ribonucleotide biosynthesis inhibitors caused a p53-dependent G0 or early G1 arrest, and p53 was proposed to be a metabolite sensor activated by depletion of ribonucleotides or their related processes [
63]. p53 was therefore suggested to play a role in inducing a quiescence-like arrest state in response to nutrient challenge and a senescence-like arrest state in response to DNA damage [
63]. In normal human fibroblasts. CTP, GTP, or UTP depletion alone was sufficient to induce cell cycle arrest [
63]. In the present study, when cells underwent folate depletion, the nucleotide shortage was less severe in cells with reduced MTHFR activity, and we postulate that the elevated p53 can help ameliorate the consequences of nucleotide depletion.
p53 was initially shown to act at the G1 checkpoint but was later also found to be important in regulating the spindle checkpoint during G2/M [
62]. The spindle assembly checkpoint is a cell-cycle regulatory pathway preventing chromosomal instability. Prolonged folate depletion in human NCM460 colon mucosal cells substantially compromises the spindle assembly checkpoint network, which predisposes cells to mitotic aberrations and chromosomal instability [
64]. Our results may suggest that shMTHFR can facilitate cell cycle progression and induce p53 expression that may effectively ameliorate folate depletion-induced mitotic aberrations and chromosomal instability, including impaired spindle assembly.
Folate adequacy protects against mutagenesis at the phosphatidylinositol glycan class A gene (Pig-a) locus and micronuclei induction in the red blood cells of mice [
1]. Mice fed a folate-deficient diet had 1.8-fold higher micronuclei (MNi) in reticulocytes, and 1.5-fold higher micronuclei in normochromic erythrocytes than mice fed the FA supplemented diet [
1]. Since micronuclei formation is generally attributed to error in DNA synthesis during the S and/or mitosis during G2/M phase in the cell-cycle [
49], it is plausible that the prolonged G2/M resulted in shMTHFR cells may assist nucleotide supply and reduce DNA damage by repairing replication errors during folate depletion. shMTHFR significantly decreased folate depletion-induced micronuclei, supporting our hypothesis.
Cells proficient in mismatch repair were highly sensitive to folate deficiency compared with cells defective in mismatch repair proteins [
54]. In vitro biochemical studies demonstrated a direct participation of mismatch repair proteins in mediating the apoptotic response induced by folate deficiency [
54]. p53 plays diverse roles to directly impact the activity of DNA-repair systems and protects cells from cancer development by maintaining genome stability [
61]. p53 was found to be required for the folate depletion-induced apoptosis process.
Loss of the DNA repair gene human Mut L homologue-1 (
hMLH1) was found to cause lymphoma, and the enforced expression of
MLH1 was found to delay tumor development driven by loss of p53 [
52]. These studies indicated that MLH1 mediated DNA repair processes are critical mediators of p53-dependent tumor suppression [
52]. The impact of shMTHFR on hMLH1 in HCC has not been studied before.
hMLH1 is one of the key proteins involved in the mismatch repair process after DNA replication. Defected hMLH1 and hMSH2 commonly occur in moderately and poorly differentiated HCC [
53]. Although it was suggested that defective DNA mismatch repair does not contribute greatly to hepatocellular carcinogenesis [
65], combined loss of expression of O6-methylguanine-DNA methyltransferase and hMLH1have been reported to accelerate the progression of HCC [
66]. Since we discovered that shMTHFR induced cell cycle arrest in the G2/M phase and promoted p53 expression, it is plausible that shMTHFR may protect cells from folate depletion-induced DNA damage via DNA repair process involved MLH1.
We found that shMTHFR not only promoted p53 but also induced hMLH1 expression in both folate depletion and repletion, consisting with the finding of reduced micronucleated binucleated cells and uracil misincorporation (
Table 6C). In addition, folate restriction decreased nuclear and cytosol hMLH1 and p53 protein abundance, and shMTHFR recovered the reduction of hMLH1, p53 (
Table 6C). These observations support our postulation that shMTHFR protects cells from folate depletion-induced DNA damage via p53 mediated MLH1 DNA repair.
