SUMOylation controls Hu antigen R posttranscriptional activity in liver cancer

SUMOylation controls Hu antigen R posttranscriptional activity in liver cancer Sofia Lachiondo-Ortega1,30, Claudia M. Rejano-Gordillo1,2,3,30, Jorge Simon1,4, Fernando Lopitz-Otsoa1, Teresa C. Delgado1, Krystyna Mazan-Mamczarz5, Naroa GoikoetxeaUsandizaga1, L. Estefanía Zapata-Pavas1, Ana Garcıía-del Río6, Pietro Guerra7, Patricia Peña-Sanfélix1, Natalia Hermán-Sánchez8, Ruba Al-Abdulla9,10, Carmen FernandezRodríguez1, Mikel Azkargorta4,11, Alejandro Velázquez-Cruz12, Joris Guyon13,14, César Martín3, Juan Diego Zalamea15, Leire Egia-Mendikute6, Arantza Sanz-Parra1, Marina Serrano-Maciá1, Irene González-Recio1, Monika Gonzalez-Lopez16, Luis Alfonso MartínezCruz1, Patrizia Pontisso7, Ana M. Aransay16, Rosa Barrio17, James D. Sutherland17, Nicola G.A. Abrescia15,18, Félix Elortza4,11, Amaia Lujambio19,20,21,22, Jesus M. Banales4,18,23,24, Raúl M. Luque8, Manuel D. Gahete8, Asıś Palazón6,18, Matias A. Avila4,25,26, Jose J. G. Marin4,27, Supriyo De5, Thomas Daubon28, Antonio Díaz-Quintana12, Irene Díaz-Moreno12, Myriam Gorospe5, Manuel S. Rodríguez29, María Luz Martínez-Chantar1,4,31,* 1Liver Disease Lab, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160 Derio, Bizkaia, Spain 2Department of Biochemistry and Molecular Biology, Faculty of Sciences, University of Extremadura, University Institute of Biosanitary Research of Extremadura (INUBE), 06071 Badajoz, Spain 3Biofisika Institute, Consejo Superior de Investigaciones Científicas (CSIC), Departamento Bioquímica y Biología Molecular, Facultad de Ciencia y Tecnología, Universidad del País Vasco (UPV/EHU), Leioa, Spain 4Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), Carlos III National Health Institute, Madrid, Spain This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). *Correspondence: [email protected]. AUTHOR CONTRIBUTIONS Conceptualization, S.L.-O., C.M.R.-G., J.S., F.L.-O., T.C.D., A.V.-C., A.D.-Q., I.D.-M., M.G., M.S.R., and M.L.M.-C.; methodology, S.L.-O. and C.M.R.-G.; software, K.M.-M., M.A., R.A.-A., J.D.Z., S.D., and A.D.Q.; formal analysis, S.L.-O., C.M.R.-G., J.S., F.L.- O., T.C.D., K.M.-M., L.E.Z.-P., R.A.-A., M.A., A.V.-C., J.G., C.M., J.D.Z., L.E.-M., S.D., T.D., and A.D.Q.; investigation, S.L.-O., C.M.R.-G., J.S., F.L.-O., T.C.D., N.G.-U., L.E.Z.-P., A.G.-d.R., P.G., P.P.-S., N.H.-S., C.F.-R., M.A., A.V.-C., J.G., C.M., J.D.Z., L.E.-M., A.S.-P., M.S.-M., I.G.-R., and M.G.-L.; resources, K.M.-M., A.G.-d.R., M.A., A.V.-C., J.G., C.M., J.D.Z., L.E.-M., M.G.-L., L.A.M.-C., P.P., A.M.A., R.B., J.D.S., N.G.A.A., F.E., A.L., J.M.B., R.M.L., M.D.G., A.P., S.D., T.D., A.D.-Q., I.D.-M., M.G., and M.S.R.; data curation, K.M.-M., M.A., A.V.-C., J.D.Z., S.D., and A.D.-Q.; writing – original draft, S.L.-O., A.V.-C., A.D.Q., I.D.-M., M.G., and M.L.M.-C.; writing – review & editing, S.L.-O., M.A., A.V.-C., P.P., J.M.B., R.M.L., M.D.G., A.P., M.A.A., J.J.G.M., T.D., A.D.-Q., I.D.-M., M.G., M.S.R., and M.L.M.-C.; visualization, S.L.-O., C.M.R.-G., J.S., F.L.-O., K.M.-M., L.E.Z.-P., P.G., P.P.-S., A.V.-C., J.G., C.M., J.D.Z., L.E.-M., A.S.-P., T.D., and A.D.-Q.; supervision, C.M., A.M.A., J.D.S., N.G.-A., F.E., M.D.G., A.P., S.D., T.D., A.D.-Q., I.D.-M., M.G., M.S.R., and M.L.M.-C.; project administration, M.L.M.-C.; funding acquisition, K.M.-M., T.C.D., C.M., L.A.M.-C., A.M.A., R.B., N.G.A.A., F.E., A.L., J.M.B., R.M.L., M.D.G., A.P., M.A.A., J.J.G.M., S.D., T.D., I.D.-M., M.G., M.S.R., and M.L.M.-C. DECLARATION OF INTERESTS The authors declare no competing interests. SUPPLEMENTAL INFORMATION Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2024.113924. HHS Public Access Author manuscript Cell Rep . Author manuscript; available in PMC 2024 April 18. Published in final edited form as: Cell Rep . 2024 March 26; 43(3): 113924. doi:10.1016/j.celrep.2024.113924. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 5Laboratory of Genetics and Genomics, National Institute on Aging (NIA), Intramural Research Program (IRP), National Institutes of Health (NIH), Baltimore, MD, USA 6Cancer Immunology and Immunotherapy Lab, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160 Derio, Bizkaia, Spain 7Unit of Internal Medicine and Hepatology (UIMH), Department of Medicine (DIMED), University of Padova, 35128 Padua, Italy 8Maimónides Institute of Biomedical Research of Córdoba (IMIBIC), Department of Cell Biology, Physiology and Immunology of University of Córdoba, Reina Sofia University Hospital, CIBER Pathophysiology of Obesity and Nutrition (CIBERobn), 14004 Córdoba, Spain 9Instituto de Investigación, Desarrollo e Innovación en Biotecnologıá Sanitaria de Elche (IDiBE), Universidad Miguel Hernández, Elche, Spain 10Institute of Medical Biochemistry and Molecular Biology, University Medicine of Greifswald, 17475 Greifswald, Germany 11Proteomics Platform, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), Carlos III Networked Proteomics Platform (ProteoRed-ISCIII), 48160 Derio, Bizkaia, Spain 12Instituto de Investigaciones Químicas (IIQ), Centro de Investigaciones Científicas Isla de la Cartuja (cicCartuja), Universidad de Sevilla, Consejo Superior de Investigaciones Científicas (CSIC), Sevilla, Spain 13University of Bordeaux, INSERM, BPH, U1219, 33000 Bordeaux, France 14CHU de Bordeaux, Service de Pharmacologie Médicale, 33000 Bordeaux, France 15Structure and Cell Biology of Viruses Lab, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160 Derio, Bizkaia, Spain 16Genome Analysis Platform, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160 Derio, Bizkaia, Spain 17Ubiquitin-likes and Development Lab, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160 Derio, Bizkaia, Spain 18Ikerbasque, Basque Foundation for Science, Bilbao, Spain 19Department of Oncological Sciences, Icahn School of Medicine at Mount Sinai, New York, NY, USA 20Liver Cancer Program, Division of Liver Diseases, Department of Medicine, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA 21The Precision Immunology Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA 22Graduate School of Biomedical Sciences at Icahn School of Medicine at Mount Sinai, New York, NY, USA Lachiondo-Ortega et al. Page 2 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 23Department of Liver and Gastrointestinal Diseases, Biodonostia Health Research Institute, Donostia University Hospital, San Sebastian, Spain 24Department of Biochemistry and Genetics, School of Sciences, University of Navarra, Pamplona, Spain 25Hepatology Program, Centro de Investigación Médica Aplicada (CIMA), University of Navarra, Pamplona, Spain 26Instituto de Investigaciones Sanitarias de Navarra (IdiSNA), Pamplona, Spain 27Experimental Hepatology and Drug Targeting (HEVEPHARM), Instituto de Investigació n Biomé dica de Salamanca (IBSAL), University of Salamanca, Salamanca, Spain 28University of Bordeaux, CNRS, IBGC, UMR 5095, Bordeaux, France 29Laboratoire de Chimie de Coordination (LCC), UPR 8241, CNRS; IPBS-University of Toulouse III-Paul Sabatier, Toulouse, France 30These authors contributed equally 31Lead contact SUMMARY The posttranslational modification of proteins critically influences many biological processes and is a key mechanism that regulates the function of the RNA-binding protein Hu antigen R (HuR), a hub in liver cancer. Here, we show that HuR is SUMOylated in the tumor sections of patients with hepatocellular carcinoma in contrast to the surrounding tissue, as well as in human cell line and mouse models of the disease. SUMOylation of HuR promotes major cancer hallmarks, namely proliferation and invasion, whereas the absence of HuR SUMOylation results in a senescent phenotype with dysfunctional mitochondria and endoplasmic reticulum. Mechanistically, SUMOylation induces a structural rearrangement of the RNA recognition motifs that modulates HuR binding affinity to its target RNAs, further modifying the transcriptomic profile toward hepatic tumor progression. Overall, SUMOylation constitutes a mechanism of HuR regulation that could be potentially exploited as a therapeutic strategy for liver cancer. In brief Lachiondo-Ortega et al. report that the RNA-binding protein HuR is posttranslationally modified by SUMOylation in the tumor tissue of patients with hepatocellular carcinoma. This mechanism of HuR regulation could be potentially exploited as a therapeutic strategy against liver cancer, thus highlighting the relevance of posttranslational modifications as disease targets. Graphical abstract Lachiondo-Ortega et al. Page 3 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript INTRODUCTION The molecular mechanisms underlying the malignant transformation of a healthy liver into hepatocellular carcinoma (HCC), the most common type of primary liver cancer, are numerous and highly heterogeneous.1 Therefore, it is currently believed that more than one signaling route would need to be drugged in order to stop the progression of the disease. At the moment, there are approved frontand second-line systemic treatments for advanced HCC available that are improving patient survival, and the number of agents found to be effective in phase 3 trials continues to grow.2 However, the empirical development of new drugs for HCC has not yielded the beneficial outcomes seen in other malignancies. To move the field forward meaningfully, we need outside-the-box approaches that adopt novel combination strategies and pursue new targets in HCC as we await the results of ongoing phase 3 clinical trials. In this context, the posttranslational modification (PTM) of proteins, which controls the specificity, timing, duration, and amplitude of virtually all physiological processes in the cell, is gaining momentum as a robust and multidimensional therapeutic strategy in cancer.3 SUMOylation is a ubiquitin-like (Ubl) PTM, conserved across eukaryotes, that consists in the covalent addition of one or multiple SUMO (small ubiquitin-like modifier) subunits or polymers to lysine residues of target proteins in a hierarchically organized process, thus contributing to the structural and functional diversity of the proteome. The SUMOylation cascade is catalyzed by a dimeric SUMO-activating enzyme E1 (SAE1/UBA2), a unique E2 Lachiondo-Ortega et al. Page 4 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript ubiquitin-conjugating enzyme 9 (UBC9), and a handful of E3 ligases, including members of the protein inhibitor of activated signal transducer and activator of transcription (PIAS) family. This modification can be reversed by the action of deSUMOylating enzymes, among which the sentrin-specific protease (SENP) family of SUMO-specific isopeptidases stand out and are also required for the maturation of precursor SUMO proteins.4 In mammals, up to five SUMO family members exist, each encoded by a separate gene. SUMO1–3 are expressed in most tissues,5 whereas SUMO46 and 57 are less abundant and restricted to specific tissues. Notably, the amino acid sequences of mature SUMO2 and SUMO3 are nearly identical but only share ~50% sequence identity with SUMO1.8 Moreover, SUMO2 and 3 tend to form polymeric chains,9,10 representing 90% of SUMO polymers.11 Mixed SUMO chains have been also reported and usually include SUMO1 at the distal end.10 SUMO1 contains an inverted SUMO motif that enables chain formation but at lower efficiency.12 In addition to covalent SUMOylation, SUMO can bind proteins non-covalently via SUMO-interacting motifs (SIMs),13 which can be found in many SUMO substrates and SUMO E3 ligases. Importantly, ~90% of SUMO-binding proteins are also covalent SUMO substrates.14 SUMOylation has an important regulatory role for most nuclear processes, including transcription, RNA processing, DNA-damage response (DDR), nucleocytoplasmic transport, cell-cycle progression, proteostasis, and nuclear body assembly.15 SUMO can regulate the activity, function, fate, and subcellular localization of target proteins by changing substrate interactions with DNA, RNA, or other proteins. SUMOylation functions during development16,17 and controls different physiological processes in adult organisms.18 Hence, its absence or dysregulation has been associated with disease.4,19 There is growing evidence that proteins implicated in the SUMOylation cascade are abundant in multiple cancers.3,20 In particular, higher expression of the genes involved in SUMOylation had been earlier detected as a pattern shared among patients with an accelerated progression of HCC.21 To date, the upregulation of SUMO1,22 SUMO2,23 SAE1,24 UBA2,25 UBC9,26–29 PIAS1,30 PIAS2,31 PIAS4,32,33 SENP1,34–36 SENP5,37 SENP6,38,39 and SENP740 in HCC and their value either as diagnostic or prognostic biomarkers have been described. Conversely, SENP2 expression is decreased in HCC and might play a role as a tumor suppressor.41 So far, the human SUMO proteome comprises more than 6,000 proteins.11 The identified SUMOylated proteins are associated with almost all cellular processes, including the main cancer hallmark functions.3 The effect of SUMO modification or removal depends on the context of the individual substrate. For example, in the background of HCC, methionine adenosyltransferase α2 (MATα2) is stabilized by SUMOylation and positively controls B cell lymphoma 2 (Bcl-2) expression, enhancing cell survival.42 SUMOylation of phosphoenolpyruvate carboxykinase 1 (PCK1) marks the protein for degradation via ubiquitination and helps human hepatoma cells grow by maintaining a glycolytic metabolism.43 Along this line, SUMOylation facilitates pyruvate kinase isoform M2 (PKM2) translocation to the plasma membrane of HCC cells and subsequent excretion via ectosomes, which accelerates macrophage differentiation by activating glycolysis and differentiation-associated transcription factors, thereby resulting in the release of cytokines/ chemokines and promoting tumor progression.44 Liver kinase B1 (LKB1) SUMOylation blocks its nucleocytoplasmic shuttling and favors its oncogenic activity in hypoxic HCC Lachiondo-Ortega et al. Page 5 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript tumors.45 Interestingly, polycomb chromobox 4 (Cbx4) SUMO E3 ligase activity controls hypoxia-inducible factor 1α (HIF-1α) SUMOylation to promote angiogenesis in HCC by increasing HIF-1 transactivation and hypoxia-induced vascular endothelial growth factor (VEGF) expression.46 Hu antigen R (HuR), also known as HuA and embryonic lethal abnormal vision-like 1 (ELAVL1), is a ubiquitous member of the ELAV/Hu family of RNA-binding proteins (RBPs). By binding through its RNA recognition motifs (RRMs) to Uand AU-rich elements (AREs) typically present in the 3′ untranslated region (UTR) of transcripts, HuR owns the posttranscriptional control of a large number of genes, enabling the protein to play pivotal roles that are dictated by the molecular functions of its target mRNAs.47,48 In turn, HuR tumorigenic effect is proposed to result from the function that it exerts on its target transcripts, which contribute to the main cancer traits (i.e., cell proliferation and survival, angiogenesis, evasion of immune recognition, invasion, and metastasis).49–51 In the past few years, the relevance of HuR in liver cancer has been extensively reviewed, as multiple signaling pathways implicated in HCC