sign in sign up
<<
Home , Epigenetics

ISSN 0233–7657. Biopolymers and . 2014. Vol. 30. N 1. P. 3–9 doi: http://dx.doi.org/10.7124/bc.000873

REVIEWS

UDC 577 : What it is about?

E. Saade1, V. V. Ogryzko2

1Faculty of Public Health, Lebanese University, P.O. Box 6573/14 Badaro, Museum, Beirut, Lebanon 2CNRS UMR8126, Paris-Sud University, Gustave Roussy Institute 114, rue Edouard Vaillant, 94805 Villejuif Cedex-France [email protected]

Epigenetics has captured the attention of scientists in the past decades, yet its scope has been continuously chan- ging. In this paper, we give an overview on how and why its definition has evolved and suggest several clarifica- tion on the concepts used in this field. Waddington coined the term in 1942 to describe interaction with each other and with their environment and insisted on dissociating these events from . Then, Holliday and others argued that epigenetic phenomena are characterized by their . However, differentiated cells can maintain their for decades without undergoing division, which points out the limitation of the «herita- bility» criterion for a particular phenomenon to qualify as epigenetic. «Epigenetic stability» encompasses traits preservation in both dividing and non dividing cells. Likewise, the use of the term «epigenetic regulation» has been misleading as it overlaps with «regulation of expression», whereas «epigenetic information» clearly distin- guishes epigenetic from genetic phenomena. Consequently, how could epigenetic information be transmitted and perpetuated? The term «epigenetic templating» has been proposed to refer to a general mechanism of perpetua- tion of epigenetic information that is based on the preferential activity of enzymes that deposit a particular epi- genetic mark on macromolecular complexes already containing the same mark. Another issue that we address is the role of epigenetic information. Not only it is important in allowing alternative interpretations of genetic infor- mation, but it appears to be important in protecting the , as can be illustrated by bacterial endonucleases that targets non methylated DNA – i. e. foreign DNA – and not the endogenous methylated DNA. Keywords: , heritability, central Dogma, DNA , genetic switch, .

Does epigenetics contradict the Central Dogma of rent cell types from the same organism can be taken from Molecular ? Epigenetic phenomena represent an organism and propagated in a , in identi- a topic of active research in current biology. This field cal environmental conditions. The genome sequence of of study is popular in part because it goes «against the these different cell types is identical (as demonstrated grain» of the Central Dogma of . If un- by the phenomenon of and iPS cell repro- derstood in an oversimplified way, the Central Dogma gramming [1, 2]), yet they exhibit very different and professes a «genocentric» view of a biological system, stable phenotypes. This observation suggests that there according to which all information necessary to define should be additional information responsible for the ma- the state of an organism (up to the due to intenance of the stable differentiated phenotypes. Un- the effects of environment) is contained in the sequence derstanding the and the mechanisms of proces- of its genome. While more or less adequate when one sing and propagation of this information represents one describes simple single-cellular organisms, such as bac- of the major topics in the epigenetic studies. teria, the «genocentric» view runs into problem when one In fact, the Central Dogma, as it was first formula- start to deal with the multicellular organisms that exhi- ted by Francis Crick in 1958 and re-stated in a Nature bit the phenomenon of . Diffe- paper published in 1970 [3] asserts: «The central dogma of molecular biology deals with the detailed residue-  Institute of Molecular Biology and , NAS of Ukraine, 2014 by-residue transfer of sequential information. It states

