DNA methylation: a historical perspective

5mC was discovered in mammals and found to have a nonrandom distribution that suggested a possible biological function.In the early 1980s, DNA methylation within 5′ promoter regions, but not elsewhere, was found to inhibit transcription of the associated gene.Throughout the 1990s and 2000s, mechanisms of gene regulation by DNA methylation were elucidated as well as its relationship with histone modifications and influence on the 3D genome organization uncovered.Over the past decade, high-throughput sequencing technologies complemented earlier single-gene efforts and ultimately provided a global understanding of DNA methylation and its dynamics in development and disease. In 1925, 5-methylcytosine was first reported in bacteria. However, its biological importance was not intuitive for several decades. After this initial lag, the ubiquitous presence of this methylated base emerged across all domains of life and revealed a range of essential biological functions. Today, we are armed with the knowledge of the key factors that establish, maintain, and remove DNA methylation and have access to a staggering and rapidly growing number of base-resolution methylation maps. Despite this, several fundamental details about the precise role and interpretation of DNA methylation patterns remain under investigation. Here, we review the field of DNA methylation from its beginning to present day, with an emphasis on findings in mammalian systems, and point the reader to select experiments that form the foundation of this field. In 1925, 5-methylcytosine was first reported in bacteria. However, its biological importance was not intuitive for several decades. After this initial lag, the ubiquitous presence of this methylated base emerged across all domains of life and revealed a range of essential biological functions. Today, we are armed with the knowledge of the key factors that establish, maintain, and remove DNA methylation and have access to a staggering and rapidly growing number of base-resolution methylation maps. Despite this, several fundamental details about the precise role and interpretation of DNA methylation patterns remain under investigation. Here, we review the field of DNA methylation from its beginning to present day, with an emphasis on findings in mammalian systems, and point the reader to select experiments that form the foundation of this field. A quarter century ago, one of the pioneers in the field of DNA methylation, Rudolf Jaenisch, outlined in the August 1997 issue of Trends in Genetics why we should bother caring about DNA methylation and speculated in which developmental contexts it might function [1.Jaenisch R. DNA methylation and imprinting: why bother?.Trends Genet. 1997; 13: 323-329Abstract Full Text PDF PubMed Scopus (313) Google Scholar]. Here, we would like to review why we still bother, what we have learned in nearly a century of research, and what we still need to address in the coming years. Since its initial discovery in bacteria in 1925, DNA methylation has been investigated in a vast range of organisms and is linked to biological topics from gene regulation and genome organization, to reproduction and development, and to disease and aging. It is the most well-studied epigenetic mechanism and is often used as the classical example of epigenetic inheritance, although recent advances have shown this modification to be more dynamic, and hence more complex, than previously thought [2.Ginno P.A. et al.A genome-scale map of DNA methylation turnover identifies site-specific dependencies of DNMT and TET activity.Nat. Commun. 2020; 11: 2680Crossref PubMed Scopus (45) Google Scholar, 3.Charlton J. et al.TETs compete with DNMT3 activity in pluripotent cells at thousands of methylated somatic enhancers.Nat. Genet. 2020; 52: 819-827Crossref PubMed Scopus (34) Google Scholar, 4.Spada F. et al.Active turnover of genomic methylcytosine in pluripotent cells.Nat. Chem. Biol. 2020; 16: 1411-1419Google Scholar]. Despite an ever-growing body of work published on DNA methylation each year, it remains difficult to pinpoint the precise function of most DNA methylation found across the genome. It is also still unresolved why DNA methylation is essential to differentiated, but not pluripotent, cells [5.Jackson-Grusby L. et al.Loss of genomic methylation causes p53-dependent apoptosis and epigenetic deregulation.Nat. Genet. 