involve the RBPs.52–55 For instance, HuR plays a crucial role in hepatocyte proliferation, dedifferentiation, and malignant transformation by promoting the stabilization and expression of cyclin A2 ( CCNA2 ), CCND1 , and MAT2A mRNAs, among others.56,57 Also, HuR decreases the translation of the death receptor FAS mRNA, shielding HCC cells from FAS-induced apoptosis and immune surveillance.58,59 Interestingly, PTMs account for the main mechanism of regulation for HuR function, allowing the protein to elicit quick changes in gene expression programs.60,61 Of note, it has been described that HuR protein abundance, subcellular localization, and RNA-binding affinity can be modulated by methylation,62 phosphorylation,63 proteolytic cleavage,64 ubiquitination,65 neddylation,66 PARylation,67 sulfhydration,68 and arginylation.69 Here, we describe that HuR is SUMOylated at a higher degree in the tumor sections collected from patients with HCC in contrast to the surrounding tissue as well as in human cell line and mouse models of the disease. SUMOylation of HuR promotes major cancer hallmarks, namely proliferation and invasion, whereas the absence of HuR SUMOylation triggers a senescent phenotype with damaged mitochondrial and endoplasmic reticulum (ER) structure and function. Regarding the mechanism of action, SUMOylation induces a structural rearrangement of the RRMs that modulates HuR binding affinity to its target RNAs, further modifying the transcriptomic profile toward hepatic tumor progression. On the one hand, understanding the effects of HuR SUMOylation in hepatocarcinogenesis will provide insights into the relatively unknown role of SUMOylation in cancer. On the other hand, this mechanism of HuR regulation may be potentially exploited as a combination therapeutic strategy for HCC, thus highlighting the importance of PTMs as disease targets. RESULTS HuR SUMOylation is increased in human HCC Considering the role of HuR in the hepatic malignant transformation, we aimed to evaluate ELAVL1 mRNA levels in HCC tumors. According to data retrieved from The Cancer Genome Atlas (TCGA) mRNA expression repository,70 ELAVL1 was found significantly upregulated in the tumor (T) tissue of patients with HCC when compared to non-tumor (NT) Lachiondo-Ortega et al. Page 6 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript tissue (Figure 1A). Moreover, high ELAVL1 mRNA expression applied to all liver cancer stages (Figure 1B) and was associated with poorer patient survival (Figure 1C). Based on ELAVL1 mRNA expression levels in patients with HCC, a gene set enrichment analysis (GSEA) was performed, and the molecular pathways showing more than a 2fold normalized enrichment score (NES) were represented (Figure 1D). It can be seen that ELAVL1 mRNA expression is mainly associated with cell cycle, TP53 activity and transcription regulation, and protein SUMOylation processes during HCC, while it seems to be inversely related to the oxidative metabolism, synthesis of bile acids and salts, metabolism of vitamins and cofactors, and compound detoxification. Focusing on SUMOylation, a significantly deregulated mRNA expression of the major components of the SUMO cycle was observed in the tumor sections of patients with HCC (Figure 1E). Along this line, Pearson correlation analyses based on an mRNA array obtained from a cohort of 86 patients with HCC revealed a stronger co-variation between ELAVL1 and the different members of the SUMOylation pathway in the T in contrast to the paired surrounding tissue (ST) (Figures 1F and S1). Given the evidence suggesting that HuR could be posttranslationally modified by SUMOylation during HCC, a protein pull-down technology based on glutathione Stransferase (GST)-tagged SUMO-binding entities (SUBEs) was used to capture the SUMO-modified proteome both from tissue and cultured cell extracts.45,71,72 These tools contain tandem-repeated SIMs from RING-finger 4 (RNF4) SUMO-targeted ubiquitin ligase (STUbL) that specifically recognize polySUMOylated substrates and enable their purification by acting as molecular traps,71 thus showing an improved capacity over the originally developed ones.73 On the one hand, SUBE-mediated protein pull-down from liver tissue samples of a cohort of patients with HCC (n = 5) in combination with western blotting analysis revealed a significant enrichment of HuR in the T in contrast with the associated ST (Figures 1G and S1A). These results were further validated in the liver tumor tissue of the MYC ; Trp53 − / − genetically engineered mosaic mouse model of HCC74 when compared with healthy control tissue (Figure 1H). On the other hand, the ratio of SUMO-interacting proteins resulting from the GST-SUBEs relative to the GST-mediated protein pull-down in combination with liquid chromatographytandem mass spectrometry (LC-MS/MS) in human liver cell lines disclosed a greater number of SUMO-modified proteins in the PLC/PRF/5 and HuH-7 hepatoma cell lines than in the NT THLE-2 cell line (Figure 1I). Approximately 6% of the trapped proteins were shared among the three cell lines, and 12% were exclusively identified in the hepatoma cell lines but not in THLE-2. Furthermore, HuR was found particularly enriched after the GST-SUBEs in contrast to the GST-mediated protein pull-down in the human hepatoma cell lines, especially in the HuH-7 cell line (Figure 1J). Interestingly, mRNA and protein expression data suggest that, in addition to HuR, the SUMOylation machinery might be more induced in the HuH-7 cell line than in PLC/PRF/5 (Figures S1B and S1C), supporting the reason why higher HuR SUMOylation levels are found in the former cell line. Lachiondo-Ortega et al. Page 7 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Altogether, we corroborate that ELAVL1 expression and SUMO dynamics are decompensated in human liver cancer tumors, which led us to describe that HuR is SUMOylated in HCC mouse and human cell line models and, more importantly, in patients. Deciphering the SUMOylation machinery and SUMOylatable lysines in HuR A molecular characterization of HuR SUMOylation was performed in the MLP-29 cell line by means of transient plasmid transfections to express different His6-tagged SUMO constructs, among others, followed by downstream nickel-histidine affinity purification and western blotting analysis of V5-tagged HuR. First, SUMOylated HuR levels were found particularly increased upon SUMO2/3 overexpression, which suggested that this could be the main HuR SUMO modifier (Figures 2A and S2A). Next, canonical SUMO E3 ligases and deSUMOylating enzyme plasmids were transfected together with wild-type (WT) HuR, UBC9, and SUMO2/3. As shown, PIAS 2β overexpression resulted in an increased V5-tagged HuR protein enrichment (Figures 2B and S2B), while SENP1, 2, and 3 overexpression caused a decrease in V5-HuR detection (Figures 2C and S2C) and were therefore designated as the main enzymes implicated in HuR SUMOylation and deSUMOylation processes, respectively. Likewise, small interfering RNA (siRNA)-mediated downregulation of Pias2b (Figure S3A) almost abolished the smear of V5-tagged HuR protein expression corresponding to its modified state (Figure S3B), whereas knockdown of Senp1 , 2 , and 3 (Figure S3A) resulted in an increased pattern of modified V5-HuR protein expression (Figure S3C) in the MLP-29 cell line transiently expressing WT HuR, UBC9, and SUMO2/3, reinforcing the findings obtained in the overexpression studies. Along this line, HuR SUMOylation status assessed through nickel-histidine affinity purification also led to a significant enrichment of V5-tagged HuR protein levels in the MLP-29 cell line transiently co-transfected with plasmids inducing the expression of WT HuR, UBC9, and SUMO2/3, in addition to Senp1 or Senp2 siRNAs (Figure S3D), further validating the enzymatic machinery regulating HuR deSUMOylation. Regarding the characterization of HuR SUMOylation in the HuH-7 human hepatoma cell line, V5-HuR was found to be particularly enriched after His6-tagged SUMO2 transient expression and downstream nickel-histidine affinity purification (Figure S4A). Since SUMO2 and SUMO3 share ~95% sequence identity, they are usually referred to as the SUMO2/3 subfamily.8 Hence, SUMO2/3 is likely to be the main SUMO form attaching to HuR not only in the MLP-29 but also in the HuH-7 cell line. Regarding the profiling of canonical SUMO E3 ligases and deconjugating enzymes in the HuH-7 cell line, HuR protein expression and its smear, which is intended to reflect its modified state, were evaluated upon transient transfection of PIAS and SENP plasmids. Thus, HuR smear was increased when PIAS 2β expression was induced (Figure S4B), as it occurred in the MLP-29 cell line (Figure 2B). Conversely, HuR smear might decrease after induction of SENP6 expression but also in the presence of SENP1, 2, and 3 (Figure S4C) and could resemble the results obtained in the case of the MLP-29 cell line (Figure 2C), despite the evident differences between the two cell lines. According to several SUMOylation site prediction tools,75–77 HuR protein sequence does not contain any forward ψ-K-X-E/D or inverted E/D-X-K (ψ is a large hydrophobic residue Lachiondo-Ortega et al. Page 8 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript and X is any amino acid) SUMO consensus motif.12,78 However, a considerable proportion of experimentally validated SUMOylation sites do not match any of these motifs.79,80 Consequently, in order to define those, HuR SUMOylation mutant plasmids were designed by modifying all lysine residues located in the RRM1 and 2 primary sequences into arginine by means of site-directed mutagenesis and were co-transfected with the UBC9 SUMO E2-conjugating enzyme and SUMO2/3 in the MLP-29 cell line. The lack of V5-HuR protein enrichment observed when mutating the lysine residues at positions 120 and 182 revealed that HuR SUMOylation could be occurring at one or both of those sites (Figures 2D and S4D), leading to the creation of the HuR K120/182R SUMOylation double-mutant construct. HuR SUMOylation sites were further confirmed in the human hepatoma HuH-7 cell line (Figures 2E and S4E). Also, considering that SUBEs preferentially retain substrates comprising polySUMO2/3 chains in non-denaturing conditions and without the need of additional transfections,71 we were able to corroborate the loss of SUMOylation of the K120/182R HuR mutant relative to WT HuR in the HuH-7 cell line by using this protein pull-down technology (Figure 2F). Additionally, HuR SUMOylation was studied in the context of hypoxia, a hallmark of many solid tumors showing an aggressive phenotype,81 in addition to a well-known inductor of protein SUMOylation.82 In this case, the HuH-7 human hepatoma cell line was transiently transfected with the constructs expressing V5tagged WT or K120/182R HuR and incubated under hypoxic conditions (1% O2) for 24 h. Immunoprecipitation of the different HuR variants using anti-V5 antibody followed by western blotting analysis revealed that hypoxia is an enhancer of HuR SUMOylation, while the lower SUMO2/3 protein enrichment levels of the mutant variant corroborate that Lys120 and 182 constitute HuR SUMOylation residues in the HuH-7 cell line (Figure S5A). Importantly, to verify that Lys120 and 182 are not ubiquitination sites, the HuH-7 cell line was transiently transfected with plasmid vectors inducing the expression of His6-tagged ubiquitin in addition to V5-tagged WT HuR and the different SUMOylation mutant variants. Protein extracts were submitted to nickel-histidine affinity purification and western blotting analysis of V5-tagged HuR, which revealed that the K120R, K182R, and K120/182R HuR mutants exhibited ubiquitination, discarding these positions as major ubiquitination sites while reinforcing their relevance as SUMOylation residues (Figure S5B). Along this line, a dose of 100 nM ML-792 SAE inhibitor for 4 h did not only lead to a significant decrease in global protein SUMOylation (Figure S5C) but also resulted in a reduction of HuR SUMOylation (Figure S5D), assessed by means of transient transfection of plasmids inducing the expression of HuR, UBC9, and SUMO2/3 in the HuH-7 cell line followed by downstream nickel-histidine affinity purification and western blotting analysis. Hence, these findings provide proof of concept that HuR is SUMOylated in this hepatoma cell line. Finally, to gain further insight into the interplay between SUMOylation and ubiquitination, protein extracts from the HuH-7 cell line transiently transfected with His6-tagged ubiquitin and WT and K120/182R HuR in addition to treatment with 100 nM ML-792 for 4 h were subjected to nickel-histidine affinity purification. Western blotting analysis of V5-tagged HuR revealed a significant reduction in WT HuR ubiquitination in the presence of the SAE inhibitor, whereas the ubiquitination levels of the K120/192R HuR variant remained mostly unaltered (Figure S5E). These results suggest that HuR ubiquitination might occur as a SUMOylation-dependent process in the HuH-7 human hepatoma cell line. Lachiondo-Ortega et al. Page 9 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript forming networks, produce more ROS, and show a lower membrane potential. Furthermore, ER function seems to be compromised in the HuH-7 cells lacking HuR SUMOylation. SUMOylation modulates HuR RNA-binding affinity to confer a tumoral phenotype in human hepatoma cells A final approach to unravel the molecular mechanisms by which HuR SUMOylation controls HCC development and progression was to identify all the differential RNA targets with which HuR interacts in the SUMOylated and non-SUMOylated state. To this end, V5-tagged HuR immunoprecipitation followed by the isolation and sequencing of the bound RNAs (RIP-seq) was performed in the HuH-7 cell line stably expressing WT and the K120/182R mutant HuR. RNA-seq of the input fraction was considered for normalization, as it constitutes a good estimation of total RNA content. First, a quality control analysis of the samples was performed. A Pearson correlation on raw counts as well as a principal component analysis (PCA) on transformed counts indicated that replicate samples within groups were homogeneous and clearly differentiated from each set of conditions. Next, a comparison of comparisons analysis was performed on RNA-seq data, which consists of the difference between the RNAs bound to K120/182R and WT HuR relative to the difference between the input RNA in K120/182R and WT HuR expressing HuH-7 cell lines. A volcano plot shows the changes in RNAs binding to the K120/182R HuR SUMOylation mutant when compared to WT HuR in the HuH-7 cell line (Figure 6A). The transcripts showing more than a 1.5-fold change enrichment and adjusted p value (padj) < 0.05 were considered to have a significantly changed interaction with the SUMOylation mutant in contrast to WT HuR. Thus, the use of this criterion rendered a list of 346 RNAs showing a reduced or enhanced binding to non-SUMOylated HuR in comparison to the WT variant, which are represented as blue and red dots, respectively in Figure 6A. A first observation was that the K120/182R HuR SUMOylation mutant is capable of strongly interacting with a greater number of RNAs than the WT version (287 and 59 