http://orcid.org/0000-0003-1840-4305
Figures
Abstract
Background
Pancreatic cancer is the third leading cause of cancer related deaths in the United States. Several dietary factors have been identified that modify pancreatic cancer risk, including low folate levels. In addition to nutrition and lifestyle determinants, folate status may be influenced by genetic factors such as single nucleotide polymorphisms (SNPs). In the present study, we investigated the association between folate levels, genetic polymorphisms in genes of the folate pathway, and pancreatic cancer.
Methods
Serum and red blood cell (RBC) folate levels were measured in pancreatic cancer and control subjects. Genotypes were determined utilizing Taqman probes and SNP frequencies between cases and controls were assessed using Fisher’s exact test. Logistic regression was used to estimate the odds ratio (OR) and corresponding 95% confidence intervals (CIs) to measure the association between genotypes and pancreatic cancer risk. The association between folate levels and SNP expression was calculated using one-way ANOVA.
Results
Mean RBC folate levels were significantly lower in pancreatic cancer cases compared to unrelated controls (508.4 ± 215.9 ng/mL vs 588.3 ± 229.2 ng/mL, respectively) whereas serum folate levels were similar. Irrespective of cancer status, several SNPs were found to be associated with altered serum folate concentrations, including the D919G SNP in methionine synthase (MTR), the L474F SNP in serine hydroxymethyl transferase 1 (SHMT1) and the V175M SNP in phosphatidyl ethanolamine methyltransferase (PEMT). Further, the V allele of the A222V SNP and the E allele of the E429A SNP in methylene tetrahydrofolate reductase (MTHFR) were associated with low RBC folate levels. Pancreatic cancer risk was found to be significantly lower for the LL allele of the L78R SNP in choline dehydrogenase (CHDH; OR = 0.29; 95% CI 0.12–0.76); however, it was not associated with altered serum or RBC folate levels.
Citation: Chittiboyina S, Chen Z, Chiorean EG, Kamendulis LM, Hocevar BA (2018) The role of the folate pathway in pancreatic cancer risk. PLoS ONE 13(2): e0193298. https://doi.org/10.1371/journal.pone.0193298
Editor: Surinder K. Batra, University of Nebraska Medical Center, UNITED STATES
Received: August 10, 2017; Accepted: February 8, 2018; Published: February 23, 2018
Copyright: © 2018 Chittiboyina et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Pancreatic cancer, the third leading cause of cancer deaths in the United States, is an aggressive cancer with median 5 year survival rates of only 8% [1]. Detection late in the disease course, rapid metastasis, and chemo-resistance contribute to the poor prognosis for pancreatic cancer [2]. In the age of personalized medicine, identification of genetic and environmental factors that affect the risk for development of pancreatic cancer may aid in prevention or lead to increased surveillance of susceptible individuals. Environmental factors including tobacco and alcohol use, exposure to selected environmental chemicals, obesity, and diet, have been postulated to play a significant role in the etiology of sporadic pancreatic cancer [3]. Deficiencies in dietary sources of methyl groups, including choline, methionine, vitamin B-12 and folate, have been associated with pancreatic dysfunction in rodents [4,5]. In addition, risk of development of various cancer types in humans, including pancreatic cancer, has been shown to increase with low dietary folate intake [6–9].
In humans, folate provides methyl groups for de novo deoxynucleotide synthesis and for intracellular methylation reactions. Methylene tetrahydrofolate reductase (MTHFR) plays a central role in folate metabolism by catalyzing the conversion of 5,10-methylenetetrahydrofolate (5,10-methylene THF) to 5-methyltetrahydrofolate (5-methyl THF), which is the main circulating form of folate in the blood and a cosubstrate for the conversion of homocysteine to methionine (Fig 1). Low levels of 5,10-methylene THF can cause an increased dUMP/dTMP ratio, which could result in the incorporation of uracil into DNA in place of thymine, leading to an increased risk for DNA mutations and DNA strand breakage [10]. In addition, low levels of 5-methyl THF can lead to decreased s-adenosylmethionine (SAM) levels, which could result in DNA hypomethylation leading to activation of cellular oncogenes, genomic instability, and DNA damage [11,12]. Low folate levels therefore would be predicted to modify cancer risk by influencing both pathways.
Depiction of folate uptake and reduction, transmethylation, and transulfuration pathways. Genes in the shaded circles were analyzed for single nucleotide polymorphisms. RFC, reduced folate carrier; PCFT, proton coupled folate transporter; DHF, dihydrofolate; THF, tetrahydrofolate; DHFR, dihydrofolate reductase; SHMT, serine hydroxymethyltransferase; MTHFR, methylene tetrahydrofolate reductase; MTHFD1, methylene tetrahydrofolate dehydrogenase 1; TYMS, thymidylate synthase; MTR, methionine synthase; MTRR, methionine synthase reductase; BHMT, betaine hydroxymethyl transferase; CHDH, choline dehydrogenase; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PEMT, phosphatidylethanolamine N-methyltransferase; SAM, S-adenosyl methionine; MAT, methionine adenosyltransferase; SAH, S-adenosyl homocysteine; AHCY, S-adeonsyl-L-homocysteine hydrolase; GSH, glutathione; CBS, cystathione ß-synthase; CTH, cystathionase; GS, glutathione synthase.
