ABSTRACT

New discoveries are increasingly demanding integration of epigenetics, molecular biology, genomic networks and physiology with evolution. This article provides a proof of concept for evolutionary transgenerational systems biology, proposed recently in the context of epigenetic inheritance in mammals. Gene set enrichment analysis of available genome-level mammalian data presented here seem consistent with the concept that: (1) heritable information about environmental effects in somatic cells is communicated to the germline by circulating microRNAs (miRNAs) or other RNAs released in physiological fluids; (2) epigenetic factors including miRNA-like small RNAs, DNA methylation and histone modifications are propagated across generations via gene networks; and (3) inherited epigenetic variations in the form of methylated cytosines are fixed in the population as thymines over the evolutionary time course. The analysis supports integration of physiology and epigenetics with inheritance and evolution. This may catalyze efforts to develop a unified theory of biology.

INTRODUCTION

Because of the inability of contemporary gene and natural selection centric evolutionary theory to fully explain heritability of phenotypic traits based solely on DNA sequence variation on the one hand, and increasing evidence of non-genetic inheritance on the other, there is a profound interest in integrating epigenetics, molecular biology, genomic networks, and physiology with the theory of evolution (Petronis, 2010; Danchin et al., 2011; Mattick, 2012; Ball, 2013; Noble, 2013,, 2015; Noble et al., 2014). A top-down approach toward this unification may begin with developing a broad conceptual mechanistic framework and testing that in a proof-of-concept analysis using empirical data. Notably, a recently proposed framework that is supported by observations reported in the literature (Sharma, 2014,, 2015a,b,c) provides an opportunity to analytically test the concept of a unified theory of biology. The proposed model explains transgenerational epigenetic inheritance and its evolutionary significance by integrating gene expression and gene networks, miRNA or other RNA, DNA methylation, histone modifications, and DNA-methylation-induced mutation, in three mechanistic steps (Fig. S1). First, heritable information about environmental effects in somatic cells is communicated to the germline by circulating RNAs, representing physiological conditions. Second, epigenetic modifications, in the form of RNA, DNA methylation and histone modifications, are propagated across generations through gene expression and gene networks. Third, inherited epigenetic variations, represented by methylated cytosines, are fixed in the population as thymines, in an evolutionary time scale. In the present analysis, these three principles are tested using available large-scale data sets related to gene expression, DNA methylation, histone modification, cytosine-methylation-induced mutation, chemical gene interaction and genetic association in diseases. The analysis is based on gene set enrichment as a measure of association. Given that the original framework was proposed in the context of mammals (Sharma, 2014,, 2015a,b,c), only mammalian data is used here for model validation.

RESULTS AND DISCUSSION

Fig. 1 illustrates evidence supporting the first principle. The circulating miRNA profiles of exosomes and body fluids closely resemble that of gonads and germ cells, in both males and females, and the miRNA profiles of germ cells resemble that of embryonic stages (Fig. 1A). The miRNA-based similarity observed between circulating factors, germ cells and developing embryo is statistically significant (Fig. 1B). miRNAs of exosome and body fluids were found to be over-represented in germ cells, as were miRNAs of gametes in the developing embryo. In contrast, number-matched control miRNAs, selected randomly from the combined set of miRNAs representing 461 organs, tissues and various other samples in the database, were not over-represented in general. These results clearly suggest that gametogenesis and development are strongly related to circulating factors in terms of miRNA profile.

Fig. 1.

Soma-to-germline communication of phenotypic information. (A) Clustering of biological samples based on presence or absence of microRNAs (miRNAs). Exosomal and body fluid profiles match closely with that of gonads and gametes, and the profiles of gametes with that of developmental stages. (B) Over-representation of exosomal and body fluid miRNAs combined in gonads and gametes, and of gametic miRNAs in developmental stages. (C) Display of the motif discovered. The motif logo, with reverse complement on the right, is shown along with the E value. (D) A subset of enriched gene ontology biological processes, all with q-value<0.05, in genes with motif-containing promoters. Log-transformed nominal P-values for individual enrichment are plotted on the y-axis. (E) Venn diagram showing enrichment of the motif-containing miRNAs in the set of nucleus-localized miRNAs. Fold enrichment (FE) and enrichment P-value are indicated. (F) Pie chart depicting enrichment of the motif in the set of nucleus-localized miRNAs excluding the initially discovered motif-containing miRNAs. Enrichment P-values, hypergeometric distribution, related to data pertaining to bar diagrams are shown above the bars. Values with 16 or more negative exponents of 10 were rounded down to zero. NS, not significant. List of miRNAs representing all the samples indicated in A,B are given in Table S1. Details of motif discovery analysis in C are provided in Table S2. Complete set of enriched gene ontology biological processes in D is shown in Table S3. Details of motif enrichment in F are given in Table S4.

