ORIGINAL RESEARCH published: 12 January 2021 doi: 10.3389/fcell.2020.613583 Extracellular Vesicles From Liver Progenitor Cells Downregulates Fibroblast Metabolic Activity and Increase the Expression of Immune-Response Related Molecules Felix Royo 1,2 , Mikel Azkargorta 1 , Jose L. Lavin 1† , Marc Clos-Garcia 1† , Ana R. Cortazar 1,3 , Monika Gonzalez-Lopez 1 , Laura Barcena 1 , Hernando A. del Portillo 4,5,6 , Edited by: María Yáñez-Mó 7 , Antonio Marcilla 8,9 , Francesc E. Borras 10,11,12 , Hector Peinado 13 , Anna Onnis, University of Siena, Italy Isabel Guerrero 14 , Mar Váles-Gómez 15 , Unai Cereijo 16 , Teresa Sardon 16 , Ana M. Aransay 1,2 , Felix Elortza 1,2 and Juan M. Falcon-Perez 1,2,17* Reviewed by: Alexander Grigoryevich 1 Center for Cooperative Research in Biosciences, Bizkaia Technology Park, Bizkaia, Spain, 2 Centro de Investigación Tonevitsky, Biomédica en Red de Enfermedades Hepáticas y Digestivas, Madrid, Spain, 3 Centro de Investigación Biomédica en Red National Research University Higher de Cáncer, Instituto de Salud Carlos III, Madrid, Spain, 4 ISGlobal, Hospital Clínic – Universitat de Barcelona, Barcelona, School of Economics, Russia Spain, 5 Health Sciences Research Institute Germans Trias i Pujol, Badalona, Spain, 6 Institució Catalana de Recerca i Olof Gidlöf, Estudis Avançats, Barcelona, Spain, 7 Departamento de Biología Molecular, Centro de Biología Molecular Severo Ochoa, Lund University, Sweden Instituto de Investigaciones Sanitarias la Princesa, Universidad Autónoma de Madrid, Madrid, Spain, 8 Àrea de Parasitologia, *Correspondence: Departament de Farmàcia i Tecnologia Farmacèutica i Parasitologia, Universitat de València, Valencia, Spain, 9 Joint Juan M. Falcon-Perez Research Unit on Endocrinology, Nutrition and Clinical Dietetics, Health Research Institute La Fe, Universitat de València, jfalcon@cicbiogune.es Valencia, Spain, 10 Department of Cell Biology, Physiology and Immunology, Autonomous University of Barcelona, Barcelona, Spain, 11 REMAR-IVECAT Group-“Germans Trias i Pujol” Health Science Research Institute (IGTP), Badalona, Spain, † Present address: 12 Nephrology Department-“Germans Trias i Pujol” University Hospital, Can Ruti Campus, Badalona, Spain, Jose L. Lavin, 13 Microenvironment and Metastasis Laboratory, Molecular Oncology Program, Spanish National Cancer Research Center, Applied Mathematics Department, Madrid, Spain, 14 Tissue and Organ Homeostasis, Centro de Biología Molecular “Severo Ochoa”, Universidad Autónoma Bioinformatics Unit NEIKER-(BRTA), de Madrid, Madrid, Spain, 15 Spanish National Centre for Biotechnology, Spanish National Research Council, Madrid, Spain, Bizkaia Technology Park, Bizkaia, 16 Anaxomics Biotech, Barcelona, Spain, 17 Ikerbasque, Basque Foundation for Science, Bilbao, Spain Spain Marc Clos-Garcia, Novo Nordisk Foundation Center Extracellular vesicles (EVs) mediate cell-to-cell crosstalk whose content can induce for Basic Metabolic Research, Faculty changes in acceptor cells and their microenvironment. MLP29 cells are mouse liver of Health and Medical Sciences, University of Copenhagen, progenitor cells that release EVs loaded with signaling cues that could affect cell Copenhagen, Denmark fate. In the current work, we incubated 3T3-L1 mouse fibroblasts with MLP29-derived Specialty section: EVs, and then analyzed changes by proteomics and transcriptomics. Results showed This article was submitted to a general downregulation of protein and transcript expression related to proliferative Membrane Traffic, and metabolic routes dependent on TGF-beta. We also observed an increase in the a section of the journal Frontiers in Cell and Developmental ERBB2 interacting protein (ERBIN) and Cxcl2, together with an induction of ribosome Biology biogenesis and interferon-related response molecules, suggesting the activation of Received: 02 October 2020 immune system signaling. Accepted: 07 December 2020 Published: 12 January 2021 Keywords: extracellular vesicles (EVs), exosomes, MLP29, fibroblast, cell crosstalk, immune response Frontiers in Cell and Developmental Biology | www.frontiersin.org 1 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity INTRODUCTION EVs Production and Purification MLP29 cells were plated in 150-mm dishes, at 10 million cells per Extracellular vesicles (EVs) play an important role in cell- dish. Cells were cultured in complete Dulbecco’s modified Eagle to-cell communication by interchanging messages between medium medium [DMEM supplemented with 10% (v/v) fetal cells through their cargo and surface proteins (Fatima and bovine serum (FBS), 0.1 mg/ml streptomycin, and 100 units/ml Nawaz, 2017). Among their cargo, there are bioactive molecules penicillin (GIBCO, Life Technologies Inc.)] for 24 h at 37◦ C capable of stimulating regenerative programs in damaged and 5% of CO2 . Then, cells were washed twice with Dulbecco’s tissues (Fatima et al., 2017). In that regard, different types modified phosphate-buffered saline (PBS) and incubated for of EVs playing a role in liver regeneration and with anti- 48 h (MLP29 cells) in 25 mM HEPES-containing complete fibrotic effect have been described. Mesenchymal stem cell DMEM medium [contaminating vesicles were first removed by (MSC)-derived EVs mediate tissue regeneration in liver fibrosis overnight centrifugation at 110,000 × g (Thery et al., 2006)]. (Katsuda and Ochiya, 2015; Damania et al., 2018). Moreover, After incubation, media were collected, and EVs were isolated they facilitate the recovery after ischemia in cardiac tissue, as previously described (Conde-Vancells et al., 2008). Briefly, shortening the wound healing time, and reducing scar formation culture supernatant was centrifuged at 1500 × g for 10 min to by modulating cellular migration, proliferation, and collagen remove lifted cells and cellular debris. The resultant supernatant synthesis (Zhao et al., 2015). was centrifuged at 10,000 × g and 100,000 × g for 30 and Liver stem cells constitute another population that secrete EVs 75 min, respectively. The resulting pellet was suspended in PBS, with regenerative effects. Liver stem cell-derived EVs ameliorated pooled, and again centrifuged at 100,000 × g for 75 min. The renal function and morphology (Herrera Sanchez et al., 2014) final pellet of EVs was suspended in PBS and stored at −80◦ C. and favor liver regeneration after hepatectomy in rats (Herrera Morphology, quantification, and the presence of typical EV et al., 2010) and after liver damage induced by non-alcoholic markers are presented in Supplementary Figure 1. As control, steatohepatitis (NASH; Bruno et al., 2020). They also show EV preparations were obtained from similar amount of media anti-tumoral effects (Brossa et al., 2020) and effectively reduce that have never been in contact with cells, to ensure that the liver injury during hypoxic conditions created by normothermic observed effect was not due to the media. machine liver perfusion associated to liver transplant, suggesting that liver stem cell-derived EVs can be employed to improve transplantation techniques (Rigo et al., 2018). EV Treatment of 3T3-L1 In the present work, we aimed to determine the effects of EVs Approximately 2 million 3T3-L1 cells were treated with the released by MLP29, a cell line with characteristics of progenitor EVs obtained from 200 million MLP29 cells. On average, this cell, isolated from fetal liver. During previous years, our group represents 20 µg of protein (measured by Bradford) contained has characterized the proteome (Conde-Vancells et al., 2008), in 8.9∗ E10 particles (measured with Nanosight LM10) per transcriptome (Royo et al., 2013), and metabolome (Royo et al., treatment. A total of six plates were treated, with three different 2019b) of the EVs released by this cell line. Compared to EVs preparations of EVs and three preparations of controls, for released by primary cultures of hepatocytes, MLP29-derived EVs 24 h, in 25 mM HEPES-containing complete DMEM medium are more resistant to mechanical stress (Royo et al., 2019b). (vesicles from tissue culture media were first removed by Moreover, they can circulate in the organism to reach different overnight centrifugation at 110,000 × g). At the end of the organs, including liver, lung, spleen, and brain, according to incubation, plates were washed twice with PBS, detached with studies of bio-distribution performed in vivo (Royo et al., 2019a). Tryple (GIBCO), and then divided equally for proteomic and Given those characteristics, we hypothesize that MLP29-derived transcriptomic analyses. An aliquot was employed to measure cell EVs are meant to have an effect over cellular parenchyma. To viability, which was higher than 90% in all the cases. Trypsin was characterize the nature of such effect, we have chosen the 3T3- removed after pelleting the cells 1000 × xg for 5 min, which were L1 fibroblasts, a cell line that can be differentiated into adipocytes frozen in dry ice and stored at −80◦ C until processing. and also has the ability to produce collagen and, therefore, mimic a fibrotic response (Li et al., 2019). We describe the crosstalk RNA Library and Sequencing observed between these cell lines, as it could give information Total RNA was extracted with RNeasy (Qiagen), including about the existing cell-to-cell communication in the organism by the step of DNase treatment for all the samples, and finally progenitor cells and fibroblasts. eluted in RNase-free water. The quantity and quality of the RNAs were evaluated using Qubit RNA HS Assay Kit (Life Technologies, Cat.# 32855) and Agilent RNA Nano Chips MATERIALS AND METHODS (Agilent Technologies, Cat.