Disruption of the folate metabolic pathway can result from either insufficient dietary intake of folate or the presence of genetic variants, such as single nucleotide polymorphisms (SNPs), in genes of the folate pathway. Several SNPs in genes of the folate pathway have been shown to result in altered serum and red blood cell (RBC) folate levels, including SNPs in the reduced folate carrier (RFC) [13], dihydrofolate reductase (DHFR) [14] and MTHFR genes [15]. An increased risk for development of pancreatic cancer has been associated with expression of the VV allele of the A222V SNP in MTHFR [16–19]. In addition, expression of the YY genotype in the methionine synthase reductase (MTRR) H595Y SNP and the 3Rc/3Rc genotype in thymidylate synthase (TYMS) also increased the risk for development of pancreatic cancer [18,20].
We have previously designed a case-control study to investigate the contribution of environment and genetics to the risk for development of pancreatic cancer, where we recruited pancreatic cancer subjects, an unrelated control who resided with the subject, as well as a genetic relative control who did not reside with the subject [21]. Initial results from this study demonstrated that direct DNA damage was significantly higher in pancreatic cancer patients as compared to healthy related and unrelated controls [21]. As the folate pathway plays a supportive role in DNA synthesis and repair pathways, the overall goal of this study was to investigate whether altered folate levels are observed in these patients and whether expression of SNPs in genes of the folate metabolic pathway are associated with pancreatic cancer risk.
Materials and methods
Ethics statement
This study was approved by the Indiana University Institutional Review Board. A signed written informed consent form was obtained from all participants at the time of recruitment.
Study population
A total of 265 subjects were recruited, of which 159 were pancreatic cancer cases (Stage I-IV), 55 were unrelated controls (spouses or partners) who shared the same environment, and 51 were related controls (children or siblings) who were expected to share a similar genetic makeup. Stratification of the controls was chosen such that the contribution of both environmental and genetic factors and their interaction in pancreatic cancer risk could be investigated. Genetically related controls were included provided that they did not reside with the case whereas unrelated controls were included only if they cohabited with the case. All subjects were Caucasian, >18yrs of age, and able to read, understand and sign a written informed consent. Information on demographics was collected through a questionnaire. Cases with any other malignancies besides pancreatic cancer, who had previously received chemotherapy, or any other treatment for cancer were excluded. Blood samples for isolation of genomic DNA and serum and RBC folate levels were collected from all participants at the time of recruitment.
DNA extraction and SNP genotyping
Whole blood was collected and stored in vacutainer vials (BD Diagnostics, Franklin Lakes, NJ) at -80°C until further analysis. Genomic DNA was isolated from peripheral blood mononuclear cells (PBMCs) using the QIAamp DNA blood midi kit (Qiagen, Valencia, CA). Allelic discrimination was performed using fluorogenic SNP Taqman probes (Applied Biosystems, Carlsbad, CA) for selected SNPs in the folate pathway (S1 Table) and analyzed on an Illumina Eco Real-Time PCR machine (Illumina, San Diego, CA). Genotypes were assigned using the analysis software provided by Illumina.
Microbiological assay for folate determination
Folate levels were measured in serum or plasma and RBCs using the Lactobacillus casei microbiological assay [22,23]. Lactobacillus casei, a folate-dependent bacterium whose growth is an indirect indicator of folate levels in a sample, was purchased from ATCC (ATCC 7946; ATCC, Manassas, VA), and cultured in L.casei broth (BD Biosciences), which contains minimal levels of folate. Bacterial stocks were prepared and cryopreserved in glycerol at -80°C. To quantify folate levels in RBCs, whole blood samples were lysed with 1% ascorbic acid and further diluted in 0.5% sodium ascorbate. For the microbiological assay, cryopreserved bacteria were suspended in 0.9% NaCl and added to folate standards and the serum/plasma or whole blood samples. After incubation in a rotating shaker at 37°C for 48 hrs, absorbance of the cultures was determined at 546 nm, which was used as a measure of bacterial growth. Serum/plasma and whole blood folate concentrations were determined using polynomial regression generated from a standard curve. To calculate RBC folate levels, hemoglobin was measured using the whole blood samples as per the manufacturer’s protocol (Hemoglobin Assay kit, Sigma). The hemoglobin concentrations were then used to calculate hematocrit (Hct) levels from which RBC folate levels were derived according to the formula:
Statistical analysis
Fisher’s exact test was used to compare SNP frequencies between patients and unrelated healthy controls. Logistic regression was used to estimate odds ratios (OR) and corresponding 95% confidence intervals to measure the association between genotypes and risk for pancreatic cancer. Hardy-Weinberg Equilibrium (HWE) was tested using χ 2 analysis. The gene dose overall, dominance and recessive were calculated using χ 2 analysis. To measure the correlation between folate levels and genotype, the independent samples t-test was used. One-way ANOVA was used to calculate overall p-value for the association of genotype with serum and RBC folate levels. Serum and RBC folate levels were compared between patients and controls (unrelated and related) using the Student’s t-test. Serum tertile distribution among patients and controls was analyzed using the Student’s t-test. The serum tertile distribution was based on the CDC definition of deficient (≤2.6 ng/mL), normal (2.6–14.2 ng/mL) and above normal (≥14.2 ng/mL) serum folate ranges (Centers for Disease Control and Prevention www.cdc.gov). For RBC folate stratification, the deficient and normal range distribution was combined because of insufficient subjects in the deficient range according to the CDC definition of normal (102.6–410.9 ng/mL), deficient (≤ 102.6 ng/mL) and above normal (≥ 410.9 ng/mL) RBC folate concentrations (Centers for Disease Control and Prevention www.cdc.gov). Age at the time of diagnosis for the cancer cases, age at the time of sample collection, and date for control groups was included as a continuous variable. Gender for the study population was included as a dichotomous variable. The resulting ORs and corresponding CIs were reported both unadjusted and adjusted for age and gender due to differences in these variables among the patients and the controls. P-values of <0.05 were considered statistically significant. All statistical analyses were performed using SPSS 16.0.