Fig. 1.

Soma-to-germline communication of phenotypic information. (A) Clustering of biological samples based on presence or absence of microRNAs (miRNAs). Exosomal and body fluid profiles match closely with that of gonads and gametes, and the profiles of gametes with that of developmental stages. (B) Over-representation of exosomal and body fluid miRNAs combined in gonads and gametes, and of gametic miRNAs in developmental stages. (C) Display of the motif discovered. The motif logo, with reverse complement on the right, is shown along with the E value. (D) A subset of enriched gene ontology biological processes, all with q-value<0.05, in genes with motif-containing promoters. Log-transformed nominal P-values for individual enrichment are plotted on the y-axis. (E) Venn diagram showing enrichment of the motif-containing miRNAs in the set of nucleus-localized miRNAs. Fold enrichment (FE) and enrichment P-value are indicated. (F) Pie chart depicting enrichment of the motif in the set of nucleus-localized miRNAs excluding the initially discovered motif-containing miRNAs. Enrichment P-values, hypergeometric distribution, related to data pertaining to bar diagrams are shown above the bars. Values with 16 or more negative exponents of 10 were rounded down to zero. NS, not significant. List of miRNAs representing all the samples indicated in A,B are given in Table S1. Details of motif discovery analysis in C are provided in Table S2. Complete set of enriched gene ontology biological processes in D is shown in Table S3. Details of motif enrichment in F are given in Table S4.

Secreted by all cell types and identified in diverse body fluids, exosomes carry miRNAs that can influence gene expression and cause physiological changes in recipient cells (Chevillet et al., 2014; Melo et al., 2014; Alexander et al., 2015). Exosomal miRNAs are also considered to play a role in male and female reproductive physiology in mammals (Belleannée et al., 2013; Santonocito et al., 2014; Barkalina et al., 2015). Evidence suggests that miRNAs are loaded into exosomes selectively, based on specific miRNA motifs and post-transcriptional modifications, and levels of miRNAs or their targets in the producer cells (Alexander et al., 2015). The preferential sorting is supported by the observation that miRNA signatures of exosomes do not directly mirror the miRNA composition of the producer cells (Alexander et al., 2015). These findings have suggested that some miRNAs may have evolved to be parceled out in exosomes to perform their biological roles (Alexander et al., 2015). Although separating exosomal miRNAs from other extracellular-vesicle-associated or vesicle-independent circulating miRNAs is technically challenging, it has recently been inferred from quantitative and stoichiometric analysis that most exosomes do not contain many copies of miRNA molecules (Chevillet et al., 2014). This has led to the hypothesis that a given miRNA is either distributed in a few exosomes in the population at low concentration or, alternatively, packaged in rare exosomes at high concentration (Chevillet et al., 2014). Furthermore, selective loss of certain miRNAs with low GC content during extraction from smaller samples has been noted (Kim et al., 2012). Given these mechanistic and technical reasons associated with quantitative analysis of extracellular miRNAs, the present qualitative analysis, based simply on the presence or absence of a given miRNA, showing profile similarity between a general pool of circulating miRNAs, and gonadal and gamete-borne miRNAs may at this time seem sufficient to support the concept that extracellular noncoding RNA can potentially mediate soma to germline transmission of heritable information.

