Mol. Cells 2023; 46(1): 48-56
Published online January 20, 2023
https://doi.org/10.14348/molcells.2023.0003
© The Korean Society for Molecular and Cellular Biology
Correspondence to : sjsj@dankook.ac.kr
This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/.
Genomic information stored in the DNA is transcribed to the mRNA and translated to proteins. The 3′ untranslated regions (3′UTRs) of the mRNA serve pivotal roles in posttranscriptional gene expression, regulating mRNA stability, translation, and localization. Similar to DNA mutations producing aberrant proteins, RNA alterations expand the transcriptome landscape and change the cellular proteome. Recent global analyses reveal that many genes express various forms of altered RNAs, including 3′UTR length variants. Alternative polyadenylation and alternative splicing are involved in diversifying 3′UTRs, which could act as a hidden layer of eukaryotic gene expression control. In this review, we summarize the functions and regulations of 3′UTRs and elaborate on the generation and functional consequences of 3′UTR diversity. Given that dynamic 3′UTR length control contributes to phenotypic complexity, dysregulated 3′UTR diversity might be relevant to disease development, including cancers. Thus, 3′UTR diversity in cancer could open exciting new research areas and provide avenues for novel cancer theragnostics.
Keywords 3′UTR diversity, alternative polyadenylation, alternative splicing, cancer, RNA alterations, transcriptome
In the nucleus, eukaryotic mRNA is synthesized from a gene in a 5′ to 3′ direction and processed into mature transcripts by 5′-capping, splicing, and 3′-end formation. Genomic information is translated to proteins once mature mRNAs are exported to the cytoplasm. 5′ and 3′ sequences flanking the coding regions are not translated, thus named as untranslated regions (UTRs). What are the roles of UTRs in regulating gene expression? 5′UTRs are the leading sites for ribosome assembly for mRNA translation. In contrast, 3′UTRs play various roles in the post-transcriptional control of gene expressions, including stability, translation, and subcellular localization of mRNAs (Mayr, 2019). We will begin by reviewing why 3′UTR diversity is important, especially in human mRNAs. This review focuses only on the length of 3′UTRs, but not on their sequences and modification, as 3′UTR diversity for simplicity.
In general, the noncoding part of the genome has increased in size and complexity during evolution (Pesole et al., 2002). The average length of 5′UTRs is similar in metazoans; whereas, the 3′UTR is significantly longer in humans compared to other species (Mayr, 2016; Pesole et al., 2002; Sood et al., 2006; Wang et al., 2019) (Fig. 1). 3′-end sequencing also revealed that at least half of human genes could generate alternative 3′UTR isoforms from the same genes, demonstrating high diversity in 3′UTR in the human transcriptome (Derti et al., 2012; Hoque et al., 2013).
3′UTRs contain cis-regulatory elements recognized by trans-acting factors. Thus, changes in 3′UTR length could alter a number of regulatory elements (Ji et al 2009). In fact, many genes have alternative 3′UTR or contain internal introns which can switch 3′UTR length. Therefore, 3′UTR length changes could affect gene expression by fine-tuning and reprogramming the mRNA regulatory landscape in human cells (Navarro et al., 2021).
3′UTRs contain cis-regulatory elements recognized by trans-acting factors. 3′UTR-mediated mRNA metabolism controls at the steps of the stability, translation, and localization of mRNA (Fig. 2). 3′UTRs form ribonucleoprotein (RNP) with many RNA binding proteins (RBPs). So we will describe representative mRNA 3′UTRs, corresponding main RBPs, and their relevant functions as follows (Table 1).
