Janin et al. Molecular Cancer (2020) 19:70 https://doi.org/10.1186/s12943-020-01192-8 REVIEW Open Access Disruption of the RNA modifications that target the ribosome translation machinery in human cancer Maxime Janin1†, Laia Coll-SanMartin1† and Manel Esteller1,2,3,4* Abstract Genetic and epigenetic changes deregulate RNA and protein expression in cancer cells. In this regard, tumors exhibit an abnormal proteome in comparison to the corresponding normal tissues. Translation control is a crucial step in the regulation of gene expression regulation under normal and pathological conditions that ultimately determines cellular fate. In this context, evidence shows that transfer and ribosomal RNA (tRNA and rRNA) modifications affect the efficacy and fidelity of translation. The number of RNA modifications increases with the complexity of organisms, suggesting an evolutionary diversification of the possibilities for fine-tuning the functions of coding and non-coding RNAs. In this review, we focus on alterations of modifications of transfer and ribosomal RNA that affect translation in human cancer. This variation in the RNA modification status can be the result of altered modifier expression (writers, readers or erasers), but also due to components of the machineries (C/D or H/ ACA boxes) or alterations of proteins involved in modifier expression. Broadening our understanding of the mechanisms by which site-specific modifications modulate ribosome activity in the context of tumorigenesis will enable us to enrich our knowledge about how ribosomes can influence cell fate and form the basis of new therapeutic opportunities. Keywords: Ribosomal RNA, Transfer RNA, Translation, Human cancer, Epitranscriptomics Introduction biogenesis steps, but the catalytic function of the ribo- Nucleotide modifications arise in most classes of RNA some in translation can also be subtly modulated by spe- and more than 160 modifications are known [1]. Initial cific posttranscriptional modifications and selectively genetic studies in yeast and E. coli demonstrated that translate certain types of messenger RNA [8], such as rRNA modifications play an important role in ribosome the case for the rRNA methyltransferase NSUN5 [9]. function [2–4], supported by further research in the area Nevertheless, translational changes in the context of hu- in the early 2000s [5]. Thenceforth, various ribosomal man cancer have been poorly studied, with most of the and transfer RNA modifications have been found to par- studies done describing how the expression of RNA ticipate in determining cellular fate in cancer [6, 7]. For modifiers could inform patient diagnosis or prognosis rRNA, modifications usually have an effect on early [10–14]. This lack of research can be explained in part by the great investment in effort and resources required * Correspondence: mesteller@carrerasresearch.org to undertake a thorough study of an rRNA or tRNA † 1 Maxime Janin and Laia Coll-SanMartin contributed equally to this work. modification —from cellular model construction and Josep Carreras Leukaemia Research Institute (IJC), Badalona, Barcelona, Catalonia, Spain validation, through RNA-sequencing to visualize the 2 Centro de Investigacion Biomedica en Red Cancer (CIBERONC), Madrid, transcriptome, to ribosome profiling and SILAC— in Spain order to appreciate the differential translational Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Janin et al. Molecular Cancer (2020) 19:70 Page 2 of 13 efficiency of specific messengers, also known as transla- Table 1 Modifications of ribosomal RNA that affect translation tomics [15]. In parallel, other methods are used to evalu- in human cancer ate global protein synthesis, including the O-propargyl- Modification Modifier Cancer References puromycin assay and isotope pulse labeling with S35- Nm Fibrillarin Breast cancer [21, 22] labeled amino acid [16]. Nm SNORD50 Colon cancer [23] Alterations in the expression levels of few RNA modi- Nm SNORD14D Leukemia [24] fiers of tRNA and rRNA have been linked to cancer as SNORD35A well as other human diseases [17–20]. However, the Ψ Dyskerin Pituitary adenoma, breast cancer [25, 26] consequences of changes affecting the core machinery of Ψ SNORA24 Hepatocellular carcinoma [27] protein translation are rarely explained. In the context of human cancer, after a brief description of the type of m6A ZCCHC4 Hepatocellular carcinoma [28] RNA and the machineries involved, we summarize the m5C NSUN5 Glioma [9] modifications that have been functionally characterized (Fig. 1; Tables 1 and 2). We also discuss which paths biogenesis, and those due to specific endogenous or exogen- look promising in cancer therapy and provide a perspec- ous factors. This heterogeneity, which also appears between tive on the possibilities for further deciphering the com- cell types, has been thoroughly reviewed by Maria Barna and plex language of RNA modifications. colleagues [48, 49]. Changes in posttranscriptional modifica- tions of rRNAs can influence translational fidelity, leading to Ribosomal RNA nonsense suppressions or amino acid misincorporations, and The ribosome is a complex system that translates the nucleo- modifying the preferential mode of translation initiation (i.e., tide code of messenger RNA into protein in cells. In humans, CAP versus internal ribosome entry site (IRES)) of key can- this complex machinery comprises four ribosomal RNAs cer genes [50, 51]. Finally, a defect in rRNA posttranscrip- (the 28S, 18S, 5.8S and 5S rRNAs) that are synthesized by tional modifications may cause specific clinical syndromes polymerases I and III, and 80 ribosomal proteins (RPs) tran- and may be associated with a higher incidence of cancer, scribed by polymerase II [46], all of which are organized in a which is the case for DKC1 mutation related to X-linked dys- large 60S subunit and a small 40S subunit. From the tran- keratosis congenita (X-DC), and for Dyskerin overexpressed scription of the ribosomal DNA to the final export into the in prostate cancer [52–55]. cytoplasm, ribosome biogenesis involves more than 400 pro- teins and small nucleolar RNAs (snoRNAs) [47]. Ribosome 2′-O-methylation (Nm) machinery and implications for heterogeneity is the result of various changes including those human cancer in the diversity of the composition of RPs and RPs mutations, The association of snoRNAs in human cancer and their abnormalities in rRNA modifications during ribosomal potential role in cancer diagnosis and therapy has recently Fig. 1 Positions of the various RNA modifications implicated in translation in human cancer in a the ribosome structure and b the transfer RNA. Structures were designed with PyMOL using 4UG0 and 5WWT identifiers from PDB Janin et al. Molecular Cancer (2020) 19:70 Page 3 of 13 Table 2 Modifications of transfer RNA that affect translation in human cancer Modification Modifier tRNA localization Cancer References m5C NSUN2 Cytosol & Mitochondria Breast cancer [29–33] Head and neck Squamous Carcinoma Gallbladder m7G METTL1 Cytosol Hepatocellular carcinoma [34–36] WDR4 mcm5U ALKBH8 Cytosol & Mitochondria Bladder cancer [37–39] 5 2 mcm s U CTU1/2 Cytosol Melanoma [40–42] ELP3 Breast cancer i6A TRIT1 Cytosol & Mitochondria Lung adenocarcinoma [43] Q QTRT1/2 Cytosol & Mitochondria Lymphoma and leukemia [44, 45] Lung cancer Ovarian carcinoma been reviewed [56]. The RNA component of the 2′-O- 62] and subsequently in models in which expression of methylation machinery is a C/D box snoRNA, an evolu- FBL was knocked down [21, 22, 50, 63]. tionarily conserved noncoding small (70–120 nucleotide- In 2009, Belin and colleagues [62] observed that breast long) nucleolar RNA [57]. The assembly and composition cancer cells with increased aggressiveness displayed non- of C/D snoRNP complexes have been extensively reviewed modified global translational activity but had IRES- [5, 19, 58], and are simply summarized here. C/D box initiated translation, notably that of p53 mRNA, which snoRNAs catalyze the methylation of ribose residues (2′- was less efficient. They also noted that control of trans- O-methylation) and can affect every type of nucleotide (A; lational fidelity was substantially reduced. This suggests U; C; G and pseudouridine). It contains two short con- that this aggressive phenotype is associated with pro- served sequence motifs (named C and D), located near the found alterations of ribosomal control, leading to poorer 5′ and 3′ ends of the snoRNA, respectively, and present quality control of translation in cancer cells. in duplicate (C′ and D’). It is the most abundant rRNA The expression of FBL gene is decreased by the direct modification in