3 SAADE E., OGRYZKO V. V. that such information cannot be transferred from pro- perimental models for this redefined field of research tein to either or nucleic acid». A careful reading were DNA methylation in [7], variega- of this definition shows that it leaves the door open for tion (PEV) in [8] and silen- other types of information that could be required to cing phenomenon in [9]. specify the state of the organism (e. g. stored in macro- How this change in the meaning of the term «epige- molecular conformations, interactions, post-translatio- netic» can be explained? Most probably it came about as nal modifications and alternative states of genetic net- the result of the shift in the meaning of the term «genetic». works) and might propagate independently from the As originally introduced, Genetics is «science of in- DNA sequence. Thus, despite a conceptual tension bet- heritance» [10]. If one strictly abides by this terminolo- ween the «genocentric» view of biological systems and gy, epigenetics cannot study heritable variations at all, the notion of epigenetic phenomena, the latter is com- simply because by definition they all should be conside- patible with the Central Dogma of Molecular Biology as red as genetic. it was formulated by Crick. However, after the solving of DNA structure by Wat- Why the definition of epigenetics changed? – Na- son and Crick and consequent deciphering of genetic tura abhorret vacuum1. The understanding of what con- code, genetic information became justifiably identified stitutes the subject of epigenetics underwent changes with a particular sequence of on DNA or from the time of its conception. The term «epigenetic» RNA. Accordingly, the meaning of the term «genetic» was coined by Conrad Waddington in 1942 referring to shifted, as all things «genetic» became firmly associa- mechanisms in which genes within an organism inter- ted with the sequences of DNA or RNA. act with each other and their environment to create a phe- On another hand, the phenomenon of stability of dif- notype (the related notion of «» dates back to ferentiated phenotypes (and related phenomena, such Aristotle (On the Generation of Animals)). [4]. One of the PEV and silencing) demonstrates existence and import- main metaphors in this early version of epigenetics was ance of heritable variations that cannot be accounted for one of «epigenetic landscape» that defines different stab- by genetic variations. To fill the conceptual void due to le trajectories that a cell can take during development and the shift in the meaning of the term «genetic», the term differentiation. «epigenetic» changed its meaning as well, motivated by In this picture, genes are responsible for generating the need to describe the class of heritable phenomena various epigenetic landscapes, which still would be con- not encoded in DNA (/RNA) sequence and thus left out- sistent with the «genocentric» view on organisms. How- side of the scope of newly redefined genetics. ever, the main focus of Waddington’s interest was on re- Is heritability essential? Epigenetic stability and lative decoupling of development from genes. A canoni- epigenetic information. The above discussion suggests cal example of such decoupling was the phenomenon of an explanation of how the modern meaning of the term phenocopy [5], i. e., a developmental abnormality that «epigenetic» came about. However, it appears that the re- is phenotypically identical to that caused by genetic mu- quirement of «heritability», currently emphasized as the tations, but induced by factors that do not change geno- criterion for a particular phenomenon to fall under the type, such as heat shock [6]. This phenomenon, together scope of epigenetic research, might be too restrictive. with the reciprocal observation that not every change in Terminally differentiated cells, such as and is reflected in , constituted the ma- muscle cells, live for decades maintaining their distinct in argument against the simple notion of development phenotypic differences in spite of environmental stresses, completely determined by genetic information. thermal noise and DNA damage/repair [11, 12]. These However, nowadays the term «epigenetics» is used cells do not proliferate, thus the term «heritable» does in a different meaning (due originally to Holliday) and not apply to their distinct and stable traits. On the other emphasizes «heritability» as a necessary requirement hand, it is reasonable to expect that maintenance and for a phenomenon to qualify as epigenetic. The first ex- propagation of information responsible for these traits employs mechanisms that are similar to those that are 1Francois Rabelais «Gargantua and Pantagruel’», 1530s. utilized in the replicating cells (e. g., DNA methylation,

