3′-untranslated area

Length and nucleotide composition

The length of the 3′-UTR may be from 60 to 4000 nucleotides . The average length of 3′-UTR in humans is about 800 nucleotides, while the average length of 5′-UTR is 200 nucleotides ^[3] . It is noteworthy that the total length of 3′-UTR in humans is more than double that in other mammals , which indicates a greater number of regulatory elements in humans than in other mammals ^[4] . The composition of the bases also varies in 3′- and 5′-UTR. So, in 5′-UTR the content of G + C is higher than in 3′-UTR. This difference is especially noticeable in mRNA of warm-blooded vertebrates, in which the G + C content in 5′-UTR is 60%, and in 3′-UTR it is 45% ^{[5] [6]} .

The length and secondary structure of 3′-UTR are largely determined by its participation in the interactions of the 5′-end of the transcript with the 3′-end (see below), and often long 3′-UTRs have a significant effect on gene expression . In 1996, it was shown that an increase in 3′-UTR of mRNA from 19 to 156 nucleotides reduced expression by 45 times, regardless of the orientation, gene, or sequence of inserted nucleotides. This suggests that the 3′-UTR length is important in mRNA expression. Another factor determining the importance of the 3′-UTR length, in addition to the interaction of 3′- and 5′-UTR, is the ability of 3'-UTR to interact with miRNAs , special regulatory RNA molecules that suppress translation (see below for more details). These interactions are carried out on special sites, which are more in long 3′-UTR, so a long 3′-UTR can have a stronger suppressive effect on translation. So, a comparison was made of the length of the 3′-UTR and the number of microRNA binding sites on it for the ribosomal protein genes and genes involved in neurogenesis . It turned out that the ribosomal 3′-UTR genes have shorter and fewer specific miRNA binding sites, while the genes involved in neurogenesis, on the contrary, have 3′-UTR longer and contain many specific miRNA binding sites. Consider another example. The gene uses alternative 3′-UTRs for flexible expression control (see below for more on this phenomenon). The longer possible 3′-UTR of this gene contains conserved binding sites for two miRNAs expressed in activated T cells . Upon activation, the relative expression of the transcript with a longer 3′-UTR decreased, and the total expression of the protein increased as mRNAs with shorter 3′-UTR were expressed that did not contain binding sites for inhibitory miRNAs. It was also shown that the 3′-UTR length depends on the presence of such regulatory elements as AU-rich elements ( ARE ) in it (see below for more details) ^[4] .

In general, long 3′-UTRs are confined to a relatively low level of expression, as was shown in experiments that compared the expression of single protein isoforms whose mRNAs differed only in 3′-UTR length. The gene expressed in two mRNAs with different 3′-UTRs, the longer of which contains an additional microRNA binding site. Functional polymorphism in this gene is associated with the appearance of endothelial dysfunction and a hereditary predisposition to hypertension . Interestingly, the allele responsible for the manifestation of these disorders usually has a longer 3′-UTR, and therefore its expression level is lower than that of the wild-type allele with a shorter 3′-UTR ^[4] .

Introns

Unlike 5′-UTR, 3′-UTR contain relatively few introns (about 5%). Some mammalian genes resulting from reverse transcription from a spliced transcript have introns in 3′-UTR that reduce the expression of these genes, directing their transcripts to the NMD pathway (i.e., to destruction). This negative effect of introns in 3′-UTR on gene expression may explain their small distribution in this region. Moreover, it was found that some transcripts are able to bind to miRNAs only in the presence of an intron in 3′-UTR, which also inhibits gene expression. This shows that different excision of introns in the 3′-UTR makes it possible for isoform-specific regulation, mediated by microRNA, which can be tissue-specific ^[7] .

Secondary Structure

Apparently, the secondary structure of the 3'-UTR is of much greater importance than previously thought. Not only the 3'-UTR length is important, but also its secondary structure, and mutations that alter it can disrupt gene expression. In 2006, a study was carried out of 83 variants of 3′-UTR associated with various diseases, and a relationship was established between the functionality of these variants and changes in the predicted secondary structure ^[8] .