eukaryotes, with two, 42 and 68 sites binding of the tumor suppressor transcription factor modified in the 5.8S, 18S and 28S subunits, respectively p53. In human mammary epithelial cells, the diminution [59]. The complex involves the methyltransferase fibril- or suppression of p53 expression leads to modifications larin (FBL), an RNA-binding protein 15.5 K (NHP2L1), in the rRNA methylation pattern at the single nucleotide and a heterodimer of two closely related proteins, NOP56 level, impairment of translational fidelity, and increased and NOP58. This complex is named C/D box snoRNP IRES-dependent translation of key cancer-related genes, (SNORD). Modification of the expression of the different such as IGF-1R, C-MYC, VEGF-A, and FGF1/2 [21]. As factors involved can directly influence the methylation sta- TP53 is frequently mutated in cancer, the authors tus of the corresponding ribose. showed that the level of FBL is significantly higher in High levels of fibrillarin (FBL) are accompanied by mutated breast cancer than in wild type cells. They ob- modifications of the rRNA methylation pattern, im- tained the same results in a retrospective statistical ana- pairment of translational fidelity, and an increase of lysis. Specifically, the authors showed that the rRNA internal ribosome entry site (IRES)-dependent trans- methylation pattern at 18 sites distributed along the lation initiation of key cancer genes [21]. Indeed, rRNAs in functionally important domains were altered FBL is overexpressed in prostatic neoplasia [60] and in cells with p53 knockdown, increasing the rRNA breast cancer [21, 22]. It was demonstrated that FBL methylation status due to the increased expression of expression correlates with RNA polymerase I tran- FBL. Functionally, the p53-mediated increase of FBL ex- scriptional activity and production of ribosomal pression leads to an increased bypass of a premature RNA, and is inversely correlated with cancer-cell stop-codon and the misincorporation of amino acids, doubling time [61]. demonstrating the combinatory roles of p53 and FBL in Chronologically, the role of 2′-O-methylation in regu- translational quality control. IRES-dependent translation lating the translation of selected mRNAs was first dem- was stronger, whereas the CAP-dependent mechanism onstrated through the observation of associations showed no significant differences. The complementary between altered 2′-O-methylation profiles and transla- experiment directly modulating FBL expression con- tion vector reporter assays in breast cancer models [21, firmed that the efficacy of IRES-dependent translation Janin et al. Molecular Cancer (2020) 19:70 Page 4 of 13 initiation is driven by rRNA methylation. Clinically, the Fibrillarin affects ribosomal RNA 2′-O-methylation al- same authors have shown that FBL overexpression is as- most globally. Nevertheless, a few studies have identified sociated with tumorigenesis and poor survival in patients particular snoRNA C/D boxes as site-specific writers with breast cancer and promotes cellular proliferation with translational consequences in human cancer. and resistance to doxorubicin chemotherapy of MCF7 SNORD50 methylates rRNA at positions 28S-C2848 breast cancer cells [21]. and 28S-G2863 in humans. Variation in SNORD50 ex- Thereafter, Su and colleagues found that FBL increased pression has been observed in colon cancer, breast car- expression in association with an increasing abundance of cinoma, prostate cancer and, finally, B-cell lymphoma. a cluster of snoRNAs. The decrease of FBL had an anti- Historically, the SNORD50 host gene was found by Ta- tumorigenic effect in vitro and in vivo, demonstrating the naka in 2000 at a chromosome breakpoint in a human importance of the controlled regulation of rRNA modifi- B-cell lymphoma [64]. Further studies have demon- cations [22]. Complementary to the study of Marcel et al. strated an association between SNORD50 and clinical [21], they revealed that FBL overexpression suppresses the outcomes [65–67]. p53 response to stress. As p53 mutation is relatively un- Regarding colon cancer, Pacilli and colleagues in 2013 common in breast cancer (around 20% of cases), this studied the expression of SNORD50 in cell lines and tu- mechanism of FBL overexpression/p53 stress response mors. They found that the level of SNORD50 expression suppression is thought to be an alternative option for p53 was lower in tumors compared with normal counter- inactivation [22]. First, the authors showed that p53 pro- parts, the difference being significant in a subgroup of tein stability was increased upon FBL knockdown and p53 low-stage tumors. They also found a significant associ- mRNAs were enriched in polysomal fractions in these ation between the level of SNORD50 expression and cells. They found that the level of IRES translation of p53 tumor stage [23]. Notably, they showed that SNORD50 and other transcripts was significantly higher in FBL expression is inversely correlated with cellular prolifera- knockdown cells and that this mechanism is driven by the tion, highlighting the fact that ribosome biogenesis re- IRES trans-acting factor PTB in the context of nucleolar quires rRNA modifications as part of the process. The stress [22]. overexpression of SNORD50 in HTC116 successfully in- More recently, in 2017, Erales et al. characterized 2′- creases both SNORD50 and C2848 methylation. They ob- O-methylation in HeLa cells using a RiboMethSeq ap- served a general decrease in CAP-dependent translation proach to quantitatively and qualitatively determine the and an increase in IRES-dependent translation for three modulation of the methylation pattern when FBL is IRESs: CrPV, HCV, and c-Myc, while EMCV IRES transla- knocked down [63]. Upon FBL knockdown, the fre- tion was not affected by changes in the methylation of this quency of 2′-O-methylation decreased at almost all sites, specific site [23]. Interestingly, the literature suggests that but not statistically significantly for all of them, and with SNORD50 expression is inversely correlated with cellular differences among sites indicating site-specific regulation proliferation and tumor progression, results that are not of 2′-O-methylation by FBL. Structurally, the affected fully consistent with the oncogenic role fibrillarin and sites were distributed throughout the ribosome struc- methylation activity play in tumors. Thus, it is possible ture, including functionally important regions like the A that it is not the increase of methylation, but the aberrant and P sites, the intersubunit bridges and the peptide exit 2′-o-methylation of specific positions which may be asso- tunnel, but in contrast, methylation sites close to the ciated to different oncogenic effects. PTC and the decoding center within 18S rRNA were not In 2016, Zhou et al., studied an underlying mechanism affected. FBL knockdown reduced global protein synthe- of self-renewal that drives leukemogenesis. This study is sis, as shown by puromycylation assay and isotope pulse an excellent example of workflow characterizing epitran- labeling with [35S]-labeled amino acids. At the mRNA- scriptomic modifications in human malignancies. The specific translational level, the authors used ribosome authors identified a pathway that links the chimeric profiling of HeLa cells expressing an inducible FBL oncogene protein AML1-ETO to enhanced snoRNA shRNA or control shRNA. They identified few transla- functions via the amino-terminal enhancer of split (AES) tionally altered genes upon FBL knockdown [63]. Using and DDX21 interaction [24]. They showed that deletion a bicistronic reporter, they saw that knockdown of FBL of two snoRNAs (SNORD14D, which methylates the induced a decrease in IRES-dependent translational initi- 18S-C462 site, and SNORD35A, which methylates the ation of FGF1, IGF-1R and the type II encephalomyocar- 28S-C4506 site close to the PTC) suppresses the clono- ditis virus (EMCV) IRES, but not of VEGFA IRES. genic potential of leukemia cells in vitro and delayed Luciferase activity/mRNA ratios, which give an indica- leukemogenesis in vivo. Briefly, they demonstrated that tion of the translation efficiency, showed a decrease in AES is essential for AML1-ETO-induced self-renewal of CAP-dependent translation, and a stronger decrease in leukemia cells in vitro and in vivo. Also in the AML1- the IRES-dependent one. ETO cells, snoRNAs were decreased by AES Janin et al. Molecular Cancer (2020) 19:70 Page 5 of 13 suppression. Indeed, microarray-based gene expression snoRNA H/ACA box or by a pseudouridine synthase profiling in a comparison of fetal liver cells overexpress- (PSU). PSUs recognize substrate and catalyze the