2001; 27: 31-39Crossref PubMed Scopus (558) Google Scholar, 6.Chen T. et al.Establishment and maintenance of genomic methylation patterns in mouse embryonic stem cells by Dnmt3a and Dnmt3b.Mol. Cell. Biol. 2003; 23: 5594-5605Crossref PubMed Scopus (560) Google Scholar, 7.Tsumura A. et al.Maintenance of self-renewal ability of mouse embryonic stem cells in the absence of DNA methyltransferases Dnmt1, Dnmt3a and Dnmt3b.Genes Cells. 2006; 11: 805-814Crossref PubMed Scopus (394) Google Scholar] and why it is altered into a distinct landscape across most cancer types [8.Baylin S.B. Jones P.A. Epigenetic determinants of cancer.Cold Spring Harb. Perspect. Biol. 2016; 8a019505Crossref PubMed Scopus (468) Google Scholar]. As typical review articles by design focus on summarizing more recent discoveries and advances around their time [9.Jones P.A. Laird P.W. Cancer epigenetics comes of age.Nat. Genet. 1999; 21: 163-167Crossref PubMed Scopus (2028) Google Scholar, 10.Bird A. DNA methylation patterns and epigenetic memory.Genes Dev. 2002; 16: 6-21Crossref PubMed Scopus (5200) Google Scholar, 11.Suzuki M.M. Bird A. DNA methylation landscapes: provocative insights from epigenomics.Nat. Rev. Genet. 2008; 9: 465-476Crossref PubMed Scopus (2109) Google Scholar, 12.Jones P.A. Liang G. Rethinking how DNA methylation patterns are maintained.Nat. Rev. Genet. 2009; 10: 805-811Crossref PubMed Scopus (537) Google Scholar, 13.Law J.A. Jacobsen S.E. Establishing, maintaining and modifying DNA methylation patterns in plants and animals.Nat. Rev. Genet. 2010; 11: 204-220Crossref PubMed Scopus (2344) Google Scholar, 14.Smith Z.D. Meissner A. DNA methylation: roles in mammalian development.Nat. Rev. Genet. 2013; 14: 204-220Crossref PubMed Scopus (1783) Google Scholar, 15.Du J. et al.DNA methylation pathways and their crosstalk with histone methylation.Nat. Rev. Mol. Cell Biol. 2015; 16: 519-532Crossref PubMed Scopus (494) Google Scholar, 16.Lyko F. The DNA methyltransferase family: a versatile toolkit for epigenetic regulation.Nat. Rev. Genet. 2018; 19: 81-92Crossref PubMed Scopus (486) Google Scholar, 17.Greenberg M.V.C. Bourc’his, D. The diverse roles of DNA methylation in mammalian development and disease.Nat. Rev. Mol. Cell Biol. 2019; 20: 590-607Crossref PubMed Scopus (533) Google Scholar, 18.Parry A. et al.Active turnover of DNA methylation during cell fate decisions.Nat. Rev. Genet. 2021; 22: 59-66Crossref PubMed Scopus (37) Google Scholar], we decided to complement this by providing a systematic review covering the entire history of the field to highlight many foundational discoveries on which our current work is built. As expected, the primary literature is vast, and we apologize for having to omit many elegant experiments as we summarize the emergence and progression of the field of DNA methylation across a century. At the turn of the 20th century, Walter Sutton (1902) and Theodore Boveri (1903) independently proposed the chromosomal theory of inheritance, linking Gregor Mendel’s (1866) long overlooked laws on gene behavior and inheritance to their own work on meiosis [19.Sutton W.S. On the morphology of the chromosome group in Brachystola magna.Biol. Bull. 1902; 4: 24-39Crossref Google Scholar,20.Sutton W.S. The chromosomes in heredity.Biol. Bull. 1903; 4: 231-251Crossref Google Scholar]. This initially controversial theory gained credence following a 1910 paper from one of its detractors, Thomas Hunt Morgan, who demonstrated that eye color in Drosophila melanogaster is determined by inheritance of a gene on the X chromosome, which provided the first decisive piece of evidence in support of this theory [21.Morgan T.H. Sex limited inheritance in Drosophila.Science. 