molecules, respectively). The RNAs showing a significantly differential enrichment after the comparison of comparisons analysis were plotted in a heatmap, which contains the abundance of each transcript expressed as normalized counts both in the V5-bound and input fractions of the HuH-7 cells stably expressing WT and K120/182R mutant HuR. This set of RNAs was further subjected to pathway analysis by using Ingenuity Pathway Analysis (IPA) software, and the top most significantly represented pathways were plotted (Figure 6B). Mostly, the retrieved pathways can be classified into three main groups, i.e., cell-cycle control and DDR, cholesterol biosynthesis via mevalonate, and Ubl-PTMs (Figure 6C). In addition, groups of transcripts related to mitochondrial membrane potential and permeability, ER stress response, and Ca2+ signaling were identified. In essence, SUMOylation compromises HuR intrinsic RNA-binding ability resulting in changes in the transcriptomic profile. Non-SUMOylated HuR shows an enhanced interaction with RNAs related to cell-cycle control and DDR, mitochondrial and ER functionality, and Ubl-PTMs, which would explain the observed senescent phenotype and regulation of the Lachiondo-Ortega et al. Page 16 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript SUMOylation process in the HuH-7 cell lines, as SUMOylated HuR would continue driving HCC progression. Xenograft tumors from human hepatoma cells lacking HuR SUMOylation sites show delayed growth and expression of senescence protein markers in mice The significance of blocking HuR SUMOylation was eventually verified in vivo by generating xenograft tumors via the subcutaneous injection of the HuH-7 human hepatoma cell line stably expressing the WT and K120/182R HuR variants in each flank of NOD scid gamma (NSG) mice (n = 6). Tumors appeared 1 week following implantation and were allowed to grow for 4 weeks, significantly increasing mouse body weight (Figure 7A). More importantly, however, monitoring tumor volume over time confirmed that tumors derived from HuH-7 cells expressing the K120/182R HuR SUMOylation mutant showed a slower growth rate compared to those derived from cells harboring WT HuR (Figure 7B). Moreover, xenograft tumor size and weight were significantly decreased in the absence of HuR SUMOylation 4 weeks after HuH-7 cell implantation (Figures 7C and 7D). Regarding the microscopic characterization, hematoxylin and eosin (H&E) staining revealed a reduced vascularization in the HuH-7 xenograft tumors expressing K120/182R HuR, possibly leading to the appearance of necrotic regions (Figure 7E). Interestingly, inhibition of HuR SUMOylation in the HuH-7 cell line resulted in tumors preserving the senescent phenotype, as per decreased CCND1 and increased p-H2AXSer139 protein expression levels (Figure 7F). In short, the absence of HuR SUMOylation sites has a proven significant inhibitory effect on HuH-7 xenograft tumor growth in mice in addition to the expression of senescence protein indicators. DISCUSSION It is well established that the expression of the RBP HuR is upregulated in many tumor types and is considered a hub in cancer because of the function that it exerts on its target RNAs, which contribute to the main hallmarks of cancer.50 Several works had already reported increased HuR protein expression in the context of liver cancer.57,58 Accordingly, in this study we detected high ELAVL1 mRNA expression levels in a cohort of patients with HCC, independently of the tumor stage, which were associated with a lower individual survival. Interestingly, along with these data, we observed that ELAVL1 mRNA expression was positively related to the Ubl-PTM SUMOylation in HCC tissue. As introduced, PTMs critically influence HuR function,60,62–69 in addition to being currently considered attractive therapeutic targets in cancer.3 Together with our earlier discovery that neddylation stabilized HuR leading to its increased abundance in HCC,66 here we demonstrate that HuR SUMOylation may also contribute to liver tumor progression by affecting its intrinsic RNAbinding affinity, further modifying the transcriptomic profile. HuR had been formerly identified as a SUMOylation target through large-scale quantitative proteomics performed in labeled HeLa cells stably expressing His6-SUMO2 in an attempt to explain the crosstalk between the SUMO cycle and the ubiquitin-proteasome Lachiondo-Ortega et al. Page 17 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript system (UPS).92 As protein-level SUMO proteomics evolved into site-specific approaches, a handful of SUMO-modified residues were disclosed for HuR in the HEK293, HeLa, and U2OS cell lines, especially under stressful conditions (e.g., heat shock and proteasome inhibition), by means of procedures based on exogenously expressed, epitope-tagged mutated SUMO variants, such as His6-SUMO2T90K93,94 or lysine-deficient His10-SUMO2Q87R,11 as well as an endogenous and native method relying on the commercially available SUMO2/3 8A2 antibody.95 However, none of these works managed to identify endogenous HuR SUMOylation in the tissue of different species or studied HuR SUMOylation individually. Thus, here not only have we described that HuR is SUMOylated in human hepatoma cell lines, the MYC ; Trp53 − / − genetically engineered mosaic mouse model of liver cancer and clinical HCC samples, but we have also established its pathophysiological relevance by elucidating the underlying mechanism relating HuR SUMOylation to tumor progression. We have comprehensively characterized HuR SUMOylation in the MLP-29 and HuH-7 cell lines as a process consisting in the covalent addition of one or multiple SUMO2/3 subunits into Lys120 and 182, both located in the RRM2, which forms a cleft with the RRM1 for RNA binding. We have additionally identified PIAS 2β and SENP1, 2, and 3 as the principal SUMO E3 ligase and deSUMOylating enzymes for HuR, respectively. Our data regarding HuR SUMOylation sites do not entirely match those experimentally proposed.11,93–95 Most HuR SUMO acceptor lysines revealed by site-specific MS-based proteomics were identified in response to stress, and different sites were preferably modified depending on the cell line and insult. This variability could also be attributed to technical differences between studies. Moreover, it is undeniable that SUMOylation co-exists with the rest of the PTMs that regulate HuR (i.e., methylation, phosphorylation, proteolytic cleavage, ubiquitination, neddylation, PARylation, sulfhydration, and arginylation). A study revealed that nearly one-quarter of the SUMOylation sites identified by MS/MS overlap with ubiquitination in the human proteome.79 Ubiquitination at Lys182 was previously reported to facilitate HuR degradation in response to heat shock in the HeLa cell line.65 Conversely, even though HuR exhibited ubiquitination in the HuH-7 human hepatoma cell line, neither Lys182 nor Lys120 was identified as a major ubiquitination site. Nevertheless, the fact that SUMO and ubiquitin can modify the same acceptor lysine does not necessarily entail competition96 or successive modifications.97 Ubiquitin, and especially SUMO, are only conjugated to a small subset of a given protein, making it possible for both modifiers to be present on the same lysine at the same time but in different subpopulations of the target proteins.98 Also, collaborative crosstalk between SUMO and ubiquitin has been described in the context of proteasomal degradation99,100 and DDR.101 The enzymes involved in this type of communication are the STUbLs and the counteracting proteases. Accordingly, we anticipated that HuR ubiquitination might depend on SUMOylation in the HuH-7 human hepatoma cell line under certain circumstances, even though a mechanistic explanation on this interplay is not provided. Apart from mixed SUMO-ubiquitin chain formation, which may additionally harbor phosphorylation and acetylation,94 SUMO can be conjugated to NEDD8, further increasing signaling complexity, but this is considered a rare event.11 In liver cancer, E3 ligase murine double minute 2 (Mdm2) mediates neddylation at Lys283, 313, and 326 of RRM3, leading to increased HuR nuclear localization and reduced Lachiondo-Ortega et al. Page 18 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript proteasomal degradation.66 Another site-specific mapping of the human SUMOylome revealed that crosstalk between SUMO and other PTMs (i.e., ubiquitination, methylation, and acetylation) may also occur by proximal modification of the same protein.11 Notably, the identified SUMO-methyl-co-modified proteins were enriched for RNA-binding properties. HuR methylation at Arg217 by co-activator-associated arginine methyltransferase 1 (CARM1) has been extensively reported to positively regulate the transcription of its target RNAs,62,102,103 and loss of HuR methylation has been observed in HCC causing increased MAT2A mRNA and protein expression and subsequent lower S -adenosylmethionine (SAMe) levels.57 Besides this, 9% of the identified human SUMOylome was reported to occur proximal to phosphorylation, and numerous SUMOylation sites were found to be fully dependent on prior phosphorylation events,11 a result in agreement with the earlier described phosphorylation-dependent SUMOylation motif (PDSM).104 Interestingly, it is well established that the checkpoint kinase 2 (Chk2)63,105 and p38 mitogen-activated protein kinase (MAPK)106,107 can modulate HuR RNA binding through the phosphorylation of Thr118, which is located close to SUMO acceptor Lys120. Overall, we highlight the potential for exploring the endogenous interplay between SUMOylation and the different PTMs that control HuR in a unified context. Apart from describing HuR SUMOylation at the molecular level, we were interested in further studying its effect on liver cancer progression. For this purpose, the HuH-7 HCC cell-line model stably expressing the WT and SUMOylation mutant HuR variants were subjected to a phenotypic characterization. On the one hand, HuH-7 cells bearing WT HuR showed higher proliferative and invasive potential than those expressing the SUMOylation mutant. Thus, the absence of HuR SUMOylation resulted in a senescent phenotype consisting of lower proliferation and invasion ratios as well as elevated β-galactosidase activity and expression of senescence protein markers, which were exacerbated after palbociclib-mediated CDK4/6 inhibition. Importantly, HuH-7 xenograft tumors expressing the K120/182R HuR SUMOylation mutant preserved the attenuated growth rate and senescent phenotype. On the other hand, mitochondrial structure and function were further examined, as they are known to be affected during cellular senescence.90 The interception of HuR SUMOylation in the HuH-7 cell line firstly revealed a rounder and more dispersed mitochondrial network. The resulting mitochondria lacked functionality and could not attain respiration as efficiently as cells expressing WT HuR. Also, ATP and NAD+/NADH levels were reduced, being consistent with a suppressed energy production in the absence of HuR SUMOylation. Furthermore, these defective mitochondria showed decreased membrane potential with high ROS generation, possibly derived from a perturbed electron transport chain (ETC). In addition to mitochondria, there was an increase in the ER mass, and its function was badly affected by the inhibition of HuR SUMOylation in human HCC cells, all conforming with the senescent phenotype.89,90 HuR being an RBP and considering that SUMOylation sites are located near or in the basic cleft involved in RNA recognition and binding, we investigated whether SUMOylation could modify HuR intrinsic RNA-binding affinity. Not surprisingly, MD simulations predicted a considerably less compact structural arrangement of the RRM1 and 2 and the concealment of the residues involved in RNA recognition upon SUMOylation. These conformational changes could modulate the binding affinity of HuR for its target mRNAs Lachiondo-Ortega et al. Page 19 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript and eventually regulate the transcriptomic profile, as was later corroborated by RIP-seq studies. Accordingly, the K120/182R HuR SUMOylation mutant showed an enhanced interaction with a larger number of transcripts than WT HuR. Importantly, the displayed in silico and in vitro data also confirmed that the observed differential RNA binding for each of the HuR SUMOylated species is not caused by the lysine-to-arginine mutations but is entirely driven by SUMOylation. The mRNAs displaying a differentially enriched interaction with the K120/182R HuR SUMOylation mutant were mainly involved in cellcycle control and DDR, cholesterol biosynthesis via mevalonate, and Ubl-PTMs, which matched with the phenotype observed along this study, as discussed later. Under physiological conditions HuR is located in the nucleus, where it participates in mRNA splicing and nuclear export. However, upon specific stimuli, HuR translocates to the cytoplasm and elicits its best-understood effects on mRNA stabilization and modulation of translation. Importantly, although cytoplasmic HuR normally promotes translation, it can also repress it.47,108 Therefore, HuR can affect the translation of the enriched mRNAs identified after the RIP-seq analysis in both directions. The loss of cell-cycle control and DNA damage are renowned hallmarks of senescence and have also been detected in the absence of HuR SUMOylation through the enrichment of a considerable subset of transcripts involved in cell-cycle regulation and DDR that significantly interact with non-SUMOylated HuR in the HuH-7 cell line, being consistent with the senescent phenotype manifested throughout this study. Moreover, the observed senescence-associated organelle damage was evidenced in the RIP-seq analysis by the enrichment of mRNAs involved in mitochondrial membrane potential and permeability, ER stress response, and Ca2+ signaling, which showed an increased interaction with the K120/182R HuR SUMOylation mutant. Interestingly, the mevalonate pathway for cholesterol biosynthesis was another significantly represented route in the RIP-seq analysis, which is known to play an important role in the progression of many types of cancer, including HCC, and may now be connected with HuR SUMOylation.109–112 Finally, changes in the binding of HuR upon SUMOylation to mRNAs related with UblPTMs are of particular relevance to this project, as they help reinforce the unequivocal relationship between HuR and SUMOylation. In this study, not only have we demonstrated that HuR is SUMOylated but we also leave open the possibility that HuR SUMOylation could be governing global SUMOylation processes through changes in RNA binding, eventually contributing to control the fate of liver tumors. On the one hand, we observed that HuR SUMOylation increased the levels of SUMO-conjugated proteins, possibly promoting protein SUMOylation and senescence escape. Complementarily, the oxidizing cellular environment observed in the absence of HuR SUMOylation could explain the reduced SUMO conjugation due to disulfide bridge formation between UBA2 and UBC9 catalytic cysteines.113 On the other hand, the differential enrichment analysis of the mRNAs bound to non-SUMOylated HuR