To examine the possibility that certain circulating miRNAs may have evolved to be selectively released into the circulation to mediate epigenetic inheritance, the exosome and body fluid miRNAs present in testis, spermatogonia, spermatocytes, spermatids or spermatozoa, as well as in the ovary, oocytes or ovum were examined for de novo discovery of sequence motifs. Interestingly, a motif was discovered in these miRNAs, not in a number-matched control set of randomly selected exosome and body fluid miRNAs (Fig. 1C). The motif was present in 19 of the total 131 miRNAs. Startlingly, enrichment analysis showed that the motif was significantly over-represented in promoters, 1000 base pair upstream to 200 base pair downstream regions, of genes associated with various gene ontology terms, including those that seem consistent with epigenetic inheritance, related to environmental factors, intercellular communication, gene expression, development, energy metabolism, and nervous system structure and function, for example (Fig. 1D). It is notable that a large proportion of examples of non-genetic inheritance reported so far in mammals relate to carbohydrate and lipid metabolism, and brain and behavior (Choi and Mango, 2014; Bale, 2015; Sharma, 2015b; Szyf, 2015; Dias et al., 2015). In addition, a recent study has found that thousands of genes that escape genome-wide DNA demethylation in human primordial germ cells (PGCs), whose complete data set is not reported, are over-represented in genes expressed in brain, and genes associated with metabolic, and neurological and neuropsychiatric disorders (Tang et al., 2015). Importantly, the potential biological significance of the identified miRNA motif and its presence in promoters was supported by examining nuclear localization of the miRNAs. The raw nuclear read counts for 8 of the 19 motif-containing miRNAs were available in a reported set of human cell line small RNA deep sequencing data, wherein a count of 10 or more was considered to indicate nuclear localization (Liao et al., 2010). Interestingly, in the data spanning 1307 unique mature miRNA sequences, the motif-containing miRNAs were over-represented in nucleus-localized miRNAs (Fig. 1E), with 7 of the 8 motif-containing miRNAs figuring within the top 21 miRNAs with highest counts in the nucleus. Furthermore, the top 145 nucleus-localized miRNAs, arbitrarily chosen from 1307 sequences, excluding the above 7 miRNAs, were found to be highly enriched for the motif, compared with the bottom 145 sequences with nuclear counts below the threshold of that for nuclear localization (Fig. 1F). Thus, profound nuclear localization of the motif-containing miRNAs and prominent presence of the motif in promoters of certain categories of genes together support the possibility that these miRNAs may regulate gene expression at transcriptional level. Indeed, examples for miRNA-mediated transcriptional regulation are known (Zhang et al., 2014). In particular, evidence exists to suggest that miRNAs with binding sites in gene promoters, located within ∼1000 base pairs without any unique feature, can modulate gene expression through epigenetic modifications of the promoter, including histone acetylation and/or methylation (Zhang et al., 2014). This is consistent with the present analysis supporting a role of circulating miRNAs in epigenetic inheritance.

A mechanism of multigenerational epigenetic inheritance mediated by Piwi-interacting RNA (piRNA), a class of small noncoding RNAs that are expressed in male and female germline and play an evolutionarily conserved role in transposon silencing, has previously been demonstrated in the nematode Caenorhabditis elegans (Ashe et al., 2012). In that study, it was shown for the first time that a piRNA-dependent foreign RNA response leads to multigenerational gene silencing involving a germline nuclear small RNA/chromatin pathway. In C. elegans, the mechanisms underlying piRNA mediated transcriptional gene silencing are considered similar to that involved in nuclear RNAi pathway in somatic tissues, with repressive histone modifications and RNA polymerase II stalling leading to silencing (Weick and Miska, 2014). The present analysis raising the possibility of miRNA, an endogenous small RNA like piRNA, playing a role in epigenetic inheritance in mammals seems very attractive because it satisfies the requirement of soma to germline communication, as envisaged in the first principle of the model under investigation.

Fig. 2 and Figs S2–S7 display evidence in support of the second principle. The common circulating and germline miRNAs are over-represented among miRNAs identified as differentially expressed in studies examining environmental effects in exposed and unexposed generations (Fig. 2A). In addition, the target genes of these common miRNAs show enrichment for genes that show differential mRNA expression (Fig. S2) and DNA methylation (Fig. S3) in gonads, gametes and various other tissues and organs in transgenerational studies. These results suggest a role of gene networks in epigenetic inheritance. Global analysis of data representing normal conditions further supports this. For example, the targets are found to over-represent imprinted genes (Fig. 2B), with imprinting representing a mode of epigenetic inheritance. The targets also enrich genes that are known to interact with a broad range of chemicals (Fig. 2C), including environmental factors known to cause transgenerational effects. Genes showing expression, regulation and differential DNA methylation, and histone modifications in gametes and developing embryo under normal conditions are also over-represented in the targets (Figs S4–S6). Furthermore, the targets enrich genes showing tissue-wide expression (Fig. 2D), differential genome-level CpG promoter density distribution (Fig. 2E) and histone modification signatures of active promoters and enhancers (Fig. 2F). The targets are also found to over-represent genes with known function in processes related to gene regulation by noncoding RNA including miRNA, biogenesis and metabolism of these RNAs, gene-specific transcription, response to chemicals and abiotic factors, embryonic development (Fig. S7) and metabolism, and brain development and function (Table S7). Finally, the reported binding regions of BLIMP1 (B lymphocyte-induced maturation protein-1; Magnúsdóttir et al., 2013), a key regulator of PGC specification involved in resetting of the epigenome towards a basal state, were found to be highly significantly enriched in the targets (fold change, 1.7; P<1×10−16), with the control showing a slight depletion (fold change, 0.86; P<0.01). Cumulatively, these results appear consistent with the second principle of the conceptual framework implicating gene networks in epigenetic inheritance.

