Mild folate deficiency induces genetic and epigenetic instability and phenotype changes in prostate cancer cells

Abstract

Background: Folate (vitamin B9) is essential for cellular proliferation as it is involved in the biosynthesis of deoxythymidine monophosphate (dTMP) and s-adenosylmethionine (AdoMet). The link between folate depletion and the genesis and progression of cancers of epithelial origin is of high clinical relevance, but still unclear. We recently demonstrated that sensitivity to low folate availability is affected by the rate of polyamine biosynthesis, which is prominent in prostate cells. We, therefore, hypothesized that prostate cells might be highly susceptible to genetic, epigenetic and phenotypic changes consequent to folate restriction.

Results: We studied the consequences of long-term, mild folate depletion in a model comprised of three syngenic cell lines derived from the transgenic adenoma of the mouse prostate (TRAMP) model, recapitulating different stages of prostate cancer; benign, transformed and metastatic. High-performance liquid chromatography analysis demonstrated that mild folate depletion (100 nM) sufficed to induce imbalance in both the nucleotide and AdoMet pools in all prostate cell lines. Random oligonucleotide-primed synthesis (ROPS) revealed a significant increase in uracil misincorporation and DNA single strand breaks, while spectral karyotype analysis (SKY) identified five novel chromosomal rearrangements in cells grown with mild folate depletion. Using global approaches, we identified an increase in CpG island and histone methylation upon folate depletion despite unchanged levels of total 5-methylcytosine, indicating a broad effect of folate depletion on epigenetic regulation. These genomic changes coincided with phenotype changes in the prostate cells including increased anchorage-independent growth and reduced sensitivity to folate depletion.

Conclusions: This study demonstrates that prostate cells are highly susceptible to genetic and epigenetic changes consequent to mild folate depletion as compared to cells grown with supraphysiological amounts of folate (2 microM) routinely used in tissue culture. In addition, we elucidate for the first time the contribution of these aspects to consequent phenotype changes in epithelial cells. These results provide a strong rationale for studying the effects of folate manipulation on the prostate in vivo, where cells might be more sensitive to changes in folate status resulting from folate supplementation or antifolate therapeutic approaches.

Figure 1

Folate, one-carbon metabolism, methionine cycle and polyamine biosynthesis overview. Dietary folate (left) is necessary for de novo synthesis of deoxythymidine monophosphate (dTMP) and adenosylmethionine (AdoMet), which are necessary for DNA synthesis and intracellular methylation reactions, respectively. AdoMet is also used for polyamine (spermidine and spermine) biosynthesis (top). dTMP can also be scavenged by DNA degradation as thymidine, while AdoMet can be synthesized independently of folate by recycling methylthioadenosine (MTA) into methionine through the methionine salvage pathway.

Figure 2

Chronic mild folate depletion affects prostate cells' phenotype. (A) Anchorage-independent colonies in soft agar. 10,000 cells well were seeded in semi-solid agar and allowed to grow for 4 weeks. Data presented is the average and SD of six replicates. (B) Folate dependent growth rates. Growth rates (average and standard deviation) are calculated normalizing to growth in control amounts of folate (2 μM) set as 100%. (C) High-performance liquid chromatography analysis of polyamine pools. Results are the average and standard deviation of two biologically independent samples. Values are normalized to those obtained in GCtrl cells, set as 100%. * = P < 0.05, ** = P < 0.01, *** = P < 0.001.

Figure 3

Mild folate depletion induces genetic damage in prostate cells. (A) High-performance liquid chromatography analysis of nucleotide pools. Results are the average and standard deviation of three independent injections. Values are relative to cells grown in control medium. (B) Single strand breaks generated by treatment with uracil-DNA-glycosidase (UDG) and Exonuclease III are quantified by random oligonucleotide-primed synthesis performed with the Klenow polymerase in the presence of P³²-deoxcytidine triphosphate (dCTP; left). The values of dCTP incorporation for all cell lines are the average of five replicates after subtraction of no enzyme treatment control background and presented as an increase over cells grown in control medium. (C) Spectral karyotype analysis reveals that mild folate depletion induced de novo chromosomal aberrations in G100. A representative metaphase from GCtrl (top), which exhibits only marker chromosomes common to both GCtrl and G100 (Table 1) and a representative metaphase from G100 (bottom) which harbours the common t(1;10) translocation along with four marker chromosomes, indicated by arrows, found frequently in G100, but never in GCtrl (Table 1). (D) Distribution of metaphase cells with frequent patterns of marker chromosomes shared between GCtrl and G100 indicates lack of clonal selection in G100. Ten different patterns (A-J) of the five marker chromosomes (a-e; Table 1) were found in two or more metaphase cells. G100 retains significant heterogeneity of patterns of marker chromosomes including two novel patterns (I and J). Pattern A represents marker chromosomes a, b, c and d; pattern B = marker a; pattern C = markers a and e; pattern D = marker b; pattern E = no marker chromosomes; pattern F = markers a, b and c; pattern G = markers b, c and d; pattern H = markers b and c; pattern I = a, b, c, d and e; pattern J = marker e. * = P < 0.05, ** = P < 0.01, *** = P < 0.001.

Figure 4

Mild folate depletion induces epigenetic damage in prostate cells. (A) High-performance liquid chromatography analysis of adenosylmethionine/adenosylhomocysteine pools. Results are the average and standard deviation of three independent injections and are presented relative to cells grown in control medium (100%). * = P < 0.05, ** = P < 0.01, *** = P < 0.001. (B) Analysis of total levels of 5-methylcytosine by liquid chromatography-mass spectrometry. Results are the average and standard deviation of two independent injections. Values are relative to cells grown in control medium. (C) Sections of restriction landmark genomic scanning (RLGS) gels including the spot 4D20, which shows decreased intensity in both D100 and G100, and the spot 4G11, which shows decreased intensity in both G100 and H100. (D) Mass array quantitative methylation analysis confirmed the methylation status of CpG identified by RLGS and associated to known loci. Peripheral blood lymphocyte DNA was used as a 0% methylation control and an in vitro methylated aliquot of the same DNA as a 100% methylated control. These DNAs were mixed at appropriate ratios to generate 75%, 50% and 25% methylated controls. The figure shows 4D20 as an example. (E) Western Blot analysis of histone acid lysates run on sodium dodecyl sulphate-page gel revealed an increase of global H3K9 methylation (17 Kda) and H3K36 di-methylation (17 Kda) upon mild folate depletion.