Telomere Length in Peripheral Blood Mononuclear Cells Is Associated with Folate Status in Men

Abstract

Human chromosomes are capped by telomeres, which consist of tandem repeats of DNA and associated proteins. The length of the telomeres is reduced with increasing cell divisions except when the enzyme telomerase is active, as in stem cells and germ cells. Telomere dysfunction has been associated with development of age-related pathologies, including cancer, cardiovascular disease, Alzheimer's disease, and Parkinson's disease. DNA damage in the telomeric region causes attrition of telomeres. Because folate provides precursors for nucleotide synthesis and thus affects the integrity of DNA, including that of the telomeric region, folate status has the potential to influence telomere length. Telomere length is epigenetically regulated by DNA methylation, which in turn could be modulated by folate status. In this study, we determined whether folate status and the 677C > T polymorphism of the methylene tetrahydrofolate reductase (MTHFR) gene are associated with the telomere length of peripheral blood mononuclear cells in healthy men. The results of our study showed that plasma concentration of folate was associated with telomere length of peripheral blood mononuclear cells in a nonlinear manner. When plasma folate concentration was above the median, there was a positive relationship between folate and telomere length. In contrast, there was an inverse relationship between folate and telomere length when plasma folate concentration was below the median. The MTHFR 677C > T polymorphism was weakly associated (P = 0.065) with increased telomere length at below-median folate status. We propose that folate status influences telomere length by affecting DNA integrity and the epigenetic regulation of telomere length through DNA methylation.

Introduction

Human chromosomes are capped by telomeres, which consist of tandem repeats of the DNA sequence TTAGGG and associated proteins. DNA polymerases replicate the termini of linear chromosomes in a 5′ to 3′ fashion in a process that requires an RNA primer for initiation and the removal of the RNA primer leads to the loss of the DNA sequence at the ends after every replication (1–4). Thus, the length of the telomeres is reduced with increasing cell divisions until they reach a critical length, at which point the cells enter senescence (5). In germ cells, embryonic stem cells, and many cancer cells, telomere length is maintained by the reverse transcriptase enzyme telomerase, which adds the telomeric repeats to the ends of the newly synthesized DNA (6,7). The single-stranded 3′ ends of the telomeres are protected by a complex of proteins collectively termed shelterin (8). The formation of the telomere-protein complex prevents single-strand DNA repair enzymes from working on the telomere ends and also restricts access to telomerase, preventing unwanted lengthening of telomeres (8).

Shortening of telomeres occurs with age, life stress, infection, and chronic diseases (3,9–11). Loss of telomere function due to attrition of the repeats makes the chromosomes susceptible to fusion with other chromosome ends and double-strand breaks, resulting in chromosomal rearrangements (12). Telomere dysfunction has been associated with development of age-related pathologies. Shortened telomeres are present in patients with Parkinson's disease (13) and Alzheimer's disease (14). The presence of short telomeres in peripheral blood cells has been shown to be a risk factor for cancer of head and neck, lung, and kidney (15) and cardiovascular disease (16,17). On the other hand, increased telomere length of peripheral lymphocytes is a risk factor and a negative prognostic factor in breast cancer (18). Elongated telomeres are present in the endometria of patients with endometriosis (19) and in hepatocellular carcinomas, long telomeres of the tumor tissues indicate poor prognosis (20). It has been shown that whereas there is a shortening of telomeres in the early stages of cancer due to increased cell proliferation, in the later stage of tumor, telomeres are elongated due to the activation of telomerase (21).

