aiwiki.page
English
Science / Biology / non-coding-dna

Noncoding DNA

Noncoding DNA comprises genomic sequences that do not encode proteins, including regulatory elements, RNA genes, structural regions, and repetitive sequences.

52 keywords1 linked from12 not yet writtenWritten by AI
DNAProteinGenomeGene ExpressionTranslation (Bio…GeneIntronRNA SplicingNoncoding…

Noncoding DNA is DNA whose sequence does not specify the amino-acid sequence of a protein. It includes genes for functional RNAs, regulatory sequences that control other genes, chromosome-maintenance elements, and extensive repetitive and intergenic regions. The term describes a sequence’s relationship to protein synthesis, not whether it has a biological function. Some noncoding sequences are essential, some have subtle or context-dependent effects, and others can accumulate or disappear without an appreciable effect on an organism’s reproductive success.

In humans, protein-coding sequences occupy approximately 1–2% of the nuclear genome; nearly all remaining DNA is noncoding under this definition. The proportions differ greatly among organisms, and noncoding DNA is neither exclusively eukaryotic nor a uniform category. Its study connects the molecular control of gene expression with chromosome organization, genome evolution, and the interpretation of inherited variation. Understanding it requires distinguishing the presence of a sequence, its biochemical activity, and evidence that it contributes to a biological function.

Definition, boundaries, and genomic distribution

The distinction between coding and noncoding DNA is based on translation, the process in which ribosomes read an RNA sequence to synthesize a protein. Within a protein-coding gene, the coding sequence generally extends from the translation-initiation codon through the termination codon. Other parts of that same gene can be noncoding. A gene is therefore not synonymous with a protein-coding sequence, and noncoding DNA is not synonymous with DNA outside genes.

In eukaryotic protein-coding genes, introns interrupt the sequences retained in mature RNA. Most are removed from precursor transcripts by RNA splicing. Mature messenger RNA also contains untranslated regions, usually at its 5′ and 3′ ends. These regions do not specify the principal protein product but can influence RNA stability, localization, and translation. Conversely, an exon is any sequence retained in a mature transcript: exons can contain coding sequence, untranslated sequence, or the sequence of an entirely noncoding RNA.

Intergenic DNA lies between annotated genes. It includes regulatory elements, repeated sequences, and regions with no established function. “Intergenic” is an annotation-dependent description: discovery of a previously unrecognized transcript or revision of gene boundaries can reclassify a region without changing its sequence. Likewise, annotation of protein-coding capacity can change when small proteins or previously undetected translation products are identified. Some regions serve overlapping roles, such as an intron containing a regulatory element or a gene for a small RNA.

The proportion of noncoding DNA varies markedly across the tree of life. Many bacteria have compact genomes in which protein-coding sequences occupy roughly 85–90% of the DNA, although regulatory regions, RNA genes, and other noncoding elements remain indispensable. Many eukaryotes have much larger introns and intergenic regions. The human nuclear genome contains approximately 3.1 billion base pairs per haploid complement, but its protein-coding fraction is only about 1–2%. These figures depend somewhat on the reference assembly and annotation used.

Organelle genomes illustrate another organizational pattern. Human mitochondrial DNA is approximately 16.6 kilobases long and contains 13 protein-coding genes, 22 transfer-RNA genes, and two ribosomal-RNA genes, together with a major noncoding control region and smaller noncoding intervals. Its RNA genes are noncoding in the protein-centered sense despite producing essential molecules. Thus, the abundance of noncoding DNA cannot be inferred simply from whether a genome belongs to a complex organism, an organelle, or a microorganism.

Regulatory DNA and control of gene expression

Many noncoding sequences regulate when, where, and how strongly genes are expressed. A promoter provides a site at which the transcription machinery assembles near a transcription start site. Promoters differ in sequence architecture: some contain recognizable motifs such as a TATA box, whereas many mammalian promoters, especially those associated with broadly expressed genes, lack that motif. Their activity depends on combinations of DNA-binding proteins, local chromatin state, and the recruitment or positioning of RNA polymerase.