The secondary structure of 3′-UTR is difficult to predict, since many protein factors that bind to it can significantly affect its spatial structure. These factors can change it due to destruction of the mRNA folding, and can interact with other factors, because of which the mRNA can become locked in a loop. The most common example of secondary structure elements that can affect expression is the hairpin , and in 3'-UTR, the RNA-binding proteins bind to the hairpins. The transcript of the brain-derived neurotrophic factor (BDNF ) contains a long hairpin responsible for the stability of mRNA in neurons in response to calcium signals. It is assumed that the hairpin is a convenient platform for the interaction of a number of RNA-binding proteins, non-coding RNAs and polyadenylation signals in response to Ca ²⁺ . At the TNFα transcript in 3′-UTR there is an ARE element forming a hairpin that is able to modulate the affinity of this site for various proteins (see below for more details). These examples show that modulation of the 3'-UTR secondary structure using proteins or other means can change its specificity for binding to various trans '' factors, thereby regulating gene expression at the post- transcriptional level ^[9] .

Alternative 3′-UTR

Alternative polyadenylation ( APA ) and alternative splicing are two mechanisms leading to the appearance of different mRNA isoforms, differing in their 3′-UTR. APA may be due to the presence of different polyadenylation sites and different terminal exons ; APA uses estimated ~ 50% of human genes. This mechanism is very convenient for complex organisms, since it allows transcripts to be expressed in the same protein, but at different levels and in different spatial localization due to differences in regulation mediated by 3′-UTR. Alternative 3′-UTRs are extremely important for tissue-specific gene expression, as well as for different expression at different stages of development . Significant changes in ARA products are characteristic of a number of cancers . ARA also plays an important role in the localization of protein isoforms. The gene protein product is an ARE-binding protein involved in the stabilization of many mRNAs containing ARE. Thanks to ARA, a number of variants of HuR protein are formed that differ in expression level, and although the vast majority of transcripts of this protein lack ARE, some still have functional AREs in 3′-UTR. These AREs are able to bind HuR, thereby regulating the type of positive feedback. Thus, the use of alternative 3′-UTR makes it possible for an even greater variety of protein products of one gene ^[10] .

Interaction with miRNAs

MicroRNAs are short single-stranded non-coding RNA molecules of endogenous origin with a length of about 20 nucleotides. They interact with mRNA targets according to the principle of complementarity and usually block the translation of the target or cause its destruction. Typically, the binding sites of miRNAs with mRNAs are localized in the latter 3′-UTR, although some of them are located in the 5′-UTR and even the coding region. MicroRNAs are often expressed differently depending on the type of tissue and stage of development, and genes involved in processes common to all genes have to selectively avoid sequences in transcripts that are partially complementary to miRNAs, i.e., to avoid the presence of miRNA binding sites. This selective avoidance process has a huge impact on the evolution of 3′-UTR ^[11] .

Stabilization of mRNA

Changing the stability of the transcript allows for rapid control of expression without changing the rate of translation. Such a mechanism is important in such vital processes as cell growth and differentiation , as well as adaptation to environmental conditions. The most well-studied regulatory elements that regulate mRNA stability are AU-rich elements ( AU-rich elements ) located in the 3′-UTR mRNA of some genes. These elements are 50 to 150 nucleotides in size and usually contain numerous copies of the AUUUA pentanucleotide ^[12] .

It was found that the sequences of AREs are different, and 3 classes of AREs are distinguished by the number and arrangement of AUUUA motifs:

• I class : 1-3 scattered AUUUA located inside the U- rich area;
• II class : many overlapping motifs AUUUA;
• Grade III : AUUUA not, but there are U-rich sites ^[12] .

