Forum
On the meaning of the word ‘epimutation’ H. Oey and E. Whitelaw La Trobe Institute for Molecular Science, Melbourne, VIC 3086, Australia
The word ‘epimutation’ is often used in a manner that can be misinterpreted. The strict definition of epimutation is a heritable change in gene activity that is not associated with a DNA mutation but rather with gain or loss of DNA methylation or other heritable modifications of chromatin. Unfortunately, there is a growing tendency in the cancer field to use the word in situations in which underlying DNA sequence changes have occurred. Epigenetic changes in cancer are well documented, often extensive, and frequently associated with genes involved in cancer progression [1]. In contemporary literature these epigenetic changes are sometimes referred to as epimutations. The term ‘epimutation’ was defined as ‘a heritable change in gene activity due to DNA modification’, not mutation, with the implication that the heritable change is aberrant [2]. By definition, the change in expression is not associated with a change in the DNA sequence, either in cis or in trans. This implies that the heritability is due to the mitotic transmission of an epigenetic mark rather than transmission of DNA variants. However, the word is sometimes used in a broader sense to include epigenetic changes driven by DNA mutation in cis or in trans, but in these cases modifying clauses or adjectives need to be added. A prime example of this can be seen in the recent study ‘Targeted p16Ink4a epimutation causes tumorigenesis and reduces survival in mice’ [3], where the authors have used the word epimutation in a manner that could be misinterpreted. While this study shows a causative link between transcriptional silencing at p16 and tumorigenesis, the process is driven by the insertion of a fragment of DNA that is likely to provide binding sites for factors involved in transcriptional repression and this fact is rarely alluded to in the text. The authors report a concomitant change in the DNA methylation state of the p16 promoter. Strictly speaking, the methylation change could be a consequence of transcriptional silencing rather than the driver of it, but either way it is not the initiating event, which is a DNA mutation (the insertion of a fragment of DNA). Horsthemke [4,5] separates epimutations into two types, primary and secondary. The former are those that occur in the absence of any DNA sequence change while the latter are those that occur secondary to a DNA mutation in a cis- or trans-acting factor. According to these definitions, the situation reported by Yu and colleagues [3] would classify as a secondary epimutation. Corresponding author: Whitelaw, E. ([email protected]). 0168-9525/ ß 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tig.2014.08.005
Epimutations have also been described as either ‘constitutional’ or ‘somatic’. The former are defined as being derived from the germline and should be present in all of the tissues of an individual and the latter arise in cells in a somatic tissue [6]. These definitions were coined to explain differing zygosity of epimutations in tumour versus adjacent non-tumour tissue of an individual. In a clinical setting, this is often all one knows about the epigenetic state. Both types of epimutation can predispose to disease and both could be primary or secondary, according to Horsthemke’s definition. It is likely that most reported epimutations in cancer represent secondary epimutations; DNA sequence aberrations leading to effects (in cis or in trans) on epigenetic marks. Only a few well documented examples of primary epimutation exist in the cancer field [7,8] because in tumour tissue it is virtually impossible to exclude functionally linked DNA mutations elsewhere in the genome. The best-described primary epimutations are outside the field of cancer and are related to aberrant transcription at imprinted genes. Imprinting, which is the parent of origin-specific inactivation of an allele via epigenetic silencing, occurs at a small number of genes in the genome. Improper silencing at these loci can cause disease, such as Angelman syndrome or Prader–Willi syndrome, and in some cases no genetic aberrations can be detected [9]. The well-studied agouti viable yellow (Avy) phenotype in mice can also be considered an example of a primary epimutation. In Avy mice, the gene agouti is aberrantly expressed due to the epigenetic status of a retrotransposable element upstream of the coding sequence. When the element is unmethylated, constitutive expression of agouti is associated with a mouse with yellow coat colour, and when the