Results
Demographic characteristics of the study population
Age and gender of the study population are shown in Table 1. Pancreatic cancer cases were significantly older than healthy related controls (p = 0.002); however, the unrelated controls exhibited a similar age range as the cases. In this study, pancreatic cancer cases were more likely to be males.
Association of serum and RBC folate levels to pancreatic cancer
Serum folate levels were not significantly different between pancreatic cancer cases (9.5 ± 4.9 ng/mL) and unrelated (9.3 ± 4.0 ng/mL) or related controls (9.6 ± 4.5 ng/mL) (Table 2). However, while not achieving statistical significance (p = 0.05), 12% of the pancreatic cancer cases exhibited deficient serum folate levels (below 2.6 ng/mL) compared with only 7% of the healthy unrelated control group (S2 Table). Pancreatic cancer cases exhibited significantly lower RBC folate levels (508.4 ± 215.9 ng/mL) compared to unrelated controls (588.3 ± 229.2 ng/mL); however, RBC folate levels in the related controls were not significantly different from that in the cases (Table 2). When RBC folate levels were stratified according to CDC criteria, 33% of the cases demonstrated deficient to normal RBC folate levels (<102.6–410.9 ng/mL) whereas 67% were observed to have higher than normal (>410.9 ng/mL) RBC folate levels. However, these values were not statistically different between cases and controls (S3 Table).
Association of SNPs in the folate pathway and pancreatic cancer
Genotypes for 15 SNPs in 12 genes of the folate metabolic pathway (Fig 1 and S1 Table) were determined in the pancreatic cancer cases and control groups (Table 3). A significant difference was observed in the genotype distribution among patients and healthy unrelated controls for the choline dehydrogenase CHDH L78R SNP (p = 0.03), as well as in the dominant gene dose model (p = 0.02) (Table 3). No other significant differences were observed in genotype frequencies for the remaining SNPs in the folate pathway. Hardy-Weinberg equilibrium (HWE) was next tested using χ2 analysis for these SNPs and polymorphisms that did not follow HWE were not analyzed further. Odds ratios were determined for the remaining 8 SNPs (Table 4). Subjects that carried the LL genotype (nucleotide TT) of the CHDH L78R SNP were found to be at decreased risk for pancreatic cancer both prior to (OR = 0.29; 95% CI 0.12–0.76) and after adjusting for age and sex (OR = 0.29; 95% CI 0.10–0.84) (Table 4). No other significant associations with pancreatic cancer risk for the remaining SNPs in the folate pathway were observed.
Association of SNPs in the folate pathway with altered serum and RBC folate levels
We observed significantly different serum folate levels associated with individuals expressing the serine hydroxymethyltransferase (SHMT1) L474F, 5-methltetrahydrofolate-homocysteine methyltransferase (MTR) D919G and the phosphatidylethanolamine N-methyltransferase (PEMT) V175M polymorphisms (Table 5). Individuals with the LF allele of the SHMT1 L474F SNP showed significantly higher serum folate levels (10.4 ± 0.4 ng/mL) compared to those of LL allele (9.1 ± 0.4 ng/mL); however a gene dose-response trend was not observed. Individuals with the DG genotype of the MTR D919G polymorphism had significantly lower serum folate concentrations (8.2 ± 0.5 ng/mL) compared to those expressing the reference genotype DD (10.3 ± 0.3 ng/mL). For the PEMT V175M polymorphism, the subjects with the VV genotype exhibited significantly lower serum folate levels (7.1 ± 1.2 ng/mL) compared to the reference genotype MM (9.6 ± 0.4 ng/mL).
Significant differences in RBC folate levels were observed in the subjects expressing A222V and E429A polymorphisms of the MTHFR gene. Subjects with the AV and VV genotypes of the MTHFR A222V SNP showed significantly lower RBC folate levels (509.6 ± 21.3 ng/mL and 370.6 ± 31.7 ng/mL respectively) compared to individuals expressing the wild type allele AA (603.5 ± 21.4 ng/mL). The EE allele of the MTHFR E429A polymorphism was also found to be significantly associated with lower RBC folate levels (496.6 ± 21.5 ng/mL) compared to the heterozygous EA and AA variant allele (560.0 ± 21.5 ng/mL and 563.8 ± 44.6 ng/mL, respectively).