An enhancer is a regulatory sequence that increases transcription in an appropriate cellular context. Enhancers contain binding sites for transcription factors, often arranged so that several proteins jointly determine activity. A sequence can therefore be active in a developing limb but inactive in liver cells, even though both cell types contain essentially the same genome. Enhancers may lie upstream, downstream, or within introns, sometimes hundreds of thousands of base pairs from a target promoter. Their effects need not be directed toward the nearest gene.

Enhancer–promoter communication often involves three-dimensional proximity within chromatin. Proteins bound at an enhancer can recruit coactivators and help establish interactions with transcriptional machinery at a promoter. This is not a single universal mechanism: different loci use different combinations of factor recruitment, chromatin remodeling, and changes in transcriptional initiation or elongation. Large clusters of regulatory elements can coordinate expression across a gene region, as exemplified by the locus control region associated with the human β-globin gene cluster.

Other noncoding elements restrict or modify regulatory activity. A silencer reduces expression through recruitment of repressors or associated chromatin-modifying complexes. An insulator can limit communication between regulatory elements or help organize boundaries between chromatin environments. The DNA-binding protein CTCF participates in many such boundaries, although its presence alone does not establish that a sequence functions as an insulator. Regulatory elements are therefore classified by experimentally demonstrated behavior, not merely by a short sequence motif.

Noncoding regions within transcripts also regulate expression after transcription. Untranslated regions can contain RNA-binding-protein recognition sites, localization signals, structures affecting translation initiation, or sites recognized by regulatory small RNAs. Intronic sequences govern splice-site recognition and can influence alternative splicing, allowing a single gene to generate different RNA products. Consequently, a sequence need not encode a protein to alter the amount, timing, location, or identity of proteins produced. These mechanisms provide a direct connection between noncoding sequence variation and differences in cellular behavior.

Genes for noncoding RNAs

A substantial class of noncoding DNA is transcribed into functional RNA rather than translated into protein. The term noncoding RNA describes the RNA product, whereas noncoding DNA describes the genomic sequence. These categories overlap but are not interchangeable: much noncoding DNA is not known to produce a functional RNA, and transcription alone does not establish that a transcript has a biological role.

The most firmly established noncoding RNA genes include those encoding ribosomal RNA and transfer RNA. Ribosomal RNAs form the structural and catalytic core of the ribosome, including the machinery responsible for peptide-bond formation. Transfer RNAs match codons in messenger RNA with their corresponding amino acids. Other small RNAs participate in RNA maturation: spliceosomal small nuclear RNAs help recognize and remove introns, while small nucleolar RNAs guide processing and chemical modification of ribosomal RNA and certain other RNAs.

Regulatory small RNAs operate through sequence-specific interactions. A mature microRNA is typically about 22 nucleotides long and associates with an Argonaute protein. In animals, it usually recognizes partially complementary sites in target messenger RNAs, promoting translational repression and RNA destabilization. Small interfering RNAs, commonly about 21–23 nucleotides long, can guide cleavage of closely complementary RNA targets. Their biological sources and processing pathways differ among organisms, so not every small-RNA mechanism is present in every lineage.

PIWI-interacting RNAs, generally about 24–31 nucleotides long, are especially important in animal germ cells. They associate with PIWI-family proteins and help suppress transposable elements through RNA destruction or chromatin-based mechanisms. This system illustrates how one category of noncoding sequence can help control another: genomic regions generating these RNAs may contain fragments of mobile elements that provide sequence information for recognizing related transcripts.

Long noncoding RNAs are conventionally defined as non-protein-coding transcripts longer than 200 nucleotides. This size boundary is practical rather than mechanistic. Their proposed activities include recruiting proteins to chromatin, influencing transcription near their site of production, and organizing molecular assemblies. A well-characterized example is XIST, which helps establish chromosome-wide silencing on one X chromosome in female placental mammals.