element is methylated, wild type expression occurs resulting in a pseudoagouti (brown) coat colour. The epigenetic status is set in the early embryo in a stochastic manner, independent of any known DNA sequence change, and is inherited mitotically. Loci of this nature are more commonly referred to as metastable epialleles because the variability in epigenetic state at the locus is common rather than rare [10]. In other words, the epigenetic variability at these genes is normal, not abnormal. While the DNA methylation change engineered by Yu and colleagues [3] does satisfy the definition of a secondary epimutation, it is important to clarify that it does not demonstrate an aberrant epigenetic event initiating the onset of cancer; rather, the cancer is driven by a DNA mutation acting in cis. The epigenetic heritability of these marks also cannot be demonstrated as they are inextricably linked to a heritable transgene. According to the constitutional versus somatic definition, this would be Trends in Genetics, December 2014, Vol. 30, No. 12
519
Forum considered constitutional because it would be present in all tissues. The issue of whether ‘true’ epigenetic events cause cancer continues to be controversial. This is more than a semantic argument; it matters because it will impact on the course of treatment. An epigenetic event implies reversibility, while a DNA mutation does not. Some drugs, such as 5-azacytidine, can decrease methylation states across the genome [11], but if an underlying DNA mutation caused the original epigenetic change (i.e., a secondary epimutation) such treatment would have to be continual. It is likely that the effect described by Yu and colleagues [3] is due to gene silencing caused by the recruitment of endogenous DNA-binding factors to the foreign DNA sequence inserted proximal to the p16 promoter. While these involve epigenetic mechanisms, they are driven by this insertional event. A similar mechanism was demonstrated recently by Serra and colleagues [12]. They showed that the promoter hypermethylation of some genes associated with the CpG island methylator phenotype (CIMP) in one form of colorectal cancer is dependent on aberrant upregulation of a transcription factor that in turn recruits transcriptional repressors to these promoters. Epimutation is a word that means different things to different people, so it should be used by authors with care. Because of the difficulty in determining the cause of epigenetic changes in cancer, we suggest that authors refrain from using this word when describing changes in an epigenetic state in these situations. Importantly, there remains little evidence in the current literature for true, or primary, epimutations driving the onset of cancer and
520
Trends in Genetics December 2014, Vol. 30, No. 12
the aberrant epigenetic events observed are typically secondary. The incorrect use of the word by cancer biologists misleads those less familiar with the field. Importantly, this is likely to include clinicians, who are at the front line of cancer care. References 1 Plass, C. et al. (2013) Mutations in regulators of the epigenome and their connections to global chromatin patterns in cancer. Nat. Rev. Genet. 14, 765–780 2 Holliday, R. (1986) The inheritance of epigenetic defects. Science 238, 163–170 3 Yu, D.H. et al. (2014) Targeted p16Ink4a epimutation causes tumorigenesis and reduces survival in mice. J. Clin. Invest. 124, 3708–3712 4 Horsthemke, B. (2006) Epimutations in human disease. Curr. Top. Microbiol. Immunol. 310, 45–59 5 Speicher, M.R. et al., eds (2010) Vogel and Motulsky’s Human Genetics. Problems and Approaches, Springer 6 Hitchins, M.P. and Ward, R.L. (2009) Constitutional (germline) MLH1 epimutation as an aetiological mechanism for hereditary non-polyposis colorectal cancer. J. Med. Genet. 46, 793–802 7 Hitchins, M.P. et al. (2007) Inheritance of a cancer-associated MLH1 germ-line epimutation. N. Engl. J. Med. 356, 697–705 8 Ward, R.L. et al. (2012) Identification of constitutional MLH1 epimutations and promoter variants in colorectal cancer patients from the Colon Cancer Family Registry. Genet. Med. 15, 25–35 9 Buiting, K. et al. (2003) Epimutations in Prader–Willi and Angelman syndromes: a molecular study of 136 patients with an imprinting defect. Am. J. Hum. Genet. 72, 571–577 10 Rakyan, V.K. et al. (2002) Metastable epialleles in mammals. Trends Genet. 18, 348–351 11 Mummaneni, P. and Shord, S.S. (2014) Epigenetics and oncology. Pharmacotherapy 34, 495–505 12 Serra, R.W. et al. (2014) A KRAS-directed transcriptional silencing pathway that mediates the CpG island methylator phenotype. Elife 3, e02313