revealed a significant number of transcripts involved in the SUMOylation pathway whose translation could be affected. Notably, the UBA2 transcript encoding one subunit of the SUMO-activating enzyme heterodimer was detected. Interestingly, UBA2 mRNA expression was significantly induced and strongly correlated with ELAVL1 in the tumor, in contrast to the paired ST of a cohort of patients with HCC. Lachiondo-Ortega et al. Page 20 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Overall, a model emerges from our results whereby HuR SUMOylation may be controlling the translation of mRNAs involved in the SUMO pathway eventually modulating the number of SUMO-conjugated proteins, including HuR itself, which could be playing a role in HCC progression. Taking our evidence together, our results unveil conceptual and functional avenues in HCC, with potential clinical implications, by demonstrating that HuR is a SUMOylation substrate in clinical HCC as well as in in vivo and cellular models of the disease. Specifically, HuR SUMOylation is likely to occur as the covalent addition of SUMO2/3 subunits into Lys120 and 182 catalyzed by PIAS 2β and reversed by the action of SENP1, 2, and 3 in the MLP-29 and HuH-7 cell lines. Importantly, we have established the pathological implication of HuR SUMOylation in liver cancer. Thus, SUMOylated HuR contributes to tumor cell proliferation and invasion, while its absence results in a senescent phenotype with damaged mitochondrial and ER structure and function. Regarding the mechanism of action, SUMOylation alters HuR intrinsic RNA-binding affinity, resulting in the modulation of the transcriptomic profile driving HCC progression. In conclusion, SUMOylation constitutes a mechanism of HuR regulation that could be potentially exploited as a therapeutic strategy for the clinical management of liver cancer, thus highlighting the value of PTMs as disease targets. Even though potent promising SAE inhibitors have been developed in the last decade,114–118 the abolition of global SUMOylation may not be entirely encouraged because of the cellular processes that this Ubl-PTM regulates independent of cancer development. Also, given the key function of SUMOylation in cell-cycle progression, exploring combination strategies with cell-cycle inhibitors has been recommended.20 Hence, a combination therapy for HCC based on HuR SUMOylation inhibition and palbociclib administration may emerge as a result of this study and could be particularly beneficial to patients, since single administration of CDK4/6 inhibitors is often suboptimal for the treatment of these malignancies.87 Furthermore, understanding the effects of HuR SUMOylation in hepatocarcinogenesis will provide insights into the relatively unknown role of SUMOylation in cancer. Limitations of the study We acknowledge that the tools to study PTMs are not yet fully developed, thereby limiting the interpretation of experimental data. For example, the use of nickel-His6-SUMO2/3 affinity purification to elucidate the canonical deSUMOylating enzyme for HuR rendered slightly vague results depending on whether this technique was preceded by upor downregulation of protein expression. The lack of HuR protein enrichment was more evident when overexpressing SENP1, 2 and 3, while the presence of SUMOylated HuR was less convincing when knocking down Senp1 and 2 expression in the MLP-29 cell line. Along this line, the analysis of the interplay between the different PTMs is also restricted. Here, we report that HuR is modified both by SUMOylation and ubiquitination in liver cancer cells even though they involve different sites. Moreover, we propose that HuR ubiquitination might occur as a SUMOylation-dependent process, but we do not further provide a mechanistic explanation for this crosstalk, encouraging others to do so. Nevertheless, we are hopeful that ongoing research will help to propel the field of PTMs. Lachiondo-Ortega et al. Page 21 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript STAR★METHODS Detailed methods are provided in the online version of this paper and include the following: RESOURCE AVAILABILITY Lead contact—Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, María Luz Martínez-Chantar ([email protected]). Materials availability—Plasmids and stably transfected cell lines generated in this study are available upon request to María Luz Martínez-Chantar ([email protected]). Data and code availability •RIP-Seq data have been deposited at Gene Expression Omnibus (GEO) and are publicly available as of the date of publication. The accession number is GSE197798 and is also listed in the key resources table. •This paper does not report original code. •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS Animals—Liver tissue samples of the MYC ; Trp53 − / − genetically engineered mosaic mouse model of HCC were kindly provided by Dr. Amaia Lujambio.74 The oncogenic xenograft murine procedure included in project P-CBG-CBBA-0722 was approved by the CIC bioGUNE Institutional Animal Care and Use Committee and the competent authority from Diputación de Bizkaia. Animal experimentation was conducted in accordance with the National Institutes of Health (NIH) guide for care and use of Laboratory animals and the guidelines of the European Research Council for animal care and use. Young 7-weeks old female NOD scid gamma (NSG) mice (614, Charles River Laboratories) were housed in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-ac-credited animal facility at CIC bioGUNE under controlled temperature (21 ± 1°C) and humidity (45 ± 10%) conditions, with 12-h light/dark cycles and ad libitum access to Teklad global 14% protein rodent maintenance diet (2014C, Envigo) and water. For the procedure, a total of 5 million HuH-7 cells stably expressing WT HuR and the K120/182R SUMOylation mutant were subcutaneously injected in each flank of mice (n = 6). Tumor size was monitored with a digital caliper every 3 days from day 10 after implantation and mice were euthanized before tumors exceeded 1,500 mm3. Tumor volume (V) was calculated by using the modified version of the elipsoidal formula: V = ½ ( Length 3 Width2 ). Tumor masses were collected for western blotting and histological analyses. Human participants—A cohort of 172 formalin-fixed paraffin-embedded (FFPE) samples consisting of paired HCC tumor and surrounding non-tumor tissue were obtained from the Córdoba Node of the Andalusian Public Health System Biobank, evaluated by liver histology and the diagnosis was confirmed by two independent, experienced Lachiondo-Ortega et al. Page 22 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript pathologists. Clinical data from patients was collected from electronic medical reports (Table S1). The study protocol was approved by the Reina Sofia University Hospital Ethics Committee, according to institutional and Good Clinical Practice guidelines (Protocol number PI17/02287) and in compliment with the declaration of Helsinki. Informed consent was obtained from all patients or their relatives. Also, a cohort of 5 patients including paired HCC tumor and surrounding tissue liver biopsies obtained during tumor resection was provided by the Basque Biobank upon informed consent and with evaluation and approval from the corresponding ethics committee. Cell lines—The THLE-2 (CRL-2706, ATCC), PLC/PRF/5 (CRL-8024, ATCC), HuH-7 (JCRB0403, JCRB Cell Bank) and MLP-29 (provided by Dr. Enzo Medico)119 cell lines were grown in the medium stated in the manufacturer’s instructions. All cell lines were maintained in culture for a maximum of 20 passages in a humidified incubator at 37°C and 5% CO2, unless otherwise stated. METHOD DETAILS Hematoxylin and eosin (H&E) staining—5 μm-thick paraffin-embedded sections of the formalin-fixed xenograft tumor samples were deparaffinized with Histo-Clear (HS-200, National Diagnostics) during 20 min and rehydrated through graded ethanol solutions (100–70%) to distilled water. Sections were incubated with Harris hematoxylin (05–06004, Bio-Optica) for 15 min, rinsed in running tap water, and differentiated with 0.5% HCl. Specimens were rinsed with distilled water and subsequently stained with aqueous Eosin Y solution (HT110232, Sigma-Aldrich) for 15 min. Finally, samples were dehydrated through graded ethanol solutions (70–100%), cleared with Histo-Clear and mounted with DPX mounting medium (06522, Sigma-Aldrich). Images were acquired with an ×10 magnification objective in a DM750 upright microscope (Leica) equipped with a ICC50W camera (Leica). Plasmid generation—The full length cDNA of WT mouse HuR was purchased from the German Resource Center for Genome Research (RZPD). The V5-HuR WT plasmid was constructed by PCR using a 5′ oligonucleotide containing the V5 tag sequence and being subcloned into a pcDNA3.3-TOPO vector (K830001, Invitrogen).66 SUMOylation sites on HuR were predicted experimentally by mutation of lysine residues into arginine. The SUMOylation mutant HuR plasmid constructs were created using the QuickChange sitedirected mutagenesis kit (200518, Stratagene), according to the manufacturer’s instructions, with two complementary oligonucleotides and with pcDNA3.3-TOPO-V5-HuR WT plasmid as template. Products were sequenced by STAB vida. Plasmid DNA was purified after bacterial transformation and amplification, using NucleoBond Xtra Midi Plus kit (740412, Macherey-Nagel), by following the manufacturer’s instructions. Plasmid DNA transfection—Cells were transfected with the different plasmids summarized in KRT,120–130 using Lipofectamine 2000 transfection reagent (11668019, Invitrogen) and Opti-MEM I reduced serum medium (31985070, Gibco), as stated in the manufacturer’s instructions. Cells were allowed to grow for additional 48 h until optimal protein expression. In order to create stable cell lines, pcDNA3.3-TOPO-V5-mHuR plasmid Lachiondo-Ortega et al. Page 23 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript constructs were digested with PvuI restriction enzyme (ER0621, Thermo Scientific) prior to transfection, by following the manufacturer’s indications. Linearization avoids unspecific cleavage and increases the chances that the vector integrates into the host cell genome without disrupting the gene of interest or other elements required for expression in mammalian cells. The transfected cells were selected and maintained in culture medium containing 1.5 mg/mL Geneticin (G418 sulfate) selective antibiotic (11811031, Gibco). Cell clones were obtained by a serial dilution process in 96-well plates and colonies were then subcultured into larger dishes. Transfection efficiency of each clone was confirmed by western blotting. siRNA transfection—Cells were transfected with Silencer Select negative control no.1 siRNA (4390843, Invitrogen) or the different Silencer Select siRNAs (4390771, Thermo Fisher Scientific) summarized in the KRT, using DharmaFECT transfection reagent (T-2001–03, Horizon Discovery) and Opti-MEM I reduced serum medium (31985070, Gibco), by following the manufacturer’s instructions. Cells were allowed to grow for additional 48 h until optimal mRNA expression knockdown, which was validated by qPCR. Hypoxia—Cells were incubated in a Bugbox M anaerobic workstation (Baker) at 1% O2 during 24 h. ML-792 treatment—ML-792 (HY-108702, MedChemExpress) was administered in vitro at 100 nM in cell culture medium during 4 h. Under no circumstances was DMSO final volume greater than 0.1%. Palbociclib treatment—Palbociclib isethionate salt (P-7766, LC Laboratories) was dissolved in DMSO and stored at −80°C. The drug was administered in vitro both as an acute and a chronic treatment. On the one hand, the acute treatment involved the administration of higher concentrations ranging from 0 to 1 μM palbociclib for 3 days. On the other hand, the chronic treatment involved the administration of lower concentrations ranging from 0 to 100 nM palbociclib for 2 weeks. During this period of time, medium and treatment were renewed every 3 days, as cells were subcultured. Under no circumstances was DMSO final volume greater than 0.1%. Bioinformatic analysis—The results published are based upon data generated by the The Cancer Genome Atlas (TCGA) Research Network (https://www.cancer.gov/tcga). Expression levels of the indicated genes in non-tumor and tumor tissue of HCC patients, as well as in the different tumor stages were expressed as RNA-Seq by Expectation Maximization (RSEM). Survival information of patients with HCC was obtained from the clinical information dataset. Patients were divided into two groups based on low and high expression levels of ELAVL1 when mRNA levels were below or above the median, respectively. Data was plotted as Kaplan-Meier curves and the Mantel-Cox test was performed for statistical comparison between the two groups. The Gene Set Enrichment Analysis (GSEA) software (https://www.gsea-msigdb.org/gsea/) was used to perform an enrichment pathway analysis based on ELAVL1 expression levels in HCC patients. By establishing the median as a cut-off value, the enriched pathways when ELAVL1 mRNA levels were low and high were designated with a normalized enrichment score (NES). In this Lachiondo-Ortega et al. Page 24 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript study, only the pathways showing a NES greater than a 2-fold change, p value <0.0001 and FDR <0.01 were considered. Total RNA isolation—Total RNA from cell lines and FFPE tissues was extracted with TRIzol reagent (15596026, Invitrogen) and the Maxwell 16 LEV RNA FFPE Purification Kit (AS1260, Promega), respectively, as per manufacturer’s instructions. RNA concentration was determined in the NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific). Reverse transcription (RT)—1–2 μg of RNA were treated with Amplification Grade DNase I (18068015, Invitrogen) by following the manufacturer’s guidelines. cDNA was synthesized with M-MLV reverse transcriptase (28025013, Invitrogen) in the presence of Random Primers (48190011, Invitrogen), dNTPs (10297018, Invitrogen), and RNaseOUT recombinant ribonuclease inhibitor (10777019, Invitrogen), by using a Veriti Dx thermal cycler (Applied Biosystems). RT conditions involved 10 min at 25°C, 3h at 37°C and 15 min at 70°C. The resulting cDNA was diluted 10-fold in nuclease-free water (W4502, Sigma-Aldrich). Quantitative PCR (qPCR)—Gene primer sequences were designed with Primer-BLAST tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) and synthesized by Sigma-Aldrich (Table S2). For conventional qPCR, 1.5 mL of cDNA were mixed with specific primers and SYBR Select master mix (4472908, Invitrogen) constituting a final volume of 6.5 mL, in MicroAmp Optical 384-Well Reaction Plates (4309849, Applied Biosystems). Each reaction was performed in triplicate using the ViiA 7 Real-Time PCR system (Applied Biosystems). qPCR conditions involved an initial denaturation step (90 s at 95°C), followed by 40 cycles of annealing (15 s at 95°C and 1 min at 59°C), and a final extension phase (15 s at 95°C, 1 min at 60°C and 15 s at 95°C). Ct values were extrapolated from the melt curve and gene expression levels were normalized with RPLP0 or Gapdh housekeeping expression by implementing the 2ΔΔCt formula. qPCR dynamic array based on microfluidic technology—A