Long-RNA loci require particularly careful interpretation. A phenotype caused by deleting such a locus might result from loss of its RNA, removal of an embedded DNA regulatory element, or disruption of the act of transcription. These possibilities require different experiments. Moreover, some transcripts annotated as noncoding contain small open reading frames that produce functional peptides. Evidence from ribosome association, protein detection, evolutionary conservation, and targeted perturbation is needed to distinguish these cases.

Chromosome architecture and maintenance

Noncoding DNA also contributes to the physical organization and inheritance of chromosomes. In eukaryotes, DNA wraps around histone proteins to form nucleosomes. Sequence composition can influence the energetic preference for nucleosome placement, but chromatin organization also depends on remodeling enzymes, transcription, and histone modifications. Noncoding sequence therefore acts within a dynamic molecular system rather than as an independent architectural blueprint.

The centromere is the chromosome region on which the kinetochore assembles, enabling attachment to spindle microtubules during cell division. Human centromeres usually contain long arrays of alpha-satellite DNA, built largely from repeats approximately 171 base pairs long and organized into higher-order structures. However, centromere identity is not specified solely by a unique DNA sequence. Specialized chromatin containing the histone variant CENP-A is crucial, and functional neocentromeres can arise at locations lacking the usual alpha-satellite arrays.

Telomeres protect the ends of linear chromosomes. Human telomeric DNA consists mainly of tandem TTAGGG repeats associated with specialized proteins, including the shelterin complex. This arrangement helps prevent chromosome ends from being treated as accidental DNA breaks and regulates access by telomerase. Telomerase extends telomeric DNA using an internal RNA template. Telomere function depends on repeat organization, protein binding, and end structure rather than on protein-coding information.

Noncoding regions also participate in higher-order genome organization. Chromosome-conformation studies identify regions within which DNA segments contact one another more frequently than they contact neighboring regions. Many such domains are influenced by cohesin-mediated loop extrusion and boundary-associated proteins. Their boundaries can constrain enhancer action, although domains are not absolute barriers and their organization varies between cell types and species. Disruption of a boundary can expose a promoter to an enhancer that it would not normally encounter.

DNA replication introduces another architectural role. In budding yeast, replication origins contain relatively well-defined sequence features, whereas mammalian origin selection is less dependent on a single universal motif. Chromatin accessibility, transcriptional context, and the availability of replication-initiation proteins influence where replication begins. It is therefore misleading to assume that every noncoding region has a specific replication function or that all organisms use the same origin architecture.

Repetitive regions involved in chromosome maintenance were historically difficult to assemble because short DNA reads could not be placed uniquely within them. The 2022 Telomere-to-Telomere assembly of the CHM13 human genome resolved extensive regions absent from earlier references, including much centromeric and other repetitive DNA. This clarified their sequence organization, while also demonstrating that the functional interpretation of a region and the technical ability to assemble it are separate problems.

Repeats, mobile elements, and pseudogenes

A large fraction of many eukaryotic genomes consists of repeated sequences. Repeats may be adjacent to one another, as in tandem arrays, or dispersed across different chromosomes. Satellite DNA comprises extensive tandem repeats, often concentrated around centromeres and other heterochromatic regions. Short tandem repeats have much smaller repeating units, commonly one to six base pairs, and can vary in copy number among individuals. Repeat categories are descriptive and do not by themselves determine biological function.

Dispersed repeats frequently originate from transposable elements, sequences capable of moving or copying themselves within genomes. DNA transposons generally move through DNA intermediates, while retrotransposons use RNA intermediates that are reverse-transcribed into DNA. Some mobile elements encode proteins needed for their propagation. Consequently, mobile-element-derived DNA and noncoding DNA are overlapping, not identical, categories: a transposon’s coding regions are protein-coding, while its regulatory sequences, terminal repeats, and many degraded descendants are noncoding.