Damage in telomeric DNA is repaired less efficiently compared with coding regions and can result in short telomeres (22,23). Folate provides precursors for nucleotide synthesis and low folate availability causes nucleotide imbalances and subsequently DNA damage (24,25). Thus, it is possible that folate can influence telomere length by its effect on integrity of telomeric DNA. A recent population study of twins showed that folate status is associated with decreased telomere length (26). Telomere length is epigenetically regulated by DNA and histone methylation of telomeric and subtelomeric regions, possibly by controlling the access of telomere elongating proteins to telomeric regions (27–29). Deficiency in the DNA methyltransferase (DNMT)⁶ enzymes DNMT1 or both DNMT3a and DNMT3b, or the histone methyltransferases suppressor of variegation 3-9 homolog or suppressor of variegation 4-20 homolog, results in longer-than–normal telomeres (27–29) without causing changes in telomerase expression. Methyl tetrahydrofolate supplies the methyl group for the vitamin B-12–dependent methylation of homocysteine (Hcy) to form methionine. S-Adenosyl methionine is the universal methyl group donor for DNA and histone methylation. Thus, it is possible that folate status can modulate the epigenetic regulation of telomere length. The 677C > T polymorphism in the methylene tetrahydrofolate reductase (MTHFR) gene reduces the enzyme activity in vivo, resulting in reduced availability of methyl tetrahydrofolate for methylation of Hcy (30). We have previously shown that homozygosity for this MTHFR polymorphism and low folate status are associated with genomic DNA hypomethylation (31). In this study, we examined the association of plasma folate status and 677C > T polymorphism of the MTHFR gene on the telomere length of peripheral blood mononuclear cells in healthy men.

Materials and Methods

Population.

Study participants (n = 195) were recruited from Milan, Italy as a part of a project to determine the interaction between folate and the 677C > T polymorphism in the MTHFR gene in healthy men. The participants were males between 40 and 68 y of age who did not take any multivitamin supplements. Eight participants were omitted from the analysis, because their telomere length could not be determined from the available DNA samples. One subject was excluded from the analysis due to extremely high plasma vitamin B-12 concentration (3953 pmol/L), suggesting the presence of potentially confounding factors. The study was conducted with approval from the Institutional Review Board of the Tufts University and the Ethical Committees of the Ospedale San Paolo Hospital at Università di Milano, Milan and Scientific Institute San Raffaele, Milan after obtaining informed consent from the participants.

Plasma analysis.

Analysis of plasma Hcy was carried out using the HPLC method described by Fermo et al. (32). Plasma concentrations of folate and vitamin B-12 were measured by chemiluminescence using the Bayer ADVIA Centaur system. The CV for assays for plasma Hcy, vitamin B-12, and folate were determined by analyzing normal pooled plasma. The within-series and between-series CV were 4.1 and 7.2% for plasma Hcy (mean = 10.8 μmol/L), 3.6 and 7.4% for vitamin B-12 (mean = 297 pmol/L), and 4.06 and 8.76% for folate (mean = 10.5 nmol/L).

DNA preparation.

DNA was isolated from the peripheral blood mononuclear cells of the participants using a QiaAmp DNA Blood Mini kit (Qiagen).

Genotyping.

The MTHFR genotype was determined using a Taqman single nucleotide polymorphism genotyping assay for the 677C > T polymorphism using a 7300 Real-Time PCR system from Applied Biosystems.

Determination of telomere length.

We determined telomere length using a real-time PCR method originally described by Cawthon (33) with the modifications of McGrath et al. (34). In this method, the telomere region is amplified using primers specific to the telomere repeats. The amplification of a single copy gene (human β-globin) is used as a control for the input DNA. Telomere length is expressed as the relative telomere length or telomere repeat:single copy gene (T:S) ratio. The reactions were carried out in a 96-well plate format using a 7300 Real-Time PCR system (Applied Biosystems). To reduce any effect of variation between plates, each assay plate carried samples from different plasma folate concentrations. A standard curve prepared from genomic DNA of cultured lymphoblastoid cells, ranging from 30 to 0.625 ng of DNA, was present in every plate and was used to quantitate the telomere or single copy gene in the samples. The reaction was linear over a range of 100–0.039 ng DNA as determined separately.

Statistical analysis.

