Access the full text.
Sign up today, get DeepDyve free for 14 days.
BioscienceHorizons Volume 8 2015 10.1093/biohorizons/hzv004 Research article Aspergillus nidulans Upf1: putative role of conserved active sites in ribosome recycling and 3′ end mRNA tagging Ryan Langley* Faculty of Health and Life Science, Department of Applied Sciences and Health, Coventry University, Priory Street, Coventry CV1 5FB, England *Corresponding author: 45 The Spinney, Finchfield, Wolverhampton WV3 9EU, England. Email: firstname.lastname@example.org Supervisor: Dr Igor Morozov, Faculty of Health and Life Sciences, Coventry University, Coventry, England. Up-frameshift protein 1 (Upf1) is a multidomain RNA helicase that is conserved from yeast to humans. Upf1 is critical for nonsense-mediated decay (NMD), a quality control mechanism that detects and eliminates aberrant transcripts harbouring a premature termination codon (PTC) and thus plays an important role in maintaining the fidelity of gene expression. Additionally, Upf1 is implicated in a broad range of cellular responses from chromosome maintenance to mRNA degradation and translational repression. Recent findings show that Upf1 also triggers 3 ′ end mRNA tagging, the addition of non-tem- plated pyrimidines (C/U) to the 3′ end of adenylated and non-adenylated (histone) mRNAs. 3′ end tagging is seen as a general precursor of mRNA degradation and has been found to occur in fungi, plants and mammals. In Aspergillus nidulans, 3′ end tagging of normal and aberrant transcripts containing PTCs occurs in an Upf1-dependent manner. Intriguingly, tagging of transcripts harbouring PTCs is not essential for transcript degradation as the disruption of either of the two enzymes that mediate 3′ end RNA tagging, CutA and CutB results in decreased efficiency of ribosome dissociation from the PTC. However, the exact role of Upf1 and its functional domains in inducing tagging and ribosome dissociation remains unknown. Therefore, the aim of this work is to propose a model for the detailed mutational analysis of A. nidulans Upf1 in relation to its role in trig- gering 3′ end tagging and translation termination. From published data of mutation analysis, active site residues of the yeast and human Upf1 proteins have been identified and aligned to their A. nidulans homologue. Analysis of the structural organi- zation of A. nidulans Upf1 reveals the presence of two major conserved domains and a number of putative actives site residues which may be crucial for 3′ end mRNA tagging, translational repression and ribosome termination. The importance of a greater understanding of the role of Upf1 in regulation of gene expression in A. nidulans, a model organism for Aspergillus species of medical and industrial importance, is discussed. Key words: upf1, 3′ end tagging, Aspergillus, nonsense-mediated decay Submitted on 18 June 2014; accepted on 30 April 2015 Introduction and Caddick, 2012). To minimize the impact of such inaccura- cies and ensure the fidelity of gene expression, cells have Gene expression comprises multiple stages and revolves evolved a number of quality control mechanisms that detect around RNA, from transcription to RNA processing and and eliminate transcripts that are no longer functional (e.g. translation. The cellular machinery that constitutes to each aberrant mRNAs) or potentially harmful (e.g. malfunctional stage is often complex, and therefore, errors may be introduced non-coding RNAs). These quality control mechanisms come in into RNA molecules at any step of its biogenesis (Morozov a form of multiple mRNA decay pathways and can require © The Author 2015. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Research article Bioscience Horizons • Volume 8 2015 active translation (Doma and Parker, 2007; Shoemaker and In this study, two conserved domains in A. nidulans Upf1 Green, 2012). One such pathway is nonsense-mediated decay (AN0646) have been identified. By reviewing published data (NMD) in which a premature termination codon (PTC), of mutation analysis of Upf1 in yeast and mammalian sys- caused by nonsense mutations or frame-shifts, is recognized by tems, numerous putative actives sites that may be implicated translating ribosomes, triggering a cascade of events leading to in 3′ end mRNA tagging, translational repression and ribo- efficient degradation of the PTC + mRNA and thus preventing some termination were also found. Detailed mutation analysis the accumulation of potentially harmful truncated proteins of A. nidulans Upf1 with respect to its role in triggering 3′ end (Imamachi, Tani and Akimitsu, 2012; Popp and Maquat, mRNA tagging and translation termination has been pro- 2013). posed. These results and their potential application in the mutational analysis of Upf1 Aspergillus species of industrial NMD primarily functions as an mRNA surveillance mech- and medical importance are discussed. anism which acts to terminate the translating ribosome at the PTC and targets PTC+ transcripts for degradation. The novel role of Upf1 in 3′ tagging Interestingly, it has been shown that mRNAs lacking nonsense in A. nidulans mutations can also be targeted by this pathway, thus suggest- ing a wider role for NMD in gene regulation (Kervestin and In eukaryotes, cytoplasmic 3′ tagging of RNA, primarily by the Jacobson, 2012; Mühlemann and Jensen, 2012). However, the addition of non-templated pyrimidines (C/U), is a general pre- mechanism by which the cell distinguishes between a normal cursor to RNA degradation (Morozov and Caddick, 2012; termination codon (NTC) and the PTC remains unknown. Hoefig and Heissmeyer, 2014 ). 