The wayward methyl group and the cascade to cancer

ABSTRACT

We propose here a hypothesis of the cause of cancer that brings together fundamental changes in methyl-group metabolism resulting in methionine dependence and global DNA hypomethylation which destabilizes the genome leading to aneuploid karyotypes which evolve and stabilize into autonomous cancer. Experimental support for this hypothesis is that methioine dependence is a general metabolic defect in caner. Methionine dependence is due to excess use of methionene for aberrant transmethylation reactions that apparently divert methyl groups from DNA. The resulting global DNA hypomethylation is also a general phenomena in cancer. Global hypomethylation leads to an unstable genomes and aneuploid karyotypes, another general phenomena in cancer. The excessive and aberrant use of methionine in cancer is strongly observed in [¹¹C]methionine PET imaging, where high uptake of [¹¹C]methionine results in a very strong and selective tumor signal compared with normal tissue background. [¹¹C]methionine is superior to [¹⁸C] fluorodeoxyglucose (FDG)-PET for PET imaging, suggesting methionine dependence is more tumor-specific than glucose dependence.

KEYWORDS:

• aneuploidy
• cancer
• chromosome instability
• DNA hypomethylation
• methionine
• methionine dependence
• unbalanced transmethylation

Methionine dependence and altered transmethylation in cancer cells

The first hint that methyl metabolism is perturbed in cancer came almost 60 years ago when Sugimura et al.¹ observed that rat tumor growth was slowed by giving the rats a defined diet depleted in methionine. Approximately 45 years ago, it was observed that L5178Y mouse leukemia cells in culture required very high levels of methionine to proliferate.² Subsequently, most cancer cell lines were found to be methionine dependent.^3,4 These cell lines were derived from various cancer types including liver, pancreatic ovarian, submaxillary, brain, lung, bladder, prostate, breast, kidney, cervical, colon, fibrosarcoma, osteosarcoma, rhabdomyosarcoma, leiomyosarcoma, neuroblastoma, glioblastoma and melanoma. The occurrence of methionine dependence among these diverse cancer types suggests that methionine dependence may be a general phenomena in cancer.

Human patient tumors, including tumors of the colon, breast, ovary, prostate, and a melanoma, were also found to be methionine dependent in Gelfoam® histoculture.⁵ Normal unestablished cell strains, thus far characterized, grow well in methionine depleted medium.^3,6

Methionine-dependent cancer cells synthesize normal amounts of methionine

Multiple lines of evidence indicated that methionine-dependent cancer cells synthesize large amounts of methionine endogenously through the reaction catalyzed by N⁵-methyltetrahydropteroyl-l-glutamate: l-homocysteine S-methyltransferase (_EC 2.1.1.13).⁷ For example, the activity of the methyltransferase involved in methionine biosynthesis was comparable in extracts of methionine-dependent malignant and normal methionine-independent cells. In addition, the uptake of radioactive label from [5−¹⁴C]methyltetrahydropteroyl-l-glutamic acid (N⁵-methyl-H₄PteGlu) was at least as great in the methionine-dependent malignant cells as in the normal methionine-independent cells and was nearly totally dependent on the addition of homocysteine, the methyl acceptor. The majority of the labeled methyl groups incorporated by methionine-dependent cancer cells was recovered as methionine. These results indicated that the methionine auxotrophy (dependence) of the malignant cells does not result simply from the inability to synthesize and incorporate methionine from homocysteine and N⁵-methyl-H₄PteGlu.⁸

Low levels of endogenous methionine and S-adenosyl methionine in cancer cells under methionine-limiting conditions

We subsequently observed that although methionine-dependent cancer cells synthesized a normal amount of endogenous methionine, the level of free methionine and S-adenosylmethionine (AdoMET), the universal methyl donor, were very low in methionine-dependent cancer cells in methionine (MET) depleted homocysteine (HCY)-supplemented medium (MET⁻HCY⁺), but is normal in MET⁺HCY⁻ medium. We determined that a low AdoMET/AdoHCY ratio probably limits growth of methionine-dependent cells in MET⁻HCY⁺ medium.⁹

Enhanced rates of transmethylation in cancer cells

We observed that cancer cells have enhanced overall rates of transmethylation compared with normal cells. Transmethylation rates were measured by blocking AdoHCY hydrolase and measuring AdoHCY, which accumulates as a result of transmethylation. The enhanced transmethylation rates may be the basis of the methionine dependence of cancer cells which explains the low levels of free methionine and the low AdoMET/AdoHCY ratio in cancer cells under methionine deprivation, despite high rates of methionine synthesis.¹⁰ The elevated methionine use in cancer cells has been termed the “Hoffman effect.”¹¹