4 EPIGENETICS: WHAT IT IS ABOUT? modifications, etc.). Thus, although one cannot to multicellular organisms and their development. How- use the criteria of «heritability», it makes good sense to ever, broadening of the scope of «epigenetic regula- qualify the variations between different cell lineages tion» complicates its demarcation from the general no- (and the underlying mechanisms) as epigenetic. The tion of «regulation of », long used in term «epigenetic stability» refers to a broader phenome- molecular biology and genetics. Genetic circuits and non that encompasses maintenance of phenotypic traits networks had been studied (not necessarily in the con- in both nonreplicating and replicating cells independent text of development) for decades in the field of molecu- from DNA sequence [13, 14]. lar genetics [18]. Genetic interactions (e. g., ) On the other extreme, some authors suggest to un- have been the subject of evolutionary genetics from its derstand epigenetic phenomena in a very broad sense, very beginning (with the studies of evolution of domi- defining them as any kind of inherited traits that do not nance being a prominent example [19]). It appears that exhibit Mendelian behavior [15]. This definition appears the term «epigenetic regulation» currently serves merely to be too loose, as it will consider as epigenetic all bac- to replace «regulation of gene expression», without any terial, organelle and viral genetics, which is based on the added value in the transaction. The introduction of new primary structure of nucleic acid, while does not con- molecular players, such as noncoding RNA, to the list of form to the rules of Mendel. It is therefore advisable to epigenetic mechanisms adds nothing but confusion. restrict the use of term epigenetic to the information A telling illustration of how the term «epigenetic re- that is not contained in nucleic acid sequence. gulation» is encroaching on the turf of classical molecu- Therefore, in what follows, epigenetic information lar genetics is a name changing game that occurred in the will be defined as the information that is required in or- field of bacteriophage lambda. The transition between der to specify the state of an organism in addition to ge- lyzogenic and lytic lifecycles of this phage was original- netic information ( sequence) and reaction ly termed «genetic switch» [20], serving as an early pa- norm. The mechanisms of processing, storage and trans- radigm for regulation of gene expression. However, cur- mission/propagation of epigenetic information are the rently the same phenomenon is often referred to as «epi- subject of molecular epigenetics, a relatively new field genetic switch» [21], also used by Ptashne himself [22], of research. including a mock change of his classic book’s cover in Is the notion of «epigenetic regulation» useful or a recent talk at Pasteur, Paris (EMBO Workshop on the confusing? The value of a scientific concept depends Model and its impact on modern molecular bio- on: (i) how well it can capture a particular class of phe- logy, 17–20 May 2011). Note that, whereas the notions nomena by clearly distinguishing it from other pheno- of genetic and epigenetic information clearly describe mena, (ii) whether it can stimulate new directions of re- distinct aspects of this system (i. e., sequence of nucleoti- search and (iii) whether it allows one to convey compli- des and alternative states of gene activity, respective- cated ideas in a succinct and lucid fashion. ly), the notions of genetic and epigenetic switch (or else Unlike the term «epigenetic information», another control, regulation, or a circuit) would have exactly the widely used notion – that of «epigenetic regulation» (or same meaning. epigenetic control) – does not meet these criteria. Given that chromatin is considered a principal car- The major problem with this notion stems from the rier of epigenetic information, one way to remedy the difficulty in clearly defining its scope. Its original inten- «encroaching problem» would be to limit the scope of tion was to describe regulation of gene expression du- epigenetic regulation to chromatin-based regulation. ring development and differentiation of multicellular However, such clarification does not seem wise, as the organisms (in line with the Waddington’s definition of fields of chromatin and epigenetics are only overlap- epigenetics). After the discovery of parallels between ping, not identical. Not every chromatin modification PEV in drosophila and silencing phenomena in yeast, (histone PTM, DNA methylation, replacement histone) as well as with finding of epigenetic-like phenomena in is heritable (or stable) and thus could serve as an epige- other [16, 17], it became evident that netic mark, and not every epigenetic change is encoded the studies of epigenetic mechanisms cannot be limited in chromatin state (as illustrated by such trans-acting epi-