3′ uridylation of human histone Translation termination of mRNAs with NTCs is a highly effi - transcripts, which are non-polyadenylated, was shown to be cient process and has been suggested to be fundamentally dif- integral to cell cycle-regulated transcript degradation (Mullen ferent to that of PTC containing mRNA in which ribosomes and Marzluff, 2008). Recent work has shown that polyadenyl- stall at a nonsense codon (Kervestin and Jacobson, 2012; ated mRNAs are also tagged with C and/or U nucleotides in Morozov et al., 2012). Furthermore, ribosome dissociation on fungi (A. nidulans and Schizosaccharomyces pombe), plants PTC+ mRNAs has been shown to be less efficient than that of (Arabidopsis thaliana) and mammals (HeLa and NIH 3T3 NTC containing mRNAs (Amrani et al., 2004). cells) (Rissland and Norbury, 2009; Morozov et al., 2010a, b, 2012; Chang et al., 2014). These 3′ end mRNA modifications Upf1 is a highly conserved RNA helicase within eukary- are conserved throughout eukaryotes, with the exception of otes, from yeast to humans (Applequist et al., 1997). It is a S. cerevisiae (Rissland and Norbury, 2009). Emerging data principal NMD factor involved in the recognition and degra- strongly argue that for functional mRNA, 3′ uridylation dation of defective mRNA harbouring a PTC. Upf1 is an defines the point at which functional transcripts are earmarked intriguing protein which in addition to NMD has been impli- for translational repression, decapping and degradation. cated in a broad range of processes, including development, chromosome maintenance, cell cycle progression and DNA Transcript uridylation has been shown to increase the affin - replication (Imamachi, Tani and Akimitsu, 2012). Recent ity of the cytoplasmic Lsm-Pat1 complex for RNA (Chowdhury, work has discovered that Upf1 plays a crucial role in 3′ end Mukhopadhyay and Tharun, 2007). Recruitment of this com- tagging of physiological and PTC+ transcripts in mammals plex leads to translational repression, transcript decapping and and Aspergillus nidulans, namely the non-canonical addition degradation involving multiple decay factors (Nissan et al., of U/C and possibly G nucleotides to the 3′ end of mRNA 2010; Parker, 2012). Consistent with this, 3′ tagging tends to (Mullen and Marzluff, 2008; Rissland and Norbury, 2009; occur concurrently with transcript degradation, and disruption Morozov et al., 2010a, 2012; Chang et al., 2014). 3′ end tag- of the process retards transcript decay (Rissland and Norbury, ging appears to be a conserved eukaryotic surveillance mecha- 2009; Morozov et al., 2010a, 2012). nism in mammalian, plant and fungi cells with Saccharomyces In A. nidulans, as well as plants and mammals, mRNA tag- cerevisiae being the prominent exception. Upf1-mediated C/U ging occurs when the poly(A) tail is degraded to ∼15–20 nucle- tagging appears to be the point at which mRNA is earmarked otides, the point at which deadenylation triggers decapping for subsequent translational suppression and degradation, (Morozov et al., 2010a, 2012; Parker, 2012; Chang et al., while G tagging is mainly associated with a long poly(A) tail 2014). 3′ end mRNA tagging in A. nidulans is known to facili- of mRNA in mammals (Chang et al., 2014) and A. nidulans tate removal of the NMD-induced termination complex (Morozov and Caddick, personal communication) and poten- in vivo in a poly(A) tail length-independent manner, suggesting tially acts as a defence pathway against mRNA degradation. that it promotes efficient mRNA decay and ribosome recycling However, the exact mechanism of 3′ end tagging and its role (Morozov et al., 2012). Mutant strains disrupted in key com- in the regulation of gene expression is not yet understood. ponents of the NMD pathway, e.g. Upf1, do not show PTC- These data underline the importance of Upf1 beyond NMD. induced tagging, consistent with it being induced by the NMD Thus, it would be of great importance to identify active sites machinery. Although 3′ tagging is not required for NMD- and conduct mutational analysis of Upf1 in A. nidulans to induced transcript degradation, its loss leads to decapping understand the mechanism by which Upf1 regulates mRNA 3′ becoming dependent on deadenylation. The role 3′ tagging end tagging and subsequent ribosome recycling of both phys- plays in the clearance of stalled/terminating ribosomes at the iological and aberrant transcripts. 2 Bioscience Horizons • Volume 8 2015 Research article PTC suggests a potentially novel function for the Lsm-Pat1 relating to functional domains of yeast and mammals were complex and associated proteins including Upf1 (Morozov identified and documented along with scores given by et al., 2012). Critically, 3′ tagging of NTC+, PTC+ and cell ClustalW in the results summary. The position of both the CH cycle-regulated histone transcripts in mammals (Mullen and and ATPase domains of Upf1 in A. nidulans were calculated Marzluff, 2008) and A. nidulans (Morozov and Caddick, per- using published data for yeast (Lasalde et al., 2014) and sonal communication) occurs in an Upf1-dependent manner. human (Imamachi, Tani and Akimitsu, 2012) Upf1. The CH domain in yeast (62–152aa) and human (123–213aa) variants The cysteine and histidine-rich (CH) domain of Upf1 has of Upf1 were aligned to one another and subsequently to the been shown to interact directly with eRF3 (Ivanov et al., 2008), CH domain of A. nidulans (71–161aa). These sequences were which is also a target for poly(A) binding protein (PAB1), pro- aligned using two protein BLAST and output using default moting efficient translational termination, ribosome dissocia - settings (Altschul et al., 1990). The helicase/ATPase domain tion and recycling (Nissan et al., 2010). As 3′ tagging of PTC+ was also aligned using two protein BLAST and corresponds in and NTC+ transcripts is induced by Upf1, which in its phos- yeast to (231–851aa), human (295–914aa) and A. nidulans phorylated state is required for efficient NMD in mammals, it (243–759aa). Whole protein sequences of yeast, human and cannot be ruled out that the phosphorylation state of Upf1 A. nidulans Upf1 were also aligned using two protein BLAST. may also play a role in tagging and/or ribosome recycling. Accession numbers of Upf1 were found for S. cerevisiae However, the role of individual domains and specific amino (NP_013797.1), Homo sapiens (AAH39817.1), A. nidulans acids (active sites) of Upf1 in 3′ tagging and ribosome recycling (AN0646), Aspergillus fumigatus (AF293), Aspergillus