Fig. 2.

Gene networks in germline transmission. (A) Over-representation of miRNAs common to exosomes, body fluids, spermatozoa, ovum and all other tissues indicated below the bars shown in Fig. 1B except fetus, total 18 in number, in the sets representing differentially expressed miRNAs reported in various studies on epigenetic inheritance. Note combined analysis, represented by the extreme right pair of bars, showing enrichment of common miRNAs in total miRNAs reported in epigenetic studies. Key to miRNA profiles: gender exposed at F0-treatment at F0-generation and sample profiled. ME, male exposure; FE, female exposure; HFD, high-fat diet; LPD, low-protein diet; T, testosterone; F, flutamide; TF, testosterone plus flutamide; B[a]P, benzo(a)pyrene; BA, bisphenol A. Mature miRNA names were used for enrichment analysis. (B) Enrichment of targets (see Fig. S2) in imprinted genes. (C) Enrichment of targets (see Fig. S2) in chemical-interacting mRNAs and proteins. (D) Proteomic profiles across human tissues. Over-representation of the targets in proteins expressed in all tissues, and depletion in proteins with tissue biased expression, are clearly observed. (E) Genome-level CpG density across human promoters. The targets are enriched and depleted in high- and low-CpG promoters, in that order. (F) Histone modification signatures of active promoters and enhancers in human cell/tissue types. The targets over-represent active regulatory signatures in cells/tissues other than embryonic stem cells (ESCs). Enrichment P-values, hypergeometric distribution, are shown above bars. Values with 16 or more negative exponents of 10 were rounded down to zero. NS, not significant. Missing bars in A,B indicate overlapping counts as either nil or excessively low, considered inappropriate for analysis. Sources of data used in all analyses are indicated in Table S5. Complete set of enriched gene ontology biological processes is provided in Table S6.

Fig. 2.

Gene networks in germline transmission. (A) Over-representation of miRNAs common to exosomes, body fluids, spermatozoa, ovum and all other tissues indicated below the bars shown in Fig. 1B except fetus, total 18 in number, in the sets representing differentially expressed miRNAs reported in various studies on epigenetic inheritance. Note combined analysis, represented by the extreme right pair of bars, showing enrichment of common miRNAs in total miRNAs reported in epigenetic studies. Key to miRNA profiles: gender exposed at F0-treatment at F0-generation and sample profiled. ME, male exposure; FE, female exposure; HFD, high-fat diet; LPD, low-protein diet; T, testosterone; F, flutamide; TF, testosterone plus flutamide; B[a]P, benzo(a)pyrene; BA, bisphenol A. Mature miRNA names were used for enrichment analysis. (B) Enrichment of targets (see Fig. S2) in imprinted genes. (C) Enrichment of targets (see Fig. S2) in chemical-interacting mRNAs and proteins. (D) Proteomic profiles across human tissues. Over-representation of the targets in proteins expressed in all tissues, and depletion in proteins with tissue biased expression, are clearly observed. (E) Genome-level CpG density across human promoters. The targets are enriched and depleted in high- and low-CpG promoters, in that order. (F) Histone modification signatures of active promoters and enhancers in human cell/tissue types. The targets over-represent active regulatory signatures in cells/tissues other than embryonic stem cells (ESCs). Enrichment P-values, hypergeometric distribution, are shown above bars. Values with 16 or more negative exponents of 10 were rounded down to zero. NS, not significant. Missing bars in A,B indicate overlapping counts as either nil or excessively low, considered inappropriate for analysis. Sources of data used in all analyses are indicated in Table S5. Complete set of enriched gene ontology biological processes is provided in Table S6.