Discussion
Pancreatic cancer is a devastating disease, with a relative 5 year survival rate of 8%. One of the reasons for the dismal prognosis is that more than half of the cases are diagnosed at late stage, where the 5-year survival rate is only 3% [1]. As such, identification of risk susceptibility profiles for pancreatic cancer could greatly impact early diagnosis and prevention strategies. In a pilot study, we have previously shown that direct DNA damage was increased in pancreatic cancer patients as compared to control subjects who resided with the patient (i.e. shared the same environment) and also as compared to controls that are genetically related, but did not share the same environment [21]. As defects in the folate metabolic pathway have been linked to DNA damage, in this study we sought to determine (1) whether serum and RBC folate levels differ between pancreatic cancer patients and controls, (2) whether SNPs in key genes of the folate pathway associate with pancreatic cancer risk and (3) whether expression of specific SNPs correlates with serum or RBC folate levels.
The relationship between dietary folate intake, measured serum and RBC folate levels, and pancreatic cancer risk is not clearly defined. Dietary folate intake has been shown to be inversely proportional to pancreatic cancer risk in several epidemiological studies in which higher dietary folate intake was associated with lower risk for pancreatic cancer [24–26]. However, in the Nurses’ Health Study and the Health Professionals Follow-up Study, no relationship between folate intake and pancreatic cancer risk was observed [27]. Low serum folate levels were associated with increased risk of pancreatic cancer in male Finnish smokers in the Alpha Tocopherol Beta Carotene Cancer prevention (ATBC) study [6]. Interestingly, the European Prospective Investigation Into Cancer study has shown a U-shaped relationship between plasma folate and pancreatic cancer risk; that is, individuals with high plasma folate (>20 nM) and those with deficient or near deficient plasma folate levels (5–10 nM) had a similar increased risk when compared to those with adequate plasma folate levels (10–15 nM) [7]. In our study, we did not find significant differences in the mean serum folate levels of the pancreatic cancer cases and the unrelated controls or related controls (9.5 ± 4.9ng/mL, 9.3 ± 4.0 ng/mL, and 9.6 ± 4.5 ng/mL, respectively) (Table 2). We did observe however, that 12% of cases fell in the deficient range (<2.6 ng/mL) for serum folate, compared to only 6% in unrelated controls and 7% in related controls (S2 Table). Since 1998, the FDA has required folate fortification of cereals and grains in the US, and folate levels in the general population have risen significantly [28], which may explain why differences in serum folate levels were not observed in our study. In addition, the ATBC study consisted exclusively of male smokers and smoking has been associated with low circulating folate levels [29].
Although RBC folate levels have been studied with respect to obesity and neural tube defects, our study is the first, to our knowledge, to determine an association of RBC folate levels with pancreatic cancer risk. We observed that RBC folate levels in pancreatic cancer cases (508.4 ± 215.9 ng/mL) were significantly lower than that of unrelated healthy controls (588.3 ± 229.2 ng/mL), but not that of related controls (531.8 ± 239.7 ng/mL), suggesting that RBC folate levels are, in part, dependent on genetics. Compared to serum folate levels, which are indicative of recent folate intake, RBC folate concentrations represent tissue folate levels, and hence are more representative of an individual’s overall folate status. While the RBC folate levels were lower in pancreatic cancer subjects in this study, the observed levels are considered to be within the CDC’s definition of folate-sufficient status. It is possible that the lower RBC folate in the cases could have been due to underlying pathologies caused by cancer, such as poor folate absorption or increased demand for DNA synthesis. On the other hand, animal studies have shown that folic acid supplementation decreased MTHFR enzymatic activity resulting in reduced methylation potential, suggesting that high tissue folate levels mimic MTHFR deficiency [30]. As 67% of the pancreatic cancer cases were observed to have higher than normal RBC folate levels according to CDC guidelines, further assessment of folate metabolism in pancreatic cancer, (e.g. determination of methyl-THF, SAM and SAH levels) appears warranted.