oryzae are currently unknown. This work reviews published data of (AORIB40), M. musculus (AAH52149.1), A. thaliana functional active sites in yeast and human Upf1 that are impli- (SwissProt Q9FJR0.2), D. melanogaster (AAF48115.2) and cated in nonsense suppression, eRF3 interaction and phos- C. elegans (SwissProt 076512.1). phorylation of Upf1. Corresponding active sites and amino acid residues in A. nidulans Upf1 have been identified and may Functional domains of Upf1 modulate Upf1 activity in 3′ end mRNA tagging, translational repression and ribosome termination. Subsequent mutational Two major domains have been found in Upf1 in yeast; the CH analysis in A. nidulans is proposed to help elucidate the molec- domain present at N-terminus and a helicase domain located ular mechanism underpinning tagging and ribosome recycling in a more central position (Lasalde et al., 2014). Both domains of transcripts targeted for degradation. A deeper understand- are conserved within human systems which also harbour an ing of how gene expression is modulated by Upf1 may be use- additional SQ motif present at C-terminus (Applequist et al., ful in understanding genetic disorders in which nonsense 1997). Alignment of human, yeast and A. nidulans homo- mutations or frame-shifts can occur (e.g. cystic fibrosis) and logues of Upf1 clearly shows conserved CH and helicase may present Upf1 as a novel target in the treatment of these domains in all three systems (Fig. 1), with varying degrees of disorders. conservation between species. The crystal structure of the human Upf1 CH domain has Bioinformatics approach: multiple revealed three zinc-binding motifs that make up two zinc sequence alignment of Upf1 homologues knuckle modules similar to the RING-box and U-box Protein sequences of human, yeast, Mus musculus, A. thali- domains often found in E3 ubiquitin ligases (Kadlec et al., ana, Drosophila melanogaster, Caenorhabditis elegans homo- 2006). In yeast, the CH domain of Upf1 is involved in interac- logues of Upf1 were retrieved using the NCBI Protein database tion with E2 Ubc3 (ubiquitin-conjugating E2 enzyme), an (Pruitt et al., 2002), while the sequence of Aspergillus species enzyme involved in the regulation of ubiquitin activity sug- were found using the Aspergillus genome database (Cerqueira gesting Upf1 per se may be an E3 ubiquitin ligase (Takahashi et al., 2013). Reciprocal BLAST was performed using the et al., 2008). Furthermore mutants tested in this study, Upf1 sequence of yeast to confirm the identity of Upf1 in dif - His94Arg, His98Arg and Cys122Ser (Table 1) failed to stimu- ferent organisms. Local BLAST of A. nidulans using Broad late NMD, and it has been proposed that the inhibition of Institute (Kostic et al., 2011) was used to identify Upf1 homo- ubiquitin ligase activity is responsible for this defect. It has logues in multiple Aspergillus species. Sequences of all studied also been suggested that Upf1 interacts with eRF3 through the Upf1 homologues were aligned in FASTA format using CH domain (Ivanov et al., 2008). Therefore, corresponding ClustalW (Larkin et al., 2007) and annotated using Jalview mutational analysis in A. nidulans (e.g. His103Arg, Table 1) (Waterhouse et al., 2009). Aligned sequences of Upf1 were may be important in understanding the role of Upf1 in ribo- analysed and corresponding amino acids in Aspergillus species some recycling and 3′ tagging. Figure 1. Schematic diagram of Upf1 in Aspergillus nidulans depicting the proposed location of the CH domain and the ATPase/Helicase domain. 3 Research article Bioscience Horizons • Volume 8 2015 Table 1. Mutation of amino acid residues in human (h) and yeast (y) Upf1 homologues and conservation and predicted phenotype with relation to ribosome recycling and 3′ tagging in Aspergillus species A. nidulans Conserved Aspergillus Predicted phenotype following Mutation corresponding amino genus, References mutational analysis in A. nidulans AA acid O/F Thr28 Ala N/A N/A N/A N/A Lasalde et al. (2014) Tagging to occur only in the presence of He, Ganesan and Cys62 Tyr Cys71 Upf2 Jacobson (2013) Tagging to occur during nonsense Weng, Czaplinski Cys65 Ser Cys74 suppres sion while NMD function remains and Peltz (1996) intact Weng, Czaplinski Tagging to occur during nonsense and Peltz (1996); Cys84 Ser Cys93 suppression while NMD function remains He, Ganesan and intact Jacobson (2013) Mutation in yeast causes defects in Upf2 Weng, Czaplinski interaction and could affect the His94 Arg His103 and Peltz (1996); mechanisms involved in 3′ tagging if Takahashi et al. (2008) disrupted to a large extent Mutation in yeast causes defects in Upf2 Weng, Czaplinski and interaction and could affect the His98 Arg His107 Peltz (1996); Takahashi mechanisms involved in 3′ tagging if et al. (2008) disrupted to a large extent Mutation in yeast causes defects in Upf2 Weng, Czaplinski interaction and could affect the Cys122 Ser Cys131 and Peltz (1996); mechanisms involved in 3′ tagging if Takahashi et al. (2008) disrupted to a large extent Mutation of interest as Upf2 interaction Weng, Czaplinski and Cys125 Ser Cys134 has not been tested in humans or yeast Peltz (1996) Disrupt NMD and lead to stalling of Kashima et al. (2006); Cys126 Ser Cys74 ribosomes at PTC Ivanov et al. (2008) Thr194 pho Ser205 Ser Ser Reduction in 3′ tagging Lasalde et al. (2014) Increased level of tagging and ribosomes unable to dissociate from mRNA, may He, Ganesan and Lys436 Glu Lys450 disrupt ribosome recycling and transla- Jacobson (2013) tion termination Increased level of tagging and ribosomes unable to dissociate from mRNA, may Czaplinski et al. (1998); Lys436 Ala Lys450 disrupt ribosome recycling and transla- Cheng et al. (2007) tion termination Ser492 pho Ser506 Reduction in 3′ tagging Lasalde et al. (2014) Increased level of tagging and ribosomes unable to dissociate from mRNA, may Lys498 Gln Lys450 Kashima et al. (2006) disrupt ribosome recycling and transla- tion termination Tyr738 pho Tyr754 Reduction in 3′ tagging Lasalde et al. (2014) Tyr738 Phe::Tyr742 Phe Tyr754:Tyr758 Reduction in 3′ tagging Lasalde et al. (2014) y y Tyr738 Glu::Tyr742 Glu Tyr754:Tyr758 Reduction in 3′ tagging Lasalde et al. (2014) y y Tyr754 pho Tyr770 Reduction in 3′ tagging Lasalde et al. (2014) Ser1073 Ala Asp1032 Asp Asp N/A Kashima et al. (2006) Ser1078 Ala Phe1037 Phe Phe N/A Kashima et al. (2006) Ser1096 Ala His1055 His His N/A Kashima et al. (2006) Ser1116 Ala Tyr1075 Tyr Tyr N/A Kashima et al. (2006) pho is indicative of a phosphorylation residue and not a mutation. AA, amino acid, An, Aspergillus nidulans, O, Aspergillus Oryzae, F, Aspergillus fumigatus. 