microfluidic-based qPCR dynamic array was used for the RNA expression analysis in samples derived from the Andalusian Public Health System Biobank cohort of patients.133–135 Specific primers for human transcripts were designed with Primer-BLAST tool (https://www.ncbi.nlm.nih.gov/ tools/primer-blast/) and synthesized by Sigma-Aldrich (Table S2). Preamplification, exonuclease treatment and qPCR dynamic array based on microfluidic technology were implemented using the Biomark System (Fluidigm) by following the manufacturer’s instructions.136 mRNA copy number of the transcripts analyzed were adjusted by normalization factor, calculated with the expression levels of ACTB , GAPDH and HPRT using geNorm 3.3 software.137 Pearson correlation analyses were computed on mRNA expression levels and reported the values of the correlation coefficient (r) and two-tailed p values. The coefficient of determination (R2) was calculated from the Pearson correlation coefficient, and the best-fit line was plotted for each correlation. Ribonucleoprotein immunoprecipitation (RIP)—In order to identify the RNAs interacting with SUMOylated and non-SUMOylated HuR, the WT and K120/182R variants were immunoprecipitated with anti-V5 coated Protein G Sepharose resin and the RNA Lachiondo-Ortega et al. Page 25 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript whereas the 5 conformations of the RNA molecule were used as input for Monte-Carlo conformation exchange during the simulation of their diffusion. Conformational exchange was allowed every 2.5 ns. A total of 75,000 diffusion trajectories (5 × 15,000) were then computed for each WT and mutant species. As the koff computations module did not allow conformation exchange, each combination of conformers was treated in a different set of independent trajectories. A total of 62,500 (25 × 2,500) trajectories were computed for each construct. Origin 2018b (OriginLab Corporation) was used for statistical analysis and data representation. Live-cell proliferation and migration imaging—Cells were seeded in tissue-culture treated 96-well plates. For the migration experiment, plates were scratched with a 96-pin WoundMaker (Essen Biosciences) when cells reached 95–100% confluence. Photomicrographs were taken every 2 h using an IncuCyte live-cell analysis system (Essen Biosciences) and confluence of the culture or wound recovery were measured using IncuCyte software (Essen Biosciences) after 132 or 60 h in culture, respectively. Cell invasion assay on a collagen I matrix—The invasive potential was assessed by evaluating cell spheroid growth on a collagen gel.157 Cell spheroid formation was achieved by seeding 5,000 cells/well in culture medium with 0.4% methyl cellulose (M0512, Sigma-Aldrich) in non-treated round-bottom 96-well plates (351177, Falcon). After a 3day incubation, spheroids were individually collected and gently washed with PBS. Each spheroid was embedded in 100 μL of a 1 mg/mL collagen I (354236, Corning) and 7.2 mM NaOH in PBS solution and carefully deposited on tissue-culture treated flat bottom 96-well plates. After incubation at 37°C for 30 min, cell culture medium was added up to a final volume of 200 μL. Pictures of the spheroids were taken after 48 h with an ×10 magnification objective of an Axio Observer Z1 inverted microscope (Zeiss). The invasion ratio was established as the total spheroid surface divided by the spheroid core, which were determined with Fiji software (https://imagej.net/software/fiji/). Annexin V staining—Cells were seeded in tissue-culture treated 6-well plates and allowed to grow until they reached 80% confluency. Apoptotic cells were identified by flow cytometry using the Annexin V FITC Apoptosis detection kit (ANXVKF, Immunostep) in combination with LIVE/DEAD fixable blue dead cell stain kit (L23105, Invitrogen), by following the manufacturer’s instructions. Culture medium was collected, cells were washed with PBS twice, detached by trypsinization and transferred to 15 mL tubes. Each cell pellet was washed with 5 mL of PBS and collected by centrifugation (600 × g , 5 min). Each pellet was resuspended in 100 μL of the LIVE/DEAD cell dye in PBS, and incubated 30 min at 4°C in the dark. The staining solution was diluted by adding 500 μL of 5% FBS-PBS per tube and cells were collected by centrifugation (600 × g , 5 min). Each cell pellet was resuspended in 100 μL of Annexin V-FITC diluted in Annexin V Binding Buffer solution and incubated 15 min at RT protected from the light. The staining solution was diluted by adding 200 μL of Annexin V Binding Buffer per tube. Annexin V-FITC fluorescence (λex = 495 nm, λem = 519 nm) was acquired in a FACSymphony flow cytometer (BD Biosciences). A set of unstained cells were used as a blank and a group of cells that had been previously treated with 1 μM staurosporin (STS) (S1421, Selleckchem) for 4 h was used as a positive Lachiondo-Ortega et al. Page 32 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript control. The results were analyzed with FlowJo v10 software (BD Biosciences) and the percentage of apoptotic cells was calculated. Crystal violet staining—Cells were seeded in tissue-culture treated 12-well plates. For proliferation studies, time points were collected on day 0 and 3. For colony formation assays, cells were collected after 2 weeks while culture medium was replaced every 3 days. Cell viability was estimated by crystal violet, which is a basic protein dye that binds to ribose-type molecules such as DNA. Cells were washed with PBS twice and fixed in ice-cold 4% paraformaldehyde solution in PBS (sc-281692, Santa Cruz Biotechnology) for 10 min at RT. Cells were washed with PBS twice and incubated with a 0.1% crystal violet (C6158, Sigma-Aldrich) solution in 20% methanol for 40 min at RT with gentle shaking. The staining was discarded and the plates were rinsed with distilled water and air-dried overnight. Crystals were resuspended in 10% acetic acid for 30–60 min at RT with gentle shaking. An approximate volume of 100 μL was transferred into 96-well clear flat bottom plates and absorbance at 595 nm was measured in a SpectraMax M2/M2e microplate reader (Molecular Devices). For higher consistency, replicates of the same well were performed when possible, and 10% acetic acid was used as blank. The absorbance data were used to calculate the percentage of proliferation relative to the initial timepoint and palbociclib IC50 values were calculated from the best-fit values of four-parameter dose-response curves with a 95% confidence interval. Caspase-3 activity assay—Cells were seeded in tissue-culture treated 6-well plates and allowed to grow until they reached 80% confluency. Apoptosis was determined by measuring the fluorescence resulting after caspase-3 mediated cleavage of a fluorogenic substrate. In order to recover possible dead cells in suspension, the medium was transferred to a set of tubes and the cell pellet was collected by centrifugation (2,000 rpm, 5 min). Cells were washed with PBS twice and any detached cell was collected by centrifugation (2,000 rpm, 5 min). Cells were lysed in 50 μL of caspase-3 reaction buffer (250 mM PIPES pH 7.4, 100 mM EDTA, 2.5% CHAPS, 125 mM DTT) and combined with the dead cell pellets. Total protein was extracted and protein concentration was determined by the Bradford assay. 40 μg of total protein were added to a mix containing 25 μM Ac-DEVD-AFC caspase-3 fluorogenic substrate (ALX-260–032, Enzo Life Sciences) in reaction buffer, constituting a final volume of 500 μL. Each sample was measured in duplicate by adding 200 μL of the reaction mixture to each well of a 96-well black flat bottom assay plate (3915, Corning). A blank without protein sample was included, and a cell lysate that had been previously treated with 1 μM STS for 4 h was used as a positive control. The reaction plate was incubated at 37°C with gentle shacking for 4 h and fluorescence (λex = 390 nm, λem = 510 nm) was measured every hour in a SpectraMax M2/M2e microplate reader (Molecular Devices). Caspase-3 activity was determined by calculating the increase in fluorescence from 0 to 4 h after background correction, and normalized with total protein. Senescence-associated β-galactosidase (SA β-gal) activity detection—Cells were seeded over 12-mm coverslips (631–1577P, VWR) previously coated with a 0.01% poly-L-lysine solution (P4707, Sigma-Aldrich), in tissue-culture treated 24-well plates and allowed to grow until they reached 50–60% confluency. SA β-gal was assayed Lachiondo-Ortega et al. Page 33 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript with the senescence detection kit (QIA117, Calbiochem), according to the manufacturer’s instructions. Culture medium was removed, cells were washed with PBS once, and incubated with 250 μL of the fixative solution per well, for 10–15 min at RT. The fixative solution was removed, rinsed with PBS twice and cells were incubated with 250 μL of the staining solution mix (12.5 μL of 10 mg/mL X-gal substrate in DMF, 2.5 μL of staining supplement and 235 μL of staining solution) per well, overnight at 37°C. PBS was added to the empty wells in the plate to avoid the evaporation of the staining solution mix. Cells were observed under the microscope for development of blue color, and the reaction was stopped by removal of the staining solution mix and rinsed with PBS three times. Coverslips were mounted in mounting medium (S3023, Agilent) and slides were observed with a Leica DM750 upright brightfield microscope equipped with a Leica ICC50W digital color camera. A minimum of five areas per coverslip were considered so that more than 200 cells per coverslip were manually counted using an ×10 magnification objective. The number of SA β-gal positive cells was normalized by the total number of cells. Tom20 immunofluorescent staining—Mitochondrial network was assessed by Tom20 immunolabeling.158 Cells were seeded over 12-mm coverslips (631–1577P, VWR) in tissueculture treated 24-well plates and allowed to grow until they reached 80% confluency. Cells were permeabilized with 0.1% Triton X-100 in PBS for 15 min and washed with PBS. Blocking was performed in 1% BSA and 2% FBS-PBS during 30 min. Next, cells were incubated with Tom20 antibody (sc-11415, Santa Cruz Biotechnology) diluted 1:200 in blocking solution for 1 h, followed by washes with PBS and incubation with Donkey anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor 488 (A-21206, Invitrogen) diluted 1:500 in blocking buffer during 30 min. DNA and actin F cytoskeleton were stained with DAPI (D1306, Invitrogen) and Rhodamine-Phalloidin (R415, Invitrogen), respectively. Finally, coverslips were rinsed three times with PBS and mounted using ProLong Gold Antifade Mountant (P10144, Molecular Probes). Images were acquired in a Nikon Ti Eclipse confocal microscope (Nikon Instruments). Hardware and image acquisition were controlled by NIS-Elements imaging software (Nikon Instruments). The pipeline analysis of mitochondrial morphology and mass was adapted in Fiji software from Koopman et al.91 Transmission electron microscopy (TEM) of epon-embedded ultrathin sections—Cells were seeded in 10 cm tissue-culture treated dishes and allowed to grow until they reached 80–100% confluency. Cells were fixed with equal parts of a 4% glutaraldehyde solution (49625, Sigma-Aldrich) in 0.24 M PBS pH 7.2 and cell culture medium during 2 h at RT. Cells were gently collected with a scraper, transferred to 15 mL tubes and centrifuged (1,000 × g , 5 min). The fixative solution was discarded and the pellet was resuspended in 1 mL of 0.12 M PBS pH 7.2 and transferred to a new tube. A compact pellet was generated by centrifugation at 5,000 × g 5 min and embedded in epoxy resins. After polymerization, 150 nm thick sections were obtained using an ultramicrotome (Leica Microsystems) and a diamond knife (Diatome), placed on 100 mesh hexagonal Cu/Pd EM grids, stained with uranyl acetate and counterstained with lead citrate to reveal and enhance contrast of cellular membranes. For morphological analysis, ultrathin epon-embedded cell sections were studied by TEM. Images were collected using Lachiondo-Ortega et al. Page 34 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript a JEOL JEM-1230 transmission electron microscope operating at 100 kV and equipped with an Ultrascan 4000S P 4 K × 4 K CCD camera (GATAN). Images were acquired at different magnifications ranging from 1,000× to 20,000X. Analysis and segmentation of mitochondria and ER were performed using images with 2,500X and 5,000× magnification. A total of 56 micrographs per condition were analyzed. Images were first subjected to contrast-limited adaptive histogram equalization (CLAHE) to ensure homogeneity in black and white balance through all micrographs using the Microscopy Image Browser (MIB) standalone version 2.7 software.159 Then segmentation masks were manually applied on the cell sites where mitochondria and ER were identified in addition to using local thresholding. Total cell area, total mitochondria area, total ER area, mitochondria major and minor axis, and number of mitochondria per visualized cell were estimated within the region of interest. MitoTracker Green staining—Cells were seeded in tissue-culture treated 6-well plates and allowed to grow until they reached 80% confluency. Mitochondrial mass was assessed by fluorescent labeling with MitoTracker Green FM probe (M7514, Invitrogen) in combination with LIVE/DEAD fixable blue dead cell stain kit (L23105, Invitrogen), by following the manufacturer’s recommendations. Culture medium was removed, cells were washed with PBS twice, detached by trypsinization and transferred to 15 mL tubes. Each cell pellet was washed with 5 mL of PBS and collected by centrifugation (600 × g , 5 min). Next, each pellet was resuspended in 100 μL of the LIVE/DEAD cell dye in PBS and incubated for 10 min in the dark at RT. The staining solution was diluted by adding 500 μL of PBS per tube and cells were collected by centrifugation (600 × g , 5 min). Each pellet was resuspended in 100 μL of 100 nM MitoTracker Green FM probe in PBS and incubated 45 min at RT protected from the light. The staining solution was diluted by adding 500 μL of PBS per tube and cells were collected by centrifugation (600 × g , 5 min). Each cell pellet was resuspended in 200 μL of ice-cold 4% paraformaldehyde solution in PBS (sc-281692, Santa Cruz Biotechnology) and fixed during 15 min in the dark at 4°C. Fixation was stopped by the addition of 1 mL of PBS supplemented with 5% FBS per tube, and cells were collected by centrifugation (600 × g , 5 min, 4°C). Each cell pellet was finally resuspended in 200 μL of 5% FBS-PBS. MitoTracker Green FM probe fluorescence (λex = 490 nm, λem = 516 nm) was assayed by flow cytometry in a BD FACSymphony system (BD Biosciences). A set of unstained cells were used as a blank. The results were analyzed with FlowJo v10 software (BD Biosciences) and the geometric mean fluorescence intensity (gMFI) was calculated. Mitochondrial energetic metabolism studies—A two-step seeding process was followed to ensure that cells were evenly distributed throughout the wells of Seahorse XF24 cell culture microplates (102070–001, Agilent). First, 100 μL of the cell suspension were seeded per