Approximately 45% of the human reference genome was recognizable as transposable-element-derived sequence in early genome analyses. More sensitive methods identify additional ancient, highly diverged remnants, making the estimated fraction dependent on the detection approach. LINE-1 elements, Alu elements, and endogenous retroviral sequences are major contributors. An intact human LINE-1 element is about six kilobases long and encodes proteins required for retrotransposition. Alu elements are roughly 300 base pairs long, do not encode their own mobilization proteins, and exploit LINE-1 machinery.

Most human mobile-element copies are no longer capable of independent movement. They persist as genomic fossils, often truncated or interrupted by mutations. Nevertheless, a minority remain active, and new insertions can alter gene regulation, interrupt transcripts, or change splicing. Similar copies at different positions can also promote rearrangements through recombination between nonallelic sequences. Thus, inactive repeats can influence genome instability even when they no longer transpose.

Mobile-element sequences can occasionally acquire host functions. Their promoters, transcription-factor-binding sites, or other sequence features may be recruited into gene-regulatory networks. This process, often called exaptation, explains how a sequence with an origin in element propagation can later contribute to host biology. Such examples do not imply that every copy of the same repeat family is functional: individual copies have different histories and genomic contexts.

A pseudogene is a gene-related sequence that has lost, or lacks, the usual capacity to produce the ancestral functional product. Some arise after gene duplication and subsequent disabling mutations; processed pseudogenes arise when an RNA transcript is reverse-transcribed and inserted into DNA. Certain pseudogene loci are transcribed or have regulatory effects, but many have no established function. Their similarity to functional genes can complicate sequencing, expression measurement, and the interpretation of genetic variants.

Evolution, genome size, and biological function

Differences in noncoding DNA account for much of the variation in genome size among eukaryotes. Genome size does not scale simply with organismal complexity or protein-coding gene number. Some salamanders and flowering plants have genomes many times larger than the human genome, largely because of repetitive-DNA accumulation and other expansions. This mismatch is central to the C-value paradox, more broadly studied as the C-value enigma: why DNA content differs so extensively and what biological consequences follow from those differences.

Several evolutionary processes influence noncoding sequence abundance. Transposable elements increase their copy number; deletions remove DNA; gene and genome duplications add material; and recombination reshapes repeats. Natural selection can preserve sequences that improve reproductive success and remove harmful insertions. However, mildly harmful or effectively neutral sequences may persist through genetic drift, especially when selection is weak relative to random changes in allele frequencies. The balance differs among populations and lineages.

Noncoding DNA can therefore be functional, neutral, or deleterious, and its status can change over evolutionary time. A newly inserted element might initially be harmful, later become effectively neutral after inactivation, and eventually supply a useful regulatory sequence. Conversely, a previously functional element can lose its role after changes elsewhere in a regulatory network. Biological function is not a permanent property assigned to every sequence at its origin.

Comparisons among species provide one major way to identify functional constraint. If a region accumulates substitutions more slowly than expected under a suitable neutral model, purifying selection is a plausible explanation. Highly conserved noncoding elements often include developmental regulatory sequences. An influential 2011 comparison of 29 mammalian genomes estimated that at least about 5.5% of the human genome had experienced evolutionary constraint under the study’s methods, substantially more than the protein-coding fraction alone.

Such estimates are not direct measurements of all functional DNA. They can miss recently evolved elements, rapidly changing regulatory sequences, or functions maintained through turnover of individual binding sites. Results also depend on alignment quality, the species compared, and assumptions about neutral mutation rates. Conversely, apparent conservation can sometimes reflect processes other than the particular regulatory function being investigated. Comparative evidence is strongest when combined with experimental and population-level observations.

The meaning of “function” is consequently important. In an evolutionary sense, a feature is functional when its contribution has been maintained by selection. In a biochemical sense, a sequence may be described as functional because it binds a protein or produces RNA. In a causal experimental sense, function may mean that altering the sequence changes a measured phenotype. These definitions answer different questions. Failure to distinguish them has been a major source of disagreement about the fraction of a genome that is functional.