4 Bioscience Horizons • Volume 8 2015 Research article Alignment of the CH domain of yeast, human and A. nidu- lans homologues of Upf1 shows identities of 64%–74% and positives of 76%–84%. Both zinc-finger motifs (I and II) were identified in A. nidulans Upf1 and aligned to respective motifs in humans and yeast (Applequist et al., 1997). Motif I of A. nidu- lans Upf1 is 67% identical to the corresponding region in human Upf1 and 60% identical to its yeast homologue. Motif II shows higher conservation than that of Motif I in A. nidulans, being 84% and 74% identical to the corresponding sequence in humans and yeast, respectively. Furthermore, 10 highly con- served cysteine residues have been identified within the CH domain of Upf1, nine of which are conserved in A. nidulans. Five highly conserved histidine residues were also identified, all of which were conserved. Therefore, the CH domain within Aspergillus species may modulate similar functional roles as its yeast and human homologues. Figure 2. Simplified model showing the phosphorylation cycle of Sequences from the ATPase/helicase domains of human, Upf1 during NMD. Adapted by permission from Macmillan Publishers yeast and A. nidulans Upf1 homologues show identities of Ltd: Nature Genetics, Holbrook et al., 2004, copyright 2004. 58%–69% and positives of 73%–82%. All seven RNA heli- case motifs common to the SF1 helicases (superfamily 1 of components, thus being a key constituent in the response to DNA/RNA helicases) (Applequist et al., 1997) are found in the aberrant mRNAs containing PTCs (Kervestin and Jacobson, same order in A. nidulans Upf1 and are highly conserved being 2012) (Fig. 2). ≥75% identical to homologous sites, five of which showed positives of 100%. Additionally, distances between each motif Initial phosphorylation of Upf1 in human cells occurs at are identical to that of its human homologue. The Q motif serine residues (e.g. S1078 ) and is catalysed by SMG1, a found in DEAD box helicases shows poor conservation how- phosphatidylinositol 3-kinase-related kinase (Okada- ever, a highly conserved glutamine constituent at position Katsuhata et al., 2012), a step believed to be important in the 427 (A. nidulans) may still act to regulate Upf1 ATP binding An recognition of the PTC. SMG1 forms a complex with two and hydrolysis (Tanner et al., 2003) which are critical for other NMD factors, SMG8 and SMG9, that prevent Upf1 NMD, for example association with the 40S ribosomal sub- phosphorylation maintaining it in an inactive form until it unit, disassembly of a terminating mRNP and recycling of has interacted with Upf2 and Upf3. Upf1 then binds via its ribosomes. helicase domain to Nuclear Cap Binding Protein Subunit 1 (NCBP1), and this initial interaction constitutes to SMG1- Alignment of Upf1 of yeast, human and A. nidulans homo- Upf1 binding to eRF1-eRF3 (eukaryotic translation termina- logues shows high sequence identities of 56%–69% and posi- tion factor 1 and 3), forming the SURF complex (Hwang tives of 72%–82%. As expected, the CH and ATPase/helicase et al., 2010). This complex promotes the interaction of Upf1 domains show a higher degree of conservation between the with the exon–exon junction complex (EJC) in human cells, three systems compared with the whole protein sequence. containing Upf2 and Upf3, leading to phosphorylation of From this, key amino acids of Upf1 that are conserved in Upf1 by SMG1 and thus activating NMD. Phosphorylated A. nidulans (Table 1) and have been found in yeast and human Upf1 binds to a number of mRNA decay factors, including homologues to affect interaction with eRF3 and eRF1 (e.g. Dcp1/2 (mRNA decapping enzyme complex) and Xrn1 (an Lys450 ), nonsense suppression (e.g. Cys71 ) and phos- An An exoribonuclease enzyme that degrades the mRNA body in phorylation (e.g. Tyr770 ) were identified. An the 5′ to 3′ direction) (Isken et al., 2008). Importantly, Upf1 can recruit the SMG7/SMG5 complex that promotes deade- The role of Upf1 as a phosphoprotein nylation independent 5′-3′ degradation (in yeast or plants) or in NMD SMG6 (in D. melanogaster), an endonuclease that initiates Phosphorylation of Upf1 is believed to be important for PTC+ mRNA degradation via internal cleavage of mRNA NMD in mammals and yeast (Yamashita et al., 2001; (Loh, Jonas and Izaurralde, 2013). SMG7/SMG5 and SMG6 Okada-Katsuhata et al., 2012; Lasalde et al., 2014) as it trig- all harbour a 14-3-3-like fold domain (Aitken, 1995; Tzivion gers conformational changes within the protein allowing and Avruch, 2002) which may target phosphorylated pro- interaction with a number of mRNA decay factors and pro- teins, including Upf1 and lead to dephosphorylation of the moting translational repression (Gleghorn and Maquat, protein (Chiu et al., 2003). Phosphorylated Upf1 can also 2011; Imamachi, Tani and Akimitsu, 2012) and thus may bind to eIF3 resulting in translational repression by prevent- initiate 3′ end tagging. NMD in mammals requires a cycle of ing the 60S ribosomal subunit from joining with the 43S phosphorylation and dephosphorylation of Upf1 which translation pre-initiation complex (Gleghorn and Maquat, involves a number of interactions with numerous NMD 2011). 