Most mRNAs undergo stability regulation through specific sequences and structures in 3′UTRs, such as AU-rich elements (AREs), GU-rich elements, CA-rich elements, and miRNA-binding sequences. Some proteins bind to AREs and recruit the degradation machinery to ARE-containing mRNAs. For example, tristetraprolin (TTP) binds AREs within the 3′UTRs of mRNAs, such as
In contrast, Hu-antigen R (HuR) protein binds AREs and stabilizes numerous mRNAs, increasing the corresponding protein level (Brennan and Steitz, 2001). HuR stabilizes mRNAs promoting tumor growth and cell survival, such as
3′UTRs also contain cis-regulatory elements related to translation control. During the development of
Different lengths of 3′UTR contribute to the translation efficiency control due to the presence or the absense of cis-regulatory elements. In the case of Uncoupling protein 1 (UCP1), the long 3′UTR isoform is predominant and contains the binding sites for the translation regulator, cytoplasmic polyadenylation element-binding protein 2 (CPEB2) (Chen et al., 2018). CPEB2 is required for associating the long 3′UTR isoform for the low-level translation of
The subcellular localization of mRNA is investigated mostly in neurons (Andreassi and Riccio, 2009). Localized mRNA is translated in specific subcellular locations and allows the precise production of proteins in spatiotemporal patterns in response to neural activity. For example, calcium/calmodulin-dependent protein kinase II alpha (CaMK2A) regulates calcium signaling in the nervous system as a critical player in activity-dependent behavioral plasticity (Bae and Miura, 2020). In
The control of mRNA localization is also investigated in
During the lifecycle of mRNAs, some RBPs bind to 3′UTRs and affect multiple steps of post-transcriptional regulation (Moore, 2005). Such multifunctional RBPs associate with 3′UTRs of many mRNAs involved in critical downstream events. Quaking (QK) is an RBP with a STAR domain involved in compact myelin formation (Vernet and Artzt, 1997). QK binds to a specific sequence in the 3′UTRs of several mRNAs that encode proteins related to myelin formation. The QK response element (QRE) is “ACUAAY” in 3′UTRs of mRNAs, such as
Some proteins are newly found RBPs implicated in multiple 3′UTR-mediated control steps. For example, β-catenin is a Wnt-activated transcription factor and an adhesion protein. In addition, it also associates with ARE in 3′UTRs of
Another example is serine/arginine-rich splicing factor 3 (SRSF3). SRSF3 is a member of the serin/arginine-rich (SR) family proteins, acting in AS mainly in the nucleus. In addition, the shuttling activity of SRSF3 is linked to AS regulation, export, and translation in the cytoplasm (Park and Jeong, 2016). SRSF3 interacts with the 3′UTR of
These multifunctional RBPs are located in both the nucleus and cytoplasm and form 3′UTR-mediated RNP to control the multiple steps of mRNA metabolism. Further research is needed to fully understand the multiple roles of RBPs in 3′UTR-mediated mRNA regulation.
Considering the important roles of 3′UTRs in post-transcriptional gene expression, 3′UTR-mediated control should be regulated by many factors, including miRNA, RBPs and RNA granules as follows.
3′UTR functions are regulated by the crosstalk of miRNAs and RBPs, and cis-regulatory elements within 3′UTRs. miRNAs are small noncoding RNA molecules that are crucially involved in regulating gene expression by binding to complementary sequences in the 3′UTRs of the target mRNA. This binding can lead to the destabilization of mRNAs or the inhibition of their translation (Baek et al., 2008; Friedman et al., 2009). For example,
The control of 3′UTR length also could affect miRNA-binding sites and leads to functional consequences (Nam et al., 2014). Thus, the dynamic regulation of 3′UTR length could play a significant role in development because it contributes to the gene expression patterns observed in different cell types and tissues. Likely, distinct sets of genes are differentially regulated by the interplay of miRNA and RBPs in a context- and condition-dependent manner (Hoffman et al., 2016).
RNA granules in the cytoplasm are emerging as essential players in mRNA localization and translation regulation (Moore, 2005; Tian et al., 2020). RNA granules are specialized structures in assembled RNAs with RNP particles. Several RNA granules include transport RNPs, stress granules, and processing bodies (P-bodies) (Kiebler and Bassell, 2006). This compartmentalization creates cellular asymmetries, may enhance biological reactions, and promote molecular interactions required for cell growth and development (Tian et al., 2020). How RNAs assemble RNA granules? 3′UTRs within mRNAs could be key components in forming RNA granules. For example, the cytoplasmic polyadenylation element (CPE) in 3′UTR and its binding protein CPEB facilitate mRNA transport to dendrites in rat hippocampal neurons (Huang et al., 2003). 3′UTR-RNP granules are formed on
Technical advances enhance cellular RNA granule visualization.
3′UTR diversity can be caused by various ways, such as genetic variations, RNA modification, alternative cleavage, alternative polyadenylation (APA), and alternative splicing (AS). These could diversify the transcriptome and contribute to epigenetic gene regulation. This section focuses on the length-diversifying mechanism, APA, and AS (Fig. 3). Databases with different 3′UTR lengths formed by APA and AS are also summarized in Table 2.
Cleavage and polyadenylation is a critical step in the maturation of 3′UTR ends of most eukaryotic mRNAs (Mitschka and Mayr, 2022). The polyadenylation process begins the recognition of specific sequences in the mRNA molecule by the polyadenylation machinery in the nucleus. The polyadenylation machinery consists of several factors, including cleavage factor I, cleavage and polyadenylation specificity factor, cleavage factor II, and cleavage stimulation factor. These complexes bind to specific sequences in mRNA 3′UTRs and facilitate the cleavage at a specific site downstream of the AAUAAA motif. This cleavage is aided by cleavage factor polyribonucleotide kinase subunit 1, followed by adding a polyadenylate tail to the mRNA molecule by the enzyme poly(A) polymerase alpha (Tian and Manley, 2013). A strong polyadenylation signal at the 3′UTR end is important for the efficient cleavage and polyadenylation of mRNAs. However, many genes also have additional polyadenylation signals at their 3′UTRs in eukaryotes, and the usage of these signals can be regulated through APA (Fig. 3A).