While the MTHFR A222V, MTRR H595Y, and TYMS 3Rc SNPs have previously been shown to be associated with pancreatic cancer [18,20], the studies did not include measurements of serum or RBC folate levels. In our study, we evaluated gene polymorphisms involved in the uptake, transmethylation and transulfuration pathways of folate metabolism (Fig 1) for their association with pancreatic cancer risk as well as association of folate levels with these gene polymorphisms. The MTHFR A222V polymorphism results from a nucleotide change at position 677 from C to T. Individuals heterozygous (CT) or homozygous (TT) for the C677T polymorphism exhibit lower enzymatic activity, 65% and 30% respectively, compared to CC homozygous individuals [9]. Decreased activity of MTHFR exhibited by the TT genotype is thus predicted to result in DNA hypomethylation and indeed, expression of the VV allele has been associated with decreased DNA methylation compared to the AA allele [31]. DNA hypomethylation, which may lead to aberrant expression of proto-oncogenes or overexpression of proteins involved in cancer progression, has been shown previously in pancreatic cancer [32]. Li et al. reported a 2-fold increased risk for pancreatic cancer for individuals with the MTHFR 677 TT versus the CC/CT polymorphisms while Wang et al. reported increased odds ratios of 2.6 and 5.1 for the CT and TT genotypes, respectively, versus the wild-type CC polymorphism [16,18]. While we did not observe an association of the MTHFR A222V or MTRR H595Y polymorphism with pancreatic cancer, we found that expression of the homozygous variant allele results in lower RBC folate levels (Table 5). Our results are consistent with other studies that observed lower RBC folate levels in individuals with the AV and VV alleles in the MTHFR A222V SNP [31,33,34]. In contrast, individuals with the EA and AA variants of the MTHFR E429A SNP had higher mean RBC folate concentrations as compared to the referent EE genotype. This is consistent with previous observations, in that the ancestral and more common alleles of the MTHFR A222V and MTHFR E429A polymorphisms exhibited opposite effects on RBC folate levels [35]. While we did not find statistically significant associations of pancreatic cancer with either of the MTHFR SNPs, the percentage of pancreatic cancer patients with the VV genotype in the A222V SNP, with the lowest RBC folate levels, is almost double that of unrelated control subjects (16% vs 9%, respectively; Table 3), and the percentage of pancreatic cancer patients with the EE genotype of the E429A SNP, with lower RBC folate levels, is also greater than that of unrelated control subjects (54% vs 44%, respectively; Table 3). A study by Keku et al. found that the AA genotype of the E429A SNP, that is associated with high RBC folate levels, was only protective when dietary folate was low, but not when folate intake was high, suggesting that determination of the genotype of an individual alone may not adequately assess risk [36]. Rather, other factors, such as dietary intake of folate, need to be considered along with genotype. As dietary folate intake data was not collected in our study, we cannot perform these same correlations in this study group. We have observed that the SHMT1 L474F, MTR D919G, and PEMT V175M polymorphisms display significant differences in serum folate levels among the alleles (Table 5). The variant alleles of the SHMT 1 L474F and MTR D919G polymorphisms have been implicated in increased risk of certain cancers [37–39]. We found that individuals expressing the LF allele of the SHMT1 L474F SNP exhibited significantly higher serum folate levels compared to those carrying the referent LL allele; however, serum folate levels in FF homozygous individuals did not significantly differ from those with the LL allele. These results are consistent with previous findings that individuals with the LL genotype exhibited lower plasma folate levels than those of the LF genotype [40]. SHMT1 catalyzes the reaction for generation of methylene-THF, which can either be utilized in thymidylate or methionine biosynthesis. While the amino acid change of leucine to phenylalanine does not appear to alter enzymatic activity, it impairs nuclear transport of SHMT1 resulting in cytosolic accumulation of the altered SHMT1 where it is not able to facilitate thymidylate synthesis [41–43]. As such, one might predict that expression of this SNP would confer a protective effect against development of pancreatic cancer; however, expression of a cytoplasmic isozyme of SHMT2, SHMT2α, appears to functionally compensate for lack of wild-type SHMT1 protein [41].
The enzyme MTR catalyzes the remethylation of homocysteine to methionine (Fig 1). The functional impact of the MTR D919G polymorphism is not known, however it has been postulated to alter enzyme activity [38]. The presence of the G allele in the MTR D919G SNP has been associated with lower plasma homocysteine levels and higher folate concentrations as compared to DD genotype in the Physicians Health Study (PHS) [44]. In contrast, we found that the G allele was associated with significantly lower serum folate levels compared to that of the DD genotype (Table 5). This could be due to the difference in the mean serum folate ranges observed in these studies, as the PHS study was conducted before folate fortification in the US. PEMT is a key enzyme in the choline metabolic pathway converting phosphatidylethanolamine (PE) to phophatidylcholine (PC) which is further converted to choline. Choline is the substrate for betaine synthesis which in turn provides the methyl group for methionine synthesis (Fig 1). PEMT therefore, plays a key role in the methylation cycle of the folate metabolic pathway. The amino acid substitution of methionine for valine at position 175 in PEMT (V175M) results in production of an enzyme with reduced activity [45]. In our study, we found that individuals with the VV genotype of the V175M PEMT SNP exhibited significantly lower serum folate levels than those with the MM genotype. In concordance with these findings, women with the rarer VV genotype were more likely to have a child with a neural tube defect [46].
