Introduction

The term “-omics” is now a very familiar term especially in the field of life sciences. This suffix seems to be first coined in the 1980s, to describe the global analysis of biological molecules. Since then, an explosion of -omics approaches has been observed in the last few decades. Although the study of a particular area suffices to constitute an “-omics” approach, e.g., phosphoproteomics, which is the systematic/global study of phosphoproteins [1]. However, most of these new “-omics” approaches are actually stemmed from the four main noticeable “-omics” approaches, i.e., genomics, transcriptomics, proteomics, and metabolomics. As from the concept of central dogma, it is not difficult to anticipate the gradual evolvement of genomics to transcriptomics and to proteomics. For metabolomics, although it is regaining attention lately, this approach could never attain such high throughput competency without the recent advancement of tandem mass spectrometry (MS/MS) [2]. Even for that, because of the complex diversity of known and unknown metabolites that might actually be present, therefore, various MS-based (e.g., GC-MS, LC-MS/MS) as well as non-MS-based methodologies (e.g., NMR-based) are still required to fulfill the objectives in each specific need in metabolomic study [3•]. Moreover, although each main approach alone is powerful, throughout these years, it becomes evident for us to realize that no single omics approach would be indeed comprehensively/conclusive enough to understand/solve/explain a particular biological problem. Therefore, in the future, it is expected that the combined utilization of multiple approaches in one study would be deemed necessary to get an even more global/comprehensive picture and facilitate in deciphering the complex biological phenomenon [4•].

Epimodifications of the Fantastic Four “-omes”

In the past half century, DNA methylation and protein phosphorylation are among the most common modifications discovered. It has been a while when histone proteins have been regarded as just the bulk materials or inert building blocks for organizing the eukaryotic genome [5], and gene expressions are primarily controlled by the degree of DNA methylation in the genome [6]. However, this concept has been changed dramatically later when histone H3 was demonstrated to be able to undergo phosphorylation and acetylation (which are closely related to cell division and participated in cellular gene regulation) [7, 8]. Since then, DNA methylation and histone modifications appear to layout the foundation of epigenetics. Indeed, it is now becoming clear that post-translational modification (PTM) of proteins happens not just in histones but virtually in every protein [9]. In here, we would like to strengthen the notion that although there should presumably still exist an unknown number and type of post-synthetic modifications on a variety of biomolecules, however, only those modifications that exert heritable changes should be considered as epimodifications, whether decorated in DNA, RNA, or proteins (although it is not clear if epimetabolites would have heritable changes to the epigenetic landscape for the moment). For instance, histone protein modifications are considered as epiproteomic modifications. Other proteins we think that they should be categorized within the epiproteome are the chromatin-associated enzymes/factors that can write, read, or erase DNA, RNA, or histone protein epi-marks [such as DNMTs (DNA writers), METTL3-METTL14-WTAP-KIAA1429 complex (RNA writers), and Aurora kinases (histone writers); methyl-CpG-binding proteins (DNA readers), YTH family proteins (RNA readers), and 14-3-3 proteins (histone readers); TETs (DNA/RNA erasers), FTO (RNA eraser), and PP1 (histone eraser)] [10••, 11, 12, 13••], as their activities could lead to chromatin remodeling/alteration of DNA methylation, histone modifications, and ncRNA expression, which subsequently influent the dynamics and operations of cellular genetics. Epimodifications of RNA and metabolites are the more recently studied field and it is intriguing to discover that RNA, similar to DNA, can also be methylated and hydroxymethylated [10••, 14]. Likewise, the modification of metabolites is gradually appreciated in the past few years and it is believed that with the use of MS/MS technology, more and more epimetabolites/oncometabolites will be identified and their functions deciphered in the near future. Nowadays, MS/MS technology is a powerful tool for the de novo identification and quantitation of epimodifications in diverse biomolecules in the field of life sciences [15, 16].

In this article, we would like to present the overall concepts of “-omes−epi-omes” interactions, i.e., the relationships among the four most noticeable “-omes” (genome, transcriptome, proteome, and metabolome) to the four “epi-omes” (epigenome, epitranscriptome, epiproteome, and epimetabolome) as well as discussing the recently identified epimodifications in humans. Lastly, the recent advances in targeting the epi-omes by various epi-drug strategies will be discussed.