AREs bind to proteins ( ARE-binding proteins, ARE-BPs ), which, as a rule, contribute to the destruction of mRNA in response to various intracellular and extracellular signals, although some of them regulate translation . AREs regulate the expression of genes encoding cytokines , growth factors , tumor suppressor genes , proto-oncogenes , as well as genes whose protein products are involved in the regulation of the cell cycle , for example, cyclin genes, enzymes , transcription factors , receptors and membrane proteins . Such a diversity of genes whose transcripts contain AREs indicates the importance of transcript stability in gene regulation ^[12] . In addition to changing the stability of mRNAs, AREs can also activate translation, although this mechanism is less common and less understood ^[13] .

Another element that regulates the stability of the transcript is the recently discovered GU-rich element (GRE) . It interacts with , an RNA-binding protein that promotes the breakdown of mRNA associated with it ^[13] .

Participation in polyadenylation

Polyadenylation is the process of adding a series of adenosines (that is, a poly (A) tail) to the 3′-end of an immature RNA transcript ^[13] . It has been established that elements regulating this process are located in the 3′-UTR. Thus, it was shown that in all polyadenylated mRNAs at a distance of 20-30 nucleotides from the 3′-end of the transcript to which the poly (A) tail attaches, the sequence contains AAUAAA - polyadenylation signal (similar sequences, such as AU, can also act as polyadenylation signals) / GUAAA or UAUAAA). Subsequently, it turned out that although the AAUAAA sequence is absolutely necessary for polyadenylation, there are other elements without which the normal attachment of a poly (A) tail is impossible. In particular, a GU-rich sequence was identified that is located immediately after the AAUAAA towards the 3 'end (also called the English downstream sequence element, DSE ), as well as a special sequence located immediately before the AAUAAA ( English upstream sequence element, USE ) These elements are largely conservative not only for mammals , but for all eukaryotes . For polyadenylation, nucleotides located at the site of cutting the 3 ′ end of the transcript are also important (after this break, a poly (A) tail will be attached to the site). Thus, 3′-UTR plays a crucial role in the polyadenylation process ^[14] .

Participation in mRNA masking

3′-UTR plays a significant role in the process of masking mRNA . Masking of mRNA occurs, for example, in the process of oogenesis and spermatogenesis , when the mRNAs synthesized during these processes are not translated into the protein, but stored in an inactive state, sometimes for a rather long time. During fertilization and during early embryogenesis , maternal mRNAs are unmasked, and the necessary proteins are synthesized from them. Masking and storage of mRNA also occurs in differentiating somatic cells of an adult organism for a long time ^[15] .

The mRNA masking phenomenon was first studied in the bivalve mollusk in 1990. It turned out that a large number of masked mRNAs encoding the small subunit of and stored in its oocytes . It was shown that when the mRNA is in the masked state, a complex of masking proteins is associated with the site in its 3′-UTR. It also turned out that masked mRNAs have a strongly shortened poly (A) tail, from 200–250 adenyl residues to 20–40. When mRNA is unmasked, masking proteins are phosphorylated , as a result of which the cap is freed from the blocking protein and polyadenylation of the mRNA is stimulated by the cytoplasmic , which restores the long poly (A) tail necessary for efficient translation ^[16] .

Selenocysteine ​​Insert

3′-UTR is sometimes involved in the inclusion of a rare but functionally important amino acid , selenocysteine, into the polypeptide chain. There is no special codon for selenocysteine, and ^Sec tRNA is attached to the terminating UGA codon, but only when there is a special sequence for the insertion of selenocysteine ​​- SECIS , which forms a characteristic element of the secondary structure. SECIS can be located at a considerable distance (up to 200 nucleotides) from the UGA, and in archaea and eukaryotes it is localized in 3′-UTR mRNA ^{[17] [18]} .

Participation in NMD

NMD ( English nonsense-mediated decay ) is an effective mechanism for the destruction of non-functional mutant transcripts. Usually, in this mechanism, the efficiency is determined by the location of the mutation relative to the junction of exons, however, 3′-UTR may also have some significance. The mechanism for termination of translation at premature stop codons depends on the distance between the termination codon and the poly (A) -binding protein . It was shown that an increase in the distance between the stop codon and the poly (A) tail leads to the launch of NMD, and changes in the spatial structure of 3'-UTR can modulate NMD ^[8] .