We used the SAS (version 9.1) software package (SAS Institute) to analyze the association between relative telomere length and the parameters tested using analysis of covariance to allow adjustments for potential confounders. This method was used for all the analyses presented. Associations with a P-value < 0.05 were considered significant. The data on relative telomere length was log transformed for statistical analysis and then back transformed for Figure 1 and Table 2. Tukey's test was used for comparisons with multiple adjustments. The graphs for the figures were prepared using Kaleidagraph (version 4.03).

Results

The characteristics of the study population are given in Table 1. Telomere length of the peripheral blood mononuclear cells is presented as relative telomere length, which is the ratio of telomere quantity:single copy gene (human β-globin) quantity. The age of the participants was inversely associated with telomere length (P = 0.005) as has been demonstrated by previous studies. The association between folate and telomere length was tested after the data were adjusted for age and plasma concentration of vitamin B-12. Comparison of mean telomere lengths in the quartiles of plasma folate concentration (Fig. 1A) showed that with the decrease in plasma folate concentration to 11.6 nmol/L (median), there was a corresponding decrease in mean telomere length. When the plasma folate concentration decreased below 11.6 nmol/L, there was an increase in telomere length. The mean telomere length of men in the lowest quartile of plasma folate concentration tended to be higher than that of the men in the highest quartile of folate (P = 0.11) (Fig. 1A).

FIGURE 1 

Relative telomere length (T:S ratio) of men according to quartiles of plasma folate (A) and Hcy (B) concentrations. In A, the data were adjusted for age and plasma concentration of vitamin B-12 and in B, for age, plasma concentrations of folate and vitamin B-12, and the MTHFR 667C > T polymorphism. The values are mean T:S ratios and 95% CI. Means without a common letter differ, P < 0.02. The plasma folate and Hcy concentrations for each quartile are means ± SD.

TABLE 1

Characteristics of the study population¹

Men, n	195
MTHFR 677C > T genotype, n
CC	56
CT	107
TT	33
Age, y	51.0 ± 6.8	50.0 (41.0, 63.0)
Plasma concentrations
Vitamin B-12, pmol/L	252.0 ± 75.0	238.0 (5.4, 19.2)
Folate, nmol/L	12.0 ± 4.9	11.6 (5.7, 19.9)
Hcy, μmol/L	9.8 ± 5.4	8.4 (5.4, 19.2)

1

Values are means ± SD or median (5th, 95th percentiles).

The MTHFR 677C > T polymorphism was associated with plasma folate status among the participants of this study. Individuals carrying the TT genotype had a lower plasma folate concentration than those with the CC genotype (P = 0.006) and tended to have a lower concentration than those with the CT genotype (P = 0.12) (Table 2). An overall analysis did not show a significant association between the MTHFR 677C > T polymorphism and telomere length (Table 2). Our previous data have shown that the effect of the MTHFR polymorphism on plasma Hcy concentration and DNA methylation occur only in people with low folate status (31). Hence, we tested the association between the MTHFR 677C > T polymorphism and telomere length with consideration of the men's plasma folate concentration (above or below the median). The telomere length in individuals with the TT genotype of the MTHFR 677C > T polymorphism tended to be longer than that of those with the CC genotype when their plasma folate concentration was below the median (≤11.6 nmol/L), after adjusting for age and the plasma vitamin B-12 concentration (P = 0.065) (Table 2). In men with plasma folate concentrations above the median, there was no association between the MTHFR polymorphism and telomere length. Among men with below median plasma folate concentration, those with the TT genotype had significantly lower plasma folate and higher plasma Hcy concentrations than those with the CC and CT genotypes (Table 2). Telomere length was longer in men in the highest quartile of plasma Hcy concentration compared with those in the 2 lowest quartiles, after adjusting for age, plasma folate and vitamin B-12 concentrations, and the 677C > T polymorphism of the MTHFR gene (Fig. 1B). The association between plasma folate and telomere length remained significant after adjusting for the 677C > T polymorphism of the MTHFR gene and plasma Hcy. The plasma vitamin B-12 concentration was not associated with telomere length.