Fig. 3 presents evidence supporting the third principle. The targets show enrichment for transcription factor binding sites created by cytosine-methylation-induced mutation in low-CpG promoter-associated genes (Fig. 3A). As CpG dinucleotides in these promoters are constitutively methylated in the germline (Weber et al., 2007), this result supports the evolutionary significance of epigenetic inheritance. Next, the targets were found to over-represent genes mutations or polymorphisms that show an association with various diseases (Fig. 3B), potentially connecting epigenetic variations, evolution and human health. This is consistent with the above-mentioned study showing over-representation of disease-associated genes in DNA-demethylation-resistant genes in human PGCs (Tang et al., 2015). As proposed (Sharma, 2015b,c), the epigenetic modifications may either discontinue to exist, persist as such or convert to genetic alterations in evolutionary time course (Fig. S1). Although transition of 5-methylcytosines to thymines per se could be a passive by-product, the resulting change can become a potential substrate for selection like any other newly arisen genetic variation does in the normal course of evolution. A recently proposed model of RNA-mediated gene evolution has underscored this possibility (Morris, 2015). With regards the overall evolutionary significance of epigenetic inheritance, it has been highlighted in a recently advanced unified theory of molecular aspects of evolution that natural selection may potentially act on acquired traits (Skinner, 2015). Cumulatively, the present results seem consistent with the third principle.

Fig. 3.

Evolutionary significance of epigenetic inheritance. (A) Representation of the targets in gene promoters associated with transcription factor binding sites created by mutation of cytosine to thymine in the context of CpG dinucleotides. Targets are clearly enriched in low-CpG promoter-associated genes. (B) Enrichment of the targets in disease-associated genes. Targets are clearly enriched. NHGRI, National Human Genome Research Institute; GAD, genetic association database; MIM, Mendelian inheritance in man; LSDB, locus-specific mutation databases; CTD, comparative toxicogenomics database. Enrichment P-values, hypergeometric distribution, are shown above bars. NS, not significant. Values with 16 or more negative exponents of 10 were rounded down to zero. Sources of data used in the analysis are indicated in Table S7.

Fig. 3.

Evolutionary significance of epigenetic inheritance. (A) Representation of the targets in gene promoters associated with transcription factor binding sites created by mutation of cytosine to thymine in the context of CpG dinucleotides. Targets are clearly enriched in low-CpG promoter-associated genes. (B) Enrichment of the targets in disease-associated genes. Targets are clearly enriched. NHGRI, National Human Genome Research Institute; GAD, genetic association database; MIM, Mendelian inheritance in man; LSDB, locus-specific mutation databases; CTD, comparative toxicogenomics database. Enrichment P-values, hypergeometric distribution, are shown above bars. NS, not significant. Values with 16 or more negative exponents of 10 were rounded down to zero. Sources of data used in the analysis are indicated in Table S7.

Evidence provided here supports the proposed (Sharma, 2014,, 2015a,b,c) model of evolutionary transgenerational systems biology. First, it demonstrates that miRNA-based similarity exists between circulating factors, gametes, developmental stages and adult tissues, under normal conditions or under environmental conditions that produce an effect across generations. This is consistent with the hypothesis that RNAs present in physiological fluids hold the potential to mediate soma to germline communication in epigenetic inheritance. Second, the results show an association between the target genes of the convergent miRNAs and regulated gene expression, DNA methylation and histone modifications in the above conditions. This is in line with the proposition that gene networks may underlie epigenetic memory propagation across generations. Third, the miRNA targets show association with cytosine-methylation-induced mutational events and disease-related mutations and polymorphisms. This supports the suggestion that epigenetic inheritance may play an evolutionarily significant role. The above analysis is not exhaustive in the sense that several sets of available data remain to be examined. For example, a recent study has revealed that several thousand genes escape genome-wide DNA demethylation in human PGCs (Tang et al., 2015). Non-availability of the complete data set, however, prevented its analysis here. But this study and several others (Magnúsdóttir et al., 2013; Smith et al., 2014; Irie et al., 2015) together provide additional data on gene expression and DNA methylation dynamics associated with gametes, and cells, tissues and organs pertaining to embryonic development in mouse and human. It will be interesting to extend the analysis to include these and newer studies in future. Nevertheless, the present results tend to unify environment, physiology, RNA, gene networks, DNA methylation and histone modifications with inheritance, development, disease and evolution. A potential role and integration of physiology, epigenetics, RNA and genetic interactions in inheritance and evolution have previously been suggested and sought for on the basis of theoretical considerations (Richards, 2008; Day and Bonduriansky, 2011; Hunter et al., 2012; Livnat, 2013; Noble, 2013,, 2015; Noble et al., 2014; Rivoire and Leibler, 2014; Jablonka and Lamb, 2015; Morris, 2015; Skinner, 2015). The data analysis presented here offers a proof of principle for a unified theory of biology. The results may prove valuable in directing future efforts toward integration.