DNA mismatch repair proteins MLH1 and PMS2 have been identified as p53 targets that may serve as a sensor in DNA repair mechanisms and a critical determinant for the decision between cell-cycle arrest and apoptosis [
67]. Both of the
hMLH1 and
hPMS2 genes were determined to be responsive to DNA damage and p53 activation in normal human fibroblasts, and have p53-response elements within their first intron [
67]. Deficiencies of DNA mismatch repair-complex proteins, including hMLH1 and hPMS2, typically result in microsatellite instability in human cancers [
68]. A normal hMLH1 protein level is important in maintaining normal levels of hPMS1 and hPMS2 proteins. Gastric and colorectal cancer cells lines with microsatellite instability lacked detectable hMLH1, and the decreased hMLH1 has been associated with markedly reduced hPMS2 and hPMS1 proteins [
68]. Yeast two-hybrid assay has been used to identify the amino acid residues in hPMS2 that interact strongly with hMLH1 [
69].
On the other hand, it was reported that the concomitant loss of Pms2 function chemosensitises p53-deficient cells to some types of anticancer agents; Pms2 positively modulates cell survival by mechanisms independent of p53, and that increased cytotoxicity is paralleled by increased apoptosis. Tumor-targeted functional inhibition of Pms2 may be a valuable strategy for increasing the efficacy of anticancer agents in the treatment of p53-mutant cancers [
70].
In the present study, we discovered that shMTHFR induced MLH1 in folate depletion, suggesting that reduced MTHFR function could be involved in MLH1 mediated DNA repair. In gastric cancer cell MKN45, knockdown of MTHFR in gastric cancer cell MKN45 decreased cell survival and resulted in cell cycle arrest at the G2 phase. Overexpression of MTHFR in human gastric cancer cell MKN45 was found to downregulate hMLH1 [
71]. These cells also had lower levels of c-myc expression, and overexpression of MTHFR increased cell proliferation and induced the downregulation of p21WAF1 and hMLH1. These data implied that c-myc and p21WAF1 could also be involved in the better DNA stability in shMTHFR of human HCC, but further studies are certainly needed. How MTHFR function regulates the interactions among p53, MLH1, and PSM2 in the DNA repair system in human HCC will be investigated in the future.
In summary, MTHFR knockdown assists liver-origin cell defense against folate depletion-induced chromosome segregation and uracil misincorporation in the DNA by prolonging the G2/M cell cycle and promoting nuclear thymidine synthesis multi-enzyme complex formation and nuclear DNA repair proteins’ MLH1/p53 expression.
4. Materials and Methods
4.1. mRNA Expression of MTHFR and Survival Rate in HCC Patients
The expression analysis of MTHFR mRNA in HCC tumors and normal tissues was examined by the GEPIA web tool (
http://gepia2.cancer-pku.cn/#index, accessed on 14 May 2021). The median of MTHFR mRNA expression level in HCC was compared to the TCGA normal data as well as to the TCGA normal and GTEx data.
When comparing gene expression between tumor and normal tissues using different datasets from public resources, one cannot rule out the possibility of systemic differences between different datasets, therefore we further compared the MTHFR gene expression in a RNA-Seq dataset containing 42 paired tumor and tumor-adjacent normal HCC tissues generated from the Cancer RNA-Seq Nexus on 14 May 2021 (CRN,
http://syslab4.nchu.edu.tw/CRN) [
36,
37]. The mean MTHFR gene expression level (fragments per kilobase per million, FPKM) of tumor tissues were compared between the HCC tumor and the adjacent normal tissues by Student’s
t-test.
The survival analysis was performed using the Pan-cancer RNA-Seq Web server (
http://kmplot.com/analysis/index.php?p=service&cancer=liver_rnaseq) on 14 May 2021) for generating Kaplan–Meier plots by auto-selecting the best cutoff values between lower and upper quartiles into high and low expression groups that included all stages, sex, race, and mutation burden [
38]. The Kaplan–Meier survival analysis was performed on the HCC RNA-seq data of the TCGA/GTEx datasets available within GEPIA2, by auto-selecting the best cutoff values into high and low expression.