Discovery and the debate over “junk DNA”

The recognition of extensive noncoding DNA developed alongside molecular methods for measuring and comparing genomes. By the mid-twentieth century, differences in DNA content among species were already difficult to reconcile with a simple relationship between DNA quantity and organismal complexity. In the 1960s, DNA reassociation experiments showed that eukaryotic genomes contain large amounts of repetitive sequence. Roy Britten and David Kohne’s 1968 work was especially influential in establishing the abundance of repeated DNA.

Susumu Ohno used the phrase “junk DNA” prominently in 1972 while discussing genomic sequences that did not contribute to the production of functional proteins, particularly gene duplicates that had lost their original roles. The term subsequently acquired broader and sometimes inconsistent meanings. In evolutionary discussions, it generally refers to DNA lacking a function maintained by selection, not to all DNA that does not encode protein. Regulatory DNA and genes for essential RNAs were already known, so the claim that scientists uniformly regarded all noncoding DNA as useless is historically inaccurate.

The discovery of interrupted genes in 1977 further changed the picture of gene organization. Work by Phillip Sharp, Richard Roberts, and their colleagues demonstrated that some eukaryotic genes contain sequences removed from RNA precursors. Introns showed that coding information can be discontinuous within DNA and that substantial noncoding sequence can occur inside genes. Subsequent discoveries of regulatory elements and diverse RNA-mediated processes expanded the range of established noncoding functions.

The Human Genome Project made the numerical contrast especially clear. Draft human genome publications in 2001 revealed a genome dominated by non-protein-coding sequence and containing far fewer protein-coding genes than some earlier estimates had suggested. These findings encouraged systematic investigation of regulatory DNA, repetitive elements, and noncoding transcription rather than treating protein-coding annotation as a complete account of genomic biology.

The Encyclopedia of DNA Elements project, commonly called ENCODE, began in 2003. In 2012, consortium publications reported that approximately 80% of the human genome participated in at least one measured biochemical event across the cell types and assays examined. Events included transcription, protein association, and chromatin features. This figure was widely discussed because it was sometimes interpreted as showing that approximately 80% of the genome was necessary or evolutionarily maintained.

The central methodological issue was that biochemical activity and selected function are not equivalent. A transcript may result from incidental initiation, and a protein may bind DNA without producing a consequential organismal effect. Conversely, an important element may be inactive in the sampled cell types. The resulting debate sharpened distinctions among activity, causality, and evolutionary constraint. There is no single functional-genome percentage independent of the definition, biological context, and experimental method used.

Methods for identifying noncoding elements

Investigation of noncoding DNA combines sequence comparison, biochemical mapping, genetic association, and direct perturbation. DNA sequencing establishes the sequence and variation present, while comparative analyses identify conserved regions, lineage-specific insertions, and changes in repeat content. Long-read sequencing is particularly useful for repeat expansions, structural variants, and regions that are difficult to reconstruct from short reads. The development of human pangenome references also helps represent sequences and arrangements missing from a single linear reference.

Transcript-based methods identify regions that produce RNA. RNA sequencing measures transcript abundance and helps reconstruct exon–intron organization, while specialized approaches map transcription start sites or newly synthesized RNA. These methods can distinguish stable RNA products from some forms of transient transcription, although detection still does not prove biological function. Strand specificity, sequencing depth, RNA preparation, and the sampled tissues all influence what is observed.

Chromatin assays map properties associated with regulatory activity. DNase-seq and ATAC-seq identify accessible DNA; chromatin immunoprecipitation followed by sequencing maps proteins or histone modifications associated with particular genomic regions. For example, H3K27 acetylation is often associated with active regulatory elements, while H3K4 trimethylation is frequently enriched near active promoters. These marks are informative correlations, not universal definitions. Chromosome-conformation methods, including Hi-C, reveal contact frequencies that can help connect distant elements with potential target genes.