APA is a widespread phenomenon in metazoan protein-coding transcripts (70%-79% of mammalian genes). In humans, >70% of genes have more than one polyadenylation site in their 3′UTRs, and approximately 50% have three or more. By contrast, in the mouse liver, >60% of expressed genes had multiple polyadenylation signals in their 3′UTRs (Tian and Manley, 2013). This can result in mRNAs with different 3′UTR ends and different regulatory and functional properties (Mitschka and Mayr, 2022). 3′UTR shortening was proposed to increase mRNA stability by reducing the accessibility of mRNA degradation mechanisms, such as RBP- or miRNA-based degradation. Conversely, 3′UTR lengthening may increase the accessibility of miRNAs, decreasing mRNA stability and translation.
In addition to regulating mRNA stability APA can play a role in mRNA–protein interactions and protein localization via alternative 3′UTRs. For example, the APA of
AS can also alter the 3′UTR length if 3′UTR contains an internal intron (Fig. 3B). It will increase the length, heterogeneity, and functional diversity of 3′UTRs (Bicknell et al., 2012; Chan et al., 2022b). The combination of APA and AS could further increase mRNA isoform variability resulting in the generation of mRNA transcripts containing various forms (Fig. 3C). The splicing of introns located within 3′UTRs was first described in
Another study demonstrated that 3′UTRs of
The 3′UTR of the
Transcriptome analyses of human cancer databases revealed that a large percentage of genes in cancers have RNA alterations (PCAWG Transcriptome Core Group et al., 2020). RNA alterations are mostly in protein-coding regions, but 3′UTR diversity also attribute to RNA alterations. Because APA is widespread and alters the regulatory potential of 3′UTRs, APA dysregulation plays a significant role in cancers (Kahles et al., 2018). Global APA within 3′UTRs is involved in the proliferation and metastasis of cancer cells and tumor tissues (Hoque et al., 2013; Mayr and Bartel, 2009). APA dysregulation is exemplified in
Based on genome and transcriptome databases, AS events in cancer tissues are estimated to be 20% higher on average than in normal tissues (Kahles et al., 2018). Recently, a pan-cancer analysis revealed that pervasive upregulation of 3′UTR splicing drives tumorigenesis (Chan et al., 2022b). In hepatocellular carcinoma, aberrant 3′UTR isoforms of
As in cancer disorders of neuronal plasticity and learning, 3′UTRs appear to be a pathological hotspot (Conne et al., 2000). Because several neuronal mRNAs depend on their 3′UTRs for appropriate subcellular targeting or translational control, perturbations in 3′UTR-mediated functions deserve attention. In addition, alterations in the secondary structure of 3′UTRs are related to the pathology of a certain disease (Reamon-Buettner et al., 2007). Mutations in the 3′UTR of
As genomic and transcriptomic data accumulated, RNA alterations emerged as important features of cancer development (PCAWG Transcriptome Core Group et al., 2020). Among many RNA alterations, 3′UTR variants received little attention in comparison with those in protein-coding regions (Bicknell et al., 2012). However, a widespread shortening of 3′UTR lengths in cancers and causal UTR variants in human diseases demonstrated the important roles of 3′UTRs in disease development (Griesemer et al., 2021; Mayr and Bartel, 2009). 3′UTR alterations can contribute to the diversification of the transcriptome and protein functions, impacting numerous biological processes.
3′UTR activities are context- and condition-dependent, dynamically affecting gene expression programs by miRNAs, RBPs, and RNA granule formation. Thus, finding rules and factors generating and modulating 3′UTR variants, such as 3′-end processing machinery, 3′UTR-binding proteome and UTR-mediated RNP granules, might be important. As functional alterations in 3′UTRs are found in various diseases, a systemic search for “3′UTR-mediated diseases” will extend the horizons in the diagnosis and treatment of these diseases (Conne et al., 2000). Specifically, functional cis-regulatory elements in 3′UTRs in pathogenic genes, interaction with disease-associated RBPs, and their pathogenic conditions should be understood to elucidate the mechanism and their consequent phenotypic changes in diseases. It could also open exciting new research areas, answering the question of what enables the biological complexity in humans during evolution.