4.2. Establishment of MTHFR Knockdown Cell-Line by RNA Interference
The MTHFR knockdown HepG2 cell lines were established using the lentiviral small hairpin RNA (shRNA) interference vectors (National RNAi Core Facility, Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan). The shRNA constructs were based on the pLKO.1-puro vector and these lentivirus-based shRNA constructs. The target sequences of various shRNAs are listed in
Table 1A and the target sites are provided in
Supplementary Figure S1.
The different shRNA lentiviruses were produced from HEK293T packaging cells that were transfected (Lipofectin from Invitrogen, Carlsbad, CA, USA) separately with either MTHFR shRNA (sh3′UTR, sh77, sh546, sh697, sh1618) or empty EGFP lentiviral plasmids, prepared by the PureYield Plasmid kit (Promega, Madison, WI, USA). To do so, virus-producer HEK 293T cells were seeded (1.5–2.0 × 10
5 cells per well in a 24-well plate) 1 day before transfection. Cells were transfected by using 1.5 μL of Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) with the plasmid DNAs. The transfected GFP-expressing cells were examined under a fluorescence microscope (
Supplemental Figure S2A–C), and the GFP fluorescence intensity was measured by flow cytometry (Cytomics FC 500; Beckman Coulter, Pasadena, CA, USA). The number of GFP-positive cells (IU: infectious unit) was divided by the volume of viral solution (mL) to calculate virus titer. The flow cytometry results of the empty lentiviral EGFP vector showed an estimation of 99% transfection efficiency (
Supplemental Figure S2D,E), proving that a successful HEK 293T packaging line was established using this lentiviral system.
After the transfection, the supernatants of different transfected cells were collected 48 h post-transfection and centrifuged at 800× g for 15 min at 4 °C to remove cell debris. Ten-fold dilution of virus solutions were used to infect HepG2 cells (Bioresource Collection and Research Center, Hsinchu, Taiwan). Forty-eight hours after the virus infection, the transduced HepG2 cells underwent puromycin selection (20 μg/mL puromycin from Sigma (Ronkonkoma, NY, USA) for at least 2 weeks to generate individual stable clones. The surviving cells were cultivated and maintained for further applications.
4.3. Cell Culture Conditions
All chemicals were purchased from Sigma-Aldrich Chemical Company (St. Louis, MO, USA) unless otherwise specified. Human hepatoma cell line HepG2 was grown in α-Modified Eagle’s Medium or RPMI Medium with 10% bovine serum, Pen-Strep-Ampho solution [100,000 units/L Penicillin, 100 mg/L Streptomycin, 0.25 mg/mL Amphotericin]. Human embryonic kidney cell line 293T (courtesy of Dr. Shih-Lan Hsu from Taichung Veterans General Hospital) was maintained in Dulbecco’s Modified Eagle’s Medium supplemented with fetal bovine serum (FBS) (TerraCell International, ON, Canada), Pen-Strep-Ampho solution [100,000 units/L Penicillin, 100 mg/L Streptomycin, 0.25 mg/mL Amphotericin]. Cells were incubated at 37 °C in a humidified atmosphere containing 5% CO2. The media were replaced every 72 h.
The interactions between shMTHFR and folate insufficiency were investigated as follows. In the cell cycle and SAM analysis, “no folate cells” were cultured in folate depletion medium (no folate RPMI-1640, Gibco Invitrogen, Carlsbad, CA, USA) for 12 days; “folate replete” cells were cultured in folate sufficient RPMI-1640 medium (with 2.2 μM colic acid) for 12 days. In the low-folate experiments for SAM analysis, cells were cultured in folate depletion medium with 36.7 μM hypoxanthine and 37.1 μM thymidine for 9 days (to accelerate the depletion of intracellular folate content that was divided in half during cell division), then treated in medium with 2 nM folinate for 3 days (to generate a steady-state of low intracellular folate) with the same hypoxanthine and thymidine levels. In the folate depletion experiments for cytosol and nucleus protein expression, “folate deplete cells” were cultured in RPMI-1640 without folate medium for 11 days; “folate replete” cells were cultured in folate-sufficient RPMI-1640 for 11 days.