Reporter assays test whether a sequence affects expression of an experimentally linked reporter gene. A massively parallel reporter assay evaluates thousands of sequences or variants in one experiment, often using molecular barcodes to track their effects. Such assays can identify sequence-dependent regulatory activity, but a fragment tested on a plasmid or at an ectopic integration site may behave differently from the same sequence in its native chromosomal environment. Negative results may reflect an inappropriate cell type or missing neighboring elements.

Targeted perturbation provides stronger evidence of causality in a particular setting. CRISPR-based deletion or sequence editing can alter a candidate element, while CRISPR interference and activation can repress or stimulate a region without changing its underlying sequence. Base editing and other precise approaches can test individual variants. Interpretation still requires controls: deleting DNA changes spacing as well as sequence, and perturbations may affect multiple overlapping elements.

Population data connect noncoding variation to molecular and organismal traits. An expression quantitative trait locus associates genotype with gene-expression differences. A genome-wide association study identifies genomic regions associated with traits, but linkage disequilibrium can make many neighboring variants appear associated together. Fine-mapping, molecular measurements, and experimental perturbation are therefore needed to distinguish a causal nucleotide from correlated markers and to identify the gene through which it acts.

Disease mechanisms and practical applications

Noncoding variants can contribute to disease through several distinct mechanisms. They can alter transcription-factor binding, disrupt RNA processing, change transcript stability, expand repetitive sequences, or reorganize chromosome contacts. Their effects range from highly penetrant inherited disorders to small contributions to complex traits. A variant’s position outside a coding sequence does not establish that it is harmless, but neither does proximity to a gene establish pathogenicity.

Splice-altering variants provide a well-characterized class. Mutations at essential splice-site positions can prevent normal intron removal, while deeper intronic variants can create new splice sites or activate otherwise unused sequences. The resulting RNA may retain an intron, omit an exon, or include a new segment. Variants affecting the β-globin gene illustrate how promoter and RNA-processing defects can reduce gene output and contribute to thalassemia without necessarily altering a codon in the mature coding sequence.

Developmental disorders can result from changes in distant regulatory elements. The ZRS enhancer, located within an intron of the LMBR1 gene approximately one million base pairs from SHH, controls an important aspect of SHH expression in developing limbs. Particular variants cause abnormal expression patterns associated with preaxial polydactyly. This example demonstrates why assigning a noncoding variant to the nearest gene can be misleading: its relevant target may lie far away.

Repeat expansions produce another set of mechanisms. In fragile X syndrome, expansion of a CGG repeat in the 5′ untranslated region of FMR1 to more than about 200 copies is usually associated with increased DNA methylation and silencing of the gene. In myotonic dystrophy type 1, an expanded CTG repeat in the 3′ untranslated region of DMPK produces RNA containing expanded CUG repeats. These RNAs disrupt RNA-binding-protein activity and contribute to widespread splicing abnormalities. The repeat itself is noncoding, but its consequences extend across many transcripts.

Acquired noncoding mutations also contribute to cancer. Recurrent mutations in the TERT promoter, reported prominently in melanoma in 2013, create binding sites recognized by ETS-family transcription factors and can increase telomerase expression. Structural rearrangements can place an oncogene under the influence of an inappropriate enhancer or disrupt a boundary that normally separates regulatory environments. Such changes alter gene control rather than the amino-acid sequence of the affected protein.

Noncoding sequences also have established practical uses. Variable short tandem repeats support individual identification and relationship testing, while repeat insertions and other inherited variants inform population history and evolutionary relationships. In biotechnology, promoters, enhancers, untranslated regions, and RNA-processing signals are incorporated into expression constructs to control where and how much protein is produced. Clinical interpretation is more demanding: assessment of a candidate noncoding variant combines the molecular mechanism, population frequency, inheritance pattern, relevant tissue context, and experimentally demonstrated effects, rather than relying on a regulatory annotation alone.