The Elements of Life
In biology, the elements of life are the essential building blocks that make up living things. They are carbon, hydrogen, nitrogen, oxygen, phosphorus, and sulfur. The first four of these are the most important, as they are used to construct the molecules that are necessary to make up living cells. These elements form the basic building blocks of the major macromolecules of life, including carbohydrates, lipids, nucleic acids and proteins. Carbon is an important element for all living organisms, as it is used to construct the basic building blocks of life, such as carbohydrates, lipids, and nucleic acids. Even the cell membranes are made of proteins. Carbon is also used to construct the energy-rich molecules adenosine triphosphate [ATP] and guanosine triphosphate [GTP]. Hydrogen is used to construct the molecules water and organic compounds with carbon. Hydrogen is also used to construct ATP and GTP. Nitrogen is used to construct the basic building blocks of life, such as amino acids, nucleic acids, and proteins. It is also used to construct ATP and GTP. Oxygen is used to construct the basic building blocks of life, such as carbohydrates, lipids, and nucleic acids. It is also used to construct ATP and GTP. Phosphorus is used to construct the basic building blocks of life, such as carbohydrates, lipids, and nucleic acids.
Introduction
Starting with the discovery of transfer RNA and ribosomal RNA in the 1950s, non-coding RNAs [ncRNAs] with biological roles have been known for close to 60 years. Even in the late 1970s and early 1980s the existence of other functional ncRNAs was known, including RNAse P [Stark et al., 1978], snRNAs [Yang et al., 1981], and 7SL [the RNA component of the signal recognition particle [Walter and Blobel, 1982]]. Later, ncRNAs that serve to regulate chromosome structure, such as Xist, were discovered [Brockdorff et al., 1992]. Since then, the number of new and putative functional ncRNAs has greatly expanded [for reviews see Wilusz et al., 2009; Wang and Chang, 2011; Ulitsky and Bartel, 2013; Rinn and Guttman, 2014]. Interest in this field was further stimulated by the finding that almost all of the mammalian genome is transcribed at some level [Carninci et al., 2005; Birney et al., 2007; Djebali et al., 2012], with some individuals speculating that much of this pervasive transcription is likely functional [Mattick et al., 2010; Ecker et al., 2012; Pennisi, 2012]. This idea was epitomized by the ENCODE consortium, which claimed to have assigned “biochemical functions for 80% of the genome” [ENCODE Project Consortium et al., 2012]. Others have disagreed, pointing out that the vast majority of these novel transcripts are present at low levels, and that the term “function” had been misappropriated [Eddy, 2012; Doolittle, 2013; Graur et al., 2013; Niu and Jiang, 2013; Palazzo and Gregory, 2014]. Despite these criticisms, the idea that the pervasive transcription of the human genome plays some role in homeostasis and/or development persists, with one group even proclaiming that they had “refuted the specific claims that most of the observed transcription across the human genome is random” [Mattick and Dinger, 2013].
At present, the distinction between functional ncRNAs and junk RNA appears to be quite vague. There has been, however, some effort to differentiate between these two groups, based on various criteria ranging from their expression levels and splicing to conservation. Ultimately these efforts have failed to bring consensus to the field.
A similar problem has plagued the investigation of whether transposable elements [TEs], which make up a significant proportion of most vertebrate genomes, have been exapted for the benefit of the host organism. Although some have claimed that many TEs are functional, a few groups have offered a much more balanced view that is in line with our current understanding of molecular evolution [de Souza et al., 2013; Elliott et al., 2014].
In this article we explain several concepts that researchers must keep in mind when evaluating whether a given ncRNA has a function at the organismal level. Importantly, the presence of low abundant non-functional transcripts is entirely consistent with our current understanding of how eukaryotic gene expression works and how the eukaryotic genome is shaped by evolution. With this in mind, researchers should take the approach that an uncharacterized non-coding RNA likely has no function, unless proven otherwise. This is the null hypothesis. If a given ncRNA has supplementary attributes that would not be expected to be found in junk RNA, then this would provide some evidence that this transcript may be functional.
The Amount of Various RNA Species in the Typical Eukaryotic Cell
As is evident from a number of sources, almost all of the human genome is transcribed. However, one must not confuse the number of different types of transcripts with their abundance in a typical cell. Many of the putative functional ncRNAs are present at very low levels and thus unlikely to be of any importance with respect to cell or organismal physiology. Importantly, the abundance of an ncRNA species roughly correlates with its level of conservation [Managadze et al., 2011], which is a good proxy for function [Doolittle et al., 2014; Elliott et al., 2014; however, see below]; thus, determining the relative abundance of a given ncRNA in the relevant cell type is an important piece of information. However, one should keep in mind that if the ncRNA has catalytic activity or if it acts as a scaffold to regulate chromosomal architecture near its site of transcription, the RNA may not need to be present at very high levels to be able to perform its task.