5 Research article Bioscience Horizons • Volume 8 2015 A number of phosphorylated sites have been found in 2006; Cheng et al., 2007; He, Ganesan and Jacobson, 2013). human Upf1. These consist of four Ser located at C-terminus Therefore, the mutation Lys450 Gln may provide a deeper An (Ser1096 ) and Thr located at N-terminus (Thr28 ) (Yamashita insight into Upf1-dependent tagging. eRF3 interaction with h h et al., 2001; Okada-Katsuhata et al., 2012). However, these upf1:Lys450 Gln should be reduced which may lead to atten- An phosphorylation sites in human Upf1 are not conserved in uation of NMD and/or enhancing of Pab1-dependent transla- both yeast and A. nidulans and thus could be organism spe- tion termination (Kashima et al., 2006). However, the cific. Recent work by ( Lasalde et al., 2014) has identified 11 possibility that Lys450 Gln mutation could affect the effi - An novel phosphorylation sites in S. cerevisiae, five of which are ciency of translation termination and ribosome recycling can- conserved within A. thaliana, D. melanogaster, C. elegans and not be ruled out. Interestingly, Upf2 co-immunoprecipitation human homologue of Upf1, as well as A. nidulans (Table 1). In with upf1:Lys498 Gln is increased along with the phosphory- yeast, Thr194 is a phosphorylation site conserved in its human lation state of Upf1, implying that the phosphorylation of homologue (Lasalde et al., 2014). However, the corresponding Upf1 requires Upf2 interaction and is independent of eRF3 amino acid in the three Aspergillus species relates to Ser and (Kashima et al., 2006). 3′ tagging of PTC+ mRNAs has been thus do not appear to be conserved. It is well known that along found to be induced by NMD machinery in A. nidulans with Tyr, Thr and Ser are also implicated in reversible phos- (Morozov et al., 2012) and upf1:Lys450 Gln and Upf2 inter- An phorylation; therefore, these sites could still be functional in action should not be disrupted, although it could be higher promoting Upf1 phosphorylation in Aspergillus. Thr194 cor- than that of WT as found in mammals (Kashima et al., 2006). responds to Ser205 , Ser198 (A. oryzae) and Ser207 Therefore, as Upf1-eRF3 interaction is thought to be crucial An Ao Af (A. fumigatus). Similar to Thr194 Ser492 is another phos- for 3′ tagging and ribosome recycling, a phenotype in which y y phorylation site which is conserved in human Upf1 and cor- ribosomes will be unable to dissociate from mRNA should be responds to Ser506 , Ser499 and Ser508 . Tyr754 is observed along with an increased level of tagging. Replacement An Ao Af y possibly the most interesting phosphorylation site found in this of Lys with Glu in yeast, Lys436 Glu, has also been shown to study due to its high conservation, suggesting importance of reduce Upf1 interaction with both eRF3 and Upf2 (He, this amino acid for Upf1 function. Therefore, mutational anal- Ganesan and Jacobson, 2013). Lys450 corresponds to An ysis of these respective sites to mimic phosphorylation and Lys445 and Lys452 in A. oryzae and A. fumigatus, respec- Ao Af dephosphorylation in Upf1 may facilitate a deeper understand- tively, and may occupy similar roles. ing into the role of phosphorylation in promoting 3′ tagging. Another mutation found to alter interaction with the release factors is Cys126 Ser. This mutation has been shown to increase Upf1-eRF interaction both eRF1 and eRF3 interaction with Upf1 while significantly It has been proposed that interaction of eRF3 with Upf1 is reducing its phosphorylation state. It has also been shown that critical for 3′ end tagging for both PTC+ and NTC+ tran- this mutation has the ability to inhibit NMD (Kashima et al., scripts with a short poly(A) tail in initiating translational 2006) and interaction between the upf1-Cys126Ser:SMG1 com- repression and degradation (Morozov et al., 2012). It has plex with either Upf2 or Upf3b is abolished (Ivanov et al., 2008). been suggested that Upf1 interacts with eRF3 via the CH The corresponding amino acid in yeast Cys65 and its replace- domain however, mutated residues within the ATPase domain ment with Ser has also been shown to substantially reduce Upf2 have also been found to alter interaction. Replacement of interaction and alter Upf1:RNA interaction as observed by gel Lys498 with Gln (Lys498 Gln) in Upf1 has been shown to shift assay (Weng, Czaplinski and Peltz, 1996). However, neither h h decrease interaction of Upf1 with eRF3 while exhibiting the mutation has been found to induce a nonsense suppression phe- same levels of eRF1 interaction as in WT. Phosphorylation of notype. Both Cys126 and Cys65 are conserved and corre- h y Upf1 and Upf1–Upf2 interaction are both increased while spond to Cys74 . Therefore, the Cys74 Ser mutation may An An ATPase activity is abolished (Kashima et al., 2006). Mutation have an opposite effect to Lys436 Glu and Lys498 Gln in that y h of the corresponding site in yeast, Lys436 Glu, shows a eRF3 interaction will increase while showing a decrease in decrease in NMD activity, leading to stabilization of PTC+ Upf1 phosphorylation. Additionally, Upf2 interaction with mRNAs. Additionally, ATPase activity is attenuated while upf1:Cys126 Ser is abolished and thus NMD activity is also translational read-through has also been found to occur. inhibited. Therefore, as 3′ tagging is induced by NMD, mutation Replacement of Lys436 with Glu has been shown to reduce of Cys74 may disrupt tagging, leading to an increase in stalled An eRF3 and Upf2 interaction with Upf1 (He, Ganesan and ribosomes on the PTC. Cys74 relates to Cys69 and Cys76 An Ao Af Jacobson, 2013) while substitution with Ala decreases Upf1 in Aspergillus homologues. interaction with eRF1. Furthermore, Lys436 Ala results in ATP being unable to dissociate from the RNA:Upf1 complex Nonsense suppression phenotype and Upf1 will not favour eRF3 binding over RNA when ATP In the presence of a PTC, Upf1 plays a key role in activating is present. Both Lys498 and Lys436 are conserved and cor- h y NMD, recognizing the nonsense mutation, enhancing transla- respond to Lys450 in A. nidulans. An tion termination and mRNA degradation. Phosphorylated Lys450 is located in the ATPase/helicase domain and cor- Upf1 interacts with translation initiation factor 3, eIF3 prevent- An responding residues in human and yeast are known to be ing the 43S pre-initiation