The Cellular Regulatory Paradigm of “-omes−epi-omes” Interactions

We would like to present the following models to demonstrate the gradual shaping of “-omes−epi-omes” interactions (the four -omes and four epi-omes) (Fig. 1). First, we start with the four noticeable “-omes”, i.e., genome (G), transcriptome (T), proteome (P), and metabolome (M) (Fig. 1a). The identification of DNA methylation in the genome has initially shaped the concept of epigenome (Epi-G), while the discovery of histone phosphorylation and acetylation in chromatin has founded the epiproteome (Epi-P) (Fig. 1b). Transcripts of the genome, whether the coding and non-coding RNA messages, constitute the transcriptome and are also subjected to various post-transcriptional modifications, which encompassed the epitranscriptome (Epi-T) (Fig. 1b). Within the constituents of the proteome (i.e., enzymes), they can perform various catabolic and anabolic reactions to produce diverse kinds of biomolecules and metabolites; the global study of all the metabolites inside the cells constitute the metabolome (Fig. 1a). Very recently, it has been proposed that the modification of metabolome can give rise to epimetabolome (Epi-M) (the dark matter of metabolism) (Fig. 1b). Since a variety of metabolites like amino acids, carbohydrates, hydroxy acids, lipids, peptides, purines, pyrimidines, and sterols can potentially be undergoing diverse epimodifications, it is highly possible that a great deal of novel epimetabolites exist and are waiting to be discovered. From this point, all eight players are now clearly emerged (Fig. 1b).

Fig. 1
figure 1

The cellular regulatory paradigm of “-omes−epi-omes” interactions. a The possible interactions among the four most noticeable “-omes”: genome (G), transcriptome (T), proteome (P), and metabolome (M). b All the possible interactions among the four “-omes” to the four “epi-omes”: epigenome (Epi-G), epitranscriptome (Epi-T), epiproteome (Epi-P), and epimetabolome (Epi-M)

Full size image

As a matter of fact, the transitions of the four -omes to four epi-omes are largely catalyzed by various enzymes/protein complexes of the proteome. For example, the modification of G by P created the Epi-G (genomic DNA being cytosine methylated by DNMTs). Indeed, constituents of P can modify itself to become Epi-P; this is not difficult to imagine, for instance, histone protein kinases phosphorylate histone proteins. Therefore, the transition of G to Epi-G and P to Epi-P is indeed a very common phenomenon in the cell. Recently, RNA was demonstrated to be modified by a variety of RNA modifying enzymes in the cells [17], which confirmed the transition of T to Epi-T. Finally, several metabolites can be further modified by some metabolic enzymes in the cells [18, 19], which also supported the M to Epi-M transition.

Inside the cells, -ome(s)/epi-ome(s) and -ome(s)/epi-ome(s) are closely interacted. We would like to cite a few examples here: (1) Epi-P→G (Ser10-phosphorylated histone H3 leads to chromatin DNA condensation) [20], (2) P↔Epi-P→G↔Epi-G (deacetylation or acetylation of chromatin-associated enzyme DNMT1 can enhance or reduce its stability, and as a result increasing or decreasing the level of methylated genomic DNA) [21], (3) T↔Epi-T→P (N6-adeonine methylated mRNA which affects its stability and translation efficiency, and as a result influent the level of protein translated) [22], and (4) T and M interaction (the discovery of metabolite-binding RNA domains in eukaryotes, in which cells are capable of using RNA for metabolite-binding and to sense cellular metabolism) [23]. Besides of the above, we think it would be tempting to investigate other novel interactions between Epi-M and the other players in this model. Suffice to say, all the four -omes to the four epi-omes are dynamically interrelated and all the possible direct/indirect effects could happen, which may ultimately explain why the cellular physiology is so complicated.

Novel Epimodifications/Epi-marks in Humans

Epimodifications are largely in part identified by MS/MS discovery/untargeted mode of identification. It is becoming usual for the identification of new epi-marks nowadays. In here, we would like to discuss some of the latest epi-marks reported (as shown in Table 1).

Table 1 An update on epimodification of biomolecules
Full size table

Epigenomic Modifications

It is generally accepted that DNA methylation by DNMTs promotes chromatin condensation and transcriptional repression, whereas DNA demethylation by DNA demethylases is associated with gene/transcriptional activation. In mammalian cells, DNA methylation is faithfully maintained and written by three major DNMTs, including DNMT1 (for maintenance of methylation), DNMT3a, and DNMT3b (for de novo methylation) [24, 25]. On the other hand, active DNA demethylation is carried by a class of enzymes called ten-eleven translocations (TETs), including TET1, TET2, and TET3, which are methylcytosine dioxygenases that convert methylated cytosine to hydroxymethylated cytosine [26].