TABLE 2

Association between relative telomere length and MTHFR 677C > T genotype according to plasma folate concentration in men¹

MTHFR 677C > T polymorphism	Participants, n	Plasma folate, nmol/L	Plasma Hcy, μmol/L	Relative telomere length (T:S ratio)2
Overall analysis
CC	56	13.4 ± 2.2a	8.19 ± 0.66a	0.55 (0.49–0.61)
CT	106	11.9 ± 2.1a,b	9.16 ± 0.48b	0.55 (0.51–0.59)
TT	33	10.2 ± 1.9b	14.86 ± 0.86c	0.62 (0.54–0.71)
Below median folate (≤11.6 nmol/L)
CC	22	9.46 ± 1.71a	8.62 ± 2.23a	0.57 (0.48–0.66)
CT	56	8.44 ± 1.81a	10.35 ± 3.64a	0.62 (0.56–0.68)
TT	20	7.19 ± 2.22b	19.1 ± 11.07b	0.74 (0.63–0.87)
Above median folate (≥11.61 nmol/L)
CC	34	15.99 ± 4.59	7.91 ± 2.49	0.54 (0.47–0.62)
CT	50	15.75 ± 4.26	7.82 ± 1.78	0.48 (0.43–0.54)
TT	13	14.73 ± 2.49	8.33 ± 1.33	0.49 (0.39–0.61)

1

Values are means ± SD or means (95% CI). Labeled means in a column without a common letter differ, P ≤ 0.03.

2

Adjusted for age and plasma vitamin B-12 concentration.

Discussion

Results of our study indicate that there is an association between folate status and telomere length in peripheral blood mononuclear cells of humans. With the decrease in plasma folate concentration to the median (11.6 nmol/L), there was a corresponding decrease in telomere length, but when plasma folate concentration decreased below the median, telomere length increased (Fig. 1A). We suggest that the nonlinear association of folate with telomere length can be explained by the dual role of folate in cellular metabolism in providing nucleotide substrates for synthesis and repair of DNA, including that of telomeres and in methylation of DNA, which in turn regulates telomere length.

Folate provides precursors for nucleotide synthesis and low folate availability induces uracil misincorporation and strand breaks in DNA (35). Telomeric DNA is repaired less efficiently compared with coding regions of the genome and damage in telomeric DNA result in shorter telomeres (22,23). Although hematopoietic stem cells express telomerase activity, it is not sufficient to prevent telomere attrition (36,37). Thus, reduced availability of folate for maintaining DNA integrity can lead to attrition of telomeres in blood cells. A recent population study on twins showed that low plasma folate status is significantly associated with reduced telomere length (26).

We propose that the lengthening of telomeres observed in the lowest folate quartiles is mediated by DNA damage response and hypomethylation of DNA due to low folate status. Shortening of telomeres beyond a critical length can trigger DNA damage response (38), which results in global decondensation of chromatin (39,40). It has been shown that in short telomeres, heterochromatic histone methylation (trimethylation of lysine residues 9 and 20 of histone 3) and subtelomeric DNA methylation are reduced, leading to an open chromatin state (41,42). The open chromatin structure of the shorter telomere makes it accessible to telomerase and proteins that are involved in telomere recombination, leading to elongation of telomeres (39). Elongation by telomerase is the major salvage pathway for short telomeres lacking the telomere-associated proteins in hematopoietic cells (43). The mechanism of telomere elongation detailed above could be engaged when telomere attrition due to low folate availability results in critically short telomeres in hematopoietic stem cells.

The elongation of short telomeres could also be influenced by DNA hypomethylation due to low folate status. Telomere length is under epigenetic regulation and methylation of the subtelomeric DNA and associated histones exerts an effect on the length of telomeres (27–29). DNA hypomethylation due to deficiency in the DNMT enzymes DNMT1 or both DNMT3a and DNMT3b leads to longer telomeres without altering the expression of telomerase (28). The elongation of telomeres in DNMT-deficient cells could be due to the increased telomeric recombination observed in these cells and/or due to increased accessibility of telomerase to the telomeric region (28). We have previously shown that low plasma folate status is associated with genomic DNA hypomethylation in the peripheral blood cells of human participants (31,44). In addition to DNA hypomethylation, telomere length is also regulated by histone methylation (trimethylation of lysine residue 9 of histone 3 and residue 20 of histone 4) (27–29). A methyl-deficient diet has been shown to reduce site-specific histone methylation (45) and alter gene expression in animals. DNA or histone hypomethylation thus adds to the factors that could cause lengthening of telomeres under conditions of very low folate status.