Methionine-dependent cancer cells arrest in S/G₂ phase of the cell cycle under methionine deprivation

Growth arrest of methionine-dependent cancer cells in MET ⁻HCY⁺ medium resulted in a reduction in the percentage of mitotic cells in the cell population. Cell cycle analysis demonstrated that the cells are arrested in the S/G₂ phases of the cell cycle in MET ⁻HCY⁺ medium.^12,13 This is in contrast to a G₁-phase accumulation of cells, which occurs only in methionine-supplemented medium at very high cell densities and is similar to the G₁ block seen in cultures of normal fibroblasts at high density.^12,13

Methionine independent revertants

Rare cells from methionine-dependent cancer cell lines regained the normal ability to grow in MET⁻HCY⁺ medium. These lines were termed methionine-independent revertants.¹⁴ Methionine-independent revertants also had much lower basal transmethylation rates than parental methionine-dependent cell lines.¹⁵ These results further suggested that methionine dependence is due to an increase in the rate of transmethylation reactions.

We then demonstrated that methionine-independent revertants cells concomitantly reverted for characteristics associated with cancer and became less malignant. Thus, the methionine-independent revertants become more normal-like indicating further a relationship between altered methionine metabolism and oncogenic transformation.¹⁶

Diversion of methyl groups in methionine-dependent cancer cells

When a methionine-dependent cancer cell line was cultured under conditions of methionine depletion, methylation of nucleic acids was decreased. However, there was increased methylation of both an endogenous substrate and Escherichia coli tRNA. These results indicated that methyl groups were being diverted from their normal sites, including DNA, in methionine-dependent cancer cells. Please see below for the implications of this abnormality.

Discovery and description of DNA hypomethylation in cancer cells

Genome-wide DNA hypomethylation of human cancer cells was first described by us in the early eighties,¹⁷ and has been found in almost all types of cancers.¹⁸ Other groups subsequently confirmed our discovery.^19-24

Perucho et al.²⁵ used a biochemical quantitative assay to estimate the percentage of methylated sites in DNA in normal and cancer gastric tissues from patients with gastric cancer and patients with high-risk gastritis (HRG) caused by H. pylori infection. The initiation and incidence and extent of somatic genome-wide hypomethylation was investigated to determine its chronology in the progression of gastric carcinogenesis.²⁵ Genome-wide hypomethylation was found in the gastric mucosa of gastric cancer patients and HRG patients as well as global DNA hypomethylation in tumors, suggesting an epigenetic field defect of DNA hypomethylation in gastric cancer, with H. pylori infection.²⁶ The degree of DNA hypomethylation in primary tumors was significantly associated with stage with the extent of invasion through the stomach layers.²⁵ One group of gastric cancer patients had much greater demethylation than the majority of the patients. The latter group was termed enhanced somatic hypomethylation (ESH). ESH tumors had a more aggressive phenotype, including invasion through the gastric wall and into adjacent lymph nodes.²⁵ Liteplo and Kerbel²⁷ found that high-metastatic variants of a melanoma had more DNA hypomethylation than the low-metastatic parental cells, similar to the pattern Perucho et al.²⁵ observed.

DNA hypomethylation and aneuploidy

Suzuki et al. observed that hypomethylation preceded diploidy loss in a significant subset of gastrointestinal cancers, and had a stronger association with genetic damage and poorer prognosis than hypermethylation.²⁸ Widespread DNA hypomethylation in gastric cancer associates with chromosomal instability and poor prognosis. Compare et al.²⁹ also observed that global demethylation is an early event in the gastritis-cancer pathway. Preneoplastic lesions had decreased DNA methylation over time despite the eradication of H. pylori. Perucho et al.²⁵ proposed that DNA hypomethylation contributes to the initial stages of the carcinogenesis process by destabilizing the genome. The association between DNA hypomethylation and aneuploidy had been previously reported.^30-34

Jaenish et al.³³ generated transgenic mice with depleted DNA methyltransferase 1 (DNMT1) which resulted in genome-wide hypomethylation in all tissues. The mutant mice developed T cell lymphomas which had a high frequency of chromosome 15 trisomy. In another study, DNMT-deficient HCT-116 colon cancer cells had a high degree of DNA hypomethylation and genome instability leading to aneuploidy, including many novel chromosomal translocations.³⁴ These results indicate that DNA hypomethylation plays a causal role in chromosomal instability, aneuploidy, and subsequent cancer.³³

We suggest here that the pre-malignant DNA hypomethylation observed by Perucho et al.²⁵ results from the unbalanced global transmethylation observed in methionine dependence.