At steady state, the vast majority of human cellular RNA consists of rRNA [∼90% of total RNA for most cells, see Table 1 and Figure 1]. Although there is less tRNA by mass, their small size results in their molar level being higher than rRNA [Figure 1]. Other abundant RNAs, such as mRNA, snRNA, and snoRNAs are present in aggregate at levels that are about 1–2 orders of magnitude lower than rRNA and tRNA [Table 1 and Figure 1]. Certain small RNAs, such as miRNA and piRNAs can be present at very high levels; however, this appears to be cell type dependent.
TABLE 1. Estimates of total RNA content in mammalian cells.
FIGURE 1. Estimate of RNA levels in a typical mammalian cell. Proportion of the various classes of RNA in mammalian somatic cells by total mass [A] and by absolute number of molecules [B]. Total number of RNA molecules is estimated at roughly 107 per cell. Other ncRNAs in [A] include snRNA, snoRNA, and miRNA. Note that due to their relatively large sizes, rRNA, mRNA, and lncRNAs make up a larger proportion of the mass as compared to the overall number of molecules.
By general convention, most other ncRNAs longer than 200 nucleotides, regardless of whether or not they have a known function, have been lumped together into a category called “long non-coding RNAs” [lncRNAs]. As a whole, these are present at levels that are two orders of magnitude less than total mRNA [Table 1]. Although the estimated number of different types of human lncRNAs has ranged from 5,400 to 53,000 [Table 2], only a small fraction have been found to be present at levels high enough to suggest that they have a function. According to ENCODE’s own estimates, fewer than 1,000 lncRNAs are present at greater than one copy per cell in the typical human tissue culture cell line [Djebali et al., 2012; Palazzo and Gregory, 2014], although some other estimates have determined that the levels may be substantially higher [Hangauer et al., 2013]. One caveat with the data collected thus far is that some of these lncRNAs may have a very restricted expression pattern; therefore until the relevant cell type is tested, we may not be in a position to judge whether it is expressed at a sufficient level to provide evidence of functionality. It is also worthwhile noting that certain annotated lncRNAs may actually encode short functional peptides [Ingolia et al., 2011, 2014; Magny et al., 2013; Bazzini et al., 2014], although in general lncRNAs are poorly translated [Bánfai et al., 2012; Guttman et al., 2013; Hangauer et al., 2013]. Finally, it is also worth pointing out that a significant fraction of these lncRNAs may actually be misannotated untranslated regions of known mRNAs [Miura et al., 2013].
TABLE 2. Estimate number of human ncRNAs from various sources.
Other short ncRNAs have been lumped into several groups, depending on their attributes. For example, several regions of the human genome that are believed to be enhancers, are transcribed into short enhancer RNAs [eRNAs]. These are thought to act as scaffolds that regulate the 3D architecture of chromosomes in the vicinity of their transcription site [Lai et al., 2013]. eRNAs are typically present at even lower levels than lncRNAs [Djebali et al., 2012; Andersson et al., 2014]; however, if these play a localized structural role, then they would be expected to be present at only a few copies per cell.
There are still other more exotic species of RNAs [Cech and Steitz, 2014], including circular RNAs [Wilusz and Sharp, 2013]. Due to their lack of free 5′ or 3′ ends, circular RNAs are quite stable and some can accumulate to levels that are comparable to mRNAs [Salzman et al., 2012; Jeck et al., 2013; Memczak et al., 2013]. However, it is likely that besides a few examples, circular RNAs represent a minute fraction of the total pool of cellular RNAs [see Table 1].
In addition to all of the mentioned species, ENCODE and other groups have found transcripts that map to the rest of the genome termed “intergenic RNA” [Djebali et al., 2012]. Most of these transcripts are present at levels that are significantly below one copy per cell [Djebali et al., 2012; Palazzo and Gregory, 2014]. Again this arbitrary division of ncRNAs has led to much confusion. It is unclear why these transcripts are considered to be intergenic if they are also functional [as in 80% of the genome is functional]; after all, if a region of DNA that is transcribed into a functional product is called a gene, then the term intergenic would automatically imply that these regions have no function.
Regardless of these concerns, it is clear that most of the ncRNAs in question [lncRNAs, eRNAs, circular RNAs, intergenic RNAs, etc.] are typically present at very low levels when compared to known functional RNAs. These observations are consistent with the idea that the eukaryotic genome produces a vast amount of spurious transcripts.
Where Do All These ncRNAs Come From?
As of spring 2014, the LNCipedia website1 [Volders et al., 2013] has compiled a list of ∼21,000 human lncRNAs, with an average length of about 1 kb [Table 2]. These would originate from