1910; 32: 120-122Crossref PubMed Google Scholar]. Levene and Jacobs’ research on nucleic acids revealed that they reside in a polymer chain of nucleotides [22.Levene P.A. Jacobs W.A. Über die hefe-nucleinsäure.Ber. Dtsch. Chem. Ges. 1909; 42: 2474-2478Crossref Scopus (0) Google Scholar] and the growing interest in the composition of these nucleic acids laid the foundation among others for the field of epigenetics, with DNA methylation as a central actor (Figure 1A ). In 1925, Johnson and Coghill isolated and crystalized nucleic acids from Mycobacterium tuberculosis in an effort to identify its pathogenic determinant. One of their candidates was 5-methylcytosine (5mC) (see Glossary), a nucleotide Johnson had postulated might occur naturally in living organisms based on his previous success with its in vitro biochemical synthesis [23.Wheeler H.L. Researches on pyrimidine derivatives.J. Am. Chem. Soc. 1904; 31: 591-606Google Scholar]. Microscopic examination of their hydrolyzed nucleic acid picrate crystals under polarized light indeed distinguished cytosine from 5mC [24.Johnson T.B. Coghill R.D. Researches on pyrimidines. C111. The discovery of 5-methyl-cytosine in tuberculinic acid, the nucleic acid of the tubercle bacillus.J. Am. Chem. Soc. 1925; 47: 2838-2844Crossref Google Scholar]. Despite this early and seemingly relevant discovery, the next report on 5mC was only published 23 years later. Using recent advances in paper chromatography [25.Vischer E. Chargaff E. The separation and characterization of purines in minute amounts of nucleic acid hydrolysates.J. Biol. Chem. 1947; 168: 781Abstract Full Text PDF PubMed Google Scholar], Hotchkiss observed a faint band near that of cytosine on his chromatograph of calf thymus DNA that behaved like cytosine, yet was slightly shifted in its migration, leading him to suggest it is cytosine but with some modification and therefore labeled it ‘epi-cytosine’ [26.Hotchkiss R.D. The quantitative separation of purines, pyrimidines, and nucleosides by paper chromatography.J. Biol. Chem. 1948; 175: 315-332Abstract Full Text PDF PubMed Google Scholar] (Figure 1B). Specifically, he noted that the epi-cytosine relates to cytosine in terms of its absorption spectrum and mobility in the same manner that thymine relates to uracil. As thymine is 5-methyluracil, Hotchkiss inferred that epi-cytosine could possibly be 5mC. Two years later, Wyatt confirmed the presence of 5mC in mammalian, insect, and plant DNA with a broad range of quantities [27.Wyatt G.R. Occurrence of 5-methyl-cytosine in nucleic acids.Nature. 1950; 166: 237-238Crossref PubMed Scopus (0) Google Scholar,28.Wyatt G.R. Recognition and estimation of 5-methylcytosine in nucleic acids.Biochem. J. 1951; 48: 581-584Crossref PubMed Google Scholar]. As nucleic acids were confirmed to be the carriers of genetic information [29.Avery O.T. et al.Studies on the chemical nature of the substance inducing transformation of pneumococcal types.J. Exp. Med. 1944; 79: 137-158Crossref PubMed Google Scholar,30.Hershey B.A.D. Chase M. Independent functions of viral protein and nucleic acid in growth of bacteriophage.J. Gen. Physiol. 1952; 36: 39-56Crossref PubMed Google Scholar], and the structure of the DNA double helix was reported [31.Watson J.D. Crick F.H.C. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid.Nature. 1953; 171: 737-738Crossref PubMed Scopus (7945) Google Scholar], interest in the field of DNA methylation grew. Sinsheimer subsequently noted that 5mC is not randomly distributed in DNA but is found specifically in the CpG dinucleotide context (Figure 1C). Interestingly, the CpG doublet was not found as frequently as expected in eukaryotic DNA [32.Smith J.D. Markham R. Polynucleotides from deoxyribonucleic acids.Nature. 1952; 170: 120-121Crossref PubMed Scopus (0) Google Scholar,33.Sinsheimer R.L. et al.The action of pancreatic desoxyribonuclease. I. Isolation of mono- and dinucleotides.J. Biol. Chem. 1954; 208: 445-459Abstract Full Text PDF PubMed Google Scholar]. Why did it take so long from its initial biological discovery before research on 5mC started to progress more rapidly? One obvious reason is the historical context of its discovery. In 1925, we did not know yet that polymer chains of nucleic acids carry genetic information. The 1928 transformation experiments by Frederick Griffith [34.Griffith F. The significance of pneumococcal types.J. Hyg. 1928; 27: 113-159Crossref PubMed Google Scholar], the 1944 Avery-MacLeod-McCarty experiment [29.Avery O.T. et al.Studies on the chemical nature of the substance inducing transformation of pneumococcal types.J. Exp. Med. 1944; 79: 137-158Crossref PubMed Google Scholar], the conclusion of the Second World War, the 1952 Hershey-Chase experiment [30.Hershey B.A.D. Chase M. Independent functions of viral protein and nucleic acid in growth of bacteriophage.J. Gen. Physiol. 