isom- ing AML1-ETO from mice wild-type (WT) for AES erization reaction of uridine to pseudouridine without (Aes+/+) or Knockout (KO) (Aes f/f) revealed that 20 of using any cofactors. They are classified into six families the top 100 candidates were snoRNAs, all of which were [68]. The catalytic center involves aspartate as the nu- severely reduced following AES deletion, and preferen- cleophile for all PSUs, even in cases where the mechan- tially C/D box snoRNAs. Conversely, forced expression ism of isomerization is poorly understood [68]. of AES in fetal liver cells induced snoRNA expression. RNA-dependent pseudouridylation involves an H/ At this point, they mapped the pseudouridylation to ACA box RNA that complexes with a set of proteins. evaluate the impact of AES in rRNA modification and Box H/ACA RNAs are noncoding RNAs that fold into a AES depletion weakly affected pseudouridylation in hairpin–hinge–hairpin–tail secondary structure. The these cells. However, two 2′-O-methylation sites guided hinge and tail regions contain evolutionarily conserved by SNORD43 and SNORD32A (methylating 18S-C1703 box H with the consensus sequence “ANANNA” (N for and 18S-G1328, respectively) showed decreased methy- any type of nucleotide) and the trinucleotide Box “ACA, lation. In line with a decrease in snoRNA expression, cell ” respectively. The two hairpins contain an internal loop size was also significantly smaller in AES KO compared known as the pseudouridylation guide pocket, which has with WT. In AML1-ETO-induced expression cells with a short specific sequence complementary to the sub- AES KO and a shRNA AES knockdown cell model, OP- strate RNA. The guiding pockets recognize the sites of Puro incorporation was significantly reduced, revealing a modifications through Watson–Crick base-pairing inter- reduction in translation efficiency following AES deple- actions with substrate RNAs, thereby positioning the tion in AML1-ETO-transduced cells [24]. Depletion of uridine to be modified at the base of the upper stem and NOP58 (a component of the C/D box complex) reduced leaving it unpaired. This brings the target site between C/D box snoRNA levels and rRNA methylation. More- 13 and 16 nucleotides upstream of either box H or box over, the colony formation of various leukemia cell lines ACA [69, 70]. The core proteins are dyskerin (encoded was significantly reduced, suggesting a crucial role for by DKC1 gene), glycine–arginine-rich protein 1 (GAR1), rRNA 2′-O-methylation in clonogenicity. Knocking out non-histone protein 2 (NHP2), and nucleolar protein 10 six SNORDs showed that reduced levels of expression of (NOP10) [69, 71]. Dyskerin is the protein that exhibits SNORD34, SNORD35A and SNORD43 impaired clono- the enzymatic activity within this complex, which is genic growth in the cell model, and curiously, depletion known as H/ACA snoRNP. This enzyme has two dis- of SNORD14D reduced colony formation without affect- tinct functions: the pseudouridylation necessary for ing 18S-C762 methylation. Depletion of SNORD14D, rRNA processing, and the stabilization of the telomerase SNORD34 (which methylates 28S-U2824), SNORD35 or RNA component that is essential for telomerase activity. SNORD43 resulted in a reduction of cell size and pro- Similar to fibrillarin for C/D box RNPs, DKC1 or dys- tein synthesis. In the context of AML1-ETO, knockout kerin regulation affects its rRNA epitranscriptomic activ- of SNORD14D or SNORD35A reduced colony forma- ity. In 1998, Heiss et al. identified the mutation tion in MV4–11 cells and delayed leukemogenesis in im- responsible for X-linked recessive dyskeratosis congenita munodeficient NSG mice. These results indicate that in a gene named XAP101 [52]. This disease is character- different snoRNAs have distinct roles and that some of ized by bone marrow failure, skin abnormalities, and an them are important for leukemogenesis beyond AML1- increased susceptibility to develop cancer [72]. Pseu- ETO-induced leukemia. As expected, fibrillarin overex- douridylation is thought to play a role in translation, pression in AES-knocked-down cells increased 2′-O- modulating the interactions between tRNA, rRNA and methylation at various sites and promoted ribosome ac- mRNA. The influence of modified nucleosides on the tivity, implying that the functions of AES in AML1-ETO local structure of the antisense loop is essential for the cells depend on both snoRNA regulation and rRNA proper binding of tRNA to the ribosome [73]. Mainten- methylation. These results are consistent with the onco- ance of the proper conformation of the three anticodon genic role fibrillarin plays in solid tumors [24]. residues helps foster correct codon–anticodon interac- tions. This may increase translational accuracy by redu- Pseudouridylation (Ψ) machinery and implications for cing the rate of peptide bond formation, thereby human cancer allowing more time to reject incorrect codon–anticodon Pseudouridylation is the most common modification in pairs [73]. RNA and the second modification found in rRNA after In 2006, Montanaro and colleagues showed that, in 2′-O-methylation. There are approximately 106 pseu- human breast cancer, low levels of expression of dys- douridines in rRNA clustered close to functionally im- kerin, as determined by immunohistochemistry and RT- portant sites [59]. These modifications are driven by a PCR, were associated with a better outcome [74]. The Janin et al. Molecular Cancer (2020) 19:70 Page 6 of 13 same year, Yoon et al. were the first to report the influence Further research is needed to fully understand the of dyskerin expression on translation, although this time it interplay between genetic alterations, regulation of epi- occurred in X-linked dyskeratosis congenita (X-DC) [75]. transcriptomics and translation in human tumorigenesis. Using a knocked-in DKC1 point-mutation mouse model, they saw that reduced rRNA pseudouridylation did not Other modifications of ribosomal RNA affecting affect total protein synthesis. They then analyzed the translation in human cancer translationally active ribosome-associated mRNAs from Only two recent studies have identified other types of steady state and activated primary splenic lymphocytes, rRNA modification impairment that influence transla- one of the hematopoietic lineages affected in X-DC. They tion in human cancer. identified three out of 1500 mRNAs whose expression was The first, by Ma and colleagues in 2018, identified the specifically decreased in polysome-associated fractions for N6-methyladenosine ZCCHC4 as being responsible for DKC1m lymphocytes: the p27 tumor suppressor, the anti- modifying the rRNA at position 4220, and this methyl- apoptotic proteins XIAP (X-linked Inhibitor of Apoptosis transferase is overexpressed in human hepatocellular Protein), both of which harbor IRES elements, and Bcl-xL carcinoma [28]. This position is not in a functionally im- showed a significant decrease in their association with portant region but, as seen before, changes in translation polysomes in DKC1m cells compared with wild-type cells are not linked to changes of modifications in these re- [75]. The decrease was observed only at the protein level, gions. Using a luciferase reporter and then l- where a specific translation in DKC1 cells was revealed. homopropargylglycine-derived metabolic labeling of Yoon et al. confirmed IRES-dependent translation of p27 newly synthesized proteins, the authors showed that and XIAP and also established that BCL-XL contains a ZCCHC4 KO in hepG2 cells displayed a lower level of previously unreported IRES element. In 2010, the same global translation than in WT. After a sucrose density group, headed by Davide Ruggero, showed that DKC1 is gradient and polysome-associated mRNA sequencing, mutated in pituitary adenoma and that this genotype they suggested that loss of m6A methylation should re- causes a defect in pseudouridylation activity and reduced sult in a reduction of the tight control of translation in expression of p27 [25]. They observed that this translation mRNA populations, and might increase the translation was reduced in the pituitary of DKC1m mice. DKC1m/ of a subset of mRNAs. Finally, gene ontology indicated p27+/− mice developed a similar pituitary malignancy to that genes affected by ZCCHC4 are involved in mem- that of p27−/− mice. Finally, they found a DKC1 mutation brane protein targeting, the mRNA catabolic process, ER in patients with pituitary adenoma that drastically reduced localization, and translation initiation. We recently stud- DKC1 expression and pseudouridylation, but also brought ied the role of NSUN5, an m5C methyltransferase, in hu- about a reduction