This work was supported by the National Research Foundation (NRF), funded by the Ministry of Science & ICT (NRF-2020R1A2C2005358 and NRF-2022M3E5F1016546).
S.J. conceived the idea, wrote the manuscript, and secured funding. D.W. performed experiments and wrote the manuscript.
The authors have no potential conflicts of interest to disclose.
List of representative mRNAs with 3’UTR-mediated control
Transcript | RNA-binding proteins | Control | Function | Reference |
---|---|---|---|---|
Pumilio Nanos Brat | Translation repression | Embryonic axis formation in | (Kuersten and Goodwin, 2003) | |
CPEB2 | Translation activation | Maintain body temperature | (Chen et al., 2018) | |
Mub | Localization translation | Control learning and memory in neuron | (Broix et al., 2021; Chen et al., 2022) | |
IGF2BP1 PAT1 | Stabilization local-translation | Constituent of the cytoskeleton for structural support and movement in neuron | (Wu et al., 2020) | |
Bruno Hrp48 | RNP granule formation Translation | Embryonic patterning and germline formation in | (Bose et al., 2022; Jambor et al., 2014) | |
CUG-BP2 | mRNA export Translation | Maintenance of the myelin sheath | (Amack and Mahadevan, 2001; Taneja et al., 1995) | |
ELAV like RNA binding protein family | ARE-mediated mRNA stabilization | Cellular growth in neuroblastoma | (Chagnovich and Cohn, 1996; Chagnovich et al., 1996) | |
SRSF3 | Translation repression | Regulator of cell cycle and senescence | (Kim et al., 2022) | |
HuR b-catenin | mRNA stabilization Translation activation | Cellular proliferation and tumorigenesis | (Kim et al., 2012; Lee and Jeong, 2006) |
List of Database related to 3′UTR divers
Tool name | Features | URL | Reference |
---|---|---|---|
Poly(A) sites databases | |||
PolyASite 2.0 | APA atlas made from 3′end sequencing data | https://polyasite.unibas.ch | (Herrmann et al., 2020) |
PolyA_DB 3 | APA atlas built from ~1.2 billion 3′end deep sequencing reads | https://exon.apps.wistar.org/PolyA_DB/v3/ | (Wang et al., 2018) |
scAPAdb | APA atlas at single-cell resolution from publications and Genomics website | http://www.bmibig.cn/scAPAdb/ | (Zhu et al., 2022) |
TREND-DB | APA database constructed from bulk RNA-seq data of potential APA regulators | http://shiny.imbei.uni-mainz.de:3838/trend-db | (Marini et al., 2021) |
3′UTR alternative splicing databases | |||
RNA structure databases | 3′UTR splicing events database from RNA-seq data of TCGA and GTEx | https://www.cbrc.kaust.edu.sa/spur/home/ | (Chan et al., 2022b) |
RNA structure databases | |||
RNAfold | Prediction tool of secondary structures for single-stranded RNA or DNA | http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi | (Lorenz et al., 2011) |
Mol. Cells 2023; 46(1): 48-56
Published online January 31, 2023 https://doi.org/10.14348/molcells.2023.0003
Copyright © The Korean Society for Molecular and Cellular Biology.
Laboratory of RNA Cell Biology, Department of Bioconvergence Engineering, Dankook University Graduate School, Yongin 16892, Korea
Correspondence to:sjsj@dankook.ac.kr
This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/.
Genomic information stored in the DNA is transcribed to the mRNA and translated to proteins. The 3′ untranslated regions (3′UTRs) of the mRNA serve pivotal roles in posttranscriptional gene expression, regulating mRNA stability, translation, and localization. Similar to DNA mutations producing aberrant proteins, RNA alterations expand the transcriptome landscape and change the cellular proteome. Recent global analyses reveal that many genes express various forms of altered RNAs, including 3′UTR length variants. Alternative polyadenylation and alternative splicing are involved in diversifying 3′UTRs, which could act as a hidden layer of eukaryotic gene expression control. In this review, we summarize the functions and regulations of 3′UTRs and elaborate on the generation and functional consequences of 3′UTR diversity. Given that dynamic 3′UTR length control contributes to phenotypic complexity, dysregulated 3′UTR diversity might be relevant to disease development, including cancers. Thus, 3′UTR diversity in cancer could open exciting new research areas and provide avenues for novel cancer theragnostics.