1952; 36: 39-56Crossref PubMed Google Scholar], and the resolution of the DNA double helix [31.Watson J.D. Crick F.H.C. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid.Nature. 1953; 171: 737-738Crossref PubMed Scopus (7945) Google Scholar] helped lay the needed foundation that enabled the exploration of the possible relevance of 5mC in DNA. Two additional factors may have caused some initial hesitation: other groups did not find 5mC in their DNA isolates of M. tuberculosis [27.Wyatt G.R. Occurrence of 5-methyl-cytosine in nucleic acids.Nature. 1950; 166: 237-238Crossref PubMed Scopus (0) Google Scholar,35.Vischer Ernst et al.Microbial nucleic acids: the desoxypentose nucleic acids of avian tubercle bacilli and yeast.J. Biol. Chem. 1949; 177: 429-438Abstract Full Text PDF PubMed Google Scholar], and the low abundance of 5mC reported by Hotchkiss and Wyatt seemed disqualifying for 5mC to have a major biological function. As an aside, it is worth mentioning that in parallel to these experimental advances, the developmental biologist Conrad Waddington coined the term ‘epigenetics’ in 1942 [36.Waddington C.H. The epigenotype.Endeavour. 1942; 1: 18-20Crossref Google Scholar] and published his widely used epigenetic landscape in 1957 [37.Waddington C.H. The Strategy of the Genes; A Discussion of Some Aspects of Theoretical Biology. Allen & Unwin, 1957Google Scholar]; however, these concepts were not linked to DNA methylation until its function became clearer over subsequent decades. The dawn of molecular biology set the stage for a more thorough investigation and appreciation of DNA methylation from plants to mammals. However, essential progress was first made by studying the methylation of nucleic acids in bacteria [38.Borek E. Srinivasan P.R. The methylation of nucleic acids.Annu. Rev. Biochem. 1966; 35: 275-298Crossref Google Scholar]. As a tractable and abundant model organism, bacteria provided major insights into the biology of 5mC in prokaryotes and thereby paved the way for its study in higher organisms (Figure 2A ). Luria, Bertani, and Weigle first demonstrated that different families of bacteriophage diverge in their ability to infect certain bacterial strains [39.Luria S.E. Mutations of bacterial viruses affecting their host range.Genetics. 1945; 30: 84-99PubMed Google Scholar,40.Bertani G. Weigle J.J. Host controlled variation in bacterial viruses.J. Bacteriol. 1953; 65: 113-121Crossref PubMed Google Scholar]. The basis for their strain specificity of viral infection was not due to a phage’s differential ability to enter the bacterial strains, but rather because once inside, incompatible phage DNA was found to be degraded in an immune-like response [41.Lederberg S. Suppression of the multiplication of heterologous bacteriophages in lysogenic bacteria.Virology. 1957; 3: 496-513Google Scholar]. A key mechanistic advance was the discovery that different bacteria have strain-specific methyltransferase activity, which raised the possibility of a role for 5mC in the defense against phages [42.Gold M. et al.The enzymatic methylation of RNA and DNA, II. On the species specificity of the methylation enzymes.Proc. Natl. Acad. Sci. U. S. A. 1963; 50: 164-169Crossref PubMed Google Scholar]. Thus, Arber proposed the restriction and modification system (R-M system) where methylation-sensitive ‘restriction enzymes’ defend the bacterial host against invading viruses by digesting their DNA. Bacterial DNA is protected from these restriction enzymes due to modifications to their DNA in the form of species-specific DNA methylation [43.Arber W. Host-controlled modification of bacteriophage.Annu. Rev. Microbiol. 1965; 19: 365-378Crossref PubMed Google Scholar]. Beyond its role in host protection, a link between bacterial DNA methylation and DNA replication was observed [44.Billen D. Hewitt R. Influence of starvation for methionine and other amino acids on subsequent bacterial deoxyribonucleic acid replication.J. Bacteriol. 