well and the plates were allowed to rest in the hood for 1 h at RT. In a second step, each well was topped up with 150 μL of culture medium and cells were allowed to grow until they reached 80% confluency. Mitochondrial respiration was assessed by performing a Cell Mito Stress Test in a Seahorse XF24 Analyzer (Agilent). The day before the assay, the XF24 sensor cartridges (102070–001, Agilent) were hydrated with Seahorse XF calibrant solution (102070–001, Agilent) overnight at 37°C in a 0% CO2 incubator. On the day of the assay, culture medium was replaced with bicarbonate-free lowLachiondo-Ortega et al. Page 35 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript buffered assay medium and cells were incubated for 1 h at 37°C in a 0% CO2 atmosphere. After establishing a respiration baseline, 6 μM oligomycin, 3 μM carbonyl cyanide 4trifluoromethoxy-phenylhydrazone (FCCP) and a combination of 0.5 μM antimycin A and 0.5 μM rotenone were sequentially injected through the reagent ports of the cartridges in order to measure changes in the oxygen consumption rate (OCR) related to ATP-linked, maximum and non-mitochondrial respiration, respectively. The OCR, expressed as pmol of O2/min, was normalized by the number of cells or protein content, which were estimated by crystal violet or Micro BCA Protein Assay Kit, respectively. MitoSOX red staining—Cells were seeded over 12-mm coverslips (631–1577P, VWR) previously coated with a 0.01% poly-L-lysine solution (P4707, Sigma-Aldrich), in tissueculture treated 24-well plates and allowed to grow until they reached 80% confluency. Mitochondrial reactive oxygen species (ROS) production was measured by MitoSOX Red mitochondrial superoxide indicator (M36008, Invitrogen), according to the manufacturer’s instructions. MitoSOX Red is a fluorogenic dye which is selectively targeted to mitochondria and exhibits red fluorescence when oxidized by superoxide anion. Cells were washed twice with warm DPBS, calcium, magnesium (14040133, Invitrogen) and incubated with 500 μL of1 μM MitoSOX Red reagent diluted in DPBS, calcium, magnesium for 10 min at 37°C protected from the light. An unstained well was used as blank. Cells were washed with DPBS, calcium, magnesium three times and fixed in ice-cold 4% paraformaldehyde solution in PBS (sc-281692, Santa Cruz Biotechnology) for 10 min at RT. The fixative solution was removed after two washes with DPBS, calcium, magnesium and coverslips were mounted in Fluoroshield with DAPI mounting medium (F6057, SigmaAldrich). MitoSOX Red fluorescence (λex = 510 nm, λem = 580 nm) was observed in the Axio Imager D1 epifluorescent microscope (Zeiss). A minimum of five areas per coverslip were assessed using an ×40 objective. MitoSOX Red fluorescence intensity was quantified using Fiji software (https://imagej.net/software/fiji/) and normalized by the number of nuclei stained with DAPI. Tetramethylrhodamine (TMRE) staining—Cells were seeded in tissue-culture treated 24-well plates and allowed to grow until they reached 80% confluency. Mitochondrial membrane potential was estimated by Tetramethylrhodamine, Ethyl Ester, Perchlorate (TMRE) (T669, Invitrogen), by following the manufacturer’s indications. TMRE is a fluorogenic cationic dye which is selectively targeted to active mitochondria. Cells were washed twice with warm DPBS, calcium, magnesium (14040133, Invitrogen) and incubated with 500 μL of 0.5 μM TMRE probe diluted in DPBS, calcium, magnesium for 30 min at 37°C protected from the light. An unstained well was used as blank. Cells were washed with DPBS, calcium, magnesium twice and maintained in 300 μL DPBS, calcium, magnesium. TMRE fluorescence (λex = 548 nm, λem = 574 nm) was measured in a SpectraMax M2/M2e microplate reader (Molecular Devices). TMRE probe fluorescence was normalized with the total protein content of each well, which was determined by the Micro BCA Protein Assay Kit. ATP levels quantification—Cells were seeded in tissue-culture treated 12-well plates and allowed to grow until they reached 80% confluency. Cellular ATP levels were Lachiondo-Ortega et al. Page 36 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript determined with the ATPlite luminescence assay system (6016943, PerkinElmer), by following the manufacturer’s instructions. The assay was performed in OptiPlate-96 white opaque 96-well microplates (6005290, PerkinElmer) and luminescence was measured in a Veritas microplate luminometer (Turner BioSystems). ATP concentration was calculated by interpolation to an ATP standard curve and subsequent normalization with the total protein content of each sample, which was determined by the Bradford assay. NAD+/NADH measurement—Cells were seeded in tissue-culture treated 6-well plates and allowed to grow until they reached 80% confluency. Cellular NAD+/NADH levels were determined with the NAD/NADH Colorimetric Assay Kit (ab65348, Abcam), according to the manufacturer’s instructions. The assay was performed in 96-well clear flat bottom plates and absorbance at 450 nm was measured in a SpectraMax M2/M2e microplate reader (Molecular Devices) after 2 h incubation at RT with NAD cycling enzyme mix. Total NAD and NADH concentrations were calculated by interpolation to a NADH standard curve. The levels of NAD+ were calculated by subtracting NADH from total NAD, and the NAD+/ NADH were represented. Cytosolic calcium (Ca2+) concentration determination—Cells were seeded over 12-mm coverslips (631–1577P, VWR) previously coated with a 0.01% poly-L-lysine solution (P4707, Sigma-Aldrich), in tissue-culture treated 24-well plates and allowed to grow until they reached 80% confluency. Cytosolic Ca2+ levels were determined by using Fura-2-AM (F1201, Invitrogen).160,161 Fura-2 is a cell-permeable ratiometric cytosolic Ca2+ indicator, whose excitation wavelength shifts from 380 nm to 340 nm as it binds to Ca2+, while its emission maximum is independent of Ca2+ concentration. Cells were washed twice with 0% FBS-culture medium and incubated with 1 μM Fura-2 dissolved in 0% FBS-culture medium for 30–45 min at 37°C. After loading, cells were incubated in 0% FBS-culture medium for 15 min at 37°C protected from the light. Coverslips were washed with 20 mM Tris-HCl pH 7.4, 2.4 mM CaCl2, 10 mM glucose solution and mounted on a thermostatized micro-perfusion chamber. Single-cell Fura-2 excitation intensity ratio (λex = 340 and 380 nm, λem = 510 nm) was measured with an ×40 oil-immersion magnification objective in an Eclipse TE 300-based microspectrofluorometer (Nikon) coupled to a DeltaRAM illumination system (Photon Technologies International). After recording a baseline for 30 s, 10 μM thapsigargin (T9033, Sigma-Aldrich) or 100 nM ATP was added to the medium to trigger the release of Ca2+ from the endoplasmic reticulum (ER) for 2 min and finally Ca2+ was added to measure the entrance of extracellular Ca2+ for additional 2 min. The ratio of excitation intensities at 340 nm and 380 nm (R) was related to Ca2+ levels by the following equation: [ Ca2 +] = Kd × Q × ( R Rmin ) / ( Rmax R ). Rmin and Rmax refer to the fluorescence intensity ratios when the probe is free or completely saturated of Ca2+ respectively, Q represents the ratio of minimum to maximum fluorescence at 380 nm (Fmin/Fmax), and Kd is the Ca2+ dissociation constant of Fura-2. The values for Fmin, Fmax, Rmin, Rmax and Kd were previously determined by means of a Fura-2 calibration curve in the presence of known Ca2+ concentrations. Lachiondo-Ortega et al. Page 37 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript QUANTIFICATION AND STATISTICAL ANALYSIS Statistical analysis was performed using Prism 8 software (GraphPad). Statistical details of experiments can be found in the figures and corresponding legends. Unless otherwise stated, experiments were reproduced at least three times. Data are represented as the mean ± standard deviation (SD) of at least three biological replicates within one representative experiment. A two-tailed t test was used to compare the differences between two groups. A p value <0.05 was considered statistically significant for all analyses and defined as *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. If not indicated otherwise, the differences were not significant (n.s.). ADDITIONAL RESOURCES Animations—Animations were created using BioRender software (https://biorender.com). Supplementary Material Refer to Web version on PubMed Central for supplementary material. ACKNOWLEDGMENTS This work was supported by grants to M.L.M.-C. from Departamento de Industria del Gobierno Vasco, Spain; Ministerio de Ciencia e Innovació n, Spain (grant no. PID2020–117116RB-I00); European Regional Development Fund (ERDF), EU; and CIBERehd, which is funded by Instituto de Salud Carlos III (ISCIII), Spain. M.L.M.-C. and J.S. received funding from Ministerio de Ciencia e Innovació n (grant no. RTC2019–007125-1) and ISCIII (grant no. DTS20/00138). M.L.M.-C. and R.M.L. acknowledge Ministerio de Ciencia e Innovación (grant no. RED2022–134397-T). M.L.M.-C. and J.M.B. were awarded with a grant from Fundació n la Caixa, Spain (grant no. HR17–00601). M.L.M.-C., J.M.B., M.A.A., and J.J.G.M. acknowledge financial support from Fundació n Cientı´fica de la Asociación Española Contra el Cáncer (AECC), Spain. M.S.R. recognizes funding from Fondo Sectorial de Investigación SRE - CONACYT, Mexico (grant no. 0280365); Horizon 2020 Research and Innovation Program funded under Marie Sk1odowska-Curie Actions, EU (grant no. 765445); and REPÈRE and Programme de Prématuration from Région Occitanie, France. M.G., S.D., and K.M.-M. were supported by the National Institute on Aging (NIA), National Institutes of Health (NIH), US (grant no. Z01-AG000511–23). I.D.-M. is grateful for the grants received from Junta de Andalucía, Spain (grant no. BIO-198, US-1254317, P18-FR-3487, and P18-HO-4091); Ministerio de Ciencia, Innovación y Universidades, Spain (grant no. PGC2018–096049-BI00); and Fundación Ramón Areces, Spain. T.D. acknowledges Fondation ARC, France (grant no. 208084). J.J.G.M. was supported by Junta de Castilla y León, Spain (grant no. SA063P17); Fundacion La Marató TV3, Spain (grant no. 201916–31); ISCIII (grant no. PI19/00819); CIBERehd; and ERDF (grant no. OLD-HEPAMARKER). M.A.A. recognizes Gobierno de Navarra, Spain (grant no. G°Na 42/21); Eurorregión Nueva Aquitania-EuskadiNavarra, Spain; Ministerio de Ciencia e Innovación (grant no. PID2019–104878RB-I00); and CIBERehd. A.P. expresses gratitude to the European Research Council (ERC), EU (grant no. 804236) for their support. M.D.G. received financial support from Junta de Andalucía (grant no. PEMP-0036–2020 and BIO-0139); Ministerio de Universidades, Spain (grant no. FPU20/03957); ISCIII (grant no. PI20/01301), Fundación Sociedad Española de Endocrinología y Nutrición (FSEEN), Spain; CIBERehd; and CIBERobn, which is also funded by ISCIII. J.M.B. acknowledges Euskadi RIS3 (grant no. 2019222054, 2020333010, and 2021333003) and Elkartek programs from Gobierno Vasco (grant no. KK-2020/00008); ISCIII (grant no. PI18/01075, CPII19/00008, and PI21/00922); CIBERehd; PSC Support, UK; AMMF The Cholangiocarcinoma Charity, UK (grant no. EU/2019/AMMFt/001); Horizon 2020 Research and Innovation Program (grant no. 825510); ERDF; and PSC Partners Seeking a Cure, US. A.L. received financial support from the Damon Runyon-Rachleff Innovation Award, US (grant no. DR52– 18) and the MERIT Award (R37) from the National Cancer Institute (NCI), NIH (grant no. R37CA230636). F.E. expresses his gratitude to ProteoRed from ISCIII (grant no. PT13/0001/0027) and CIBERehd. N.G.A.A. was funded by Ministerio de Ciencia, Innovación y Universidades (grant no. RTI2018–095700-B-I00). R.B. acknowledges financial support from Gobierno Vasco (grant no. IT1165–19); Ministerio de Economía, Industria y Competitividad, Spain (grant no. SAF2017–90900-REDT); Ministerio de Econom´ıa, Industria y Competitividad, ERDF (grant no. BFU2017–84653-P); Ministerio de Ciencia e Innovació n (grant no. PID2020–114178GB-I00); and Horizon 2020 funded under Marie Sk1odowska-Curie Actions (grant no. 765445-EU). A.M.A. acknowledges CIBERehd. L.A.M.-C. obtained grants from Ministerio de Economía y Competitividad (grant no. CSD2008– 00005); Ministerio de Economía, Industria y Competitividad (grant no. BFU2016–77408-R); ISCIII; and EJP RD, EU (grant no. EJPRD19–040). I.G.-R. was supported by Ministerio de Economía, Industria y Competitividad (grant no. BES-2017–080435 ). M.S.-M. is grateful to the AECC, Sede de Bizkaia, Spain for the financial support. J.D.Z. Lachiondo-Ortega et al. Page 38 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript was awarded with a grant from Ministerio de Economía, Industria y Competitividad (grant no. SEV-2016–0644-18– 2). C.M. acknowledges Gobierno Vasco (grant no. IT-1264–19) and Ministerio de Ciencia e Innovación (grant no. PID2022–136788OB-I00). A.V.-C. was supported by Ministerio de Educación, Cultura y Deporte, Spain (grant no. FPU016/01513). C.F.-R. thanks Tekniker, Spain and CIC bioGUNE, Spain for financial support. A.G.-d.R. was funded by Bikaintek program from Gobierno Vasco (grant no. 48-AF-W1–2019-00012). N.G.-U. obtained a grant from Gobierno Vasco. T.C.D. expresses gratitude to AECC. J.S. received financial support from CIBERehd. C.M.R.-G. was supported by Ayudas a la Recualificación Margarita Salas from Universidad de Extremadura, Ministerio de Universidades financed by Next-GenerationEU. S.L.-O. is thankful to Gobierno Vasco. We thank the Genome Analysis platform at CIC bioGUNE for performing the sequencing of the RIP experiment. We are also grateful to the Electron Microscopy Service at Centro de Biología Molecular Severo Ochoa (CBMSO) for technical assistance in ultrathin cell sectioning and the EM platform at CIC bioGUNE for their support in TEM imaging. Last but not least, we would like to extend our sincere gratitude to Begoña Rodríguez-Iruretagoyena from the Liver Disease Lab at CIC bio-GUNE for her assistance in the completion of this study. REFERENCES 1. Craig AJ, von Felden J, Garcia-Lezana T, Sarcognato S, and Villanueva A (2020). Tumour evolution in hepatocellular carcinoma. Nat. Rev. Gastroenterol. Hepatol. 17, 139–152. 10.1038/ s41575-019-0229-4. [PubMed: 31792430] 2. Faivre S, Rimassa L, and Finn RS (2020). Molecular therapies for HCC: Looking outside the box. J. Hepatol. 72, 342–352. 10.1016/j.jhep.2019.09.010. [PubMed: 31954496] 3. Seeler J-S, and Dejean A (2017). SUMO and the robustness of cancer. Nat. Rev. Cancer 17, 184– 197. 10.1038/nrc.2016.143. [PubMed: 28134258] 4. Vertegaal ACO (2022). Signalling mechanisms and cellular functions of SUMO. Nat. Rev. Mol. Cell Biol. 23, 715–731. 10.1038/s41580-022-00500-y. [PubMed: 35750927] 5. Bouchard D, Wang W, Yang W-C, He S, Garcia A, and Matunis MJ (2021). SUMO paralogue– specific functions revealed through systematic analysis of human knockout cell lines and gene expression data. Mol. Biol. Cell 32, 1849–1866. 10.1091/mbc.E21-01-0031. [PubMed: 34232706] 6. Owerbach D, McKay EM, Yeh ETH, Gabbay KH, and Bohren KM (2005). A proline-90 residue unique to SUMO-4 prevents maturation and sumoylation. Biochem. Biophys. Res. Commun. 337, 517–520. 