Keywords: 3′UTR diversity, alternative polyadenylation, alternative splicing, cancer, RNA alterations, transcriptome
In the nucleus, eukaryotic mRNA is synthesized from a gene in a 5′ to 3′ direction and processed into mature transcripts by 5′-capping, splicing, and 3′-end formation. Genomic information is translated to proteins once mature mRNAs are exported to the cytoplasm. 5′ and 3′ sequences flanking the coding regions are not translated, thus named as untranslated regions (UTRs). What are the roles of UTRs in regulating gene expression? 5′UTRs are the leading sites for ribosome assembly for mRNA translation. In contrast, 3′UTRs play various roles in the post-transcriptional control of gene expressions, including stability, translation, and subcellular localization of mRNAs (Mayr, 2019). We will begin by reviewing why 3′UTR diversity is important, especially in human mRNAs. This review focuses only on the length of 3′UTRs, but not on their sequences and modification, as 3′UTR diversity for simplicity.
In general, the noncoding part of the genome has increased in size and complexity during evolution (Pesole et al., 2002). The average length of 5′UTRs is similar in metazoans; whereas, the 3′UTR is significantly longer in humans compared to other species (Mayr, 2016; Pesole et al., 2002; Sood et al., 2006; Wang et al., 2019) (Fig. 1). 3′-end sequencing also revealed that at least half of human genes could generate alternative 3′UTR isoforms from the same genes, demonstrating high diversity in 3′UTR in the human transcriptome (Derti et al., 2012; Hoque et al., 2013).
3′UTRs contain cis-regulatory elements recognized by trans-acting factors. Thus, changes in 3′UTR length could alter a number of regulatory elements (Ji et al 2009). In fact, many genes have alternative 3′UTR or contain internal introns which can switch 3′UTR length. Therefore, 3′UTR length changes could affect gene expression by fine-tuning and reprogramming the mRNA regulatory landscape in human cells (Navarro et al., 2021).
3′UTRs contain cis-regulatory elements recognized by trans-acting factors. 3′UTR-mediated mRNA metabolism controls at the steps of the stability, translation, and localization of mRNA (Fig. 2). 3′UTRs form ribonucleoprotein (RNP) with many RNA binding proteins (RBPs). So we will describe representative mRNA 3′UTRs, corresponding main RBPs, and their relevant functions as follows (Table 1).
Most mRNAs undergo stability regulation through specific sequences and structures in 3′UTRs, such as AU-rich elements (AREs), GU-rich elements, CA-rich elements, and miRNA-binding sequences. Some proteins bind to AREs and recruit the degradation machinery to ARE-containing mRNAs. For example, tristetraprolin (TTP) binds AREs within the 3′UTRs of mRNAs, such as
In contrast, Hu-antigen R (HuR) protein binds AREs and stabilizes numerous mRNAs, increasing the corresponding protein level (Brennan and Steitz, 2001). HuR stabilizes mRNAs promoting tumor growth and cell survival, such as
3′UTRs also contain cis-regulatory elements related to translation control. During the development of
Different lengths of 3′UTR contribute to the translation efficiency control due to the presence or the absense of cis-regulatory elements. In the case of Uncoupling protein 1 (UCP1), the long 3′UTR isoform is predominant and contains the binding sites for the translation regulator, cytoplasmic polyadenylation element-binding protein 2 (CPEB2) (Chen et al., 2018). CPEB2 is required for associating the long 3′UTR isoform for the low-level translation of
The subcellular localization of mRNA is investigated mostly in neurons (Andreassi and Riccio, 2009). Localized mRNA is translated in specific subcellular locations and allows the precise production of proteins in spatiotemporal patterns in response to neural activity. For example, calcium/calmodulin-dependent protein kinase II alpha (CaMK2A) regulates calcium signaling in the nervous system as a critical player in activity-dependent behavioral plasticity (Bae and Miura, 2020). In
The control of mRNA localization is also investigated in
During the lifecycle of mRNAs, some RBPs bind to 3′UTRs and affect multiple steps of post-transcriptional regulation (Moore, 2005). Such multifunctional RBPs associate with 3′UTRs of many mRNAs involved in critical downstream events. Quaking (QK) is an RBP with a STAR domain involved in compact myelin formation (Vernet and Artzt, 1997). QK binds to a specific sequence in the 3′UTRs of several mRNAs that encode proteins related to myelin formation. The QK response element (QRE) is “ACUAAY” in 3′UTRs of mRNAs, such as
Some proteins are newly found RBPs implicated in multiple 3′UTR-mediated control steps. For example, β-catenin is a Wnt-activated transcription factor and an adhesion protein. In addition, it also associates with ARE in 3′UTRs of
Another example is serine/arginine-rich splicing factor 3 (SRSF3). SRSF3 is a member of the serin/arginine-rich (SR) family proteins, acting in AS mainly in the nucleus. In addition, the shuttling activity of SRSF3 is linked to AS regulation, export, and translation in the cytoplasm (Park and Jeong, 2016). SRSF3 interacts with the 3′UTR of
These multifunctional RBPs are located in both the nucleus and cytoplasm and form 3′UTR-mediated RNP to control the multiple steps of mRNA metabolism. Further research is needed to fully understand the multiple roles of RBPs in 3′UTR-mediated mRNA regulation.