1966; 92: 609-617Google Scholar]. Billen found that during normal Escherichia coli growth, DNA methyltransferase activity was evident behind the replication fork where 5mC was exclusively placed on the unmethylated nascent strand of DNA (Figure 2B). DNA replication in the absence of methionine, the methyl donor, led to the synthesis of an unmethylated nascent strand, which retained the ability to get methylated after S phase when methionine was added back into the media [45.Billen D. Methylation of the bacterial chromosome: an event at the “replication point”?.J. Mol. Biol. 1968; 31: 477-486Crossref PubMed Google Scholar]. However, it seemed that the unmethylated nascent strand of DNA cannot serve as template DNA in the subsequent round of replication [46.Lark C. Studies on the in vivo methylation of DNA in Escherichia coli 15T.J. Mol. Biol. 1968; 31: 389-399Crossref PubMed Google Scholar] and strains deficient for the methyl-donor showed DNA degradation [47.Lark C. Arber W. Host specificity of DNA produced by Escherichia coli.J. Mol. Biol. 1970; 52: 337-348Crossref PubMed Google Scholar]. In their 1964 review on nucleic acid modifications, Srinivasan and Borek speculated that because 5mC plays a defining role in bacteria, similar mechanisms might act in eukaryotes that could underlie their cell type diversity [48.Srinivasan P.R. Borek E. Enzymatic alteration of nucleic acid structure: enzymes put finishing touches characteristic of each species on RNA and DNA by insertion of methyl groups.Science. 1964; 145: 548-553Crossref PubMed Google Scholar]. Four years later, they reported DNA methyltransferase activity in the nuclear extracts of different tissues of embryonic as well as adult rat and tested their ability to methylate DNA from various species. Interestingly, these experiments showed that some extracts, such as from kidney or liver, harbor more potent methyltransferase activity than brain or spleen extracts. Based on these observations, they suggested that different tissues from the same organism might have different 5mC content [49.Sheid B. et al.Deoxyribonucleic acid methylase of mammalian tissues.Biochemistry. 1968; 7: 280-285Crossref PubMed Google Scholar]. The first biological roles for DNA methylation were gleaned from studies on the basics of bacterial immunity and DNA replication, though it remained unclear whether any of these functions would be conserved in higher organisms. A key advance was based on the discovery that enzymes are responsible for adding the methyl group to cytosines in nucleic acid polymers. This suggested that DNA methylation could be regulated, thus providing a path for specific target modification. In particular, the possible tissue-specific roles of 5mC in rodents were intriguing, but the data were too sparse to draw more meaningful conclusions yet. Once it became clear that 5mC, despite its relatively low abundance, does have a biological function in bacteria, the possibility that DNA methylation could also play a more general regulatory role across species gained credibility. As in many fields, important technological advances were needed to enable a thorough and informative investigation of the theoretical models that emerged in this decade. The presence of 5mC in bacteria, plants, and mammals indicated that 5mC is a widespread DNA modification, which led to further exploration of methylation content using mass spectrometry. In the early 1970s, Vanyushin quantified 5mC levels present in different cell types of many animals, including sponges, mollusks, sea urchins, bony fish, amphibians, reptiles, and mammals [50.Vanyushin B.F. et al.Rare bases in animal DNA.Nature. 1970; 225: 948-949Crossref PubMed Scopus (290) Google Scholar,51.Vanyushin B.F. et al.The content of 5-methylcytosine in animal DNA: the species and tissue specificity.Biochim. Biophys. Acta. 1973; 299: 397-403Crossref PubMed Scopus (0) Google Scholar]. These analyses showed that while both GC and 5mC content can differ between species, they are often more similar between closely related species and generally comparable between different tissues. Interestingly, Vanyushin later found 5mC in sequence contexts other than the CpG dinucleotide in plants and in varying quantities across different plant species [52.Guseinov V.A. Vanyushin B.F. Content and localisation of 5-methylcytosine in DNA of healthy and wilt-infected cotton plants.Biochim. Biophys. Acta (BBA) - Nucleic Acids and Protein Synth. 