Our study is the first to report a significant association between the L78R CHDH SNP and the risk of pancreatic cancer, where we found that the LL genotype of the polymorphism L78R CHDH was significantly protective (OR = 0.29; CI 0.12–0.76) against pancreatic cancer when compared with the referent RR allele. The CHDH enzyme catalyzes the synthesis of betaine from choline. In addition to 5,10 methylene THF, betaine is an alternative substrate for methionine synthesis, which would aid in maintaining sufficient pools of SAM for DNA methylation reactions. While detailed investigation of the effect of this SNP on CHDH enzyme activity has not been conducted, several studies suggest that the L78R substitution results in a functional change of the protein. Expression of the minor T allele, which codes for expression of leucine, has been associated with development of clinical symptoms of choline deficiency [47]. In addition, infertility linked to altered sperm function was found in men with the GT or TT genotype, which was attributed to decreased ATP production by dysmorphic mitochondria [48]. Further, CHDH protein levels in sperm and hepatocytes in individuals with either the GT and TT genotype was decreased in comparison to that of the GG individuals [48]. As folate levels were not different among genotypes in the L78R SNP (Table 5), the mechanism by which the TT allele confers protection against pancreatic cancer is likely not due to modulation of folate metabolism. CHDH has recently been found to stimulate mitophagy, a form of autophagy, following mitochondrial damage [49]. As autophagy has been shown to promote pancreatic cancer [50], we postulate that the inability of the L78 form of CHDH to stimulate autophagy may lead to protection against pancreatic cancer development. Further studies are required to validate this hypothesis.
One limitation of our study is the small size of the study population, which may have affected the statistical power of our study and may explain inconsistencies seen with previous larger epidemiological studies. Another limitation is the lack of information on lifestyle factors such as smoking, alcohol consumption, and dietary folate intake which can influence folate status. In conclusion, in this study, we have explored several SNPs in genes in the folate pathway for their association with both folate status and pancreatic cancer, and have identified that the LL variant of the L78R SNP in CHDH is associated with a decreased risk for pancreatic cancer. Further studies are needed to verify and extend these findings in larger pancreatic cancer cohorts.
Supporting information
S1 Table. SNPs in the folate pathway examined in this study.
https://doi.org/10.1371/journal.pone.0193298.s001
(DOCX)
S2 Table. Serum folate tertile distribution among patients and controls.
https://doi.org/10.1371/journal.pone.0193298.s002
(DOCX)
S3 Table. Red blood cell folate tertile distribution among patients and controls.
https://doi.org/10.1371/journal.pone.0193298.s003
(DOCX)
References
- 1. Siegel RL, Miller KD, Jemal A (2017) Cancer Statistics, 2017. CA Cancer J Clin 67: 7–30. pmid:28055103
- 2. Hezel AF, Kimmelamn AC, Stanger BZ, Bardeesy N, D R.A. (2006) Genetics and biology of pancreatic ductal adenocarcinoma. Genes Dev 20: 1218–1249. pmid:16702400
- 3. Maisonneuve P, Lowenfels A (2010) Epidemiology of pancreatic cancer: An update. Dig Dis 28: 645–656. pmid:21088417
- 4. Pogribny I, Ross S, Wise C, Pogribna M, Jones E, Tryndyak VP, et al. (2006) Irreversible global DNA hypomethylation as a key step in hepatocarcinogenesis induced by dietary methyl deficiency. Mutat Res 593: 80–87. pmid:16144704
- 5. Zeisel S (2012) Dietary choline deficiency causes DNA strand breaks and alters epigenetic marks on DNA and histones. Mutat Res 733: 34–38. pmid:22041500
- 6. Stolzenberg-Solomon R, Albanes D, Nieto F, Hartman T, Tangrea J, Rautalahti M, et al. (1999) Pancreatic cancer risk and nutrition-related methyl-group availability indicators in male smokers. J Natl Cancer Inst 91: 535–541. pmid:10088624
- 7. Chuang S, Stolzenberg-Solomon R, Ueland P, Vollset S, Midttun Ø, Olsen A, et al. (2011) A U-shaped relationship between plasma folate and pancreatic cancer risk in the European Prospective Investigation into Cancer and Nutrition. Eur J Cancer 47: 1808–1816. pmid:21411310
- 8. Sanchez G, Weinstein S, Stolzenberg-Solomon R (2012) Is dietary fat, vitamin D, or folate associated with pancreatic cancer? Mol Carcinog 51: 19–27.