MATERIALS AND METHODS

Data available in various public databases and published papers along with associated supplementary information was used, as provided. The papers were identified through PubMed search using appropriate key words, and the relevant data sets included in the analysis without any bias. For miRNA and mRNA clustering or enrichment analysis, all the human miRNAs in the organ-miRNA interaction database (http://www.umm.uni-heidelberg.de/apps/zmf/mirwalk) or all the human genes in gene ontology (http://geneontology.org/) were used as the total miRNA or mRNA population, in that order. To obtain controls for enrichment analysis, matching numbers of miRNAs or mRNAs were randomly selected (https://www.random.org/sequences/) from the aforesaid populations, as appropriate. A documented set of validated miRNA target genes was used (http://zmf.umm.uni-heidelberg.de/apps/zmf/mirwalk2/index.html). The statistical significance of enrichment was computed using hypergeometric distribution probability, with 0.05 as the nominal P-value cut-off. Over-representation analysis of gene ontology biological processes was carried out using 0.05 as Benjamini-Hochberg adjusted P-value cut-off (http://david.abcc.ncifcrf.gov/). Chi-square test for homogeneity was used to check equality of two different populations. The MEME suite 4.10.1 (Bailey et al., 2009) was used for de novo discovery (meme-suite.org/tools/meme) and enrichment (meme-suite.org/tools/fimo) of miRNA motif, and its association with promoters of genes linked to gene ontology terms (meme-suite.org/tools/gomo), all under default settings. For motif discovery, the complete set of human miRNAs in miRBase (http://www.mirbase.org/) was used as background (Kozomara and Griffiths-Jones, 2014). A 0-order Markov model was assumed for the background, and the minimum motif width set at 6 nucleotides.

Footnotes

Funding

This research received a grant from the BSC0122 network project of the Council of Scientific and Industrial Research, India.