On the meaning of the word ‘epimutation’ H. Oey and E. Whitelaw La Trobe Institute for Molecular Science, Melbourne, VIC 3086, Australia
The word ‘epimutation’ is often used in a manner that can be misinterpreted. The strict definition of epimutation is a heritable change in gene activity that is not associated with a DNA mutation but rather with gain or loss of DNA methylation or other heritable modifications of chromatin. Unfortunately, there is a growing tendency in the cancer field to use the word in situations in which underlying DNA sequence changes have occurred. Epigenetic changes in cancer are well documented, often extensive, and frequently associated with genes involved in cancer progression [1]. In contemporary literature these epigenetic changes are sometimes referred to as epimutations. The term ‘epimutation’ was defined as ‘a heritable change in gene activity due to DNA modification’, not mutation, with the implication that the heritable change is aberrant [2]. By definition, the change in expression is not associated with a change in the DNA sequence, either in cis or in trans. This implies that the heritability is due to the mitotic transmission of an epigenetic mark rather than transmission of DNA variants. However, the word is sometimes used in a broader sense to include epigenetic changes driven by DNA mutation in cis or in trans, but in these cases modifying clauses or adjectives need to be added. A prime example of this can be seen in the recent study ‘Targeted p16Ink4a epimutation causes tumorigenesis and reduces survival in mice’ [3], where the authors have used the word epimutation in a manner that could be misinterpreted. While this study shows a causative link between transcriptional silencing at p16 and tumorigenesis, the process is driven by the insertion of a fragment of DNA that is likely to provide binding sites for factors involved in transcriptional repression and this fact is rarely alluded to in the text. The authors report a concomitant change in the DNA methylation state of the p16 promoter. Strictly speaking, the methylation change could be a consequence of transcriptional silencing rather than the driver of it, but either way it is not the initiating event, which is a DNA mutation (the insertion of a fragment of DNA). Horsthemke [4,5] separates epimutations into two types, primary and secondary. The former are those that occur in the absence of any DNA sequence change while the latter are those that occur secondary to a DNA mutation in a cis- or trans-acting factor. According to these definitions, the situation reported by Yu and colleagues [3] would classify as a secondary epimutation. Corresponding author: Whitelaw, E. ([email protected]). 0168-9525/ ß 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tig.2014.08.005
Epimutations have also been described as either ‘constitutional’ or ‘somatic’. The former are defined as being derived from the germline and should be present in all of the tissues of an individual and the latter arise in cells in a somatic tissue [6]. These definitions were coined to explain differing zygosity of epimutations in tumour versus adjacent non-tumour tissue of an individual. In a clinical setting, this is often all one knows about the epigenetic state. Both types of epimutation can predispose to disease and both could be primary or secondary, according to Horsthemke’s definition. It is likely that most reported epimutations in cancer represent secondary epimutations; DNA sequence aberrations leading to effects (in cis or in trans) on epigenetic marks. Only a few well documented examples of primary epimutation exist in the cancer field [7,8] because in tumour tissue it is virtually impossible to exclude functionally linked DNA mutations elsewhere in the genome. The best-described primary epimutations are outside the field of cancer and are related to aberrant transcription at imprinted genes. Imprinting, which is the parent of origin-specific inactivation of an allele via epigenetic silencing, occurs at a small number of genes in the genome. Improper silencing at these loci can cause disease, such as Angelman syndrome or Prader–Willi syndrome, and in some cases no genetic aberrations can be detected [9]. The well-studied agouti viable yellow (Avy) phenotype in mice can also be considered an example of a primary epimutation. In Avy mice, the gene agouti is aberrantly expressed due to the epigenetic status of a retrotransposable element upstream of the coding sequence. When the element is unmethylated, constitutive expression of agouti is associated with a mouse with yellow coat colour, and when the