in p27 expression. Therefore, DKC1 is a man gliomagenesis [9]. NSUN5 methylates the rRNA at tumor suppressor that controls translation by its direct position 28S-C3761 (old nomenclature 3782) and influ- role in rRNA modifications [25]. In parallel, they also ences global translation, since, under conditions of established that p53 IRES-dependent translation is im- stress, reduced OP-puro and H3-leucine incorporations paired in DCK1m cells during oncogene-induced senes- were observed in cell lines depleted in NSUN5 [9]. This cence in X-DC [76]. Normally, during this senescence condition also leads to a specific translation in response transition, IRES translation is facilitated, promoting the to stress, as determined for some proteins regulated only translation of p53. This mechanism in X-DC is close to at the translational level. We chose to validate NQO1 as that controlling fibrillarin overexpression, which inhibits a relevant candidate for therapeutic targeting [9]. p53 expression [21]. In 2013, Rocchi et al. found that Other rRNA modifiers have been linked to human angiogenesis in human breast epithelial cells was pro- cancer, but the underlying mechanisms driving transla- moted by enhanced VEGF-IRES-mediated translation as- tional change require further investigation. NAT10, for sociated to the lack of dyskerin [26]. example, is a N-acetyltransferase involved in histone and A H/ACA box RNP has been identified as playing a microtubule modification, pre-rRNA processing, but also role in tumorigenesis by regulating translation. A recent in the acetylation of rRNA at positions C1337 and study by McMahon and colleagues has shown the regu- C1842, and the acetylation of leucine/serine tRNA [77, lation of SNORA24 in human hepatocellular carcinoma 78]. NAT10 has been found upregulated in hepatocellu- and its collaboration with RAS mutation to promote lar carcinoma (HCC) [79], ovarian cancer [80] and AML cancer [27]. SNORA24 drives two pseudouridylations [13], and correlates with poor survival of patients. This in the 18S rRNA at positions 609 and 863. Ribosomes upregulation also promotes metastasis of HCC cells [81, lacking the corresponding modifications exhibited per- 82]. Also, sub-cellular redistribution of NAT10 in colo- turbations in the aminoacyl-transfer RNA (aa-tRNA) rectal cancer is able to promote cell motility through selection and altered pre-translocation ribosome com- modifying cytoskeleton dynamics [83]. Another modifier, plex dynamics [27]. NOP2 (also named NSUN1 or p120), a probable rRNA Janin et al. Molecular Cancer (2020) 19:70 Page 7 of 13 methyltransferase for position C4447 of the human 28S RNA polymerase III, aminoacyl-tRNA synthetases and RNA [84], has been described as overexpressed in lung other proteins involved in tRNA maturation steps are adenocarcinoma [14, 85]. In HCC, Wang and colleagues known to be altered in a wide range of tumors. This can have described that LncRNA-hPVT1 stabilizes NOP2, be explained by the need for tRNA availability and trans- thereby increasing cell cycle rate, and promoting both lation efficiency to supply the higher level of protein syn- cell proliferation and the acquisition of stem cell-like thesis that occurs in cancer cells [91, 92]. properties [86]. The cytosine methylated by NOP2 is During different steps of their maturation, modifica- part of the A-loop of the peptidyl transferase center. tions are anchored to tRNA, allowing the correct folding This is a critical position within an area responsible for of their secondary (cloverleaf) and tertiary (L-shaped) the initial interaction with the incoming peptidyl-tRNAs, structures through correct base-pairing, codon recogni- as well as the presentation of tRNAs in the correct con- tion and binding. These ensure the fidelity of translation, formation for the optimal peptidyl transfer reaction [84]. structural stability and integrity [29, 37, 93]. Distur- This mechanism could have major impact on protein bances in the regulation of these steps can create oppor- synthesis in cancer cells with NOP2 expression changes. tunities for cancer to arise. Overall, the role of rRNA modifications in the tumori- genic mechanisms driven by these two modifiers will Transfer RNA modifications in human cancer need further investigation. 5-methylcytidine (m5C) In summary, few studies have clearly identified 5-methylcytidine is incorporated at different positions of changes in translation in human cancer, but there are tRNA by NSUN2, NSUN3, NSUN6, and TRDMT1. enough of them to reveal the importance of this regula- NSUN2 and TRDMT1 are the best functionally char- tion in human tumorigenesis (Fig. 1a; Table 1). acterized epitranscriptomic enzymes. TRDMT1 methyla- tion prevents tRNA cleavage [94, 95]. tRNA fragment accumulation and loss of cytosine-5 methylation in Transfer RNA tRNAs was associated to decreased ribosome density in Transfer RNAs are small RNAs that carry amino acids mRNAs and increased ribosome density in 5′ UTRs to ribosomes that translate an mRNA. This allows amino [96]. In mice, methylation also confers stability to their acid incorporation into the synthesized peptide accord- tRNA substrates as loss of 5-methylcytidine causes tRNA ing to the corresponding codon. In humans, tRNAs can degradation, and loss of Trdmt1 and Nsun2 leads to a carry up to 21 different amino acids. tRNAs present five global decrease of protein synthesis [29]. Upregulation of functional arms or loops. The aminoacylation arm is NSUN2 expression was identified in one-third of pri- where the aminoacyl synthase incorporates the corre- mary breast tumors and breast cancer cell lines. Also, sponding amino acid, whereas the anticodon loop is the high levels of NSUN2 expression were detected in head region containing the three nucleosides that pair with and neck carcinoma. In these studies, this aberrant ex- the mRNA codon. There are also the D-loop, so called pression significantly increases mortality [30–32]. More- because of the dihydrouridines it contains, the T-loop over, the oncogene RLP6 has been found to partially for the conserved TψC sequence, and the variable arm, regulate NSUN2 translation and so tumorigenesis func- whose length differs in each tRNA [87]. tion in human gallbladder carcinoma [33]. tRNAs are synthesized by RNA polymerase III and all pre-tRNAs undergo 5′ and 3′ end processing, and m7G addition of CCA to the processed 3′ end. Some tRNAs METTL1 and WDR4 are both necessary for the introduc- also require intron cleavage and subsequent ligation. tion of the 7-methylguanosine at position 46 [34]. In 2018, tRNA biogenesis occurs in the nucleus and mitochon- Lin et al. have reported that METTL1 knockdown reduces dria, depending where they are codified. Those that are the m7G fraction, which causes tRNA destabilization and synthesized in the nucleus are exported to the cytoplasm ribosome pausing at m7G tRNA-dependent codons that as a final step [88, 89]. Aminoacylation is the covalent leads to a decrease in translation [35]. Overexpression of attachment of amino acids to their corresponding tRNA, METTL1 in hepatocellular carcinoma has been shown to the reaction being catalyzed by aminoacyl-tRNA synthe- promote cell proliferation and migration, and to lead to tases. There is at least one enzyme for each amino acid, poor prognosis in HCC [36]. and they can function in the cytoplasm, the mitochon- dria, or in both locations [90]. Throughout this process, mcm5U tRNA can undergo nucleoside modifications, some of 5-methoxycarbonylmethyluridine (mcm5U) is a common which are introduced into all tRNAs, whereas others are modification in the wobble position. ALKBH8, with tRNA-specific. Modifications can also vary between TRM112 as a partner, catalyzes the methylation of 5- cytoplasmic and mitochondrial tRNA. carboxymethyluridine (cm5U) to mcm5U [37, 38]. Janin et al. Molecular Cancer (2020) 19:70 Page 8 of 13 tRNASec is one of the targets of ALKBH8, this tRNA Queuosine (Q) transfers the amino acid selenocysteine during seleno- Queuosine (7-(((4,5-cis-dihydroxy-2-cyclopenten-1-yl)a- protein translation. A large number of selenoproteins mino)methyl)-7-deazaguanosine) is a derivative of the 7- are involved in the detoxification of ROS. It has been deazagunine located in the wobble position. This modifi- shown that ALKBH8 deficiency causes a decrease in cation is introduced by the tRNA-guanine transglycosy- selenocysteine protein translation and consequently ROS lase (TGT) heterodimer formed by QTRT1 (catalytic stress that triggers oxidative DNA damage [38, 97]. It subunit) and QTRT2 [44]. Eukaryotes cannot synthesize has been reported both in vitro and in vivo that queuine but they