10.1016/j.bbrc.2005.09.090. [PubMed: 16198310] 7. Liang Y-C, Lee C-C, Yao Y-L, Lai C-C, Schmitz ML, and Yang W-M (2016). SUMO5, a Novel Poly-SUMO Isoform, Regulates PML Nuclear Bodies. Sci. Rep. 6, 26509. 10.1038/srep26509. [PubMed: 27211601] 8. Saitoh H, and Hinchey J (2000). Functional heterogeneity of small ubiquitin-related protein modifiers SUMO-1 versus SUMO-2/3. J. Biol. Chem. 275, 6252–6258. 10.1074/jbc.275.9.6252. [PubMed: 10692421] 9. Tatham MH, Jaffray E, Vaughan OA, Desterro JM, Botting CH, Naismith JH, and Hay RT (2001). Polymeric Chains of SUMO-2 and SUMO-3 Are Conjugated to Protein Substrates by SAE1/SAE2 and Ubc9. J. Biol. Chem. 276, 35368–35374. 10.1074/jbc.M104214200. [PubMed: 11451954] 10. Matic I, van Hagen M, Schimmel J, Macek B, Ogg SC, Tatham MH, Hay RT, Lamond AI, Mann M, and Vertegaal ACO (2008). In vivo identification of human small ubiquitin-like modifier polymerization sites by high accuracy mass spectrometry and an in vitro to in vivo strategy. Mol. Cell. Proteomics 7, 132–144. 10.1074/mcp.M700173-MCP200. [PubMed: 17938407] 11. Hendriks IA, Lyon D, Young C, Jensen LJ, Vertegaal ACO, and Nielsen ML (2017). Site-specific mapping of the human SUMO proteome reveals co-modification with phosphorylation. Nat. Struct. Mol. Biol. 24, 325–336. 10.1038/nsmb.3366. [PubMed: 28112733] 12. Matic I, Schimmel J, Hendriks IA, van Santen MA, van de Rijke F, van Dam H, Gnad F, Mann M, and Vertegaal ACO (2010). Site-Specific Identification of SUMO-2 Targets in Cells Reveals an Inverted SUMOylation Motif and a Hydrophobic Cluster SUMOylation Motif. Mol. Cell 39, 641–652. 10.1016/j.molcel.2010.07.026. [PubMed: 20797634] 13. Hecker C-M, Rabiller M, Haglund K, Bayer P, and Dikic I (2006). Specification of SUMO1and SUMO2-interacting Motifs. J. Biol. Chem. 281, 16117–16127. 10.1074/jbc.M512757200. [PubMed: 16524884] 14. González-Prieto R, Eifler-Olivi K, Claessens LA, Willemstein E, Xiao Z, Talavera Ormeno CMP, Ovaa H, Ulrich HD, and Vertegaal ACO (2021). Global non-covalent SUMO interaction networks Lachiondo-Ortega et al. Page 39 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript reveal SUMO-dependent stabilization of the non-homologous end joining complex. Cell Rep. 34, 108691. 10.1016/j.celrep.2021.108691. [PubMed: 33503430] 15. Zhao X (2018). SUMO-Mediated Regulation of Nuclear Functions and Signaling Processes. Molecular cell 71, 409–418. 10.1016/j.molcel.2018.07.027. [PubMed: 30075142] 16. Nacerddine K, Lehembre F, Bhaumik M, Artus J, Cohen-Tannoudji M, Babinet C, Pandolfi PP, and Dejean A (2005). The SUMO Pathway Is Essential for Nuclear Integrity and Chromosome Segregation in Mice. Dev. Cell 9, 769–779. 10.1016/j.devcel.2005.10.007. [PubMed: 16326389] 17. Wang L, Wansleeben C, Zhao S, Miao P, Paschen W, and Yang W (2014). SUMO2 is essential while SUMO3 is dispensable for mouse embryonic development. EMBO Rep. 15, 878–885. 10.15252/embr.201438534. [PubMed: 24891386] 18. Demarque MD, Nacerddine K, Neyret–Kahn H, Andrieux A, Danenberg E, Jouvion G, Bomme P, Hamard G, Romagnolo B, Terris B, et al. (2011). Sumoylation by Ubc9 Regulates the Stem Cell Compartment and Structure and Function of the Intestinal Epithelium in Mice. Gastroenterology 140, 286–296. 10.1053/j.gastro.2010.10.002. [PubMed: 20951138] 19. Chang H-M, and Yeh ETH (2020). SUMO: From Bench to Bedside. Physiol. Rev. 100, 1599–1619. 10.1152/physrev.00025.2019. [PubMed: 32666886] 20. Kroonen JS, and Vertegaal ACO (2021). Targeting SUMO Signaling to Wrestle Cancer. Trends Cancer 7, 496–510. 10.1016/j.trecan.2020.11.009. [PubMed: 33353838] 21. Lee J-S, and Thorgeirsson SS (2004). Genome-scale profiling of gene expression in hepatocellular carcinoma: Classification, survival prediction, and identification of therapeutic targets. Gastroenterology 127, S51–S55. 10.1053/j.gastro.2004.09.015. [PubMed: 15508103] 22. Guo WH, Yuan LH, Xiao ZH, Liu D, and Zhang JX (2011). Overexpression of SUMO-1 in hepatocellular carcinoma: a latent target for diagnosis and therapy of hepatoma. J. Cancer Res. Clin. Oncol. 137, 533–541. 10.1007/s00432-010-0920-x. [PubMed: 20502916] 23. Chen J, Chen C, Lin Y, Su Y, Yu X, Jiang Y, Chen Z, Ke S, Lin S, Chen L, et al. (2021). Down-regulation of SUMO2 inhibits hepatocellular carcinoma cell proliferation, migration and invasion. FEBS Open Bio 11, 1771–1784. 10.1002/2211-5463.13173. 24. Ong JR, Bamodu OA, Khang NV, Lin Y-K, Yeh C-T, Lee W-H, and Cherng Y-G (2021). SUMOActivating Enzyme Subunit 1 (SAE1) Is a Promising Diagnostic Cancer Metabolism Biomarker of Hepatocellular Carcinoma. Cells 10. 10.3390/cells10010178. 25. Tu J, Chen Y, Cai L, Xu C, Zhang Y, Chen Y, Zhang C, Zhao J, Cheng J, Xie H, et al. (2015). Functional Proteomics Study Reveals SUMOylation of TFII-I is Involved in Liver Cancer Cell Proliferation. J. Proteome Res. 14, 2385–2397. 10.1021/acs.jpro-teome.5b00062. [PubMed: 25869096] 26. Tomasi ML, Tomasi I, Ramani K, Pascale RM, Xu J, Giordano P, Mato JM, and Lu SC (2012). Sadenosyl methionine regulates ubiquitin-conjugating enzyme 9 protein expression and sumoylation in murine liver and human cancers. Hepatology 56, 982–993. 10.1002/hep.25701. [PubMed: 22407595] 27. Tsai CL, Liu WL, Hsu FM, Yang PS, Yen RF, Tzen KY, Cheng AL, Chen PJ, and Cheng JCH (2018). Targeting histone deacetylase 4/ubiquitin-conjugating enzyme 9 impairs DNA repair for radiosensitization of hepatocellular carcinoma cells in mice. Hepatology 67, 586–599. [PubMed: 28646552] 28. Yang H, Gao S, Chen J, and Lou W (2020). UBE2I promotes metastasis and correlates with poor prognosis in hepatocellular carcinoma. Cancer Cell Int. 20, 234. 10.1186/s12935-020-01311-x. [PubMed: 32536822] 29. Fang S, Qiu J, Wu Z, Bai T, Guo W, Fang S, Qiu J, Wu Z, Bai T, and Guo W (2017). Down-regulation of UBC9 increases the sensitivity of hepatocellular carcinoma to doxorubicin. Oncotarget 8, 49783–49795. 10.18632/ONCOTARGET.17939. [PubMed: 28572537] 30. Li J, Zeng M, Yan K, Yang Y, Li H, and Xu X (2020). IL-17 promotes hepatocellular carcinoma through inhibiting apoptosis induced by IFN-γ. Biochem. Biophys. Res. Commun. 522, 525–531. 10.1016/j.bbrc.2019.11.134. [PubMed: 31784083] 31. Morofuji N, Ojima H, Hiraoka N, Okusaka T, Esaki M, Nara S, Shimada K, Kishi Y, and Kondo T (2016). Antibody-based proteomics to identify an apoptosis signature for early recurrence of hepatocellular carcinoma. Clin. Proteonomics 13, 28. 10.1186/s12014-016-9130-0. Lachiondo-Ortega et al. Page 40 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 32. Liu Q, Zhou B, Liao R, Zhou X, and Yan X (2020). PIAS4, upregulated in hepatocellular carcinoma, promotes tumorigenicity and metastasis. J. Cell. Biochem. 121, 3372–3381. 10.1002/ jcb.29610. [PubMed: 31943317] 33. Zheng L, You N, Huang X, Gu H, Wu K, Mi N, and Li J (2019). COMMD7 Regulates NF-kB Signaling Pathway in Hepatocellular Carcinoma Stem-like Cells. Mol. Ther. Oncolytics 12, 112– 123. 10.1016/j.omto.2018.12.006. [PubMed: 30719501] 34. Zhang W, Sun H, Shi X, Wang H, Cui C, Xiao F, Wu C, Guo X, and Wang L (2016). SENP1 regulates hepatocyte growth factor-induced migration and epithelial-mesenchymal transition of hepatocellular carcinoma. Tumour Biol. 37, 7741–7748. 10.1007/s13277-015-4406-y. [PubMed: 26695141] 35. Cui C-P, Wong CC-L, Kai AK-L, Ho DW-H, Lau EY-T, Tsui Y-M, Chan L-K, Cheung T-T, Chok KS-H, Chan ACY, et al. (2017). SENP1 promotes hypoxia-induced cancer stemness by HIF-1α deSUMOylation and SENP1/HIF-1α positive feedback loop. Gut 66, 2149–2159. 10.1136/gutjnl-2016-313264. [PubMed: 28258134] 36. Tao Y, Li R, Shen C, Li J, Zhang Q, Ma Z, Wang F, and Wang Z (2020). SENP1 is a crucial promotor for hepatocellular carcinoma through deSUMOylation of UBE2T. Aging 12, 1563–1576. 10.18632/aging.102700. [PubMed: 31969492] 37. Jin Z-L, Pei H, Xu Y-H, Yu J, and Deng T (2016). The SUMO-specific protease SENP5 controls DNA damage response and promotes tumorigenesis in hepatocellular carcinoma. Eur. Rev. Med. Pharmacol. Sci. 20, 3566–3573. [PubMed: 27649656] 38. Qian J, Luo Y, Gu X, and Wang X (2013). Inhibition of SENP6-Induced Radiosensitization of Human Hepatocellular Carcinoma Cells by Blocking Radiation-Induced NF-κB Activation. Cancer Biother. Radiopharm. 28, 196–200. 10.1089/cbr.2012.1288. [PubMed: 23461386] 39. Stefanska B, Cheishvili D, Suderman M, Arakelian A, Huang J, Hallett M, Han Z-G, Al-Mahtab M, Akbar SMF, Khan WA, et al. (2014). Genome-Wide Study of Hypomethylated and Induced Genes in Patients with Liver Cancer Unravels Novel Anticancer Targets. Clin. Cancer Res. 20, 3118–3132. 10.1158/1078-0432.CCR-13-0283. [PubMed: 24763612] 40. Ji P, Liang S, Li P, Xie C, Li J, Zhang K, Zheng X, Feng M, Li Q, Jiao H, et al. (2018). Speckle-type POZ protein suppresses hepatocellular carcinoma cell migration and invasion via ubiquitin-dependent proteolysis of SUMO1/sentrin specific peptidase 7. Biochem. Biophys. Res. Commun. 502, 30–42. 10.1016/j.bbrc.2018.05.115. [PubMed: 29777712] 41. Tang X, Liu B, Zhang C, Tang W, Liang S, Xiao Y, Deng R, and Li Z (2021). SENP2 Reduces Hepatocellular Carcinoma Stemness and Improves Sorafenib Sensitivity Through Inactivating the AKT/GSK3β/CTNNB1 Pathway. Front. Oncol. 11. 42. Tomasi ML, Ryoo M, Ramani K, Tomasi I, Giordano P, Mato JM, Lu SC, Lauda Tomasi M, Ryoo M, Ramani K, et al. (2015). Methionine adenosyltransferase a2 sumoylation positively regulate Bcl-2 expression in human colon and liver cancer cells. Oncotarget 6, 37706–37723. 10.18632/ ONCOTARGET.5342. [PubMed: 26416353] 43. Bian XL, Chen HZ, Yang PB, Li YP, Zhang FN, Zhang JY, Wang WJ, Zhao WX, Zhang S, Chen QT, et al. (2017). Nur77 suppresses hepatocellular carcinoma via switching glucose metabolism toward gluconeogenesis through attenuating phosphoenolpyruvate carboxykinase sumoylation. Nat. Commun. 8, 14420. 10.1038/ncomms14420. [PubMed: 28240261] 44. Hou PP, Luo LJ, Chen HZ, Chen QT, Bian XL, Wu SF, Zhou JX, Zhao WX, Liu JM, Wang XM, et al. (2020). Ectosomal PKM2 Promotes HCC by Inducing Macrophage Differentiation and Remodeling the Tumor Microenvironment. Mol. Cell 78, 1192–1206.e10. 10.1016/j.molcel.2020.05.004. [PubMed: 32470318] 45. Zubiete-Franco I, García-Rodríguez JL, Lopitz-Otsoa F, Serrano-Macia M, Simon J, FernándezTussy P, Barbier-Torres L, Fernández-Ramos D, Gutiérrez-de-Juan V, López de Davalillo S, et al. (2019). SUMOylation regulates LKB1 localization and its oncogenic activity in liver cancer. EBioMedicine 40, 406–421. 10.1016/j.ebiom.2018.12.031. [PubMed: 30594553] 46. Li J, Xu Y, Long XD, Wang W, Jiao HK, Mei Z, Yin QQ, Ma LN, Zhou AW, Wang LS, et al. (2014). Cbx4 governs HIF-1α to potentiate angiogenesis of hepatocellular carcinoma by its SUMO E3 ligase activity. Cancer Cell 25, 118–131. 10.1016/j.ccr.2013.12.008. [PubMed: 24434214] Lachiondo-Ortega et al. Page 41 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 144. Schmidt MW, Baldridge KK, Boatz JA, Elbert ST, Gordon MS, Jensen JH, Koseki S, Matsunaga N, Nguyen KA, Su S, et al. (1993). General atomic and molecular electronic structure system. J. Comput. Chem. 14, 1347–1363. 10.1002/jcc.540141112. 145. Seminario JM (1996). Calculation of intramolecular force fields from second-derivative tensors. Int. J. Quant. Chem. 60, 1271–1277. 10.1002/(SICI)1097-461X(1996)60:7<1271::AIDQUA8>3.0.CO;2-W. 146. Berendsen HJC, Postma JPM, van Gunsteren WF, DiNola A, and Haak JR (1984). Molecular dynamics with coupling to an external bath. J. Chem. Phys. 81, 3684–3690. 10.1063/1.448118. 147. Ryckaert J-P, Ciccotti G, and Berendsen HJ (1977). Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput. Phys. 23, 327–341. 10.1016/0021-9991(77)90098-5. 148. Roe DR, and Cheatham TE 3rd. (2013). PTRAJ and CPPTRAJ: Software for Processing and Analysis of Molecular Dynamics Trajectory Data. J. Chem. Theor. Comput. 9, 3084–3095. 10.1021/ct400341p. 149. Díaz-Moreno I, Muñoz-López FJ, Frutos-Beltrán E, De la Rosa MA, and Díaz-Quintana A (2009). Electrostatic strain and concerted motions in the transient complex between plastocyanin and cytochrome f from the cyanobacterium Phormidium laminosum. Bioelectrochemistry 77, 43–52. 10.1016/j.bioelechem.2009.06.003. [PubMed: 19616485] 150. Hayward S, and Lee RA (2002). Improvements in the analysis of domain motions in proteins from conformational change: DynDom version 1.50. J. Mol. Graph. Model. 21, 181–183. 10.1016/s1093-3263(02)00140-7. [PubMed: 12463636] 151. Hayward S, and Berendsen HJ (1998). Systematic analysis of domain motions in proteins from conformational change: new results on citrate synthase and T4 lysozyme. Proteins 30, 144–154. [PubMed: 9489922] 152. Gabdoulline RR, and Wade RC (1998). Brownian dynamics simulation of protein-protein diffusional encounter. Methods 14, 329–341. 10.1006/meth.1998.0588. [PubMed: 9571088] 153. Elcock AH, Gabdoulline RR, Wade RC, and McCammon JA (1999). Computer simulation of protein-protein association kinetics: acetylcholinesterase-fasciculin11Edited by B. Honig. J. Mol. Biol. 291, 149–162. 10.1006/jmbi.1999.2919. [PubMed: 10438612] 154. Gabdoulline RR, and Wade RC (2009). On the Contributions of Diffusion and Thermal Activation to Electron Transfer between Phormidium laminosum Plastocyanin and Cytochrome f: Brownian Dynamics Simulations with Explicit Modeling of Nonpolar Desolvation Interactions and Electron Transfer Event. J. Am. Chem. Soc. 131, 9230–9238. 10.1021/ja809567k. [PubMed: 19518050] 155. Humphrey W, Dalke A, and Schulten K (1996). VMD: visual molecular dynamics. J. Mol. Graph. 14, 33–38. 10.1016/0263-7855(96)00018-5. [PubMed: 8744570] 156. Baker NA, Sept D, Joseph S, Holst MJ, and McCammon JA (2001). Electrostatics of nanosystems: application to microtubules and the ribosome. USAUSA 98, 10037–10041. 10.1073/pnas.181342398. 157. Daubon T, Léon C, Clarke K, Andrique L, Salabert L, Darbo E, Pineau R, Guérit S, Maitre M, Dedieu S, et al. (2019). Deciphering the complex role of thrombospondin-1 in glioblastoma development. Nat. Commun. 10, 1146. 10.1038/s41467-019-08480-y. [PubMed: 30850588] 158. Guyon J, Fernandez-Moncada I, Larrieu CM, Bouchez CL, Pagano Zottola AC, Galvis J, Chouleur T, Burban A, Joseph K, Ravi VM, et al. (2022). Lactate dehydrogenases promote glioblastoma growth and invasion via a metabolic symbiosis. EMBO Mol. Med. 14. 10.15252/ EMMM.202115343/SUPPL_FILE/EMMM202115343-SUP-0004-SDATAEV.ZIP. 159. Belevich I, Joensuu M, Kumar D, Vihinen H, and Jokitalo E (2016). Microscopy Image Browser: A Platform for Segmentation and Analysis of Multidimensional Datasets. PLoS Biol. 14, e1002340. [PubMed: 26727152] 160. Martín C, Gómez-Bilbao G, and Ostolaza H (2010). Bordetella Adenylate Cyclase Toxin Promotes Calcium Entry into Both CD11b+ and CD11b− Cells through cAMP-dependent Ltype-like Calcium Channels. J. Biol. Chem. 285, 357–364. 10.1074/jbc.M109.003491. [PubMed: 19875442] Lachiondo-Ortega et al. Page 48 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript 161. Uribe KB, Martín C, Etxebarria A, González-Bullón D, Gómez-Bilbao G, and Ostolaza H (2013). Ca2+ influx and tyrosine kinases trigger Bordetella adenylate cyclase toxin (ACT) endocytosis. Cell physiology and expression of the CD11b/CD18 integrin major determinants of the entry route. PLoS One 8, e74248. 