Considering the important roles of 3′UTRs in post-transcriptional gene expression, 3′UTR-mediated control should be regulated by many factors, including miRNA, RBPs and RNA granules as follows.
3′UTR functions are regulated by the crosstalk of miRNAs and RBPs, and cis-regulatory elements within 3′UTRs. miRNAs are small noncoding RNA molecules that are crucially involved in regulating gene expression by binding to complementary sequences in the 3′UTRs of the target mRNA. This binding can lead to the destabilization of mRNAs or the inhibition of their translation (Baek et al., 2008; Friedman et al., 2009). For example,
The control of 3′UTR length also could affect miRNA-binding sites and leads to functional consequences (Nam et al., 2014). Thus, the dynamic regulation of 3′UTR length could play a significant role in development because it contributes to the gene expression patterns observed in different cell types and tissues. Likely, distinct sets of genes are differentially regulated by the interplay of miRNA and RBPs in a context- and condition-dependent manner (Hoffman et al., 2016).
RNA granules in the cytoplasm are emerging as essential players in mRNA localization and translation regulation (Moore, 2005; Tian et al., 2020). RNA granules are specialized structures in assembled RNAs with RNP particles. Several RNA granules include transport RNPs, stress granules, and processing bodies (P-bodies) (Kiebler and Bassell, 2006). This compartmentalization creates cellular asymmetries, may enhance biological reactions, and promote molecular interactions required for cell growth and development (Tian et al., 2020). How RNAs assemble RNA granules? 3′UTRs within mRNAs could be key components in forming RNA granules. For example, the cytoplasmic polyadenylation element (CPE) in 3′UTR and its binding protein CPEB facilitate mRNA transport to dendrites in rat hippocampal neurons (Huang et al., 2003). 3′UTR-RNP granules are formed on
Technical advances enhance cellular RNA granule visualization.
3′UTR diversity can be caused by various ways, such as genetic variations, RNA modification, alternative cleavage, alternative polyadenylation (APA), and alternative splicing (AS). These could diversify the transcriptome and contribute to epigenetic gene regulation. This section focuses on the length-diversifying mechanism, APA, and AS (Fig. 3). Databases with different 3′UTR lengths formed by APA and AS are also summarized in Table 2.
Cleavage and polyadenylation is a critical step in the maturation of 3′UTR ends of most eukaryotic mRNAs (Mitschka and Mayr, 2022). The polyadenylation process begins the recognition of specific sequences in the mRNA molecule by the polyadenylation machinery in the nucleus. The polyadenylation machinery consists of several factors, including cleavage factor I, cleavage and polyadenylation specificity factor, cleavage factor II, and cleavage stimulation factor. These complexes bind to specific sequences in mRNA 3′UTRs and facilitate the cleavage at a specific site downstream of the AAUAAA motif. This cleavage is aided by cleavage factor polyribonucleotide kinase subunit 1, followed by adding a polyadenylate tail to the mRNA molecule by the enzyme poly(A) polymerase alpha (Tian and Manley, 2013). A strong polyadenylation signal at the 3′UTR end is important for the efficient cleavage and polyadenylation of mRNAs. However, many genes also have additional polyadenylation signals at their 3′UTRs in eukaryotes, and the usage of these signals can be regulated through APA (Fig. 3A).
APA is a widespread phenomenon in metazoan protein-coding transcripts (70%-79% of mammalian genes). In humans, >70% of genes have more than one polyadenylation site in their 3′UTRs, and approximately 50% have three or more. By contrast, in the mouse liver, >60% of expressed genes had multiple polyadenylation signals in their 3′UTRs (Tian and Manley, 2013). This can result in mRNAs with different 3′UTR ends and different regulatory and functional properties (Mitschka and Mayr, 2022). 3′UTR shortening was proposed to increase mRNA stability by reducing the accessibility of mRNA degradation mechanisms, such as RBP- or miRNA-based degradation. Conversely, 3′UTR lengthening may increase the accessibility of miRNAs, decreasing mRNA stability and translation.