5 SAADE E., OGRYZKO V. V. genetic factors as alternative states of genetic networks, Me NNCGNNNCGNN , structural/cortical inheritance [16, 18, 23, 24]). NNGCNNNGCNN In contrast to «epigenetic regulation», the scope of Me «epigenetic information» is, by construction, clearly de- fined – namely, as something additional to (and hence, Semiconservative different from) «genetic information». Moreover, after DNAreplication Me distinguishing between the two notions, the time tested NNCGNNNCGNN NNCGNNNCGNN conceptual framework established to study genetic infor- NNGCNNNGCNN NNGCNNNGCNN Me mation can be now transferred to the newer field of epi- Maintenancemethyltransferase genetics, immediately suggesting plethora of questions activityathemimethylated and directions of research. How epigenetic information CpGdinucleotides Me Me is reproduced and transmitted? How is it recorded de no- NNCGNNNCGNN NNCGNNNCGNN vo and how it is read and/or erased? Does the notion of NNGCNNNCGNN NNGCNNNGCNN «epigenetic information damage» make any sense? If so, Me Me Fig. 1. Methylation maintenance. After replication, only semimethyla- can it be repaired and/or are there mechanisms of epige- ted bases are converted into fully methylated netic damage response, such as specialized checkpoints [12]? Further one, consistent with the recognized role of In addition to the clearly defined scope, the value of chromatin as a principal carrier of epigenetic informa- the term «epigenetic information» is evident from the tion, very similar mechanisms have been suggested for new questions it opens and the avenues of research it perpetuation of post-translational histone modifications. stimulates. Some histone acetyl transferase (HAT) complexes could Epigenetic templating – «like draws to like». How be preferentially recruited to acetylated chromatin due epigenetic information can be propagated and maintai- to the presence of bromodomains, recognizing acetyla- ned? Studies of DNA methylation (an earliest recognized ted , whereas the chromodomain containing his- epigenetic mark) provided an early clue on one mecha- tone methyl transferase (HMT) complexes could be like- nism, based on a different behavior of the so called main- wise recruited to methylated chromatin [26–28]. Also, tenance methylase enzyme towards un-methylated ver- domain structure of some protein kinases suggests that sus semi-methylated DNA (Fig. 1). The semiconserva- similar mechanism could operate not only on the level tive DNA replication of a methylated double strand gi- of chromatin, but in controlling the organization of cy- ves rise to semimethylated bases comprising a parental toplasm as well. Tyrosine kinases often contain SH2 do- strand and a newly synthesized one. mains, which recognize phosphotyrosine, whereas some enzymes bind to the semimethylated sites where they serine/threonine kinases contain FHA domains that re- methylate the new strand. The modifying machinery is cognize phosphoserine/phosphothreonine. If some of not recruited to un-methylated strands. the targets of these kinases form dimers (or oligomers), Not surprisingly, this mechanism differs from that their status can be perpetuated according of replication of genetic information. Instead of Wat- to the Crick’s original proposal. son-Crick base complementarity, it is rather based on the To emphasize the difference of this way to propaga- ancient principle «like draws to like». Later on, Francis te information from the canonical Watson-Crick base- Crick proposed a similar mechanism as the molecular base complementarity mechanism, Vasily Ogryzko’s basis for neurobiological , essentially an epige- group used the term «epigenetic templating» [13, 14] to netic phenomenon [25]. He postulated that (i) the strength refer to a general mechanism of perpetuation of epige- of a synapse is determined by phosphorylation of a pro- netic information that is based on the preferential activi- tein molecule, (ii) this protein can form dimers (or oligo- ty of enzymes that deposit a particular epigenetic mark mers) and (iii) the kinase responsible for the modifica- on macromolecular complexes already containing the tion will only modify monomer in a dimer that has se- same mark (Fig. 2). They tested whether this model can cond monomer already phosphorylated. also apply to variant , putative epigenetic marks