References

Alexander
,
M.
,
Hu
,
R.
,
Runtsch
,
M. C.
,
Kagele
,
D. A.
,
Mosbruger
,
T. L.
,
Tolmachova
,
T.
,
Seabra
,
M. C.
,
Round
,
J. L.
,
Ward
,
D. M.
and
O'Connell
,
R. M.
(
2015
).
Exosome-delivered microRNAs modulate the inflammatory response to endotoxin
.
Nat. Commun.
6
,
7321
.
Ashe
,
A.
,
Sapetschnig
,
A.
,
Weick
,
E.-M.
,
Mitchell
,
J.
,
Bagijn
,
M. P.
,
Cording
,
A. C.
,
Doebley
,
A.-L.
,
Goldstein
,
L. D.
,
Lehrbach
,
N. J.
,
Le Pen
,
J.
, et al. 
(
2012
).
piRNAs can trigger a multigenerational epigenetic memory in the germline of C. elegans
.
Cell
150
,
88
-
99
.
Bailey
,
T. L.
,
Boden
,
M.
,
Buske
,
F. A.
,
Frith
,
M.
,
Grant
,
C. E.
,
Clementi
,
L.
,
Ren
,
J.
,
Li
,
W. W.
and
Noble
,
W. S.
(
2009
).
MEME SUITE: tools for motif discovery and searching
.
Nucleic Acids Res.
37
,
W202
-
W208
.
Bale
,
T. L.
(
2015
).
Epigenetic and transgenerational reprogramming of brain development
.
Nat. Rev. Neurosci.
16
,
332
-
344
.
Ball
,
P.
(
2013
).
DNA: celebrate the unknowns
.
Nature
496
,
419
-
420
.
Barkalina
,
N.
,
Jones
,
C.
,
Wood
,
M. J. A.
and
Coward
,
K.
(
2015
).
Extracellular vesicle-mediated delivery of molecular compounds into gametes and embryos: learning from nature
.
Hum. Reprod. Update
21
,
627
-
639
.
Belleannée
,
C.
,
Calvo
,
É.
,
Caballero
,
J.
and
Sullivan
,
R.
(
2013
).
Epididymosomes convey different repertoires of microRNAs throughout the bovine epididymis
.
Biol. Reprod.
89
,
30
.
Chevillet
,
J. R.
,
Kang
,
Q.
,
Ruf
,
I. K.
,
Briggs
,
H. A.
,
Vojtech
,
L. N.
,
Hughes
,
S. M.
,
Cheng
,
H. H.
,
Arroyo
,
J. D.
,
Meredith
,
E. K.
,
Gallichotte
,
E. N.
, et al. 
(
2014
).
Quantitative and stoichiometric analysis of the microRNA content of exosomes
.
Proc. Natl. Acad. Sci. USA
111
,
14888
-
14893
.
Choi
,
Y.
and
Mango
,
S. E.
(
2014
).
Hunting for Darwin's gemmules and Lamarck's fluid: transgenerational signaling and histone methylation
.
Biochim. Biophys. Acta
1839
,
1440
-
1453
.
Danchin
,
É.
,
Charmantier
,
A.
,
Champagne
,
F. A.
,
Mesoudi
,
A.
,
Pujol
,
B.
and
Blanchet
,
S.
(
2011
).
Beyond DNA: integrating inclusive inheritance into an extended theory of evolution
.
Nat. Rev. Genet.
12
,
475
-
486
.
Day
,
T.
and
Bonduriansky
,
R.
(
2011
).
A unified approach to the evolutionary consequences of genetic and nongenetic inheritance
.
Am. Nat.
178
,
E18
-
E36
.
Dias
,
B. G.
,
Maddox
,
S. A.
,
Klengel
,
T.
and
Ressler
,
K. J.
(
2015
).
Epigenetic mechanisms underlying learning and the inheritance of learned behaviors
.
Trends Neurosci.
38
,
96
-
107
.
Hunter
,
B.
,
Hollister
,
J. D.
and
Bomblies
,
K.
(
2012
).
Epigenetic inheritance: what news for evolution?
Curr. Biol.
22
,
R54
-
R56
.
Irie
,
N.
,
Weinberger
,
L.
,
Tang
,
W. W. C.
,
Kobayashi
,
T.
,
Viukov
,
S.
,
Manor
,
Y. S.
,
Dietmann
,
S.
,
Hanna
,
J. H.
and
Surani
,
M. A.
(
2015
).
SOX17 is a critical specifier of human primordial germ cell fate
.
Cell
160
,
253
-
268
.
Jablonka
,
E.
and
Lamb
,
M. J.
(
2015
).
The inheritance of acquired epigenetic variations
.
Int. J. Epidemiol.
Kim
,
Y.-K.
,
Yeo
,
J.
,
Kim
,
B.
,
Ha
,
M.
and
Kim
,
V. N.
(
2012
).
Short structured RNAs with low GC content are selectively lost during extraction from a small number of cells
.
Mol. Cell
46
,
893
-
895
.
Kozomara
,
A.
and
Griffiths-Jones
,
S.
(
2014
).
miRBase: annotating high confidence microRNAs using deep sequencing data
.
Nucleic Acids Res.
42
,
D68
-
D73
.
Liao
,
J.-Y.
,
Ma
,
L.-M.
,
Guo
,
Y.-H.
,
Zhang
,
Y.-C.
,
Zhou
,
H.
,
Shao
,
P.
,
Chen
,
Y.-Q.
and
Qu
,
L.-H.
(
2010
).
Deep sequencing of human nuclear and cytoplasmic small RNAs reveals an unexpectedly complex subcellular distribution of miRNAs and tRNA 3′ trailers
.
PLoS ONE
5
,
e10563
.
Livnat
,
A.
(
2013
).
Interaction-based evolution: how natural selection and nonrandom mutation work together
.
Biol. Direct
8
,
24
.
Magnúsdóttir
,
E.
,
Dietmann
,
S.
,
Murakami
,
K.
,
Günesdogan
,
U.
,
Tang
,
F.
,
Bao
,
S.
,
Diamanti
,
E.
,
Lao
,
K.
,
Gottgens
,
B.
and