obtain it from gut microbes or food. ALKBH8 promotes the expansion, survival and invasion Tuorto et al. suggested that G34-containing tRNAs of bladder cancer [39, 98]. Cancer cells use AKBH8 ex- cause a translational delay that induces an accumulation pression as an antioxidant system against the ROS stress of misfolded proteins and hence, endoplasmic reticulum generated during tumorigenic transformation [97]. stress and an unfolded protein response. In order to Msm5U is necessary for the formation of some deriva- avoid this situation, Q modification prevents the elong- tives, such as 5-methoxycarbonylmethyl-2-thiouridine ation from slowing down [105]. Reduced Q content has (mcm5s2U). CTU1, CTU2 and ELP3 are the thiolation been detected in a wide range of cancers [45]. effectors with URM1 as the sulfur donor [37, 40, 99]. This overview of the modifications affecting transfer The mcm5s2U complex is heavily implicated in cancer. RNAs (Fig. 1b; Table 2) summarizes the characteristics On one hand, its enzymes levels have been detected at that influence cellular fate and that can drive tumori- higher levels in human tumor biopsies and in cultured genic processes. melanoma cells with BRAFV600E mutation. An in silico analysis revealed that this kind of melanoma also has Conclusions high levels of hypoxia-induced factor 1α (HIF1α), which Regulation of translation is crucial for cell proliferation is known for its involvement in tumorigenesis progres- and cell fate in human malignancies. There is increasing sion. The mRNA sequence of HIF1α is rich in U34 co- evidence that epitranscriptional modifications affect dons that require tRNA modification for correct translation by modifying the affinity or the stability of translation. Thus, U34-related enzyme overexpression mRNAs, tRNAs, or the interactions between them inside triggers increased HIF1α translation and thereby cancer the ribosome. transformation [41]. Similarly, U34-related enzymes Here, we have reviewed the various modifications of were found to be overexpressed in breast cancer pa- transfer and ribosomal RNAs that are known to have tients. The IRES trans-acting factor (ITAF) protein DEK consequences for translation in human cancer (Fig. 2). has U34 codon-dependent translation regulation. DEK We chose to exclude modifications affecting other types promotes the translation through the IRES sequence of of RNAs and alterations in the expression and activity of LEF1, whose expression is associated with a higher risk specific translation factors, not only because these have of metastases [42]. already been extensively reviewed [106, 107], but also to present a consistent overview of the role of RNA modifi- cations in the core functions of translation. Isopentenyladenosine (i6A) The role of epigenetics in cancer is now very firmly TRIT1 introduces a dimethylallyl pyrophosphate at pos- established, to the extent that corresponding therapies ition 37, generating an isopentenylated adenosine in exist. However, the recent expansion of our knowledge both cytoplasmic and mitochondrial tRNA [100–102]. In about epitranscriptomics has not yet resulted in any particular, it has been related to selenoprotein expres- clinical applications, although there is evidence of prog- sion since tRNASec is one of its targets [103]. Expression nostic markers and possible treatments. In this sense, of some selenoproteins are influenced by TRIT1 deple- the exhaustive characterization and classification of tion [43]. Spinola et al. observed downregulation of RNA modifiers in each relevant cell type as tumor sup- TRIT1 expression in lung adenocarcinoma patients and pressor or oncogene will provide us with tools to identify noted that expression of TRIT1 in a lung adenocarcin- new targets and design specific treatments that disrupt oma cell line diminishes tumorigenesis. Its tumor sup- the carcinogenic pathways [108, 109]. Certainly, it is very pressor effect is indicated by the role of some important to identify the cellular background, as for a selenoproteins in diminishing oxidative stress [104]. specific cancer type, different environments can be bene- However, as demonstrated in the case of ALKBH8, ficial or lead to apoptosis [110]. which modifies mcm5U34, cancer can also use seleno- Studies highlight that fibrillarin is frequently overex- proteins to maintain oxidative stress at sublethal levels. pressed in human cancer, and its expression controls the Similarly, i6A is needed for ALKBH8 modification to fine-tuning of ribose 2′-O-methylation of the rRNA in take place [43]. human cancer. FBL depletion leads to the regulation of Janin et al. Molecular Cancer (2020) 19:70 Page 9 of 13 Fig. 2 Schematic representation of translational effects triggered by RNA modification defects at rRNA and tRNA in the context of cancer cells IRES translation globally, but some specific mRNAs are emphasizes its potential as a therapeutic target that altered in a different way, such as p53, whose expression could lower the genotoxic effects of anti-cancer treat- increases when FBL is knocked down, and also revealed ment. This could be achieved by exploiting the associ- by ribosome profiling [21]. Further studies would cer- ation between low levels of FBL expression and better tainly confirm the qualitative and quantitative role of breast cancer survival, and its functions in ribosome bio- FBL in human cancer translation and would produce genesis and p53 regulation [111]. Other 2′-O-methyla- substantial information about identifying the down- tion modifiers have shown therapeutic potential, such as stream molecular and cellular functions affected by this SNORD44 in colorectal cancer and SNORD47 in glio- mechanism. Cellular context is a critical factor as a blastoma [112, 113]. stress influence on the translational response to FBL Few studies have highlighted the relevance of dyskerin modulation [21]. Clinically, fibrillarin has not yet been in driving cancer development. Their findings can ex- exploited as an oncological target, but a recent review plain in part the increased risk of cancer observed in X- Janin et al. Molecular Cancer (2020) 19:70 Page 10 of 13 DC. In some other studies, the translational impact of like drugs could specifically target the altered RNAs in pseudouridylation defects is not clear. This is the case cancer. In this regard, in 2013 Begley and colleagues found for SNORA42 in lung cancer [10], in which SNORA42 is that tRNA methyltransferase 9-like protein (hTRM9L) overexpressed in non-small cell lung cancer and its sup- was downregulated in breast, bladder, colorectal, cervix pression inhibits cell growth, proliferation and tumori- and testicular carcinomas and this downregulation was as- genicity of cancer cells by inducing p53-dependent sociated with altered tRNA modification levels [118]. Im- apoptosis. These results were partially confirmed 2 years portantly, hTRM9L-deficient tumors were highly sensitive later [114], but a translational shift driven by SNORA42 to aminoglycoside antibiotics [118]. Regarding tRNAs, remains to be conclusively demonstrated. Strikingly, Bel- NSUN2 overexpression indicates a shorter overall survival lodi et al., found that SNORA42 expression is increased in head and neck carcinomas [32]. Another example is in DKC1(ΔL37)-mutated lymphocytes, which raises the ALKBH8, which is crucial in bladder cancer transform- possibility that certain snoRNAs may be selectively in- ation. It has been proposed that an intravesical injection creased as a compensatory mechanism for perturbations of ALKBH8 siRNA may reverse the invasive character of in subsets of H/ACA snoRNAs [115]. They also ob- the neoplasm in patients [38, 98]. It has also been de- served an unexpected heterogeneity of expression of the scribed that a deficit of ALKBH8 increases the sensitivity H/ACA snoRNAs in cells of patients with X-DC harbor- to DNA-damaging agents and could complement the pre- ing distinct DKC1 mutations. This is in line with the un- vious strategy [38]. Moreover, Rapino et al. proposed that expected 2′-O-methylation alteration observed in X-DC. targeting wobble uridine 34 (U34) tRNA (U34 enzymes) Very recently, a study linked the rRNA 2′-O-methyla- can avoid the resistance to Vemurafenib in BRAFV600E tion alteration caused by NPM1 mutations to the devel- melanoma [41]. opment of congenital dyskeratosis [116]. These results RNA modifications have a promising future. The rela- demonstrate the complex etiology of such diseases and tively small number of studies, whether considering how a common phenotype can arise as the result of dif- tRNA or rRNA, that completely describe the transla- ferent alterations. tional role of