In the folate-replete/deplete stable isotope labeling experiments, control cells were cultured in folate-replete RPMI-1640 medium for 11 days. Folate-depleted cells (-FA) were cultured in RPMI-1640 without folate medium for 11 days. Folinate has been shown to be effective in rescuing certain impaired 1C metabolic pathways induced by methotrexate [
45] or folate depletion [
47]. The folate-replete and -deplete cells (-FA + folinate group) were cultured in no-folate medium for 8 days and then supplemented with 10 nM folinate for 3 days. These levels were based on our previous experiments and were used to generate mildly low intracellular folate [
26,
48].
In the folate deficiency-induced micronuclei and uracil misincorporation experiments, “folate repleted” cells were cultured in folate-sufficient medium for 12 days; “low folate” cells were cultured in folate depletion medium for 9 days and then in 2 nM folinate for 3 days; “no folate” cells were cultured in no-folate RPMI medium for 12 days, which was modified based on our previous experiment [
8].
4.4. MTHFR Gene Expression in HepG2 Cells with shMTHFR
Five different homologous human MTHFR shRNA target sequences were used to generate different lentivirus clones termed (1) sh3′UTR, (2) sh77, (3) sh546, (4) sh697, (5) sh1618 that represented the target site on the MTHFR cDNA sequence. The shGFP clone was used as the negative control cell-line (Neg) as it underwent the same lentiviral transfection procedure but it did not target a specific human gene sequence. The relative efficiency of different target shRNA sequences on the reduction of MTHFR expression was determined by real-time PCR.
To determine the efficiency of each shRNA on MTHFR mRNA expression, total RNA form was extracted by TRIZOL reagent (Invitrogen, Carlsbad, CA, USA) and integrity was checked by electrophoresis [
40]. RNA was then reversely transcribed with random primers following the manufacturer’s protocol and cDNA from each cell lines were used as templates for quantitative PCR using SYBR green gene expression assay with predesigned primers for human MTHFR. Real-time quantitative PCR was performed using Prism 7000 (Applied Biosystem Inc., Foster City, CA). The MTHFR mRNA expression was calculated by normalizing the threshold cycle value of the target gene to that of the control housekeeping gene (18sRNA). The relative MTHFR expression was given by the formula: 2
−ΔΔCT, where ΔΔCT = ΔC
TRNAi clones—ΔC
T negative control; with ΔCT = ΔCT
MTHFR—ΔCT
18S. Relative efficiency of different target shRNA sequences on the reduction of MTHFR compared to wildtype (WT, as 100%) are shown in
Table 1. Clone sh77 that had the most significant reduction (by ~63%) in MTHFR mRNA expression was used to further study the efficiency on MTHFR protein expression, MTHFR enzyme activity, and its impacts in combination with folate depletion.
4.5. Effects of siRNA on MTHFR Protein Expression
Effects of siRNA on MTHFR protein expression were examined by Western blot. Cell lysates were prepared [
72]. Protein lysates were resolved on 10% SDS polyacrylamide gel, electro-transferred to polyvinylidene fluoride membranes, and blocked in 5% non-fat dry milk in Tris-buffered saline, pH 7.5 (100 mM NaCl, 50 mM Tris, and 0.1% Tween-20). Membranes were immunoblotted overnight at 4 °C with anti-MTHFR polyclonal antibody (Santa Cruz, CA. USA), and anti-β-actin antibody (Sigma-Aldrich, St. Louis, MO, USA). Washed blots were incubated with Immobilon Western Chemiluminescent HRP Substrate (Millipore) according to the manufacturer’s instructions, followed by the detection with a Koda fluorescence scanner. Values are expressed as mean ± SD (
n = 3). The data were analyzed by one-way ANOVA. Different letters indicate statistical differences (
p < 0.05) among different groups.