Homozygosity for the T allele of the 677C > T polymorphism of the MTHFR gene was weakly associated with longer telomere length when folate status was below median (Table 2). This could be because of the increased DNA hypomethylation due to the interaction of the TT genotype with low folate status (31). It is possible that this association between the MTHFR genotype and telomere length would be stronger in a larger study population. The longer telomeres present in the participants in the highest quartile of plasma Hcy (16.2 ± 7.6 μmol/L) (Fig. 1B) could also be explained by the inhibition of DNA methylation by a high concentration of S-adenosyl Hcy. The association between Hcy and telomere length was significant after adjusting for age, plasma concentrations of folate and vitamin B-12, and the MTHFR 677C > T polymorphism. Thus, it seems that factors other than folate or vitamin B-12 status, including polymorphisms in the genes other than MTHFR that affect Hcy concentration in plasma also influence telomere length. In our study participants, telomere length did not decrease with an increase in plasma Hcy concentration, unlike the findings of Richards et al. (26). One of the main differences between the 2 populations is that the cohort studied by Richards et al. (26) was predominantly female (91.5%) and our study participants were all male. Telomere length attrition in lymphocytes has been shown to be influenced by gender (46), which could be one of the reasons for the differences in the results between the 2 studies. Another factor that could have influenced our results is the higher concentration of plasma Hcy in our study participants (16.2 ± 7.6 μmol/L) compared with that in the Richards et al. (26) study (11.8 ± 4.2 μmol/L) in the highest Hcy category, which possibly results in extensive inhibition of DNA or histone methylation reactions that regulate telomere length.

The association between folate status and telomere length remained significant after adjusting for plasma Hcy concentration, which reflects the methylation status of the cell. Thus, our data suggests that the variations of telomere length under low folate status could be primarily due to DNA damage and the resulting response and, to a lesser extent, due to changes DNA methylation affecting the epigenetic control of telomere length. Changes in telomere length due to reduced folate availability appear similar to that in cancer cells. In cancer, there is a shortening of telomeres when telomere maintenance mechanisms are outpaced by their high proliferation (47). Once the tumor progresses to an advanced stage, however, activation of telomerase and alternative telomere lengthening by telomere sister chromatid exchange averts further attrition of telomeres (5,20,21,48) and in the case of some cancers results in longer telomeres than the normal tissue (20). This prevents the senescence or apoptosis of the cells that would otherwise occur due to critically shortened telomeres.

In summary, low folate status is associated with altered telomere length in a nonlinear manner. There is a shortening of telomeres with decreased plasma folate concentration to the median, but telomere length increases with decreased plasma folate concentration below the median. Abnormal regulation of telomere length can lead to chromosome rearrangements and instability, potentially leading to cancer. Increased telomeric recombination has been demonstrated in cells with elongated telomeres due to decreased DNA methylation (28) or histone 4 methylation (27). Chromosome rearrangements as indicated by nucleoplasmic bridges have been reported under low folate conditions in cell culture (49,50), which was hypothesized to be due to the elongation of telomeres resulting from hypomethylation of the subtelomeric region and associated proteins (5,51). We are in the process of investigating the association of folate with telomere length in other cohorts and identifying the mechanism behind the changes in telomere length under low folate status.

Abbreviations

• DNMT

  DNA methyltransferase

• Hcy

  homocysteine

• MTHFR

  methylene tetrahydrofolate reductase

• T:S

  ratio of telomere repeat:single copy gene