1975; 395: 229-238Google Scholar]. As more reports of methylation profiling by mass spectrometry accumulated, several groups speculated about the possible role of 5mC in higher organisms, including that: (i) 5mC may play no role in eukaryotic development [53.Adams R.L.P. The relationship between synthesis and methylation of DNA in mouse fibroblasts.Biochim. Biophys. Acta. 1971; 254: 205-212Crossref PubMed Google Scholar,54.Adams R.L.P. Delayed methylation of DNA in developing sea urchin embryos.Nat. New Biol. 1973; 244: 27-29Crossref PubMed Google Scholar]; (ii) 5mC may guide DNA mutations, which at the time were thought to be required for transcriptional changes [55.Scarano E. The control of gene function in cell differentiation and in embryogenesis.Adv. Cytopharmacolo. 1971; 1: 13-24PubMed Google Scholar]; and (iii) 5mC may act as a transcriptional activator [56.Comings D.E. Methylation of euchromatic and heterochromatic DNA.Exp. Cell Res. 1972; 74: 383-390Crossref PubMed Scopus (0) Google Scholar] (Box 1). In 1975, three notable reviews were published that each provided unique frameworks for contemplating and investigating the biological effects of DNA methylation [57.Holliday R. Pugh J.E. DNA modification mechanisms and gene activity during development.Science. 1975; 187: 226-232Crossref PubMed Scopus (1302) Google Scholar, 58.Riggs A.D. X inactivation, differentiation, and DNA methylation.Cytogenet. Cell Genet. 1975; 14: 9-25Crossref PubMed Google Scholar, 59.Sager R. Kitchin R. Selective silencing of eukaryotic DNA.Science. 1975; 189: 426-433Crossref PubMed Google Scholar]. While each review differed in its specific, well rationalized mechanisms, they all fundamentally agreed that 5mC would play a role in regulating gene expression and orchestrating development.Box 1Early theories on the function of DNA methylationWork prior to the 1970s led several scientists to propose formal hypotheses about the function of DNA methylation in eukaryotes. In the late 1960s, Scarano and colleagues observed that 90% of 5mC in sea urchin DNA is found in the CG context and is thus not randomly distributed in DNA, leading them to speculate about a role for 5mC in differentiation [277.Scarano E. et al.The heterogeneity of thymine methyl group origin in DNA pyrimidine isostichs of developing sea urchin embryos.Proc. Natl. Acad. Sci. U. S. A. 1967; 57: 1394-1400Crossref PubMed Google Scholar,278.Grippo P. et al.Methylation of DNA in developing sea urchin embryos.J. Mol. Biol. 1968; 36: 195-208Crossref PubMed Google Scholar]. In 1971, Scarano proposed that spontaneous deamination of 5mC, which generates a C→T conversion, could lead to heritable changes in the DNA sequence. A popular theory from the 1960s up until the early 1980s was that sequence mutations in genes direct differentiation. Scarano therefore speculated that 5mC-guided mutation could direct cellular differentiation during embryogenesis [55.Scarano E. The control of gene function in cell differentiation and in embryogenesis.Adv. Cytopharmacolo. 1971; 1: 13-24PubMed Google Scholar].In the same year, Adams’ work on 5mC patterning following DNA replication in mouse fibroblasts revealed that early replicating DNA is methylated quickly, while late replicating DNA takes several hours to become fully methylated. The observation that active DNA methylation occurs predominantly in S-phase led Adams to conclude that 5mC must not play a role in controlling transcription [53.Adams R.L.P. The relationship between synthesis and methylation of DNA in mouse fibroblasts.Biochim. Biophys. Acta. 1971; 254: 205-212Crossref PubMed Google Scholar]. His conclusions might also have been influenced by the bacterial studies by Billen and Lark that implicated 5mC as a regulator of DNA replication [45.Billen D. Methylation of the bacterial chromosome: an event at the “replication point”?.J. Mol. Biol. 1968; 31: 477-486Crossref PubMed Google Scholar,46.Lark C. Studies on the in vivo methylation of DNA in Escherichia coli 15T.J. Mol. Biol. 1968; 31: 389-399Crossref PubMed Google Scholar]. Interestingly, in 1972 Comings came to a different conclusion looking at Chinese hamster ovarian