Azim Surani
,
M.
(
2013
).
A tripartite transcription factor network regulates primordial germ cell specification in mice
.
Nat. Cell Biol.
15
,
905
-
915
.
Mattick
,
J. S.
(
2012
).
Rocking the foundations of molecular genetics
.
Proc. Natl. Acad. Sci. USA
109
,
16400
-
16401
.
Melo
,
S. A.
,
Sugimoto
,
H.
,
O'Connell
,
J. T.
,
Kato
,
N.
,
Villanueva
,
A.
,
Vidal
,
A.
,
Qiu
,
L.
,
Vitkin
,
E.
,
Perelman
,
L. T.
,
Melo
,
C. A.
, et al. 
(
2014
).
Cancer exosomes perform cell-independent microRNA biogenesis and promote tumorigenesis
.
Cancer Cell
26
,
707
-
721
.
Morris
,
K. V.
(
2015
).
The theory of RNA-mediated gene evolution
.
Epigenetics
10
,
1
-
5
.
Noble
,
D.
(
2013
).
Physiology is rocking the foundations of evolutionary biology
.
Exp. Physiol.
98
,
1235
-
1243
.
Noble
,
D.
(
2015
).
Evolution beyond neo-Darwinism: a new conceptual framework
.
J. Exp. Biol.
218
,
7
-
13
.
Noble
,
D.
,
Jablonka
,
E.
,
Joyner
,
M. J.
,
Müller
,
G. B.
and
Omholt
,
S. W.
(
2014
).
Evolution evolves: physiology returns to centre stage
.
J. Physiol.
592
,
2237
-
2244
.
Petronis
,
A.
(
2010
).
Epigenetics as a unifying principle in the aetiology of complex traits and diseases
.
Nature
465
,
721
-
727
.
Richards
,
E. J.
(
2008
).
Population epigenetics
.
Curr. Opin. Genet. Dev.
18
,
221
-
226
.
Rivoire
,
O.
and
Leibler
,
S.
(
2014
).
A model for the generation and transmission of variations in evolution
.
Proc. Natl. Acad. Sci. USA
111
,
E1940
-
E1949
.
Santonocito
,
M.
,
Vento
,
M.
,
Guglielmino
,
M. R.
,
Battaglia
,
R.
,
Wahlgren
,
J.
,
Ragusa
,
M.
,
Barbagallo
,
D.
,
Borzì
,
P.
,
Rizzari
,
S.
,
Maugeri
,
M.
, et al. 
(
2014
).
Molecular characterization of exosomes and their microRNA cargo in human follicular fluid: bioinformatic analysis reveals that exosomal microRNAs control pathways involved in follicular maturation
.
Fertil. Steril.
102
,
1751
-
1761.e1
.
Sharma
,
A.
(
2014
).
Bioinformatic analysis revealing association of exosomal mRNAs and proteins in epigenetic inheritance
.
J. Theor. Biol.
357
,
143
-
149
.
Sharma
,
A.
(
2015a
).
Transgenerational epigenetic inheritance requires a much deeper analysis
.
Trends Mol. Med.
21
,
269
-
270
.
Sharma
,
A.
(
2015b
).
Transgenerational epigenetic inheritance: resolving uncertainty and evolving biology
.
Biomol. Concepts
6
,
87
-
103
.
Sharma
,
A.
(
2015c
).
Transgenerational epigenetic inheritance: emerging concepts and future prospects
.
J. Reprod. Health Med.
1
,
60
-
63
.
Skinner
,
M. K.
(
2015
).
Environmental epigenetics and a unified theory of the molecular aspects of evolution: a Neo-Lamarckian concept that facilitates Neo-Darwinian evolution
.
Genome Biol. Evol.
7
,
1296
-
1302
.
Smith
,
Z. D.
,
Chan
,
M. M.
,
Humm
,
K. C.
,
Karnik
,
R.
,
Mekhoubad
,
S.
,
Regev
,
A.
,
Eggan
,
K.
and
Meissner
,
A.
(
2014
).
DNA methylation dynamics of the human preimplantation embryo
.
Nature
511
,
611
-
615
.
Szyf
,
M.
(
2015
).
Nongenetic inheritance and transgenerational epigenetics
.
Trends Mol. Med.
21
,
134
-
144
.
Tang
,
W. W. C.
,
Dietmann
,
S.
,
Irie
,
N.
,
Leitch
,
H. G.
,
Floros
,
V. I.
,
Bradshaw
,
C. R.
,
Hackett
,
J. A.
,
Chinnery
,
P. F.
and
Surani
,
M. A.
(
2015
).
A unique gene regulatory network resets the human germline epigenome for development
.
Cell
161
,
1453
-
1467
.
Weber
,
M.
,
Hellmann
,
I.
,
Stadler
,
M. B.
,
Ramos
,
L.
,
Pääbo
,
S.
and
Rebhan
,
M.
and
Schübeler
,
D.
(
2007
).
Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome
.
Nat. Genet.
39
,
457
-
466
.
Weick
,
E.-M.
and
Miska
,
E. A.
(
2014
).
piRNAs: from biogenesis to function
.
Development
141
,
3458
-
3471
.
Zhang
,
Y.
,
Fan
,
M.
,
Zhang
,
X.
,
Huang
,
F.
,
Wu
,
K.
,
Zhang
,
J.
,
Liu
,
J.
,
Huang
,
Z.
,
Luo
,
H.
,
Tao
,
L.
, et al. 
(
2014
).
Cellular microRNAs up-regulate transcription via interaction with promoter TATA-box motifs
.
RNA
20
,
1878
-
1889
.

Competing interests

The author declares no competing or financial interests.

Supplementary information