- 9. Larsson S, Giovannucci E, Wolk A (2006) Folate intake, MTHFR polymorphisms, and risk of esophageal, gastric, and pancreatic cancer: A meta-analysis. Gastroenterology 131: 1271–1283. pmid:17030196
- 10. Kim YI, Pogribny IP, Basnakian AG, Miller JW, Selhub J, James SJ, et al. (1997) Folate deficiency in rats induces DNA strand breaks and hypomethylation within the p53 tumor suppressor gene. Am J Clin Nutr 65: 46–52. pmid:8988912
- 11. Das PM, Singal R (2004) DNA methylation and cancer. J Clin Oncol 22: 4632–4642. pmid:15542813
- 12. Sibani S, Melnyk S, Pogribny IP, Wang W, Hiou-Tim F, Deng L, et al. (2002) Studies of methionine cycle intermediates (SAM, SAH), DNA methylation and impact of folate deficiency on tumor numbers in Min mice. Carcinogenesis 23: 61–65. pmid:11756224
- 13. Dufficy LM, Naumovski N, N X., Blades B, Yates Z, Travers C, et al. (2006) G80A reduced folate carrier SNP influences the absorption and cellular translocation of dietary folate and its association with blood pressure in an elderly population. Life Sci 79: 957–966. pmid:16750224
- 14. Stanisławska-Sachadyn A, Brown KS, Mitchell LE, Woodside JV, Young IS, Scott JM, et al. (2008) An insertion/deletion polymorphism of the dihydrofolate reductase (DHFR) gene is associated with serum and red blood cell folate concentrations in women. Hum Genet 123: 289–295. pmid:18247058
- 15. Jacques PF, Bostom AG, Williams RR, Ellison RC, Eckfeldt JH, Rosenberg IH, et al. (1996) Relation between folate status, a common mutation in methylenetetrahydrofolate reductase, and plasma homocysteine concentrations. Circulation 93: 7–9. pmid:8616944
- 16. Li D, Ahmed M, Li Y, Jiao L, Chou T-H, Wolff RA, et al. (2005) 5,10-Methylenetetrahydrofolate reductase polymorphisms and the risk of pancreatic cancer. Cancer Epidemiol Biomarkers Prev 14: 1470–1476. pmid:15941958
- 17. Suzuki T, Matsuo K, Sawaki A, Mizuno N, Hiraki A, Kawase T, et al. (2008) Alcohol drinking and one-carbon metabolism-related gene polymorphisms on pancreatic cancer risk. Cancer Epidemiol Biomarkers Prev 17: 2742–2747. pmid:18843018
- 18. Wang L, Miao X, Tan W, Lu X, Zhao P, Zhao X, et al. (2005) Genetic polymorphisms in methelenetetrahydrofolate reductase and thymidylate synthase and risk of pancreatic cancer. Clin Gastroenterol Hepatol 3: 743–751. pmid:16234002
- 19. Matsubayashi H, Skinner HG, Iacobuzio-Donahue C, Abe T, Sato N, Riall TS, et al. (2005) Pancreaticobiliary cancers with deficient methylenetetrahydrofolate reductase genotypes. Clin Gastroenterol Hepatol 3: 752–760. pmid:16234003
- 20. Ohnami S, Sato Y, Yoshimura K, Ohnami S, Sakamoto H, Aoki K, et al. (2008) His595Tyr polymorphism in the methionine synthase reductase (MTRR) gene is associated with pancreatic cancer risk. Gastroenterology 135: 477–488. pmid:18515090
- 21. Hocevar BA, Kamendulis LM, Pu X, Perkins SM, Wang Z-Y, Johnston EL, et al. (2014) Contribution of environment and genetics to pancreatic cancer susceptibility. PLoS ONE 9: e90052. pmid:24651674
- 22. O'Broin S, Kelleher B (1992) Microbiological assay on microtitre plates of folate in serum and red cells. J Clin Pathol 45: 344–347. pmid:1577973
- 23. Wilson SD, Horne DW (1982) Use of glycerol-cryoprotected Lactobacillus casei for microbiological assay of folic acid. Clin Chem 28: 1198–1200. pmid:6804123
- 24. Larsson S, Håkansson N, Giovannucci E, Wolk A (2006) Folate intake and pancreatic cancer incidence: a prospective study of Swedish women and men. J Natl Cancer Inst 98: 407–413. pmid:16537833
- 25. Lin HL, An QZ, Wang QZ, Liu CX (2013) Folate intake and pancreatic cancer risk: an overall and dose-response meta-analysis. Public Health 127: 607–613. pmid:23769243
- 26. Oaks B, Dodd K, Meinhold C, Jiao L, Church T, Stolzenberg-Solomon RZ (2010) Folate intake, post-folic acid grain fortification, and pancreatic cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Am J Clin Nutr 91: 449–455. pmid:20007302
- 27. Skinner H, Michaud D, Giovannucci E, Rimm E, Stampfer M, Willett WC, et al. (2004) A prospective study of folate intake and the risk of pancreatic cancer in men and women. Am J Epidemiol 160: 248–258. pmid:15257998
- 28. Yetley EA, Johnson CL (2011) Folate and vitamin B-12 biomarkers in NHANES: history of their measurements and use. Am J Clin Nutr 94: 1S–10S.