ribosomal complex from joining to involved in interaction with eRF3 and eRF1 (Kashima et al., the 60S and promoting eIF3-dependent translational repression 6 Bioscience Horizons • Volume 8 2015 Research article Table 2. Sequence similarity of Aspergillus Upf1 homologues (Gleghorn and Maquat, 2011). Mutations in Upf1 have been found to disrupt translational repression during NMD and induce translational read-through (Weng, Czaplinski and Peltz, UPF1 sequences aligned Score 1996). Cys62 Ser has been shown to reduce NMD activity Yeast: Human 48.51 while almost abolishing Upf1 interaction with Upf2. A non- sense suppression phenotype is also displayed; however, this Yeast: A. nidulans 49.74 defect is rescued when Upf2 is present (He, Ganesan and Human: A. nidulans 54.5 Jacobson, 2013). This suggests that Upf2 can suppress the defects caused by Cys62 Ser. Moreover, a series of substitu- A. nidulans: A. oryzae 92.54 tions in the CH-domain of yeast Upf1 have been found to dis- play similar phenotypes. Cys62 Ser, His94 Arg, His98 Arg, y y y A. nidulans: A. fumigatus 92.29 Cys 122Ser and Cys125 Ser result in a nonsense suppression y y A. oryzae: A. fumigatus 93.75 phenotype while decreasing NMD activity. Additionally, these mutations reduce helicase activity and alter RNA interaction a Scores calculated by the number of identities between two sequences divided by properties of Upf1 (Weng, Czaplinski and Peltz, 1996). the length of the alignment in amino acids and expressed as a percentage. Furthermore, His94 Arg, His98 Arg and Cys122 Ser also y y y decrease both Upf2 and Upf3 interaction with Upf1 and show a decrease in E3 ligase activity (Takahashi et al., 2008). As Residues within the helicase domain of Upf1 have also these mutations failed to stimulate NMD, it has been proposed been found to induce translational read-through. The double that the inhibition of ubiquitin ligase activity is responsible for mutation Tyr738 Phe:Tyr742 Phe has been shown to suppress y y this defect. These mutations, however, lead to reduced interac- NMD while displaying a nonsense suppression phenotype. tion of Upf1 with Upf2, a protein known to be a key constitu- ATPase activity and phosphorylation of mutated Upf1 are ent in NMD (Weng, Czaplinski and Peltz, 1996). also decreased (Lasalde et al., 2014). Tyr738 and Tyr742 y y correspond to Tyr754 and Tyr758 , respectively. An An The Cys62 Ser mutation has been shown to inhibit NMD and is likely due to the abolished interaction of Upf1 with It has been found that Tyr738 is a phosphorylation site Upf2 and thus maintaining Upf1 in inactive form. This defect while Tyr742 may only be phosphorylated in some circum- in Upf2 interaction could possibly affect 3′ tagging in A. nidu- stances and thus the phosphorylated state of Upf1 is lans; however, the presence of Upf2 is known to rescue non- decreased upon mutation of both sites. Upf2 interaction with sense suppression and could potentially restore tagging (He, upf1:Tyr738Phe:Tyr742Phe has not been tested; however, the Ganesan and Jacobson, 2013). Therefore, this substitution reduced phosphorylation state of Upf1 is believed to account may be unsuitable for mutational analysis in A. nidulans. for the defect in NMD (Lasalde et al., 2014). Therefore, this His94 Arg, His98 Arg and Cys122 Ser have also been found mutation may be useful in testing 3′ tagging of A. nidulans as y y y to induce translational read-through while having a negative interaction with key factors involved in RNA tagging is effect on NMD. Although Upf1–Upf2 interaction is not abol- believed to remain intact. ished in these strains, a decrease in Upf2 and Upf3 interaction Two interesting mutations within the CH domain of yeast with Upf1 is displayed along with a decrease in E3 ligase Upf1, Cys65 Ser and Cys84 Ser have been shown to result in y y activity compared with WT (Takahashi et al., 2008). The heli- a nonsense suppression phenotype and decreased Upf2 co- case activity of the three mutants is also decreased while inter- immunoprecipitation. The formation of the RNA:Upf1 com- action with RNA is altered as observed by gel shift assay plex is also affected following the mutation of either (Weng, Czaplinski and Peltz, 1996). Hence, mutation of cor- Cys65 Ser or Cys84 Ser as found using a gel shift assay y y responding sites in A. nidulans may also be unsuitable in (Weng, Czaplinski and Peltz, 1996). However, unlike other assessing the mechanisms involved in tagging if Upf2 interac- mutations affecting translational read-through and Upf2 tion is disrupted to a large extent. The mutation Cys125 Ser interaction, NMD activity remains unaffected in these has also been found to suppress the nonsense phenotype while mutants (Weng, Czaplinski and Peltz, 1996; He, Ganesan and inhibiting NMD. As in the His94 Arg, His98 Arg and y y Jacobson, 2013). These mutations can therefore provide a Cys122 Ser, the helicase activity and RNA interaction of Upf1 great insight into NMD-induced tagging with regards to non- are altered; however, its interaction with Upf2 has not been sense suppression while eliminating any defects critical for tested so far. Therefore, it would be of great interest to test mRNA degradation. Cys65 /Cys84 relate to Cys74 / y y An whether this mutation would have any effect on tagging and Cys93 , Cys67 /Cys86 and Cys76 /Cys95 , respectively. An Ao Ao Af Af ribosome clearance in A. nidulans. Cys125 corresponds to Cys134 , Cys129 and Cys136 . An Ao Af Upf1 conservation in Aspergillus species Mutations of Lys436 Glu and Lys436 Ala are also known y y to result in a nonsense suppression phenotype; however, Aspergillus nidulans is a tractable genetic model system in due to their pleiotropic effect on Upf1, these mutations are which key eukaryotic pathways are conserved (Caddick et al., likely to affect other activities and required for 3′ end tagging 2006). Pioneering work in A. nidulans has facilitated research and thus may be unsuitable for analysis in A. nidulans. into industrially relevant fungi such as A. oryzae (used in the 7 Research article Bioscience Horizons • Volume 8 2015 Figure 3. Multiple sequence alignment of Upf1 homologues using ClustalW (Larkin et al., 2007) and annotated using Jalview (Waterhouse et al., 2009). Residues are coloured according to the percentage of residues in each column that agree with the consensus sequence. Numbers indicate amino acid position. 