a specific modification in this context high- Modifications can directly influence the level of tRNAs lights the significant gap that remains to be filled if we by modulating their stability. The by-products formed are to better understand the complex mechanism of by angiogenin cleavage regulate translation in human tumorigenesis (Fig. 1; Tables 1 and 2). In this puzzle, the cancer. There are many examples in the literature of ef- addition or removal of a piece (RNA modifications) can forts that have been made to improve our understanding effectively change the final picture. of the subtle influence of proteomic modifications and Abbreviations the consequent tumorigenesis. Another aspect emerging FBL: Fibrillarin; IRES: Internal ribosome entry site; i6A: Isopentenyladenosine; from these studies is the difficulty of appreciating the ITAF: IRES trans-acting factor; m5C: 5-methylcytidine; mcm5U: 5- enzyme-site-specific role. Indeed, this is the case for methoxycarbonylmethyluridine; m7G: 7-methylguanosine; PSU: Pseudouridine synthase; Q: Queuosine; RPs: Ribosomal proteins; rRNA: Ribosomal RNA; ribosomal RNA modifiers also, lots of enzymes have a SILAC: Stable Isotope Labeling by/with Amino acids in Cell culture; panel of activities, and the modulation of its expression SNORD: C/D box snoRNP; snoRNAs: Small nucleolar RNAs; TGT: tRNA-guanine can have side effects that also play a role in translational transglycosylase; tRNA: Transfer RNA; X-DC: X-linked dyskeratosis congenita changes. For example, TRIT1 and ALKBH8 functions Acknowledgements are deeply related as mcm5U have been described as i6A Not applicable dependent, although they are described as a tumor sup- Authors’ contributions pressor and oncogene respectively, some hypotheses Maxime Janin and Laia Coll-SanMartin have contributed equally to this art- exist as other functions of the genes but more research icle. All authors were involved in writing the manuscript. The authors read are needed to clarify it. This is also the case for NSUN5, and approved the final manuscript. we related the translational changes to the modification Funding on the rRNA as it logically seems to have the most dir- We thank the CERCA Programme / Generalitat de Catalunya for institutional ect role on ribosome function. Nevertheless, considering support. ME is an ICREA Research Professor. Research at ME lab is supported by the Secretariat for Universities and Research of the Ministry of Business the clinical implications of these modifiers, it is certain and Knowledge of the Government of Catalonia (2017SGR1080); MCIU/AEI/ that the global phenotype takes precedence over the sin- FEDER, UE (RTI2018–094049-B-I00 and SAF2014–55000) and Olga Torres gle modification. A nuance is the relevance of a cancer- Foundation and from Instituto de Salud Carlos III (ISCIII), co-financed by the European Development Regional Fund, ‘A way to achieve Europe’ ERDF specific translation in drug development and therapeutic (grant CB16/12/00312). opportunities [117]. For example, in 1990, the discovery that eIF4E overexpression could drive tumor initiation Availability of data and materials has led to the development of strategies to decrease its Not applicable expression. To go further, ribosomal RNA modifications Ethics approval and consent to participate can lead to antibiotic resistance in yeast, so antibiotic- Not applicable Janin et al. Molecular Cancer (2020) 19:70 Page 11 of 13 Consent for publication 20. Wei F-Y, Tomizawa K. tRNA modifications and islet function. Diabetes Obes All authors consent to publication. Metab. 2018;20:20–7. 21. Marcel V, Ghayad SE, Belin S, Therizols G, Morel A-P, Solano-Gonzàlez E, et al. p53 acts as a safeguard of translational control by regulating Fibrillarin Competing interests and rRNA methylation in Cancer. Cancer Cell. 2013;24(3):318–30. ME is a consultant of Ferrer International and Quimatryx. The other authors 22. Su H, Xu T, Ganapathy S, Shadfan M, Long M, Huang TH-M, et al. Elevated declare that they have no conflict of interest. snoRNA biogenesis is essential in breast cancer. Oncogene. 2014;33(11):1348–58. 23. Pacilli A, Ceccarelli C, Treré D, Montanaro L. SnoRNA U50 levels are regulated Author details by cell proliferation and rRNA transcription. Int J Mol Sci. 2013;14(7):14923–35. 1 Josep Carreras Leukaemia Research Institute (IJC), Badalona, Barcelona, 24. Zhou F, Liu Y, Rohde C, Pauli C, Gerloff D, Köhn M, et al. AML1-ETO requires Catalonia, Spain. 2Centro de Investigacion Biomedica en Red Cancer enhanced C/D box snoRNA/RNP formation to induce self-renewal and (CIBERONC), Madrid, Spain. 3Physiological Sciences Department, School of leukaemia. Nat Cell Biol. 2017;19(7):844–55. Medicine and Health Sciences, University of Barcelona (UB), Barcelona, 25. Bellodi C, Krasnykh O, Haynes N, Theodoropoulou M, Peng G, Montanaro L, Catalonia, Spain. 4Institucio Catalana de Recerca i Estudis Avançats (ICREA), et al. Loss of function of the tumor suppressor DKC1 perturbs p27 Barcelona, Catalonia, Spain. translation control and contributes to pituitary tumorigenesis. Cancer Res. 2010;70(14):6026–35. Received: 5 January 2020 Accepted: 24 March 2020 26. Rocchi L, Pacilli A, Sethi R, Penzo M, Schneider RJ, Treré D, et al. Dyskerin depletion increases VEGF mRNA internal ribosome entry site-mediated translation. Nucleic Acids Res. 2013;41(17):8308–18. 27. McMahon M, Contreras A, Holm M, Uechi T, Forester CM, Pang X, et al. A References single H/ACA small nucleolar RNA mediates tumor suppression downstream 1. Boccaletto P, Machnicka MA, Purta E, Piątkowski P, Bagiński B, Wirecki TK, of oncogenic RAS. eLife. 2019;8:e48847. et al. MODOMICS: a database of RNA modification pathways. 2017 update. 28. Ma H, Wang X, Cai J, Dai Q, Natchiar SK, Lv R, et al. N6-Methyladenosine Nucleic Acids Res. 2018;46(D1):D303–7. methyltransferase ZCCHC4 mediates ribosomal RNA methylation. Nat Chem 2. Poldermans B, Bakker H, Van Knippenberg PH. Studies on the function of Biol. 2019;15(1):88–94. two adjacent N 6 , N 6 -dimethyladenosines near the 3′ end of 16 S ribosomal RNA of Escherichia coli. IV. The effect of the methylgroups on 29. Tuorto F, Liebers R, Musch T, Schaefer M, Hofmann S, Kellner S, et al. RNA ribosomal subunit interaction. Nucleic Acids Res. 1980;8(1):143–52. cytosine methylation by Dnmt2 and NSun2 promotes tRNA stability and 3. Sirum-Connolly K, Mason T. Functional requirement of a site-specific ribose protein synthesis. Nat Struct Mol Biol. 2012;19(9):900–5. methylation in ribosomal RNA. Science. 1993;262(5141):1886–9. 30. Yi J, Gao R, Chen Y, Yang Z, Han P, Zhang H, et al. Overexpression of 4. Sirum-Connolly K, Peltier JM, Crain PF, McCloskey JA, Mason TL. Implications NSUN2 by DNA hypomethylation is associated with metastatic progression of a functional large ribosomal RNA with only three modified nucleotides. in human breast cancer. Oncotarget. 2017;8(13):20751–65. Biochimie. 1995;77(1–2):30–9. 31. Frye M, Dragoni I, Chin S-F, Spiteri I, Kurowski A, Provenzano E, et al. 5. Decatur WA, Fournier MJ. rRNA modifications and ribosome function. Genomic gain of 5p15 leads to over-expression of Misu (NSUN2) in breast Trends Biochem Sci. 2002;27(7):344–51. cancer. Cancer Lett. 2010;289(1):71–80. 6. Popis MC, Blanco S, Frye M. Posttranscriptional methylation of transfer and 32. Lu L, Zhu G, Zeng H, Xu Q, Holzmann K. High tRNA Transferase NSUN2 ribosomal RNA in stress response pathways, cell differentiation, and cancer. gene expression is associated with poor prognosis in head and neck Curr Opin Oncol. 2016;28(1):65–71. squamous carcinoma. Cancer Investig. 2018;36(4):246–53. 7. Nachtergaele S, He C. The emerging biology of RNA post-transcriptional 33. Gao Y, Wang Z, Zhu Y, Zhu Q, Yang Y, Jin Y, et al. NOP2/Sun RNA modifications. RNA Biol. 2017;14(2):156–63. methyltransferase 2 promotes tumor progression via its interacting partner 8. Motorin Y, Helm M. RNA nucleotide methylation. Wiley Interdiscip Rev RNA. RPL6 in gallbladder carcinoma. Cancer Sci. 2019;110(11):3510–9. 2011;2(5):611–31. 34. Bahr A, Hankeln T, Fiedler T, Hegemann J, Schmidt ER. Molecular analysis of 9. Janin M, Ortiz-Barahona V, de Moura MC, Martínez-Cardús A, Llinàs-Arias P, METTL1, a novel human methyltransferase-like gene with a high degree of Soler M, et al. Epigenetic loss of RNA-methyltransferase NSUN5 in glioma phylogenetic conservation. Genomics. 1999;57(3):424–8. targets ribosomes to drive a stress adaptive translational program. Acta 35. Lin S, Liu Q, Lelyveld VS, Choe J, Szostak JW, Gregory RI. Mettl1/Wdr4-Mediated Neuropathol. 