Results from Duesberg and colleagues^35-41 suggest that autonomous cancers result from speciation based on selection of existing aneuploid karyotypes that convert mortal somatic cells to immortality and eventually to independence and cancer. It should be noted that Boveri⁴² more than 100 years ago implicated altered karyotypes in the generation of malignancy.

Deprivation of methionine leads to cancer

Copeland and Salmon observed the development of liver, lung, and other cancers in a significant percentage of rats on a choline-deficient diet which results in depleted methionine.⁴² Subsequently, Ghoshal and Farber⁴³ observed that Fischer 344 male rats fed a choline-methionine-deficient diet developed a 100% incidence of preneoplastic hepatocyte nodules, whereby 51% of the rats developed hepatocellular carcinoma. Supplement of the diet with 0.8% choline chloride prevented the development of both the precancerous nodules and subsequent cancer.⁴⁴ The cancers resulting from methionine/choline-deprived diets in the rats were probably due to methyl shortage and subsequent DNA hypomethylation, since methionine is the source of methyl groups for DNA methylation. DNA hypomethylation appears to be an early event in choline/methionine-deficient diets, which is irreversible^44-47 and mediated by a deficiency in S-adenosylmethionine.⁴⁸

From perturbed methionine metabolism to DNA hypomethylation to a destabilized genome to aneuploidy to cancer (Fig. 1)

We suggest here that the pre-malignant DNA hypomethylation such as that observed by Perucho et al.²⁵ results from the unbalanced excess transmethylation which diverts methyl groups into where they are not available for normal methylation processes such as DNA methylation. The excess transmethylation rate explains the “methionine dependence” of cancer cells. Hypomethylated DNA then results in the initiation of aneuploidy and subsequent speciation to autonomous cancer.

Figure 1. The wayward methyl group and the cascade to cancer.

Thus, we propose that carcinogenesis is initiated by perturbation of methionine metabolism, for example, rats on a chronic methionine-depleted diet, and altered transmethylation (“metabolic reprogramming”) resulting in DNA hypomethylation thereby destabilizing the karyotype, which sets off a chain reaction of aneuploidizations as Duesberg proposes, which generate ever-more abnormal karyotypes and eventually cancer-specific combinations.⁴⁹

Every hypothesis should be testable and be able to be negated. The present hypothesis predicts that premalignancy such as observed in the stomach with H. pylori infection or SV40 infection, other infectious agents such as human papiloma virus, or Epstein-Barr virus,⁵⁰ or carcinogens such as nitrosourea or cigarette smoke, should have an elevated methionine requirement and excess and altered transmethylation leading to DNA hypomethylation as shown in the stomach after H. pylori infection by Perucho et al.²⁵ The hypothesis predicts that chronic perturbation of methionine metabolism should be carcinogenic, such as a chronic methionine-depleted diet.^42,43

The current hypothesis predicts that deprivation of methionine should be therapeutic for cancer and in mouse models, this is the case.^7,51-55

Our hypothesis accommodates the role of altered genes that may influence the behavior of cancers, but with possibly rare exceptions, individual gene alterations do not account for the global changes of cancer that methionine dependence, global DNA hypomethylation and altered karotype evolution do.

The hypothesis also explains why the use of [¹¹C] methionine is so effective in positron emission tomography (MET-PET) imaging, since the cancers use excessive methionine for their aberrant excess transmethylation and therefore take up excess [¹¹C] methionine, compared to normal tissue.^56-60

In a comparison of MET-PET and fluorodeoxyglucose (FDG)-PET, MET-PET was found to be superior for glioma⁶¹ suggesting that cancer may have a greater abnormal requirement for methionine than glucose.

Recently, a paper appeared with the title “The new anticancer era: tumor metabolism targeting.⁶² The present review suggests this “new anticancer era” started in 1959 with the observation of Sugimura et al.¹ that depriving cancer of methionine arrested tumor growth. It is our hope that this era will continue and lead to more effective cancer treatments.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.