10.1371/journal.pone.0074248. [PubMed: 24058533] Lachiondo-Ortega et al. Page 49 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Highlights •HuR is SUMOylated in preclinical and clinical hepatocellular carcinoma tumors •SUMOylation affects HuR binding affinity to its target RNAs •HuR SUMOylation promotes major cancer hallmarks, namely proliferation and invasion •Lack of HuR SUMOylation results in a senescent phenotype Lachiondo-Ortega et al. Page 50 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 1. HuR SUMOylation is increased in human HCC (A and B) ELAVL1 mRNA expression levels (A) in the T (n = 368) and NT (n = 50) tissue of patients with HCC, and (B) at the different stages (I–IV) of the disease. (C) Survival curve of liver cancer patients with high (n = 180) and low (n = 190) ELAVL1 mRNA expression levels. (D) Enriched molecular processes after performing a GSEA according to ELAVL1 mRNA expression in HCC patients. Lachiondo-Ortega et al. Page 51 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript (E) mRNA expression levels of the main components of the SUMO pathway in the T (n = 370) of HCC patients relative to the NT tissue (n = 50). (F) Heatmap representing R2 values obtained from Pearson correlation studies on ELAVL1 mRNA expression and the canonical SUMOylation pathway members in paired T and ST liver samples from a cohort of patients with HCC (n = 86). (G–J) Enrichment, identification, and quantification of the SUMO-interacting proteome from non-tumoral and tumoral (G) human (n = 5) and (H) mouse liver tissue (n = 10) and (I and J) human cell line protein extracts by means of GST-SUBEs pull-down technology in combination with (G and H) western blotting and (I and J) LC-MS/MS. Data in (A–E) were obtained from TCGA Research Network. Data in (A), (B), (E), (I), and (J) are presented as the mean ± SD of at least three biological replicates within one representative experiment. *p < 0.05, **p < 0.01, and ***p < 0.001, two-tailed t test vs. NT (A, B, and E), THLE-2 (I), or GST (J). If not indicated otherwise, the differences were not significant. In (H), western blots are representative of at least three biological replicates. See also Figure S1. Lachiondo-Ortega et al. Page 52 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 2. HuR is mainly modified by SUMO2/3 at lysines in position 120 and 182 inducing a structural rearrangement of the RRM 1 and 2 so as to modulate its intrinsic RNA-binding ability (A–D) Modified V5-HuR protein enrichment after transient transfection of plasmids expressing the different (A) SUMO, (B) PIAS, and (C) SENP isoforms as well as (D) all the lysine-to-arginine HuR mutants contained in the RRM1–2 domains, and subsequent nickel-histidine affinity purification, relative to total V5-HuR protein expression levels in the mouse liver progenitor MLP-29 cell line. (E) Modified V5-HuR protein enrichment after co-transfection of plasmids expressing UBC9, SUMO2/3, and WT HuR or the SUMOylation double mutant, and subsequent Lachiondo-Ortega et al. Page 53 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript nickel-histidine affinity purification, relative to total V5-HuR protein expression levels in the human hepatoma HuH-7 cell line. (F) SUMOylated HuR enrichment after transient transfection of WT HuR and the different SUMOylation mutants and subsequent GST-SUBEs protein pull-down, relative to total V5-HuR protein expression levels in the HuH-7 cell line. (G) Ribbon and surface representations of the HuR RRM1–2 protein construct and the different single and doubly SUMOylated species (K120, K182, and K120/182). RRM2 domains were kept in a fixed position in all representations to show them in the same orientation. The rotation angles of RRM1 relative to RRM2 (using the WT conformation as a reference) are depicted on the left of SUMOylated HuR models, and SUMOylated sites are highlighted with dashed circles on the ribbon structures. Models correspond to the structures with the lowest root-mean-square deviation (RMSD) with respect to the last 10-ns average coordinates in each MD trajectory. In (A–F), western blots are representative of at least three biological replicates within one representative experiment. In (D) and (E), the entire blot image was digitally processed to eliminate irrelevant lanes. See also Figures S2–S6 and Table S3. Lachiondo-Ortega et al. Page 54 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 3. SUMOylation of HuR promotes the main cancer hallmarks and avoids palbociclibinduced senescence in human hepatoma cells (A and B) Cell (A) proliferation and (B) scratch-wound healing process of HuH-7 cell lines stably expressing WT HuR and the different SUMOylation mutants analyzed in the IncuCyte system. (C) Representative pictures of 3D spheroids from HuH-7 cell lines stably expressing WT HuR and the SUMOylation mutant species embedded on a collagen type I matrix for 48 h, and quantification of the relative invasive area. Lachiondo-Ortega et al. Page 55 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript (D) Percentage of apoptosis detected in the WT HuR and the indicated SUMOylation mutant HuH-7 cell variants by flow-cytometry analysis after FITC-annexin V and viability stainings, relative to STS-treated HuH-7 cells. (E) Quantification of cell proliferation in the HuH-7 cell lines stably expressing WT HuR and the different SUMOylation mutants after an acute 3-day treatment with a range of palbociclib concentrations analyzed by crystal violet staining. (F) Quantification of relative senescence in the HuH-7 cell lines stably expressing the WT and the K120/182R, K120R, K182R HuR mutant species after chronic 2-week treatment with a range of palbociclib concentrations analyzed by β-galactosidase staining. (G) Crystal violet staining of colonies from HuH-7 cell lines stably expressing WT HuR and the SUMOylation mutants treated with the indicated doses of palbociclib for 10 days, and generation of dose-response curves for calculation of IC50 values. In (A–F), data are presented as the mean ± SD of at least three biological replicates within one representative experiment. *p < 0.05, **p < 0.01, and ***p < 0.001, two-tailed t test vs. HuH-7WT HuR. If not indicated otherwise, the differences were not significant. In (C) and (G), images are representative of at least three biological replicates within one representative experiment. See also Figures S7–S9. Lachiondo-Ortega et al. Page 56 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 4. SUMOylated HuR evades palbociclib-mediated senescence by increasing HuR and global SUMOylation levels in human hepatoma cells (A) Enrichment of SUMOylated HuR in the HuH-7WT HuR and HuH-7HuR [K120/182R] cells treated with 5 nM palbociclib for 2 weeks by means of protein pull-down with GST control and SUBEs in combination with western blotting analysis. (B and C) (B) SUMO1 and (C) SUMO2/3 protein expression levels and quantification in the HuH-7cell lines stably expressing WT HuR and the K120/182R, K120R, K182R HuR SUMOylation mutant species treated with 5 nM palbociclib for 2 weeks. Lachiondo-Ortega et al. Page 57 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript In (C), (E), and (F), images are representative of at least three biological replicates within one representative experiment. Scale bars represent 1 cm in (C) and 200 μm in (E). In (A), (B), (D), and (F), data are presented as the mean ± SD of at least three biological replicates within one representative experiment. *p < 0.05, **p < 0.01, and ***p < 0.001, two-tailed t test. If not indicated otherwise, the differences were not significant. Lachiondo-Ortega et al. Page 64 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Lachiondo-Ortega et al. Page 65 KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies HuR/ELAV1 (3A2) Santa Cruz Biotechnology Cat#sc-5261; RRID:AB_627770 SUMO1 76–86 Developmental Studies Hybridoma Bank (DSHB) N/A SUMO-2 8A2 Developmental Studies Hybridoma Bank (DSHB) N/A UBC9 Proteintech Cat#14837-1-AP; RRID:AB_2272479 V5 tag Invitrogen Cat#R960-25; RRID: N/A His Cytiva Cat#27-4710-01; RRID:AB_771435 FLAG M2 Sigma-Aldrich Cat#F1804; RRID:AB_262044 RGSHis Qiagen Cat#34610; RRID: N/A Cyclin D1 (92G2) Cell Signaling Technology Cat#2978; RRID:AB_2259616 Cyclin A2 (BF683) Cell Signaling Technology Cat#4656; RRID:AB_2071958 Phospho-Rb (Ser780) (D59B7) Cell Signaling Technology Cat#8180; RRID:AB_10950972 Rb (4H1) Cell Signaling Technology Cat#9309; RRID:AB_823629 gamma H2A.X (phospho S139) [EP854(2)Y] Abcam Cat#ab81299; RRID:AB_1640564 β-Actin Sigma-Aldrich Cat#A5441; RRID:AB_476744 Anti-mouse IgG, HRP-linked Cell Signaling Technology Cat#7076; RRID:AB_330924 Anti-rabbit IgG, HRP-linked Cell Signaling Technology Cat#7074; RRID:AB_2099233 Biological samples Human HCC tumor and surrounding non-tumor tissue Córdoba Node of the Andalusian Public Health System Biobank http://www.biobancosspa.com Human HCC tumor and surrounding non-tumor tissue Basque Biobank https://www.biobancovasco.org MYC;Trp53−/− mouse liver tissue Amaia Lujambio74 N/A Chemicals, peptides, and recombinant proteins Geneticin (G418 sulfate) selective antibiotic Gibco Cat#11811031 ML-792 SUMO-Activating Enzyme inhibitor MedChemExpress Cat#HY-108702 Staurosporin (STS) Selleckchem Cat#S1421 Palbociclib, isethionate salt, >99% LC Laboratories Cat#P-7766 Glutathione-agarose, lyophilized powder Sigma-Aldrich Cat#G4510 Low density nickel-agarose beads ABT Cat#6BCL-QLNi-25 Protein A/G PLUS-Agarose Santa Cruz Biotechnology Cat#2003 Protein G Sepharose 4 Fast Flow Cytiva Cat#GE17-0618-01 PR-619 DUB inhibitor V Calbiochem Cat#662141 Complete mini EDTA-free protease inhibitor cocktail tablets Roche Cat#11836170001 Critical commercial assays Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Lachiondo-Ortega et al. Page 66 REAGENT or RESOURCE SOURCE IDENTIFIER QuickChange site-directed mutagenesis kit Stratagene Cat#200518 Lipofectamine 2000 transfection reagent Invitrogen Cat#11668019 DharmaFECT 1 transfection reagent Horizon Discovery Cat#T-2001-03 Maxwell 16 LEV RNA FFPE purification kit Promega Cat#AS1260 SYBR Select master mix Invitrogen Cat#4472908 TruSeq Stranded Total RNA human/mouse/rrat kit Illumina Cat#RS-122-2201 Annexin V FITC Apoptosis detection kit Immunostep Cat#ANXVKF LIVE/DEAD fixable blue dead cell stain kit Invitrogen Cat#L23105 Ac-DEVD-AFC caspase-3 fluorogenic substrate Enzo Life Sciences Cat#ALX-260-032 Senescence detection kit Calbiochem Cat#QIA117 MitoTracker Green FM dye Invitrogen Cat#M7514 MitoSOX Red mitochondrial superoxide indicator Invitrogen Cat#M36008 Tetramethylrhodamine, ethyl ester, perchlorate (TMRE) Invitrogen Cat#T669 ATPlite luminescence assay system Perkin Elmer Cat#6016943 NAD/NADH colorimetric assay kit Abcam Cat#ab65348 Fura-2, AM, cell permeant Invitrogen Cat#F1201 Deposited data RIP-Seq This paper GEO: GSE197798 Experimental models: Cell lines Human: THLE-2 cells ATCC CRL-2706 Human: PLC/PRF/5 cells ATCC CRL-8024 Human: HuH-7 cells JCRB Cell Bank JCRB0403 Human: HuH-7 WT HuR cells This paper N/A Human: HuH-7 HuR [K120R] cells This paper N/A Human: HuH-7 HuR [K182R] cells This paper N/A Human: HuH-7 HuR [K120/182R] cells This paper N/A Mouse: MLP-29 cells Enzo Medico119 N/A Experimental models: Organisms/strains Mouse: NOD SCID gamma Charles River Laboratories Strain code: 614 Oligonucleotides Primers for qPCR, See Table S2 This paper N/A siRNA targeting sequence: Mouse Pias2b Sense: 5’- CCUCCUAUGUUUUUGGAUAtt-3’ Antisense: 5’- UAUCCAAAAACAUAGGAGGac-3’ Thermo Fisher Scientific Custom siRNA targeting sequence: Mouse Senp1 Sense: 5’- AGAAAGGUGGGUUAACAAAtt-3’ Antisense: 5’- UUUGUUAACCCACCUUUCUca-3’ Thermo Fisher Scientific Cat#n383576 siRNA targeting sequence: Mouse Senp2 Sense: 5’- GAUUAGGUACUACAUCUUUtt-3’ Antisense: 5’- AAAGAUGUAGUACCUAAUCtt-3’ Thermo Fisher Scientific Cat#n251444 Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Lachiondo-Ortega et al. Page 67 REAGENT or RESOURCE SOURCE IDENTIFIER siRNA targeting sequence: Mouse Senp3 Sense: 5’- CCGACCCUCUCAUAGAAAAtt-3’ Antisense: 5’- UUUUCUAUGAGAGGGUCGGag-3’ Thermo Fisher Scientific Cat#s96092 Primer: HuR RRM1-2 K120R Forward: 5’- ACCATGACCCAGAGGGACGTAGAAGAC-3’ STAB vida Custom Primer: HuR RRM1-2 K120R Reverse: 5’ - GTCTTCTACGTCCCTCTGGGTCATGGT-3’ STAB vida Custom Primer: HuR RRM1-2 K182R Forward: 5’- CATCACAGTGAGGTTTGCAGCCAA-3’ STAB vida Custom Primer: HuR RRM1-2 K182R Reverse: 5’- GTTGGCTGCAAACCTCACTGTGATG-3’ STAB vida Custom Recombinant DNA Plasmid: pcDNA3-His6-Ub Manuel S Rodríguez120 N/A Plasmid: pcDNA3-His6-SUMO-1 Manuel S Rodríguez121,122 N/A Plasmid: pcDNA3-His6-SUMO-2 Manuel S Rodríguez121,122 N/A Plasmid: pcDNA3-His6-SUMO-3 Manuel S Rodríguez121,122 N/A Plasmid: pGEX-2T-Ubc9 Manuel S Rodríguez123 N/A Plasmid: pCMV-FLAG-hPIAS1 Manuel S Rodríguez124 N/A Plasmid: pCMV-FLAG-hPIASxαArora et al125 Addgene #15209 Plasmid: pCMV-FLAG-hPIASxβArora et al.125 Addgene #15210 Plasmid: pCMV-FLAG-hPIASγLiu etal.126 Addgene #15208 Plasmid: FLAG-hSENPI Manuel S Rodríguez124 N/A Plasmid: pFLAG-CMV-hSENP2 Kang et al.127 Addgene #18047 Plasmid: pcDNA3-RGS-hSENP3 Gong et al.128 Addgene #18048 Plasmid: pcDNA3-RGS-hSENP5 Gong et al.128 Addgene #18053 Plasmid: pFLAG-CMV-hSENP6 Dou etal.129 Addgene #18065 Plasmid: p3xFLAG-CMV-10-hSENP7 Bawa-Khalfe et al.130 Addgene #42886 Plasmid: pcDNA3.3-TOPO Invitrogen Cat# K830001 Plasmid: pcDNA3.3-TOPO-V5-mHuR WT This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K50R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K72R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR[ K89R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR[ K92R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K104R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K120R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K156R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K182R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K191R] This paper N/A Plasmid: pcDNA3.3-TOPO-V5-mHuR [K120/182R] This paper N/A Plasmid: pGEX-4T2-His6-HuR RRM1-2 WT Irene Diaz-Moreno131 N/A Plasmid: pGEX-4T2-His6-HuR RRM1-2 [K120R] This paper N/A Plasmid: pGEX-4T2-His6-HuR RRM1-2 [K182R] This paper N/A Cell Rep . Author manuscript; available in PMC 2024 April 18. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Lachiondo-Ortega et al. Page 68 REAGENT or RESOURCE SOURCE IDENTIFIER Plasmid: pGEX-4T2-His6-HuR RRM1-2 [K120/182R] This paper N/A Software and algorithms The Cancer Genome Atlas (TCGA) National Cancer Institute https://www.cancer.gov/tcga The Gene Set Enrichment Analysis (GSEA) UC San Diego and Broad Institute https://www.gsea-msigdb.org/ gsea/ Ingenuity Pathway Analysis (IPA) Qiagen https:// digitalinsights.qiagen.com/ PEAKS Studio Bioinformatics Solutions https://www.bioinfor.com/peaksstudio/ Perseus Max Planck Institute of Biochemistry https://www.maxquant.org/ perseus/ The Amber 16 Molecular Dynamics Package Case et al.132 https://ambermd.org/ GetAmber.php Origin 2018b OriginLab https://www.originlab.com/2018b Chimera UC San Francisco http://www.rbvi.ucsf.edu/chimera FlowJo version 10 BD Biosciences https://www.flowjo.com/ solutions/flowjo/ Microscopy Image Browser (MIB) version 2.7 University of Helsinki https://mib.helsinki.fi/ Agilent Seahorse Analytics Agilent https:// seahorseanalytics.agilent.com/ ImageJ National Institutes of Health https://imagej.net/software/fiji/ Prism version 8 GraphPad https://www.graphpad.com/ BioRender Science Suite Inc. https://www.biorender.com Cell Rep . Author manuscript; available in PMC 2024 April 18.