2019;138;1053–74. m7G tRNA Methylome Is Required for Normal mRNA Translation and Embryonic 10. Mei Y-P, Liao J-P, Shen J, Yu L, Liu B-L, Liu L, et al. Small nucleolar RNA 42 Stem Cell Self-Renewal and Differentiation. Mol Cell. 2018;71(2):244–255.e5. acts as an oncogene in lung tumorigenesis. Oncogene. 2012;31(22):2794– 36. Tian Q-H, Zhang M-F, Zeng J-S, Luo R-G, Wen Y, Chen J, et al. METTL1 804. overexpression is correlated with poor prognosis and promotes 11. Braicu C, Zimta A-A, Harangus A, Iurca I, Irimie A, Coza O, et al. The function hepatocellular carcinoma via PTEN. J Mol Med Berl Ger. 2019;97(11):1535–45. of non-coding RNAs in lung cancer tumorigenesis. Cancers. 2019;11(5):605. 37. Songe-Møller L, van den Born E, Leihne V, Vågbø CB, Kristoffersen T, Krokan 12. Yang X, Li Y, Li L, Liu J, Wu M, Ye M. SnoRNAs are involved in the HE, et al. Mammalian ALKBH8 possesses tRNA methyltransferase activity progression of ulcerative colitis and colorectal cancer. Dig Liver Dis. 2017; required for the biogenesis of multiple wobble uridine modifications 49(5):545–51. implicated in translational decoding. Mol Cell Biol. 2010;30(7):1814–27. 13. Liang P, Hu R, Liu Z, Miao M, Jiang H, Li C. NAT10 upregulation indicates a 38. Fu D, Brophy JAN, Chan CTY, Atmore KA, Begley U, Paules RS, et al. Human poor prognosis in acute myeloid leukemia. Curr Probl Cancer. 2019;2: AlkB homolog ABH8 is a tRNA methyltransferase required for wobble uridine 100491. https://doi.org/10.1016/j.currproblcancer.2019.06.006. modification and DNA damage survival. Mol Cell Biol. 2010;30(10):2449–59. 14. Uchiyama B, Saijo Y, Kumano N, Abe T, Fujimura S, Ohkuda K, et al. 39. Shimada K, Nakamura M, Anai S, De Velasco M, Tanaka M, Tsujikawa K, et al. Expression of nucleolar protein p120 in human lung cancer: difference in A novel human AlkB homologue, ALKBH8, contributes to human bladder histological types as a marker for proliferation. Clin Cancer Res. 1997;3(10): cancer progression. Cancer Res. 2009;69(7):3157–64. 1873–7. 40. Liu Y, Vinyard DJ, Reesbeck ME, Suzuki T, Manakongtreecheep K, Holland PL, 15. Zhao J, Qin B, Nikolay R, Spahn CMT, Zhang G. Translatomics: the global et al. A [3Fe-4S] cluster is required for tRNA thiolation in archaea and view of translation. Int J Mol Sci. 2019;20(1):212. eukaryotes. Proc Natl Acad Sci U S A. 2016;113(45):12703–8. 16. Iwasaki S, Ingolia NT. The growing toolbox for protein synthesis studies. 41. Rapino F, Delaunay S, Rambow F, Zhou Z, Tharun L, De Tullio P, et al. Trends Biochem Sci. 2017;42(8):612–24. Codon-specific translation reprogramming promotes resistance to targeted 17. Jonkhout N, Tran J, Smith MA, Schonrock N, Mattick JS, Novoa EM. The RNA therapy. Nature. 2018;558(7711):605–9. modification landscape in human disease. RNA. 2017;23(12):1754–69. 42. Delaunay S, Rapino F, Tharun L, Zhou Z, Heukamp L, Termathe M, et al. Elp3 18. Kadumuri RV, Janga SC. Epitranscriptomic code and its alterations in human links tRNA modification to IRES-dependent translation of LEF1 to sustain disease. Trends Mol Med. 2018;24(10):886–903. metastasis in breast cancer. J Exp Med. 2016;213(11):2503–23. 19. Therizols G, Laforêts F, Marcel V, Catez F, Bouvet P, Diaz J-J. Ribosomal RNA 43. Fradejas N, Carlson BA, Rijntjes E, Becker N-P, Tobe R, Schweizer U. methylation and cancer. In: Epigenetic cancer therapy: Elsevier; 2015. p. Mammalian Trit1 is a tRNA([Ser]sec)-isopentenyl transferase required for full 115–39. [cited 2019 Nov 5]. selenoprotein expression. Biochem J. 2013;450(2):427–32. Janin et al. Molecular Cancer (2020) 19:70 Page 12 of 13 44. Chen Y-C, Kelly VP, Stachura SV, Garcia GA. Characterization of the human 70. Yu Y-T, Terns RM, Terns MP. Mechanisms and functions of RNA-guided RNA tRNA-guanine transglycosylase: confirmation of the heterodimeric subunit modification. In: Grosjean H, editor. Fine-Tuning of RNA Functions by structure. RNA N Y N. 2010;16(5):958–68. Modification and Editing. Berlin: Springer Berlin Heidelberg; 2004. p. 223–62. 45. Vinayak M, Pathak C. Queuosine modification of tRNA: its divergent role in [cited 2019 Nov 11]. cellular machinery. Biosci Rep. 2009;30(2):135–48. 71. Penzo M, Montanaro L. Turning Uridines around: role of rRNA 46. Dalla Venezia N, Vincent A, Marcel V, Catez F, Diaz J-J. Emerging role of Pseudouridylation in ribosome biogenesis and ribosomal function. eukaryote ribosomes in translational control. Int J Mol Sci. 2019;20(5):1226. Biomolecules. 2018;8(2):38. 47. Bohnsack KE, Bohnsack MT. Uncovering the assembly pathway of human 72. Marrone A, Dokal I. Dyskeratosis congenita: molecular insights into telomerase ribosomes and its emerging links to disease. EMBO J 2019;38(13). [cited function, ageing and cancer. Expert Rev Mol Med. 2004;6(26):1–23. 2019 Nov 11]. 73. Gray MC, Michael W. Pseudouridine in RNA: what, where, how, and why. 48. Genuth NR, Barna M. Heterogeneity and specialized functions of translation IUBMB Life Int Union Biochem Mol Biol Life. 2000;49(5):341–51. machinery: from genes to organisms. Nat Rev Genet. 2018;19(7):431–52. 74. Montanaro L, Brigotti M, Clohessy J, Barbieri S, Ceccarelli C, Santini D, et al. 49. Xue S, Barna M. Specialized ribosomes: a new frontier in gene regulation Dyskerin expression influences the level of ribosomal RNA pseudo- and organismal biology. Nat Rev Mol Cell Biol. 2012;13(6):355–69. uridylation and telomerase RNA component in human breast cancer. J 50. Basu A, Das P, Chaudhuri S, Bevilacqua E, Andrews J, Barik S, et al. Pathol. 2006;210(1):10–8. Requirement of rRNA methylation for 80S ribosome assembly on a cohort 75. Yoon A. Impaired control of IRES-mediated translation in X-linked of cellular internal ribosome entry sites. Mol Cell Biol. 2011;31(22):4482–99. Dyskeratosis Congenita. Science. 2006;312(5775):902–6. 51. Ruggero D. Translational control in Cancer etiology. Cold Spring Harb 76. Bellodi C, Kopmar N, Ruggero D. Deregulation of oncogene-induced Perspect Biol. 2013;5(2):a012336. senescence and p53 translational control in X-linked dyskeratosis congenita. 52. Heiss NS, Knight SW, Vulliamy TJ, Klauck SM, Wiemann S, Mason PJ, et al. X- EMBO J. 2010;29(11):1865–76. linked dyskeratosis congenita is caused by mutations in a highly conserved 77. Sleiman S, Dragon F. Recent advances on the structure and function of RNA gene with putative nucleolar functions. Nat Genet. 1998;19(1):32–8. Acetyltransferase Kre33/NAT10. Cells. 2019;05:8(9). 53. Ruggero D. Dyskeratosis Congenita and Cancer in mice deficient in 78. Sharma S, Langhendries J-L, Watzinger P, Kötter P, Entian K-D, Lafontaine ribosomal RNA modification. Science. 2003;299(5604):259–62. DLJ. Yeast Kre33 and human NAT10 are conserved 18S rRNA cytosine 54. Sieron P, Hader C, Hatina J, Engers R, Wlazlinski A, Müller M, et al. DKC1 acetyltransferases that modify tRNAs assisted by the adaptor Tan1/ overexpression associated with prostate cancer progression. Br J Cancer. THUMPD1. Nucleic Acids Res. 2015;43(4):2242–58. 2009;101(8):1410–6. 79. Li Q, Liu X, Jin K, Lu M, Zhang C, Du X, et al. NAT10 is upregulated in 55. Alawi F, Lin P. Dyskerin is required for tumor cell growth through hepatocellular carcinoma and enhances mutant p53 activity. BMC Cancer. mechanisms that are independent of its role in telomerase and only 2017;17(1):605. partially related to its function in precursor rRNA processing. Mol Carcinog. 80. Tan TZ, Miow QH, Huang RY, Wong MK, Ye J, Lau JA, et al. Functional 2011;50(5):334–45. genomics identifies five distinct molecular subtypes with clinical relevance 56. Liang J, Wen J, Huang Z, Chen X, Zhang B, Chu L. Small Nucleolar RNAs: and pathways for growth control in epithelial ovarian cancer. EMBO Mol insight into their function in Cancer. Front Oncol. 2019;9:587. Med. 2013;5(7):1051–66. 57. Monaco P, Marcel V, Diaz J-J, Catez F. 2′-O-methylation of ribosomal RNA: 81. Ma R, Chen J, Jiang S, Lin S, Zhang X, Liang X. Up regulation of NAT10 towards an Epitranscriptomic control of translation? Biomolecules. 2018;8(4): promotes metastasis of hepatocellular carcinoma cells through epithelial-to- 106. mesenchymal transition. Am J Transl Res. 2016;8(10):4215–23. 58. Watkins NJ, Bohnsack MT. The box C/D and H/ACA snoRNPs: key players in 82. Zhang X, Chen J, Jiang S, He S, Bai Y, Zhu L, et al. N-Acetyltransferase 10 the modification, processing and the dynamic folding of ribosomal RNA: enhances doxorubicin resistance in human hepatocellular carcinoma cell box C/D and H/ACA snoRNPs. Wiley Interdiscip Rev RNA. 2012;3(3):397–414. lines by promoting the epithelial-to-Mesenchymal transition. Oxidative Med 59. Taoka M, Nobe Y, Yamaki Y, Sato K, Ishikawa H, Izumikawa K, et al. Cell Longev. 2019;2019:1–14. Landscape of the complete RNA chemical modifications in the human 80S 83. Zhang H, Hou W, Wang H-L, Liu H-J, Jia X-Y, Zheng X-Z, et al. GSK-3 ribosome. Nucleic Acids Res. 2018;46(18):9289–98. -regulated N-Acetyltransferase 10 is involved in colorectal Cancer invasion. 