8 Bioscience Horizons • Volume 8 2015 Research article food industry and primarily aids in the fermentation of soya helping me to develop not only as a scientist but also as a beans), as well as pathogenic fungi, including the human oppor- person. Very special thanks to my beloved girlfriend Zoe for tunistic pathogen A. fumigatus (Martinelli and Kinghorn, always believing in me and providing copious amounts of 1994). Therefore, a greater understanding of functional motivation, inspiration and support. domains of Upf1 in A. nidulans may be valuable for other Aspergillus species of great industrial (e.g. for optimisation of Funding gene expression) and medical importance (e.g. discovery new therapeutic targets), including A. oryzae and A. fumigatus, This work was supported by the Department of Applied respectively. Sciences & Health, Coventry University. The length of the CH and ATPase/Helicase domains in human and yeast Upf1 remains similar between species, likely References due to important functional activities they modulate Aitken, A. (1995) 14-3-3 proteins on the MAP, Trends in Biochemical (Culbertson and Leeds, 2003). For example, the Upf1 CH Sciences, 20, 95–97. domain in the absence of Upf2 can interact with its helicase domain, yielding a closed conformation (Chakrabarti et al., Altschul, S., Gish, W., Miller, W. et al. (1990) Basic local alignment search 2011); this in turn increases RNA binding by the helicase tool, Journal of Molecular Biology, 215, 403–410. domain and decreases ATPase and helicase activities. Both Amrani, N., Ganesan, R., Kervestin, S. et al. (2004) A faux 3′-UTR promotes yeast and human Upf1 amino acid sequences have been shown aberrant termination and triggers nonsense-mediated mRNA decay, to exhibit high levels of conservation between one another Nature, 432, 112–118. (Applequist et al., 1997); however, both sequences show higher percentage identities when aligned with A. nidulans. Applequist, S., Selg, M., Raman, C. et al. (1997) Cloning and characteriza- tion of HUPF1, a human homolog of the Saccharomyces cerevisiae As shown in Table 2, Upf1 shows a high conservation nonsense mRNA-reducing UPF1 protein, Nucleic Acids Research, 25, between Aspergillus species (sharing 92%–94% similarity) 814–821. and all residues documented in this study that have been found to affect Upf1 activity in humans and yeast were pres- Caddick, M., Dobson, C., Morozov, I. et al. (2006) Gene regulation in ent in A. fumigatus and A. oryzae sequences if initially con- Aspergillus: from genetics to genomics, Medical Mycology Month, 44, served in A. nidulans (Table 1; Fig. 3). S1–S4. Cerqueira, G., Arnaud, M., Inglis, D. et al. (2013) The Aspergillus Genome Conclusion Database: multispecies curation and incorporation of RNA-Seq data to improve structural gene annotations, Nucleic Acids Research, Upf1 could be a potential target in Aspergillus species of 42(Database issue), D705–D710. industrial (e.g. to improve and optimize the gene expression of recombinant proteins) and biomedical (e.g. to target pathoge- Chakrabarti, S., Jayachandran, U., Bonneau, F. et al. (2011) Molecular nicity) importance. The presented work can pave the way for mechanisms for the RNA-dependent ATPase activity of Upf1 and its further research into the novel mechanisms involved in 3′ tag- regulation by Upf2, Molecular Cell, 41, 693–703. ging in fungi, mammals and plants. In addition, this could Chang, H., Lim, J., Ha, M. et al. (2014) TAIL-seq: genome-wide determina- facilitate discovery of novel targets in the treatment of Upf1- tion of poly(A) tail length and 3′ end modifications, Molecular Cell, 53, mediated genetic disorders in which genes harbouring non- 1044–1052. sense or frame-shift mutations induce PTC, thus leading to diseases such as cystic fibrosis or β-thalassemia. Cheng, Z., Muhlrad, D., Lim, M. et al. (2007) Structural and functional insights into the human Upf1 helicase core, The EMBO Journal, 26, Author’s biography 253–264. Chiu, S., Serin, G., Ohara, O. et al. (2003) Characterization of human R.L. carried out this research as part of his undergraduate dis- Smg5/7a: a protein with similarities to Caenorhabditis elegans SMG5 sertation, which attributed towards the degree of Bsc (Hons) and SMG7 that functions in the dephosphorylation of Upf1, RNA, 9, in Biomedical Science at Coventry University. The project was 77–87. chosen as he has a particular interest in protein structure and function, an area in which he would like to further explore in Chowdhury, A., Mukhopadhyay, J. and Tharun, S. (2007) The decapping postgraduate study. Other areas of interest include Immuno- activator Lsm1p-7p-Pat1p complex has the intrinsic ability to distin- logy and Pathogen-Host Cell Biology. guish between oligoadenylated and polyadenylated RNAs, RNA, 13, 998–1016. Acknowledgements Culbertson, M. and Leeds, P. (2003) Looking at mRNA decay pathways I would like to express my sincere gratitude to Dr Igor through the window of molecular evolution, Curr Opin Genet Dev, 13, Morozov for his continued support throughout this work and 297–314. 