- 29. Piyathilake CJ, Macaluso M, Hine RJ, Richards EW, Krumdieck CL (1994) Local and systemic effects of cigarette smoking on folate and vitamin B-12. Am J Clin Nutr 60: 559–566. pmid:8092091
- 30. Christensen KE, Mikael LG, Leung K-Y, Levesque N, Deng L, Wu Q, et al. (2015) High folic acid consumption leads to pseudo-MTHFR deficiency, altered lipid metabolism, and liver injury in mice. Am J Clin Nutr 101: 646–658. pmid:25733650
- 31. Friso S, Choi SW, Girelli D, Mason JB, Dolnikowski GG, Bagley PJ, et al. (2002) A common mutation in the 5,10-methylenetetrahydrofolate reductase gene affects genomic DNA methylation through an interaction with folate status. Proc Natl Acad Sci U S A 99: 5606–5611. pmid:11929966
- 32. Sato N, Maitra A, N F, van Heck N, Matsubayashi H, Iacobuzio-Donahue CA, et al. (2003) Frequent hypomethylation of multiple genes overexpressed in pancreatic ductal adenocarcinoma. Cancer Res 63: 4158–1466. pmid:12874021
- 33. Zittoun J, Tonetti C, Bories D, Pignon JM, Tulliez M (1998) Plasma homocysteine levels related to interactions between folate status and methylenetetrahydrofolate reductase: a study in 52 healthy subjects. Metabolism 47: 1413–1418. pmid:9826223
- 34. Narayanan S, McConnell J, Little J, Sharp L, Piyathilake CJ, Powers H, et al. (2004) Associations between two common variants C677T and A1298C in the methylenetetrahydrofolate reductase gene and measures of folate metabolism and DNA stability (strand breaks, misincorporated uracil, and DNA methylation status) in human lymphocytes in vivo. Cancer Epidemiol Biomarkers Prev 13: 1436–1443. pmid:15342443
- 35. Parle-McDermott A, Mills JL, Molloy AM, Carroll N, Kirke PN, Cox C, et al. (2006) The MTHFR 1298CC and 677TT genotypes have opposite associations with red cell folate levels. Mol Genet Metab 88: 290–294. pmid:16621645
- 36. Keku T, Millikan R, Worley K, Winkel S, Eaton A, Biscocho L, et al. (2002) 5,10-Methylenetetrahydrofolate reductase codon 677 and 1298 polymorphisms and colon cancer in African Americans and whites. Cancer Epidemiol Biomarkers Prev 11: 1611–1621. pmid:12496052
- 37. Goode EL, Potter JD, Bigler J, Ulrich CM (2004) Methionine synthase D919G polymorphism, folate metabolism, and colorectal adenoma risk. Cancer Epidemiol Biomarkers Prev 13: 157–162. pmid:14744749
- 38. Skibola CF, Smith MT, Hubbard A, Shane B, Roberts AC, Law GR, et al. (2002) Polymorphisms in the thymidylate synthase and serine hydroxymethyltransferase genes and risk of adult acute lymphocytic leukemia. Blood 99: 3786–3791. pmid:11986237
- 39. Galbiatti ALS, Ruiz MT, Biselli-Chicote PM, Raposo LS, Maniglia JV, Pavarino-Bertelli EC, et al. (2010) 5-Methyltetrahydrofolate-homocysteine methyltransferase gene polymorphism (MTR) and risk of head and neck cancer. Braz J Med Biol Res 43: 445–450. pmid:20490431
- 40. Heil SG, Van der Put NM, Waas ET, den Heijer M, Trijbels FJ, Blom HJ (2001) Is mutated serine hydroxymethyltransferase (SHMT) involved in the etiology of neural tube defects? Mol Genet Metab 73: 164–172. pmid:11386852
- 41. Anderson DD, Stover PJ (2009) SHMT1 and SHMT2 are functionally redundant in nuclear de novo thymidylate biosynthesis. PLoS ONE 4: e5839. pmid:19513116
- 42. Herbig K, Chiang EP, Lee LR, Hills J, Shane B, Stover PJ (2002) Cytoplasmic serine hydroxymethyltransferase mediates competition between folate-dependent deoxyribonucleotide and S-adenosylmethionine biosyntheses. J Biol Chem 277: 38381–38389. pmid:12161434
- 43. Woeller CF, Anderson DD, Szebenyi DM, Stover PJ (2007) Evidence for small ubiquitin-like modifier-dependent nuclear import of the thymidylate biosynthesis pathway. J Biol Chem 282: 17623–17631. pmid:17446168
- 44. Chen J, Stampfer MJ, Ma J, Selhub J, Malinow MR, Hennekens CH, et al. (2001) Influence of a methionine synthase (D919G) polymorphism on plasma homocysteine and folate levels and relation to risk of myocardial infarction. Atherosclerosis 154: 667–672. pmid:11257268
- 45. Song J, da Costa KA, Fischer LM, Kohlmeier M, Kwock L, Wang S, et al. (2005) Polymorphism of the PEMT gene and susceptibility to nonalcoholic fatty liver disease (NAFLD). FASEB J 19: 1266–1271. pmid:16051693
- 46. Mills JL, Fan R, Brody LC, Liu A, Ueland PM, Wang Y, et al. (2014) Maternal choline concentrations during pregnancy and choline-related genetic variants as risk factors for neural tube defects. Am J Clin Nutr 100: 1069–1074. pmid:25240073
- 47. da Costa K, Niculescu M, Craciunescu C, Fischer L, Zeisel S (2006) Choline deficiency increases lymphocyte apoptosis and DNA damage in humans. Am J Clin Nutr 84: 88–94. pmid:16825685
- 48. Johnson AR, Lao S, Wang TH, Galanko JA, Zeisel SH (2012) Choline dehydrogenase polymorphism rs12676 is a functional variation and is associated with changes in human sperm cell function. PLoS ONE 7: e36047. pmid:22558321
- 49. Park S, Choi S-G, Yoo S-M, Son JH, Jung Y-K (2014) Choline dehydrogenase interacts with SQSTM1/p62 to recruit LC3 and stimulates mitophagy. Autophagy 10: 1906–1920. pmid:25483962
- 50. Gomez VE, Giovannetti E, Peters GF (2015) Unraveling the complexity of autophagy: potential therapeutic applications in pancreatic ductal adenocarcinoma. Semin Cancer Biol 35: 11–19. pmid:26408419