6 EPIGENETICS: WHAT IT IS ABOUT? on chromatin different from post-translational histone D modifications. R What are the roles of epigenetic information and their evolutionary relations? Why needs epige- T netic information? As was implied above, one of its ro- les is in facilitating maintenance of differentiated cell phenotypes during development of multicellular orga- nisms, that is, in supporting alternative interpretations of genetic information. However, one might argue that from evolutionary perspective, a more fundamental and primary role could have been protection of genetic in- I II III formation. Fig. 2. Scheme of epigenetic templating [14]. In the case I, the modify- An important aspect of functioning of genetic infor- ing machinery is recruited to sites exhibiting a particular mark and the- mation is its formal («mechanical») character. The rep- refore it amplifies the information. In the case II, if the mark is absent, the machinery is not recruited. In III, the recruitment occurs but no addi- lication, , recombination and other enzyma- tional modification takes place because target sites are already modified tic machineries are designed to treat all nucleotide sequ- ences equally, regardless of their meaning, i. e., of what the mismatch repair machinery to correctly identify do they . However, this «equal treatment» allows which base of a mismatch pair has to be removed. viruses and other parasitic genetic elements to take ad- The mismatch repair mechanism illustrates how the vantage of the cellular resources and propagate their own notion of «epigenetic information» allows one to convey genetic information. In order to protect themselves, cells a complicated idea in an economical fashion, demonstra- have developed ways to label their own genetic informa- ting an additional value of this concept. Indeed, the dis- tion so that it could be recognized as their own and distin- crimination between the parental and daughter strands guished from the foreign one. One can consider an exam- can work only if there is a time window when the semi- ple of restriction-methylation system, a defense mecha- methylated DNA has not yet been converted to the fully nism developed by bacteria. This system contains a rest- methylated state. This feature can be formulated in the riction endonuclease and a methylase enzymes that can following elegant way: the increased accuracy of repli- target the same DNA sequence, depending on its methy- cation of genetic information in bacteria is ensured by lated status (Fig. 2). A foreign DNA invading the cell is the difference between the rates of replication of gene- not methylated and thus becomes a target of the endonuc- tic and epigenetic information. Incidentally, this makes lease. However, the same target sequence present origi- mismatch repair an example of a general kinetic proof- nally in the cell is fully methylated, and the methylated reading scheme [29]. state is maintained after replication due to the action of In the above examples, epigenetic information acts maintenance methylase (as described in the previous as a part of protection system serving to differentiate the section). Thus, cellular machinery that helps it to pro- «self» from the «other» and the «old» from the «new». tect its own genetic information is based on recognition Thus, at least two distinct roles of epigenetic informa- of epigenetic information associated with DNA. tion can be distinguished: interpretation of genetic infor- Moreover, the same mechanism of replication of mation and its protection. What could be the evolutio- methylated status of DNA serves additional protective nary relation between these two functions? It seems re- purpose, by helping bacteria to prevent errors during asonable to propose that the protection of genetic infor- replication of genetic information. In order to correctly mation had emerged first in evolution, since it would be remove the erroneously incorporated nucleotide, mis- beneficial already for the single cellular organisms, which match repair mechanisms need to recognize the pa- typically have much less need for cell differentiation. rental strand of the newly duplicated DNA. While the Only later, epigenetic mechanisms could be recruited to newly replicated DNA is in the semimethylated state, play a role in cell differentiation. Supporting this idea is the parental strand is labeled with methyl, which allows the evidence that many epigenetic mechanisms appear to