9 Research article Bioscience Horizons • Volume 8 2015 Czaplinski, K., Ruiz-Echevarria, M., Paushkin, S. et al. (1998) The surveil- Loh, B., Jonas, S. and Izaurralde, E. (2013) The SMG5–SMG7 heterodimer lance complex interacts with the translation release factors to directly recruits the CCR4–NOT deadenylase complex to mRNAs con- enhance termination and degrade aberrant mRNAs, Genes and taining nonsense codons via interaction with POP2, Genes and Development, 12, 1665–1677. Development, 27, 2125–2138. Doma, M. and Parker, R. (2007) RNA quality control in eukaryotes, Cell, Martinelli, S. and Kinghorn, J. (1994) Aspergillus: 50 Years On, Elsevier, 131, 660–668. Amsterdam, 29, p. 851. Gleghorn, M. and Maquat, L. (2011) UPF1 learns to relax and unwind, Morozov, I. and Caddick, M. (2012) Cytoplasmic mRNA 3′ tagging in Molecular Cell, 41, 621–623. eukaryotes: does it spell the end? Biochemical Society Transactions, 40, 810–814. He, F., Ganesan, R. and Jacobson, A. (2013) Intra- and intermolecular regula- tory interactions in Upf1, the RNA helicase central to nonsense- Morozov, I., Jones, M., Razak, A. et al. (2010a) CUCU modification of mediated mRNA decay in yeast, Molecular and Cell Biology, 33, mRNA promotes decapping and transcript degradation in Asperigllus 4672–4684. nidulans, Molecular Cell Biology, 30, 460. Hoefig, K. and Heissmeyer, V. (2014) Degradation of oligouridylated his - Morozov, I., Jones, M., Spiller, D. et al. (2010b) Distinct roles for Caf1, Ccr4, tone mRNAs: see UUUUU and goodbye, Wiley Interdisciplinary Edc3 and CutA in the co-ordination of transcript deadenylation, Reviews: RNA. doi:10.1002/wrna.1232. decapping and P-body formation in Aspergillus nidulans, Molecular Microbiology, 76, 503. Holbrook, J., Gabriele, N., Hentze, M. et al. (2004) Nonsense-mediated decay approaches the clinic, Nature Genetics, 36, 801–808. Morozov, I., Jones, M., Gould, P. et al. (2012) mRNA 3′ tagging is induced by nonsense-mediated decay and promotes ribosome dissociation, Hwang, J., Sato, H., Tang, Y. et al. (2010) UPF1 association with the cap- Molecular and Cell Biology, 32, 2585. binding protein, CBP80, promotes nonsense-mediated mRNA decay at two distinct steps, Molecular Cell, 39, 396–409. Mühlemann, O. and Jensen, T. (2012) mRNP quality control goes regula- tory, Trends in Genetics, 28, 70–77. Imamachi, N., Tani, H. and Akimitsu, N. (2012) Up-frameshift protein 1 (UPF1): multitalented entertainer in RNA decay, Drug Disc Mullen, T. and Marzluff, W. (2008) Degradation of histone mRNA requires Therapeutics, 6, 55–61. oligouridylation followed by decapping and simultaneous degradation of the mRNA both 5′ to 3′ and 3′ to 5′, Genes and Development, 22, 50. Isken, O., Kim, Y., Hosoda, N. et al. (2008) Upf1 phosphorylation triggers translational repression during nonsense-mediated mRNA decay, Nissan, T., Rajyaguru, P., She, M. et al. (2010) Decapping activators in Cell, 133, 314–327. Saccharomyces cerevisiae act by multiple mechanisms, Molecular Cell, 39, 773. Ivanov, P., Gehring, N., Kunz, J. et al. (2008) Interactions between UPF1, eRFs, PABP and the exon junction complex suggest an inte - Okada-Katsuhata, Y., Yamashita, A., Kutsuzawa, K. et al. (2012) N- and grated model for mammalian NMD pathways, The EMBO Journal, C-terminal Upf1 phosphorylations create binding platforms for 27, 736–747. SMG-6 and SMG-5:SMG-7 during NMD, Nucleic Acids Research, 40, 1251–1266. Kadlec, J., Guilligay, D., Ravelli, R. et al. (2006) Crystal structure of the UPF2-interacting domain of nonsense-mediated mRNA decay factor Parker, R. (2012) RNA degradation in Saccharomyces cerevisae, Genetics, UPF1, RNA, 12, 1817–1824. 191, 671–702. Kashima, I., Yamashita, A., Izumi, N. et al. (2006) Binding of a novel SMG- Popp, M. and Maquat, L. (2013) Organizing principles of mammalian non- 1-Upf1-eRF1-eRF3 complex (SURF) to the exon junction complex sense-mediated mRNA decay, Annual Review of Genetics, 47, 139–165. triggers Upf1 phosphorylation and nonsense-mediated mRNA Pruitt, K., Brown, G, Tatusova, T et al. (2002) Chapter 18, The Reference decay, Genes and Development, 20, 355–367. Sequence (RefSeq) Project in The NCBI Handbook, National Library of Kervestin, S. and Jacobson, A. (2012) NMD: a multifaceted response to Medicine (US), National Center for Biotechnology Information, premature translational termination, Nature Reviews Molecular Cell Bethesda, MD. http://www.ncbi.nlm.nih.gov/books/NBK21091/ Biology, 13, 700–712. (accessed 2 Jan 2015). Kostic, A., Ojesina, A., Pedamallu, C. et al. (2011) PathSeq: software to Rissland, O. and Norbury, C. (2009) Decapping is preceded by 3′ uri- identify or discover microbes by deep sequencing of human tissue, dylation in a novel pathway of bulk mRNA turnover, Nature Structural Nature Biotechnology, 29, 393–396. and Molecular Biology, 16, 616–623. Larkin, M., Blackshields, G., Brown, N. et al. (2007) ClustalW and ClustalX Shoemaker, C. and Green, R. (2012) Translation drives mRNA quality con- version 2, Bioinformatics, 23, 2947–2948. trol, Nature Structural and Molecular Biology, 19, 594–601. Lasalde, C., Rivera, A., León, A. et al. (2014) Identification and functional Takahashi, S., Araki, Y., Ohya, Y. et al. (2008) Upf1 potentially serves as a analysis of novel phosphorylation sites in the RNA surveillance pro- RING-related E3 ubiquitin ligase via its association with Upf3 in yeast, tein Upf1, Nucleic Acids Research, 42, 1916–1920. RNA, 14, 1950–1958. 10 Bioscience Horizons • Volume 8 2015 Research article Tanner, N., Cordin, O., Banroques, J. et al. (2003) The Q motif: a newly Weng, Y., Czaplinski, K. and Peltz, S. (1996) Identification and identified motif in DEAD box helicases may regulate ATP binding and characterization of mutations in the UPF1 gene that affect hydrolysis, Molecular Cell, 11, 127–138. nonsense suppression and the formation of the Upf protein complex but not mRNA turnover, Molecular and Cell Biology, 16, Tzivion, G. and Avruch, J. (2002) 14-3-3 proteins: active cofactors in cel- 5491–5506. lular regulation by serine/threonine phosphorylation, The Journal of Yamashita, A., Ohnishi, T., Kashima, I. et al. (2001) Human SMG-1, a novel Biological Chemistry, 277, 3061–3064. phosphatidylinositol 3-kinase-related protein kinase, associates with Waterhouse, A., Procter, J., Martin, D. et al. (2009) Jalview Version 2-a mul- components of the mRNA surveillance complex and is involved in the tiple sequence alignment editor and analysis workbench, regulation of nonsense-mediated mRNA decay, Genes and Development, Bioinformatics, 25, 1189–1191. 15, 2215–2228.
Bioscience Horizons – Oxford University Press
Published: Jun 16, 2015
Access the full text.
Sign up today, get DeepDyve free for 14 days.