Similar to nuclear DNA (nDNA), mitochondrial DNA (mtDNA) can also be subjected to cytosine methylation (m5C)/hydroxymethylation (hm5C) [27]. Lately, it has also been reported that instead of CpG methylation, mitochondrial DNA can undergo GpC methylation (mediated by a mitochondria-targeted DNMT1 transcript variant, mtDNMT1) as potential regulator of mitochondrial gene expression [28•]. Moreover, recent findings also indicate that dysfunctional mtDNA methylation could underlie aging and diseases [29].

Besides of the above, other less familiar DNA epimodifications can also occur, including cytosine formylation (5fC, a demethylation intermediate and epigenetic mark), cytosine carboxylation (5caC, a demethylation intermediate and possible epigenetic mark), and N6-adeonine methylation (m6A, a potential epigenetic mark) [14, 30]. For a more detailed perspective of the related information of other DNA epimodifications, we refer the reader to the following online DNA modification database (DNAmod) [31•].

Epitranscriptomic Modifications

Since the first discovery of N6-adeonine methylation (m6A) as the most abundant reversible post-transcriptional modification on mRNAs and lncRNAs in eukaryotes [32], an explosion of over 100 different types of chemical modifications have been found in coding and non-coding RNAs [including mRNAs, tRNAs, rRNAs, snRNAs, and ncRNAs (lncRNAs, miRNAs, and snoRNAs)] so far with the use of the latest MS and sequencing technologies [10••, 33,34,35, 36•]. As mentioned, similar to DNA, RNA can also be cytosine methylated/hydroxymethylated. In addition, other less familiar RNA epimodifications can also occur, including N1-adeonine methylation (m1A), pseudouridylation (Ψ), and 2 ′-O-me t h y l a t i o n ( 2 ′-O-Me ) [All these epimodifications are closely related to RNA stability, structure, translational efficiency, and viral replication] [10••, 37,38,39]. More recently, it has been suggested the important role of reversible RNA modifications in meiosis and pluripotency [40] as well as in memory formation [41]. For a more detailed perspective of the related information of other RNA epimodifications, we refer the reader to the following online RNA modification databases (RMBase, MODOMICS, and RNAMDB) [42, 43, 44•].

Epiproteomic Modifications

Acetylation, phosphorylation, methylation, and ubiquitylation are among the most abundant and well-known reversible histone PTM marks [45••]. Recently, diverse histone acyl lysine modifications that use intermediates in metabolism have been gradually reported, including lysine β-hydroxybutyrylation (bhb) [46], crotonylation (cr) [47], 2-hydroxyisobutyrylation (hib) [48], malonylation (ma) [49], and succinylation (su) [49] (which are closely related to cellular metabolism and played important roles in regulating histone structure and functions). Other less familiar epiproteomic modifications include histone O-glcNAcylation (O-glc) (important for cell cycle transition) [50], cysteine glutathionylation (glu) (responsible for redox sensing and regulation of chromatin structure) [51], and the latest lysine deacetylimination, which converts the acetyllysine to allysine (as firstly demonstrated in transcription factor STAT3) [52]. As mentioned earlier, besides histones, chromatin-associated enzymes/factors that can write, read, or erase DNA, RNA, or histone protein epi-marks are likely exerting epigenetic effects; therefore, the epimodification status (such as ubiquitylation) and turnover of these chromatin-associated enzymes/factors would likely influent the cellular genetics. Here, we want to address on the complexity of protein ubiquitylation that at such not only eight different types of ubiquitin (ub) linkage can occur (at M1, K6, K11, K27, K29, K33, K48, and K63 position of ub), but the degree of ub-conjugation can also be variable (i.e., whether the target molecule is monoubiquitylated, multi-monoubiquitylated, or homotypically/heterotypically polyubiquitylated) [53••]. For example, the K48 homotypic polyubiquitylation, the most studied type of ub-conjugation, targets protein to proteasomal degradation [54], while K63 homotypic polyubiquitylation is linked to DNA damage/stress response, translation, and lysosomal targeting [55, 56]. Likewise, although not as complex as protein ubiquitylation, lysine can undergo mono/di/tri-methylation (Kme/Kme2/Kme3) [57]; while arginine can undergo mono/dimethylation (Rme/Rme2). Arginine dimethylation can occur in a symmetrical (Rme2s) or asymmetrical (Rme2a) manner [58]. Therefore, close and cautious examination should be undertaken for epiproteomic studies involving ubiquitylation and methylation.