In addition to regulating mRNA stability APA can play a role in mRNA–protein interactions and protein localization via alternative 3′UTRs. For example, the APA of
AS can also alter the 3′UTR length if 3′UTR contains an internal intron (Fig. 3B). It will increase the length, heterogeneity, and functional diversity of 3′UTRs (Bicknell et al., 2012; Chan et al., 2022b). The combination of APA and AS could further increase mRNA isoform variability resulting in the generation of mRNA transcripts containing various forms (Fig. 3C). The splicing of introns located within 3′UTRs was first described in
Another study demonstrated that 3′UTRs of
The 3′UTR of the
Transcriptome analyses of human cancer databases revealed that a large percentage of genes in cancers have RNA alterations (PCAWG Transcriptome Core Group et al., 2020). RNA alterations are mostly in protein-coding regions, but 3′UTR diversity also attribute to RNA alterations. Because APA is widespread and alters the regulatory potential of 3′UTRs, APA dysregulation plays a significant role in cancers (Kahles et al., 2018). Global APA within 3′UTRs is involved in the proliferation and metastasis of cancer cells and tumor tissues (Hoque et al., 2013; Mayr and Bartel, 2009). APA dysregulation is exemplified in
Based on genome and transcriptome databases, AS events in cancer tissues are estimated to be 20% higher on average than in normal tissues (Kahles et al., 2018). Recently, a pan-cancer analysis revealed that pervasive upregulation of 3′UTR splicing drives tumorigenesis (Chan et al., 2022b). In hepatocellular carcinoma, aberrant 3′UTR isoforms of
As in cancer disorders of neuronal plasticity and learning, 3′UTRs appear to be a pathological hotspot (Conne et al., 2000). Because several neuronal mRNAs depend on their 3′UTRs for appropriate subcellular targeting or translational control, perturbations in 3′UTR-mediated functions deserve attention. In addition, alterations in the secondary structure of 3′UTRs are related to the pathology of a certain disease (Reamon-Buettner et al., 2007). Mutations in the 3′UTR of
As genomic and transcriptomic data accumulated, RNA alterations emerged as important features of cancer development (PCAWG Transcriptome Core Group et al., 2020). Among many RNA alterations, 3′UTR variants received little attention in comparison with those in protein-coding regions (Bicknell et al., 2012). However, a widespread shortening of 3′UTR lengths in cancers and causal UTR variants in human diseases demonstrated the important roles of 3′UTRs in disease development (Griesemer et al., 2021; Mayr and Bartel, 2009). 3′UTR alterations can contribute to the diversification of the transcriptome and protein functions, impacting numerous biological processes.
3′UTR activities are context- and condition-dependent, dynamically affecting gene expression programs by miRNAs, RBPs, and RNA granule formation. Thus, finding rules and factors generating and modulating 3′UTR variants, such as 3′-end processing machinery, 3′UTR-binding proteome and UTR-mediated RNP granules, might be important. As functional alterations in 3′UTRs are found in various diseases, a systemic search for “3′UTR-mediated diseases” will extend the horizons in the diagnosis and treatment of these diseases (Conne et al., 2000). Specifically, functional cis-regulatory elements in 3′UTRs in pathogenic genes, interaction with disease-associated RBPs, and their pathogenic conditions should be understood to elucidate the mechanism and their consequent phenotypic changes in diseases. It could also open exciting new research areas, answering the question of what enables the biological complexity in humans during evolution.
This work was supported by the National Research Foundation (NRF), funded by the Ministry of Science & ICT (NRF-2020R1A2C2005358 and NRF-2022M3E5F1016546).
S.J. conceived the idea, wrote the manuscript, and secured funding. D.W. performed experiments and wrote the manuscript.
The authors have no potential conflicts of interest to disclose.