60. Koh CM, Gurel B, Sutcliffe S, Aryee MJ, Schultz D, Iwata T, et al. Alterations in Clin Cancer Res. 2014;20(17):4717–29. Nucleolar structure and gene expression programs in prostatic Neoplasia 84. Sharma S, Yang J, Watzinger P, Kötter P, Entian K-D. Yeast Nop2 and Rcm1 are driven by the MYC oncogene. Am J Pathol. 2011;178(4):1824–34. methylate C2870 and C2278 of the 25S rRNA, respectively. Nucleic Acids 61. Derenzini M, Trerè D, Pession A, Montanaro L, Sirri V, Ochs RL. Nucleolar Res. 2013;41(19):9062–76. function and size in cancer cells. Am J Pathol. 1998;152(5):1291–7. 85. Sato G, Saijo Y, Uchiyama B, Kumano N, Sugawara S, Fujimura S, et al. 62. Belin S, Beghin A, Solano-Gonzàlez E, Bezin L, Brunet-Manquat S, Textoris J, Prognostic value of Nucleolar protein p120 in patients with resected lung et al. Dysregulation of ribosome biogenesis and translational capacity is adenocarcinoma. J Clin Oncol. 1999;17(9):2721. associated with tumor progression of human breast Cancer cells. PLoS One. 86. Wang F, Yuan J-H, Wang S-B, Yang F, Yuan S-X, Ye C, et al. Oncofetal long 2009;4(9):e7147. noncoding RNA PVT1 promotes proliferation and stem cell-like property of 63. Erales J, Marchand V, Panthu B, Gillot S, Belin S, Ghayad SE, et al. Evidence hepatocellular carcinoma cells by stabilizing NOP2: WANG ET AL. for rRNA 2′-O-methylation plasticity: control of intrinsic translational Hepatology. 2014;60(4):1278–90. capabilities of human ribosomes. Proc Natl Acad Sci. 2017;114(49):12934–9. 87. Lorenz C, Lorenz C, Lünse CE, Mörl M. tRNA modifications: impact on 64. Tanaka R, Satoh H, Moriyama M, Satoh K, Morishita Y, Yoshida S, et al. structure and thermal adaptation. Biomolecules. 2017;7(2):E35. Intronic U50 small-nucleolar-RNA (snoRNA) host gene of no protein-coding 88. Lopes RRS, Kessler AC, Polycarpo C, Alfonzo JD. Cutting, dicing, healing and potential is mapped at the chromosome breakpoint t(3;6)(q27;q15) of sealing: the molecular surgery of tRNA. Wiley Interdiscip Rev RNA. 2015;6(3): human B-cell lymphoma. Genes Cells. 2000;5(4):277–87. 337–49. 65. Dong X-Y, Rodriguez C, Guo P, Sun X, Talbot JT, Zhou W, et al. SnoRNA U50 is 89. Maraia RJ, Lamichhane TN. 3′ processing of eukaryotic precursor tRNAs. a candidate tumor-suppressor gene at 6q14.3 with a mutation associated with Wiley Interdiscip Rev RNA. 2011;2(3):362–75. clinically significant prostate cancer. Hum Mol Genet. 2008;17(7):1031–42. 90. Rajendran V, Kalita P, Shukla H, Kumar A, Tripathi T. Aminoacyl-tRNA 66. Dong X-Y, Guo P, Boyd J, Sun X, Li Q, Zhou W, et al. Implication of snoRNA synthetases: structure, function, and drug discovery. Int J Biol Macromol. U50 in human breast cancer. J Genet Genomics. 2009;36(8):447–54. 2018;111:400–14. 67. Mourtada-Maarabouni M, Pickard MR, Hedge VL, Farzaneh F, Williams GT. 91. Park SG, Schimmel P, Kim S. Aminoacyl tRNA synthetases and their GAS5, a non-protein-coding RNA, controls apoptosis and is downregulated connections to disease. Proc Natl Acad Sci U S A. 2008;105(32):11043–9. in breast cancer. Oncogene. 2009;28(2):195–208. 92. Goodarzi H, Nguyen HCB, Zhang S, Dill BD, Molina H, Tavazoie SF. 68. Rintala-Dempsey AC, Kothe U. Eukaryotic stand-alone pseudouridine Modulated expression of specific tRNAs drives gene expression and Cancer synthases – RNA modifying enzymes and emerging regulators of gene progression. Cell. 2016;165(6):1416–27. expression? RNA Biol. 2017;14(9):1185–96. 93. Helm M, Giegé R, Florentz C. A Watson-crick base-pair-disrupting methyl 69. De Zoysa MD, Yu Y-T. Posttranscriptional RNA Pseudouridylation. In: The group (m1A9) is sufficient for cloverleaf folding of human mitochondrial Enzymes: Elsevier; 2017. p. 151–67. [cited 2019 Nov 11]. tRNALys. Biochemistry. 1999;38(40):13338–46. Janin et al. Molecular Cancer (2020) 19:70 Page 13 of 13 94. Schaefer M, Pollex T, Hanna K, Tuorto F, Meusburger M, Helm M, et al. RNA 117. Chu J, Pelletier J. Therapeutic opportunities in eukaryotic translation. Cold methylation by Dnmt2 protects transfer RNAs against stress-induced Spring Harb Perspect Biol. 2018;10(6):a032995. cleavage. Genes Dev. 2010;24(15):1590–5. 118. Begley U, Sosa MS, Avivar-Valderas A, Patil A, Endres L, Estrada Y, et al. A 95. Tuorto F, Herbst F, Alerasool N, Bender S, Popp O, Federico G, et al. The human tRNA methyltransferase 9-like protein prevents tumour growth by tRNA methyltransferase Dnmt2 is required for accurate polypeptide regulating LIN9 and HIF1-α. EMBO Mol Med. 2013;5(3):366–83. synthesis during haematopoiesis. EMBO J. 2015;34(18):2350–62. 96. Blanco S, Bandiera R, Popis M, Hussain S, Lombard P, Aleksic J, et al. Stem cell function and stress response are controlled by protein synthesis. Nature. Publisher’s Note 2016;534(7607):335–40. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 97. Endres L, Begley U, Clark R, Gu C, Dziergowska A, Małkiewicz A, et al. Alkbh8 regulates Selenocysteine-protein expression to protect against reactive oxygen species damage. PLoS One. 2015;10(7):e0131335. 98. Ohshio I, Kawakami R, Tsukada Y, Nakajima K, Kitae K, Shimanoe T, et al. ALKBH8 promotes bladder cancer growth and progression through regulating the expression of survivin. Biochem Biophys Res Commun. 2016; 477(3):413–8. 99. Schlieker CD, Van der Veen AG, Damon JR, Spooner E, Ploegh HL. A functional proteomics approach links the ubiquitin-related modifier Urm1 to a tRNA modification pathway. Proc Natl Acad Sci U S A. 2008;105(47): 18255–60. 100. Golovko A, Hjälm G, Sitbon F, Nicander B. Cloning of a human tRNA isopentenyl transferase. Gene. 2000;258(1–2):85–93. 101. Lamichhane TN, Blewett NH, Maraia RJ. Plasticity and diversity of tRNA anticodon determinants of substrate recognition by eukaryotic A37 isopentenyltransferases. RNA N Y N. 2011;17(10):1846–57. 102. Schweizer U, Bohleber S, Fradejas-Villar N. The modified base isopentenyladenosine and its derivatives in tRNA. RNA Biol. 2017;14(9):1197– 208. 103. Lamichhane TN, Mattijssen S, Maraia RJ. Human cells have a limited set of tRNA anticodon loop substrates of the tRNA isopentenyltransferase TRIT1 tumor suppressor. Mol Cell Biol. 2013;33(24):4900–8. 104. Spinola M, Galvan A, Pignatiello C, Conti B, Pastorino U, Nicander B, et al. Identification and functional characterization of the candidate tumor suppressor gene TRIT1 in human lung cancer. Oncogene. 2005;24(35):5502– 9. 105. Tuorto F, Legrand C, Cirzi C, Federico G, Liebers R, Müller M, et al. Queuosine-modified tRNAs confer nutritional control of protein translation. EMBO J. 2018;37(18):e99777. 106. Silvera D, Formenti SC, Schneider RJ. Translational control in cancer. Nat Rev Cancer. 2010;10(4):254–66. 107. Robichaud N, Sonenberg N, Ruggero D, Schneider RJ. Translational control in Cancer. Cold Spring Harb Perspect Biol. 2019;11(7):a032896. 108. Okamoto M, Fujiwara M, Hori M, Okada K, Yazama F, Konishi H, et al. tRNA modifying enzymes, NSUN2 and METTL1, determine sensitivity to 5- fluorouracil in HeLa cells. PLoS Genet. 2014;10(9):e1004639. 109. Lewinska A, Adamczyk-Grochala J, Kwasniewicz E, Deregowska A, Semik E, Zabek T, et al. Reduced levels of methyltransferase DNMT2 sensitize human fibroblasts to oxidative stress and DNA damage that is accompanied by changes in proliferation-related miRNA expression. Redox Biol. 2018;14:20– 34. 110. Pelicano H, Carney D, Huang P. ROS stress in cancer cells and therapeutic implications. Drug Resist Updat Rev Comment Antimicrob Anticancer Chemother. 2004;7(2):97–110. 111. El Hassouni B, Sarkisjan D, Vos JC, Giovannetti E, Peters GJ. Targeting the ribosome biogenesis key molecule fibrillarin to avoid chemoresistance. Curr Med Chem. 2018;26 [cited 2019 Nov 13]. 112. Yuan S, Wu Y, Wang Y, Chen J, Chu L. An Oncolytic adenovirus expressing SNORD44 and GAS5 exhibits antitumor effect in colorectal Cancer cells. Hum Gene Ther. 2017;28(8):690–700. 113. Xu B, Ye M-H, Lv S-G, Wang Q-X, Wu M-J, Xiao B, et al. SNORD47, a box C/D snoRNA, suppresses tumorigenesis in glioblastoma. Oncotarget. 2017;8(27): 43953–66. 114. Mannoor K, Shen J, Liao J, Liu Z, Jiang F. Small nucleolar RNA signatures of lung tumor-initiating cells. Mol Cancer. 2014;13(1):104. 115. Bellodi C, McMahon M, Contreras A, Juliano D, Kopmar N, Nakamura T, et al. H/ACA small RNA dysfunctions in disease reveal key roles for noncoding RNA modifications in hematopoietic stem cell differentiation. Cell Rep. 2013; 3(5):1493–502. 116. Nachmani D, Bothmer AH, Grisendi S, Mele A, Bothmer D, Lee JD, et al. Germline NPM1 mutations lead to altered rRNA 2′-O-methylation and cause dyskeratosis congenita. Nat Genet. 2019;51(10):1518–29.