4.6. Effects of siRNA on MTHFR Enzyme Activity
Specific MTHFR activity was compared between WT, Neg, and shMTHFR cells using the
14C-labeled methyltetrahydrofolate (
14C-CH
3THF)–menadione oxidoreductase assay [
73] with modifications. Cell extracts were incubated for 60 min at 37 °C in a reaction mixture containing 0.18 mol/L phosphate buffer, 3.5 mmol/L menadione, 1.4 mmol/L EDTA, 7.6 mmol/L ascorbic acid, 70 µmol/L FAD, and 300 µmol/L [
14C]CH
3THF in a total volume of 143 µL. The reaction was terminated by the addition of 125 µL 0.6 mol/L sodium acetate, pH 4.5. After the addition of 50 µL 100 mmol/L formaldehyde and 75 µL 0.4 mol/L dimedone, the mixture was boiled for 12 min and subsequently cooled on ice. For each sample, 2.5 mL toluene was added and the tubes were vigorously vortexed twice for 15 s. After centrifugation, the formation of the radio-labeled [
14C] formaldehyde-dimedone adduction was quantified by scintillation counting of the supernatant. Enzyme activity was expressed as nanomoles of formaldehyde formed per hour per milligram of protein.
4.7. Cell Cycle Analyses
A total of 5 × 10
6 cells were WT, Neg, and shMTHFR HepG2 cells harvested, washed, and resuspended in cold PBS with ice-cold ethanol, and then washed with Ca2+/Mg2+ free HBSS containing 1% BSA [
74]. The cells were then incubated in 50 mg/mL propidium iodide (PI; Sigma, St. Louis, MO, USA) containing 1 mg/mL sodium citrate, 100 mg/mL RNase I and 0.1% Triton X-100 for 30 min at 37 °C, and analyzed by flow cytometry in fluorescence-activated cell sorter (Epics XL.MCL, Beckman Coulter, Inc, Fullerton, CA, USA) using the EXPO32 software.
4.8. Determination of SAM and SAH Contents
The impacts of folate restriction on SAM were examined separately in each cell line and compared to the same cell line under folate-replete conditions. After the incubation period, cells were harvested, washed, and pelleted by centrifugation. Cell extracts for SAM and SAH analysis were prepared by a modification of the previously described procedure. Cells were centrifuged and washed with cold PBS twice while being kept on ice. PBS was carefully aspirated and cell pellets were resuspended in 0.4 M ice-cold perchloric acid. Pellets were hand-homogenized on ice with a hand-held mini pestle [
26]. Homogenates were centrifuged at 4 °C and supernatants were collected and stored at −80 °C for analysis by HPLC as described by Fell [
75]. The supernatant of each sample was filtered through 0.45 μ M and then loaded onto a C18 column (250 × 4.6 mm), fitted with a matched guard column operated by a Hitachi L-7100 intelligent pump connected to an L-7400 UV detector. The absorption of eluted compounds was monitored using ex = 254 nm [
40].
SAM and SAH values were normalized to total cell numbers [
5]. In folate depletion experiments, cells were grown without folate-developed macrocytosis, so the calculation was normalized to protein contents instead of cell number [
8]. Cellular protein contents were determined by the bicinchoninic assay (Pierce, Rockford, IL, USA).
4.9. Western Blot Analyses
After cells were harvested, nuclear and cytosol proteins were isolated and fractionated using the nuclear/cytosol fractionation Kit (Nuclear/Cytosol Fractionation Kit, Catalog #K266, BioVision, Milpitas, CA, USA, San Francisco) following the manufacturer’s instructions [
76]. Protein was quantified using a BCA assay kit (Pierce, Rockford, IL, USA, Winnebago). For Western blots, proteins were separated by SDS-PAGE gel, transferred to polyvinylidene difluoride membrane, and blotted with a designated antibody according to the manufacturer’s instructions. Equal loading and/or purity of nuclear fractions was confirmed through the detection of Lamin A using α-Lamin A (GeneTex, Irvine, CA, USA, Orange, 1∶1000 dilution) [
77], and tubulin using α-tubulin (Abcam, Cambridge, non-metropolitan county, UK, Cambridgeshire, 1∶5000 dilution). No cross-contamination between the nuclear and cytoplasmic fractions was observed. Thirty micrograms of proteins (from folate depletion experiments) or 16 µg of proteins (from folate repletion experiments) were separated by SDS-PAGE (12% gel) and subsequently transferred to a polyvinylidene difluoride membrane and blotted with primary antibodies, including anti-Thymidylate synthase (1:1000; sc-376161; Santa Cruz, Dallas, TX, USA), anti-DHFR (1:1000; ab133546; Abcam, Cambridge, non-metropolitan county, UK, Cambridgeshire), anti-a-Tubulin (1:5000; NB100-690; Novus, Littleton, CO, USA), anti-MTHFD1 (1:500; sc-134732; Santa Cruz, Dallas, TX, USA), anti-MLH1 (1:1000; sc-271978; Santa Cruz, Dallas, TX, USA), anti-SHMT1 (1:1000; #80715; Cell Signaling, Danvers, MA, USA, Essex), anti-Actin (1:5000; NB600-501; Novus, Littleton, CO, USA), anti–Lamin A/C (1:1000; GTX101127; GeneTex, Irvine, CA, USA, Orange), anti-p53 (1:1000; #9282; Cell Signaling, Danvers, MA, USA, Essex), followed by incubation with HRP-conjugated polyclonal secondary antibody (1:2000; ab6721; Abcam, Cambridge, non-metropolitan county, UK). All Western blots were visualized using the enhanced plus chemiluminescence assay kit (Adv, San Jose, CA, USA, San Francisco), according to the manufacturer’s protocol. Protein expression levels were quantified by ImageJ software (Analytik Jena US LLC, Upland, CA, USA, San Bernardino).