5′-UTR and 3′-UTR Interaction

It is known that mRNA is capable of being locked into a ring (circularization) due to the interaction of special proteins that bind to the poly (A) tail , which contribute to the binding of factor eIF4F to the cap . As a result, mRNA takes on a closed form, translation initiation is stimulated, and translation efficiency is increased. However, in some cases, 5′-UTR and 3′-UTR of the same mRNA can bind to each other. So, the mRNA of the human p53 gene has regions in 5′-UTR and 3′-UTR that are complementary to each other. By binding to each other and to the translational factor , they thereby increase the translation efficiency of p53 protein in response to DNA damage ^[8] .

Analysis of the mRNA of various human genes showed that a 5′-UTR contains a motif that specifically interacts with the 3′-ends of microRNAs, and many of these mRNAs have a site at the 5′-end complementary to 3′-UTR. Further studies showed that binding of the 5′-UTR to miRNAs facilitates the binding of the 5′-end of mRNA to the 3′-end, and mRNAs whose activity is significantly determined by miRNAs have predictable binding sites at both UTRs. Such mRNAs are called miBridge. It was further established that the loss of these binding sites reduced the repression of transcript translation driven by miRNAs. Thus, it was found that the binding sites of UTR to each other are necessary to suppress translation of mRNA. This suggests that a complementary interaction of 5′-UTR and 3′-UTR is necessary for the precise regulation of gene expression ^[9] .

Bacteria

The bacterial mRNA also contains 5′- and 3′-untranslated regions ^{[19] [20]} .

Unlike eukaryotes, long 3′-UTRs are rare in bacteria and poorly understood. Nevertheless, it is known that some bacteria, in particular, Salmonella enterica , have mRNAs with long 3′-UTR similar to eukaryotic ones (in S. enterica it is hilD mRNA). It is assumed that 3′-UTR hilD perform various functions, in particular, affect the circulation of their mRNAs, since a deletion of these regions caused an increase in the number of corresponding mRNAs ^[21] .

Archaea

Untranslated regions are also found in the mRNA of many archaea . In particular, in the 5′- and 3′-UTR mRNAs of the methanogenic archaea (as in other representatives of the Methanopyrales and Methanococcales orders), the SECIS element is responsible for the insertion of the amino acid selenocysteine into the polypeptide chain ^[22] .

It was established that the mRNA of most haloarchae , as well as and lack pronounced 5′-UTR, but mRNAs of archaehemetanogens have long 5′-UTRs. In this regard, it is assumed that the mechanism of translation initiation of methanogenic archaea can be different from that of other representatives of this domain ^[23] . Nevertheless, there are 3′-UTRs in haloarchae mRNAs and their 3′-ends do not undergo post-transcriptional modification. What is surprising is the fact that in those transcripts of haloarchae that have 5′-UTR, there is no Shine – Dalgarno sequence. The length of the 3′-UTR of the haloarchae ranged from 20 to 80 nucleotides; no conservative structural motifs and sequences, except for the penta-U nucleotide in the field of translation termination, were detected ^[24] .

Viruses

У многих вирусов инициация трансляции происходит по кэп -независимому механизму и осуществляется через элементы IRES , локализованные в 5′-UTR ^[25] . Тем не менее, у вирусов обнаружен и другой кэп-независимый механизм инициации трансляции , не связанный с IRES. Такой механизм имеется у многих вирусов растений . В этом случае имеется особый кэп-независимый трансляционный элемент ( англ. cap-independent translation element (CITE) ), расположенный в 3′-UTR. Нередко CITE связывает факторы трансляции, например, комплекс eIF4F, и затем комплементарно взаимодействует с 5′-концом, доставляя факторы инициации трансляции к месту её начала ^[26] .