Epimetabolomic Modifications

More recently, studies of identification and elucidation of the biological functions of epimetabolites are emerging. Epimetabolites are produced by modification of metabolites or repairing of damaged metabolites but without creating new metabolic pathways [59]. For instance, several categories of epimetabolites were discovered, including lipid epimetabolites [fatty esters of monohydroxy fatty acids (FAHFAs) with anti-diabetic and anti-inflammatory effects], methylated epimetabolites [N-methylglycine, an oncometabolite; 1-methylnicotinamide (1MNA), acts as a methylation sink in naive embryonic stem cells preventing deposit of H3K27me3 marks; symmetric and asymmetric dimethylarginine (SDMA and ADMA), in which SDMA can be used as a urine biomarker for renal insufficiency, while ADMA levels are significantly associated with an increased risk of coronary artery disease], and isomeric variants of epimetabolite 2-hydroxyglutarate (2HG), an oncometabolite, which exists in L- or D-enantiomer forms [18, 19, 60,61,62].

From the above, although the presence of diverse epimodifications/epi-marks have been proven to be existing in each of the four -omes, however, intensive researches would still be needed to dissect the inter-omes–epi-omes relationship as well as their biological function and significance. Since epimodification of the constituents within a particular -ome might directly/indirectly affect the constituents of the other -ome(s)/epi-ome(s), therefore, it is deemed necessary to have the whole epimodification profile of the four -omes in order for us to be able to comprehend the physiological status of the cells at a particular time/event.

Rise of the Epi-drugs

Drugs that target the epi-omes have potential over conventional cancer therapeutic approaches. Many epi-drugs are being developed and undergoing clinical trials. It can be seen that drugs that target the DNMT and various epigenetic modifiers (HDAC, Aurora B, EZH2, IDH1/2, LSD1, SETD2, NSD2, SWI/SNF complex, SMARCA2/4, BRD4, and DOT1L) are being deployed as a strategy for battling against various human diseases [63]. Although genome editing by zinc-finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), or a (clustered regularly interspaced short palindromic repeats)/Cas9 (CRISPR-associated 9) system can remove or insert genetic elements within the genome [64••], however, the effect is somewhat permanent and is difficult to be reverted. Therefore, epigenetic editing by artificial transcription factors (such as zinc finger-based artificial transcription factors or fusion of designer DNA binding domains to epigenetic writers/erasers) have been examined as novel therapeutics in killing of cancer cells as in general the genomic DNA sequence in these cells are not altered but only the epimodifications/epi-marks in the target chromatin region are manipulated which ultimately fine-tune the gene expression levels by the self, endogenous gene promoters (i.e., whether re-expressing of selected epigenetically silenced tumor suppressor genes or silencing of pro-metastatic genes/oncogenes) [65,66,67, 68•]. Moreover, natural dietary compounds such as astaxanthin (AST) [69•], phenethyl isothiocyanate (PEITC) [70], curcumin (CUR) [71], sulforaphane (SFN) [72], resveratrol (RSVL) [73], and epigallocatechin-3-gallate (EGCG) [74] have also been demonstrated to have the ability for epigenetically suppressing cancer cell invasion or proliferation, reactivating tumor suppressors, and enhancing cellular anti-oxidative stress responses, which suggested the beneficial role of dietary intake of these phytochemicals. Last but not the least, epi-drugs in combination with immunotherapeutics has been proposed as a new avenue to improve anticancer efficacy [75]. To sum up, all these studies have provided the potentials of epi-drugs that laid out the foundation and paved the way for future epi-drug research and development.

Conclusions

With the advancement of MS and sequencing technologies, novel epimodifications/epi-marks are gradually revealed in recent years. Given the fact that a variety of epimodifications can occur in DNA/RNA, proteins, and metabolites, it is possible that an unknown number of epimodifications/epi-marks might exist and are yet to be discovered. Nevertheless, it will be an exciting quest for scientists for the ongoing discovery of novel epimodifications and hopefully one day we will conquer these uncharted epi(c) territories and be able to monitor the global epi-omic signature of individuals. In this regard, the ability to decipher and manipulate the epi-omes might present new avenues in drug design for procuring better treatment of various human diseases. Suffice to say, the era of epi-omics research and epi-drug development is on its way.