4.10. Stable Isotope Labeling Experiments
To investigate the impacts of shMTHFR on purine and thymidylate synthesis under different folates conditions, kinetic experiments were conducted based on our previous studies [
5,
26,
45] with modifications for the present study. Stable isotopic tracer [3-
13C]-serine (Cambridge Isotope Laboratories, Woburn, MA, USA) was used to trace the 1-carbon flow in the subsequent experiments as the β-carbon of serine provides the major 1-carbon source in folate metabolic pathways.
After the treatment period, cells were harvested and total genomic DNA was extracted using a standard phenol/chloroform/isoamyl alcohol procedure with RNase treatment to remove RNA. DNA was dried under nitrogen and hydrolyzed in formic acid under vacuum. The dried DNA samples were then converted into trimethylsilane-base derivatives which were then separated on an HP-5MS column. Isotopic enrichments in the nucleotides were determined in positive ionization mode for nucleotides and negative ionization mode for amino acids by GC/MS using a model 6890 gas chromatography and model 5973 mass spectrometer (Agilent, Palo Alto, CA, USA). Selected ion monitoring was conducted at a m/z 255–257 for thymine, m/z 280–283 for adenine, and m/z 368–371 for guanine.
4.11. Determination of Micronuclei
After the designated treatment period, cells were fixed with 70% ethanol and then stained using EtBr and visualized under a digital fluorescence microscope for the identification of MNi. Nuclei were considered as MNi according to established criteria including the following: (1) diameter of the MN less than one third of the main nucleus; (2) MNi was separated from the main nuclei with distinguishable nuclear boundary, and (3) MNi exhibited similar staining intensity as the main nucleus [
1].
MN frequency was obtained by counting the number of MN-positive cells among 500 total cells in each treatment group, and finally, the percentage of MN was calculated from the data obtained.
4.12. Determination of Uracil Content in the DNA
Genomic DNA was extracted using a standard phenol/chloroform/isoamyl alcohol procedure with RNase treatment to remove RNA. DNA samples were then incubated with uracil DNA glycosylase (New England Biolab, Ipswich, MA, USA). An internal standard of [15N2]uracil (kindly provided by Professor Stover at Texas A&M University) was added before the uracil was extracted and derivatized. Analysis of uracil-3,5-bis(trifluoromethyl)benzyl bromide was carried out by 6890-GC coupled with 5975C-MS17, 39 (Agilent Technology, Palo Alto, CA, USA). Separation of derived uracil was achieved on an HP-5MS (30 m, 0.25 mm) column. Ionization was achieved using the NCI mode and monitoring at
m/z 337 for uracil and 339 for [15N2]uracil. The amount of uracil in DNA was presented as pg uracil/µg DNA [
8]
4.13. Statistical Analysis
For cell culture data analyses, the differences amongst cells with different genotypes were examined by one-way ANOVA, then the comparisons of means between each of the two groups were determined using post hoc analyses. All statistical analyses were performed with SYSTAT 11.0 for WindowsTM (Systat Software Inc., Richmond, CA, USA). For all analyses, the results were considered statistically significant if p-values were <0.05.