У вирусов, геном которых представлен одноцепочечной молекулой РНК положительной , 3′-UTR не только оказывает влияние на трансляцию, но также задействована в репликации : именно с неё начинается репликация вирусного генома ^[27] .

Вирус кори (род семейства Paramyxoviridae ) имеет геном, представленный одноцепочечной молекулой РНК отрицательной полярности. Для его генов М и F был установлен интересный механизм. мРНК этих генов имеют длинные UTR, на их долю приходится ~6,4 % всей мРНК. Хотя эти гены непосредственно не участвуют в репликации , 3′-UTR мРНК гена М увеличивает экспрессию белка М и тем самым запускает репликацию генома. В то же время 5′-UTR мРНК гена F снижает образование белка F и тем самым подавляет репликацию ^[28] .

При изучении структуры и функций 3′-UTR ученые используют несколько различных методов. Даже если для данной 3′-UTR показано наличие её в определённой ткани, для получения полного представления о её функциях необходимо проанализировать эффекты её различной локализации, выяснить срок функционирования, описать взаимодействия с транс-регуляторными белками, влияние на эффективность трансляции ^[29] . С помощью методов биоинформатики на основании анализа первичной структуры (то есть последовательности нуклеотидов) можно искать элементы ARE и сайты связывания с микроРНК в данной 3′-UTR. Экспериментальными методами устанавливаются последовательности, взаимодействующие с теми или иными транс-регуляторными белками, и в настоящий момент на основании данных секвенирования и экспериментальных данных возможно находить сайты взаимодействия с определёнными белками в данном транскрипте ^[30] . Искусственно индуцируя мутации в 3′-UTR, например, затрагивающие терминаторный кодон, сигнал полиаденилирования или вторичную структуру 3′-UTR, можно установить, как мутации в этих участках могут приводить к нарушениям трансляции и появлению болезней (подробнее о заболеваниях, ассоциированных с 3′-UTR, см. ниже) ^[31] . Итак, с помощью всех этих методов мы можем развивать наши представления о структуре и функциях цис-регуляторных элементов в 3′-UTR, а также взаимодействующих с 3′-UTR белков.

Мутации, затрагивающие 3′-UTR, имеют важное значение, поскольку одна такая мутация может сказаться на экспрессии многих генов. Хотя на уровне транскрипции мутации влияют на конкретный аллель и физически сцепленные гены, поскольку связывающиеся с 3′-UTR белки также принимают участие в процессинге и экспорте МРНК из ядра. Таким образом, мутация может влиять на несвязанные гены ^[32] . Так, мутации, произошедшие в ARE, приводят к сбою в работе ARE-связывающих белков, в результате чего могут развиться такие заболевания, как злокачественные перерождения кроветворных органов и лейкемия ^{[33] [34]} . Повышенное содержание тринуклеотида CTG в 3′-UTR гена вызывает . Вставка ретротранспозона длиной 3 килобаз, состоящего из тандемных повторов , в 3′-UTR гена белка , связана с врожденной мышечной дистрофией типа Фукуяма ^[29] . Изменения в элементах, локализованных в 3′-UTR, связаны с развитием таких заболеваний человека, как острый миелоидный лейкоз , , нейробластома , кератинопатия, аниридия , , врождённые пороки сердца ^[31] . Связь некоторых из этих заболеваний с конкретными элементами 3′-UTR представлена на схеме ниже.

• Spirin A.S. Molecular Biology. Ribosomes and protein biosynthesis. - M .: Publishing Center "Academy", 2011. - 496 p. - ISBN 978-5-7695-6668-4 .
• Konichev A.S., Sevastyanova G.A. Molecular biology. - Publishing Center "Academy", 2012. - 400 p. - ISBN 978-5-7695-9147-1 .
• Lucy W. Barrett, Sue Fletcher, Steve D. Wilton. Untranslated Gene Regions and Other Non-coding Elements . - Springer Briefs in Biochemistry and Molecular Biology, 2013 .-- 57 p. - ISBN 978-3-0348-0679-4 .