7 SAADE E., OGRYZKO V. V.

Funding. This work was supported by «La Ligue Eco RI methylase Contre le » (9ADO1217/1B1-BIOCE), the «Insti- tut National du Cancer» (247343/1B1-BIOCE) and ARC Unmethylated 5' GAA* TTC 3' foundation (SFI20121205936). DNA 3' CTTAAG 5' * Acknowledgements. The authors thank Drs. Marc Restriction Lipinski, Murat Saparbaev and Alexander Ishchenko for ensyme Eco RIwillnot many fruitful discussions. Eco RI cleavemethylated DNA CH3 Що ж таке епіге не ти ка? Methylated 5' GAATTC 3' Е. Саад, В. В. Огриз ко DNA 3' CTTAAG 5' Ре зю ме CH3 Fig. 3. Restriction endonucleases are a natural part of the bacterial de- Епіге не ти ка при вер тає ува гу вче них уже декілька де ся тиліть, fense system але зна чен ня вкла де но го в неї сен су постійно змінюється. Ми об - го во рюємо при чи ни та ких змін і про по нуємо декількя по яс ню валь - них ко мен тарів. Уод дин гтон увів термін «епіге не ти ка» в 1942 ро- be related to mechanisms of suppression of parasitic ge- ці для опи сан ня взаємодії генів, важ ли вих для роз вит ку організму, netic elements, as illustrated by RNA-directed DNA me- між со бою та з на вко лишнім се ре до ви щем. На далі Холлідей та thylation [30, 31], the tendency of eukaryotic cells to he- інші на по ля га ли на по нятті «успад ко ву ва ності» як не обхідній ха- ракте рис тиці епіге не тич них явищ. Однак ди фе ренційо вані кліти- terochromatinize tandem repeats, in either RNA depen- ни зберіга ють свої фе но ти пи про тя гом де ся тиліть і при цьо му dent or independent way [32, 33], etc. This observation не ділять ся, що вка зує на об ме женість «успад ко ву ва ності» як allows us to suggest that the epigenetic machinery was кри терію того, щоб пев не яви ще роз гля да ти як епіге не тич не. «Епіге не тич на стабільність» за про по но ва на як більш за галь не established by Life for «genome protection» purposes по нят тя, яке озна чає збе ре жен ня ха рак те рис тик в обох кліти - first. In the further course of evolution, however, the нах: які ділять і не ділять ся. З іншо го боку, термін «епіге не тич на mechanisms of processing of epigenetic information – ре гу ляція» при зво дить до плу та ни ни, оскільки його зна чен ня сут- тєво пе ре кри вається або навіть повністю збігається за змістом which allow recognition of different epigenetic marks з ви ра зом «ре гу ляція експресії генів», тоді як «епіге не тич на ін- and channeling the signals encoded in these marks along формація» чітко роз ме жо вує епіге не тичні і ге не тичні яви ща. І по- стає пи тан ня, яким чи ном епіге не тич на інфор мація може від- appropriate response pathways – were recruited for other тво рю ва ти ся? Ми за про по ну ва ли термін «epigenetic templating» purposes, such as for stabilizing different alternative sta- для виз на чен ня ме ханізму відтво рен ня епіге не тич ної інфор мації, tes of the same organism. за сно ва но го на тому, що фер менті, які став лять пев ну епіге не - тич ну мітку, відда ють пе ре ва гу мак ро мо ле ку ляр ним суб стра - Conclusion. With the ongoing progress in chroma- там, які вже містать подібну мітку. На решті ми тор каємось tin research, the development of iPS cell technologies and пи тан ня щодо ролі епіге не тич ної інфор мації. Вона потрібна не the emergence of evo-devo paradigm in evolutionary лише для аль тер на тив них інтер пре тацій ге не тич ної інфор мації, але й для за хис ту ге но му, як це проілюс тро ва но нами на при кладі studies, research in epigenetics remains on the forefront бак терійних ен до нук ле аз, які ата ку ють не ме тиль о ва ну (тоб то of modern biology. As it happens with fast-developing чу жорідну) ДНК і не по шкод жу ють ме тиль о ва ної (тоб то влас - fields, the scope of epigenetics tends to widen. An incre- ної) ДНК. Ключові сло ва: хро ма тин, спад ковість, цен траль на дог ма, ме- asing number of molecular, cell and evolutionary biolo- тилю ван ня ДНК, ге не тич ний пе ре ми кач, ево люція. gists become motivated to position their research as epi- genetic, both to keep up with the fashion and to have a Что же та кое эпи ге не ти ка? better chance in funding or publishing in trendy journals. Э. Саад, В. В. Огрыз ко The many various views on what constitutes the proper scope of epigenetics call for development of a unified Ре зю ме conceptual framework for this field. We hope that the Эпи ге не ти ка при влек a ет вни ма ние уче ных уже не сколь ко де сяти- ле тий, но зна че ние вкла ды ва е мо го в нее смыс ла по сто ян но ме ня - proposed clarifications of the notions of «epigenetic sta- ет ся. Мы об суж да ем при чи ны та ких из ме не ний и пред ла га ем не- bility», «epigenetic information» and «epigenetic temp- сколько по яс ня ю щих ком мен та ри ев. Уод дин гтон ввел тер мин «эпи- lating», as well as our discussion of changing roles of гене ти ка» в 1942 году для опи са ния вза и мо де йствия ге нов, важ - ных для раз ви тия орга низ ма, друг с дру гом и с окру жа ю щей сре - epigenetic mechanisms in evolution will contribute into дой. Поз же Хол ли дей и дру гие на ста и ва ли на по ня тии «на сле дуе- this worthy endeavor. мос ти» как не об хо ди мой ха рак те рис ти ке эпи ге не ти чес ких яв ле -