List of representative mRNAs with 3’UTR-mediated control
Transcript | RNA-binding proteins | Control | Function | Reference |
---|---|---|---|---|
Pumilio Nanos Brat |
Translation repression | Embryonic axis formation in |
(Kuersten and Goodwin, 2003) | |
CPEB2 | Translation activation | Maintain body temperature | (Chen et al., 2018) | |
Mub | Localization translation | Control learning and memory in neuron | (Broix et al., 2021; Chen et al., 2022) | |
IGF2BP1 PAT1 |
Stabilization local-translation | Constituent of the cytoskeleton for structural support and movement in neuron | (Wu et al., 2020) | |
Bruno Hrp48 |
RNP granule formation Translation |
Embryonic patterning and germline formation in |
(Bose et al., 2022; Jambor et al., 2014) | |
CUG-BP2 | mRNA export Translation |
Maintenance of the myelin sheath | (Amack and Mahadevan, 2001; Taneja et al., 1995) | |
ELAV like RNA binding protein family | ARE-mediated mRNA stabilization | Cellular growth in neuroblastoma | (Chagnovich and Cohn, 1996; Chagnovich et al., 1996) | |
SRSF3 | Translation repression | Regulator of cell cycle and senescence | (Kim et al., 2022) | |
HuR b-catenin |
mRNA stabilization Translation activation |
Cellular proliferation and tumorigenesis | (Kim et al., 2012; Lee and Jeong, 2006) |
List of Database related to 3′UTR divers
Tool name | Features | URL | Reference |
---|---|---|---|
Poly(A) sites databases | |||
PolyASite 2.0 | APA atlas made from 3′end sequencing data | https://polyasite.unibas.ch | (Herrmann et al., 2020) |
PolyA_DB 3 | APA atlas built from ~1.2 billion 3′end deep sequencing reads | https://exon.apps.wistar.org/PolyA_DB/v3/ | (Wang et al., 2018) |
scAPAdb | APA atlas at single-cell resolution from publications and Genomics website | http://www.bmibig.cn/scAPAdb/ | (Zhu et al., 2022) |
TREND-DB | APA database constructed from bulk RNA-seq data of potential APA regulators | http://shiny.imbei.uni-mainz.de:3838/trend-db | (Marini et al., 2021) |
3′UTR alternative splicing databases | |||
RNA structure databases | 3′UTR splicing events database from RNA-seq data of TCGA and GTEx | https://www.cbrc.kaust.edu.sa/spur/home/ | (Chan et al., 2022b) |
RNA structure databases | |||
RNAfold | Prediction tool of secondary structures for single-stranded RNA or DNA | http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi | (Lorenz et al., 2011) |
. List of representative mRNAs with 3’UTR-mediated control.
Transcript | RNA-binding proteins | Control | Function | Reference |
---|---|---|---|---|
Pumilio Nanos Brat | Translation repression | Embryonic axis formation in | (Kuersten and Goodwin, 2003) | |
CPEB2 | Translation activation | Maintain body temperature | (Chen et al., 2018) | |
Mub | Localization translation | Control learning and memory in neuron | (Broix et al., 2021; Chen et al., 2022) | |
IGF2BP1 PAT1 | Stabilization local-translation | Constituent of the cytoskeleton for structural support and movement in neuron | (Wu et al., 2020) | |
Bruno Hrp48 | RNP granule formation Translation | Embryonic patterning and germline formation in | (Bose et al., 2022; Jambor et al., 2014) | |
CUG-BP2 | mRNA export Translation | Maintenance of the myelin sheath | (Amack and Mahadevan, 2001; Taneja et al., 1995) | |
ELAV like RNA binding protein family | ARE-mediated mRNA stabilization | Cellular growth in neuroblastoma | (Chagnovich and Cohn, 1996; Chagnovich et al., 1996) | |
SRSF3 | Translation repression | Regulator of cell cycle and senescence | (Kim et al., 2022) | |
HuR b-catenin | mRNA stabilization Translation activation | Cellular proliferation and tumorigenesis | (Kim et al., 2012; Lee and Jeong, 2006) |
. List of Database related to 3′UTR divers.
Tool name | Features | URL | Reference |
---|---|---|---|
Poly(A) sites databases | |||
PolyASite 2.0 | APA atlas made from 3′end sequencing data | https://polyasite.unibas.ch | (Herrmann et al., 2020) |
PolyA_DB 3 | APA atlas built from ~1.2 billion 3′end deep sequencing reads | https://exon.apps.wistar.org/PolyA_DB/v3/ | (Wang et al., 2018) |
scAPAdb | APA atlas at single-cell resolution from publications and Genomics website | http://www.bmibig.cn/scAPAdb/ | (Zhu et al., 2022) |
TREND-DB | APA database constructed from bulk RNA-seq data of potential APA regulators | http://shiny.imbei.uni-mainz.de:3838/trend-db | (Marini et al., 2021) |
3′UTR alternative splicing databases | |||
RNA structure databases | 3′UTR splicing events database from RNA-seq data of TCGA and GTEx | https://www.cbrc.kaust.edu.sa/spur/home/ | (Chan et al., 2022b) |
RNA structure databases | |||
RNAfold | Prediction tool of secondary structures for single-stranded RNA or DNA | http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi | (Lorenz et al., 2011) |
Christopher J. Occhiuto, Jessica A. Moerland, Ana S. Leal, Kathleen A. Gallo, and Karen T. Liby
Mol. Cells 2023; 46(3): 176-186 https://doi.org/10.14348/molcells.2023.2191Aryatara Shakya, Nicholas W. McKee, Matthew Dodson, Eli Chapman, and Donna D. Zhang
Mol. Cells 2023; 46(3): 165-175 https://doi.org/10.14348/molcells.2023.0005Yajing Hao, Ting Cai, Chang Liu, Xuan Zhang, and Xiang-Dong Fu
Mol. Cells 2023; 46(1): 57-64 https://doi.org/10.14348/molcells.2023.2176