element is methylated, wild type expression occurs resulting in a pseudoagouti (brown) coat colour. The epigenetic status is set in the early embryo in a stochastic manner, independent of any known DNA sequence change, and is inherited mitotically. Loci of this nature are more commonly referred to as metastable epialleles because the variability in epigenetic state at the locus is common rather than rare [10]. In other words, the epigenetic variability at these genes is normal, not abnormal. While the DNA methylation change engineered by Yu and colleagues [3] does satisfy the definition of a secondary epimutation, it is important to clarify that it does not demonstrate an aberrant epigenetic event initiating the onset of cancer; rather, the cancer is driven by a DNA mutation acting in cis. The epigenetic heritability of these marks also cannot be demonstrated as they are inextricably linked to a heritable transgene. According to the constitutional versus somatic definition, this would be Trends in Genetics, December 2014, Vol. 30, No. 12
519
Forum considered constitutional because it would be present in all tissues. The issue of whether ‘true’ epigenetic events cause cancer continues to be controversial. This is more than a semantic argument; it matters because it will impact on the course of treatment. An epigenetic event implies reversibility, while a DNA mutation does not. Some drugs, such as 5-azacytidine, can decrease methylation states across the genome [11], but if an underlying DNA mutation caused the original epigenetic change (i.e., a secondary epimutation) such treatment would have to be continual. It is likely that the effect described by Yu and colleagues [3] is due to gene silencing caused by the recruitment of endogenous DNA-binding factors to the foreign DNA sequence inserted proximal to the p16 promoter. While these involve epigenetic mechanisms, they are driven by this insertional event. A similar mechanism was demonstrated recently by Serra and colleagues [12]. They showed that the promoter hypermethylation of some genes associated with the CpG island methylator phenotype (CIMP) in one form of colorectal cancer is dependent on aberrant upregulation of a transcription factor that in turn recruits transcriptional repressors to these promoters. Epimutation is a word that means different things to different people, so it should be used by authors with care. Because of the difficulty in determining the cause of epigenetic changes in cancer, we suggest that authors refrain from using this word when describing changes in an epigenetic state in these situations. Importantly, there remains little evidence in the current literature for true, or primary, epimutations driving the onset of cancer and
520
Trends in Genetics December 2014, Vol. 30, No. 12
the aberrant epigenetic events observed are typically secondary. The incorrect use of the word by cancer biologists misleads those less familiar with the field. Importantly, this is likely to include clinicians, who are at the front line of cancer care. References 1 Plass, C. et al. (2013) Mutations in regulators of the epigenome and their connections to global chromatin patterns in cancer. Nat. Rev. Genet. 14, 765–780 2 Holliday, R. (1986) The inheritance of epigenetic defects. Science 238, 163–170 3 Yu, D.H. et al. (2014) Targeted p16Ink4a epimutation causes tumorigenesis and reduces survival in mice. J. Clin. Invest. 124, 3708–3712 4 Horsthemke, B. (2006) Epimutations in human disease. Curr. Top. Microbiol. Immunol. 310, 45–59 5 Speicher, M.R. et al., eds (2010) Vogel and Motulsky’s Human Genetics. Problems and Approaches, Springer 6 Hitchins, M.P. and Ward, R.L. (2009) Constitutional (germline) MLH1 epimutation as an aetiological mechanism for hereditary non-polyposis colorectal cancer. J. Med. Genet. 46, 793–802 7 Hitchins, M.P. et al. (2007) Inheritance of a cancer-associated MLH1 germ-line epimutation. N. Engl. J. Med. 356, 697–705 8 Ward, R.L. et al. (2012) Identification of constitutional MLH1 epimutations and promoter variants in colorectal cancer patients from the Colon Cancer Family Registry. Genet. Med. 15, 25–35 9 Buiting, K. et al. (2003) Epimutations in Prader–Willi and Angelman syndromes: a molecular study of 136 patients with an imprinting defect. Am. J. Hum. Genet. 72, 571–577 10 Rakyan, V.K. et al. (2002) Metastable epialleles in mammals. Trends Genet. 18, 348–351 11 Mummaneni, P. and Shord, S.S. (2014) Epigenetics and oncology. Pharmacotherapy 34, 495–505 12 Serra, R.W. et al. (2014) A KRAS-directed transcriptional silencing pathway that mediates the CpG island methylator phenotype. Elife 3, e02313