8 EPIGENETICS: WHAT IT IS ABOUT?

ний. Одна ко диф фе рен ци ро ван ные клет ки со хра ня ют свои фе но - 13. Viens A, Mechold U, Brouillard F, Gilbert C, Leclerc P, Ogryz- ти пы на про тя же нии де ся ти ле тий и при этом не де лят ся, что ko V. Analysis of human histone H2AZ deposition in vivo argues указы ва ет на огра ни чен ность «на сле ду е мос ти» как кри те рия то- against its direct role in epigenetic templating mechanisms. Mol го, что бы опре де лен ное яв ле ние рас смат ри вать как эпи ге не ти - Cell Biol. 2006; 26(14):5325–35. чес кое. «Эпи ге не ти чес кая ста биль ность» пред ло же на как бо лее 14. Ogryzko VV. Erwin Schroedinger, Francis Crick and epigenetic об щее по ня тие, озна ча ю щее со хра не ние ха рак те рис тик в об е их stability. Biol Direct. 2008; 3:15. де ля щих ся и не де ля щих ся клет ках. С дру гой сто ро ны, тер мин 15. Lederberg J. The meaning of epigenetics. The Scientist. 2001; «эпи ге не ти чес кая ре гу ля ция» ве дет к пу та ни це, по сколь ку его зна- 15(18):6. чение силь но пе ре кры ва ет ся или даже со впа да ет по смыс лу с вы - 16. Beisson J, Sonneborn TM. Cytoplasmic inheritance of the orga- ра же ни ем «ре гу ля ция экс прес сии ге нов», в то вре мя как «эпи ге не- nization of the cell cortex in aurelia. Proc Natl тичес кая ин фор ма ция» чет ко раз гра ни чи ва ет эпи ге не ти чес кие и Acad Sci USA. 1965; 53(2):275–82. ге нетичес кие яв ле ния. Да лее сле ду ет воп рос, ка ким об ра зом эпи - 17. Kleine Borgmann LA, Hummel H, Ulbrich MH, Graumann PL. ге нетичес кая ин фор ма ция мо жет вос про из во дить ся? Мы пред ло - SMC condensation centers in Bacillus subtilis are dynamic struc- жи ли тер мин «epigenetic templating» для об озна че ния ме ха низ ма tures. J Bacteriol. 2013; 195(10):2136–45. воспро из ве де ния эпи ге не ти чес кой ин фор ма ции, осно ван но го на 18. Kauffman SA. Metabolic stability and epigenesis in randomly том, что фер мен ты, ста вя щие опре де лен ную эпи ге не ти чес кую constructed genetic nets. J Theor Biol. 1969; 22(3):437–67. мет ку, пред поч та ют мак ро мо ле ку ляр ные суб стра ты, уже со дер- 19. Fisher RA. The possible modification of the response of the wild type to recurrent . Am Nat. 1928; 62(679):115–26. жащие по до бную мет ку. На ко нец, мы ка са ем ся воп ро са о роли эпи- rd ге не ти чес кой ин фор ма ции. Она нуж на не толь ко для аль тер на - 20. Ptashne M. A genetic switch: Phage lambda revisited (3 ed.). тив ных интер пре та ций ге не ти чес кой ин фор ма ции, но и для за щи - Cold Spring Harbor: Cold Harbor Spring Lab. Press; 2004. ты ге но ма, как это про ил люс три ро ва но нами на при ме ре бак те- 21. Zhu XM, Yin L, Hood L, Ao P. Calculating biological behaviors риальных эн до нук ле аз, ата ку ю щих не ме ти ли ро ван ную (то есть of epigenetic states in the phage lambda life cycle. Funct Integr чу же род ную) ДНК и не по вреж да ю щих ме ти ли ро ван ной (то есть . 2004; 4(3):188–95. сво ей) ДНК. 22. Ptashne M. Binding reactions: epigenetic switches, signal trans- Клю че вые сло ва: хроматин, наследствен ность, централь ная дог- duction and cancer. Curr Biol. 2009; 19(6):R234–41. 