cells, where he found that late replicating AT-rich DNA is undermethylated to a greater extent than would be expected from its base composition, while early replicating GC-rich DNA is highly methylated. Comings speculated that if DNA methylation is needed in high amounts in euchromatic DNA where it might play a role in active transcription, then spontaneous deamination of 5mC leading to CG→TA mutations would be actively selected against in euchromatin [56.Comings D.E. Methylation of euchromatic and heterochromatic DNA.Exp. Cell Res. 1972; 74: 383-390Crossref PubMed Scopus (0) Google Scholar]. To Comings, the idea that 5mC is enriched in active DNA regions in eukaryotes implicated it as a transcriptional activator.In 1973, Adams demonstrated that sea urchin DNA is twice as methylated at the pluteus stage than at the morula stage [54.Adams R.L.P. Delayed methylation of DNA in developing sea urchin embryos.Nat. New Biol. 1973; 244: 27-29Crossref PubMed Google Scholar], in agreement with earlier findings suggesting that methylation in sea urchins does not occur until gastrulation [278.Grippo P. et al.Methylation of DNA in developing sea urchin embryos.J. Mol. Biol. 1968; 36: 195-208Crossref PubMed Google Scholar,279.Comb D.G. Methylation of nucleic acids during sea urchin embryo development.J. Mol. Biol. 1965; 11: 851-855Crossref PubMed Google Scholar]. Adams’ report was the first to quantify such changes at each developmental stage, which led him to revise his previous theory that DNA methylation does not regulate gene expression and to instead propose the new idea that DNA methylation could function to ‘switch off’ genes after contributing to their specific function in early development. Work prior to the 1970s led several scientists to propose formal hypotheses about the function of DNA methylation in eukaryotes. In the late 1960s, Scarano and colleagues observed that 90% of 5mC in sea urchin DNA is found in the CG context and is thus not randomly distributed in DNA, leading them to speculate about a role for 5mC in differentiation [277.Scarano E. et al.The heterogeneity of thymine methyl group origin in DNA pyrimidine isostichs of developing sea urchin embryos.Proc. Natl. Acad. Sci. U. S. A. 1967; 57: 1394-1400Crossref PubMed Google Scholar,278.Grippo P. et al.Methylation of DNA in developing sea urchin embryos.J. Mol. Biol. 1968; 36: 195-208Crossref PubMed Google Scholar]. In 1971, Scarano proposed that spontaneous deamination of 5mC, which generates a C→T conversion, could lead to heritable changes in the DNA sequence. A popular theory from the 1960s up until the early 1980s was that sequence mutations in genes direct differentiation. Scarano therefore speculated that 5mC-guided mutation could direct cellular differentiation during embryogenesis [55.Scarano E. The control of gene function in cell differentiation and in embryogenesis.Adv. Cytopharmacolo. 1971; 1: 13-24PubMed Google Scholar]. In the same year, Adams’ work on 5mC patterning following DNA replication in mouse fibroblasts revealed that early replicating DNA is methylated quickly, while late replicating DNA takes several hours to become fully methylated. The observation that active DNA methylation occurs predominantly in S-phase led Adams to conclude that 5mC must not play a role in controlling transcription [53.Adams R.L.P. The relationship between synthesis and methylation of DNA in mouse fibroblasts.Biochim. Biophys. Acta. 1971; 254: 205-212Crossref PubMed Google Scholar]. His conclusions might also have been influenced by the bacterial studies by Billen and Lark that implicated 5mC as a regulator of DNA replication [45.Billen D. Methylation of the bacterial chromosome: an event at the “replication point”?.J. Mol. Biol. 1968; 31: 477-486Crossref PubMed Google Scholar,46.Lark C. Studies on the in vivo methylation of DNA in Escherichia coli 15T.J. Mol. Biol. 1968; 31: 389-399Crossref PubMed Google Scholar]. Interestingly, in 1972 Comings came to a different conclusion looking at Chinese hamster ovarian cells, where he found that late replicating AT-rich DNA is undermethylated to a greater extent than would be expected from its base composition, while early replicating GC-rich DNA is highly methylated. Comings speculated that if DNA methylation