23. Halfmann R, Jarosz DF, Jones SK, Chang A, Lancaster AK, Lind- ма, ме ти ли ро ва ние ДНК, ге не ти чес кий пе ре клю ча тель, эво лю ция. quist S. Prions are a common mechanism for phenotypic inheri- tance in wild . Nature. 2012; 482(7385):363–8. REFERENCES 24. Holmes DL, Lancaster AK, Lindquist S, Halfmann R. Heritable remodeling of yeast multicellularity by an environmentally respon- 1. Gurdon JB, Elsdale TR, Fischberg M. Sexually mature indivi- sive . Cell. 2013; 153(1):153–65. duals of laevis from the transplantation of single soma- 25. Crick F. Memory and molecular turnover. Nature. 1984; 312 tic nuclei. Nature. 1958; 182(4627):64–5. (5990):101. 2. Takahashi K, Yamanaka S. Induction of pluripotent stem cells 26. Owen DJ, Ornaghi P, Yang JC, Lowe N, Evans PR, Ballario P, from mouse embryonic and adult fibroblast cultures by defined Neuhaus D, Filetici P, Travers AA. The structural basis for the re- factors. Cell. 2006; 126(4):663–76. cognition of acetylated histone H4 by the bromodomain of histo- 3. Crick F. Central dogma of molecular biology. Nature. 1970; 227 ne acetyltransferase gcn5p. EMBO J. 2000; 19(22):6141–9. (5258):561–3. 27. Jenuwein T, Allis CD. Translating the . Science. 2001; 293(5532):1074–80. 4. Waddington CH. The epigenotype. Endeavour 1942; 1:18–20. 28. Ogryzko VV. Mammalian histone acetyltransferases and their 5. Goldschmidt R. Gen und ausseneigenschaft. I Zeitschr ind Abstl. complexes. Cell Mol Life Sci. 2001; 58(5–6):683–92. 1935; 69:38–69. 29. Hopfield JJ. Kinetic proofreading: a new mechanism for redu- 6. Waddington CH, Pantelouris EM. Transplantation of nuclei in cing errors in biosynthetic processes requiring high specificity. newt's eggs. Nature. 1953; 172(4388):1050–1. Proc Natl Acad Sci USA. 1974; 71(10):4135–9. 7. Bird AP. CpG-rich islands and the of DNA methyla- 30. Bender J. RNA-directed DNA methylation: getting a grip on me- tion. Nature. 1986; 321(6067):209–13. chanism. Curr Biol. 2012; 22(10):R400–1. 8. Wallrath LL, Elgin SC. Position effect variegation in Drosophila 31. Wassenegger M, Heimes S, Riedel L, Sanger HL. RNA-directed is associated with an altered chromatin structure. Genes Dev. de novo methylation of genomic sequences in plants. Cell. 1994; 1995; 9(10):1263–77. 76(3):567–76. 9. Grewal SI, Klar AJ. Chromosomal inheritance of epigenetic sta- 32. Martienssen RA. Maintenance of by RNA inter- tes in fission yeast during and . Cell. 1996; 86 ference of tandem repeats. Nat Genet. 2003; 35(3):213–4. (1):95–101. 33. Wang F, Koyama N, Nishida H, Haraguchi T, Reith W, Tsuka- 10. Mayr E. The growth of biological thought: Diversity, evolution, moto T. The assembly and maintenance of heterochromatin ini- and inheritance. Cambridge: Belknap Press, 1982. tiated by transgene repeats are independent of the RNA interfe- 11. Green CM, Almouzni G. When repair meets chromatin. First in rence pathway in mammalian cells. Mol Cell Biol. 2006; series on chromatin dynamics. EMBO Rep. 2002; 3(1):28–33. 26(11):4028–40. 12. Ikura T, Ogryzko VV. Chromatin dynamics and DNA repair. Front Biosci. 2003; 8:s149–55. Received 10.12.13

9