# 5067-1511), respectively. Sequencing libraries were prepared using “TruSeq Stranded Cell Lines mRNA LT sample prep kit” (Illumina Inc., Cat.# RS-122- All media and reagents for tissue culture were purchased from 2101 or RS-122-2102), following the “TruSeq Stranded mRNA R GIBCO (Life Technologies Inc.). MLP29 is a murine liver Sample Preparation LS Protocol (Part # 15031058 Rev. E). progenitor cell line (Medico et al., 1996). The embryonic murine In brief, starting from 1000 ng of total RNA, mRNA was fibroblast cell line 3T3-L1 was purchased from ATCC (CL-173) purified, fragmented, and primed for cDNA synthesis. cDNA first as a murine embryonic fibroblast cell line. strand was synthesized with SuperScript-II Reverse Transcriptase Frontiers in Cell and Developmental Biology | www.frontiersin.org 2 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity (Life Technologies, Cat.#18064-014) for 10 min at 25◦ C, 15 min specifically selected for CID. Peptides were excluded from further at 42◦ C, 15 min at 70◦ C, and paused at 4◦ C. cDNA second analysis during 60 s using the dynamic exclusion feature. strand was synthesized with Illumina reagents at 16◦ C for 1 h. Then, A-tailing and adaptor ligation were performed. Finally, Database Search enrichment of libraries was achieved by PCR (30 s at 98◦ C; 15 Database searching was performed using MASCOT 2.2.07 cycles of 10 s at 98◦ C, 30 s at 60◦ C, 30 s at 72◦ C; 5 min at 72◦ C (Matrix Science, London, United Kingdom) against a and paused at 4◦ C). Libraries were quantified and visualized on UNIPROT/Swissprot database filled only with entries an Agilent 2100 Bioanalyzer using Agilent High Sensitivity DNA corresponding to Homo sapiens (without isoforms). For kit (Agilent Technologies, Cat.# G2938-90320) and Qubit dsDNA protein identification, the following parameters were adopted: HS DNA Kit (Life Technologies, Cat.# 32851) and single-read carbamidomethylation of cysteines (C) as fixed modification, sequenced to obtain 30 million 50-nt reads. oxidation of methionines (M) as variable modifications, 10 ppm of peptide mass tolerance, 0.5 Da fragment mass tolerance, and Alignment and Quantification of up to two missed cleavage points, peptide charges of +2 and +3. Transcriptome Progenesis LC-MS Software Analysis for Label Free Quality Control of sequenced samples was performed by Differential Expression Analysis FASTQC software.1 Reads were mapped against the mouse Progenesis LC-MS (version 2.0.5556.29015, Non-linear (mm10) reference genome by STAR (Dobin et al., 2013) program Dynamics) was used for the label-free differential protein to account for spliced junctions. The resulting BAM alignment expression analysis. One of the runs was used as the reference files for the samples were then used to generate a table of raw to which the precursor masses in all other samples were aligned counts by Rsubread (Liao et al., 2013). Raw counts table was the to. Only features comprising charges of 2+ and 3+ were input for the differential expression (DE) analysis, carried out by selected. The raw abundances of each feature were automatically DESeq2 (Love et al., 2014), to detect differentially expressed genes normalized in a logarithmic scale against the reference run. among the different conditions. Samples were grouped in accordance to the comparison being performed, and ANOVA analyses were performed. A peak list Proteomic Analysis and Quantification containing the information of all the features was generated and In Solution Digestion exported to the Mascot search engine (Matrix Science Ltd.). Samples were digested following the filter-aided FASP protocol This file was searched against a UniProt/Swiss-Prot database described by Wisniewski et al. (2009) with minor modifications. under the conditions stated in the previous section, and the list Trypsin was added to a trypsin:protein ratio of 1:10, and the of identified peptides was imported back to Progenesis LC-MS. mixture was incubated overnight at 37◦ C, dried out in a RVC2 Protein quantitation was performed based on the three most 25 speedvac concentrator (Christ), and resuspended in 0.1% intense non-conflicting peptides (peptides occurring in only formic acid (FA). one protein), except for proteins with only two non-conflicting peptides. The significance of expression changes was tested at Liquid Chromatography protein level, and proteins identified with at least two peptides Liquid chromatography (LC) was performed using a and an ANOVA p value ≤ 0.05 were selected for further analyses. NanoAcquity nano-HPLC (Waters), equipped with a Waters BEH C18 nano-column (200 mm × 75 µm ID, 1.8 µm). A chromatographic ramp of 120 min [5–60% acetonitrile (ACN)] Annotation, Interaction, and Enrichment was used with a flow rate of 300 nl/min. Mobile phase A was Pathway Analysis water containing 0.1% v/v formic acid, while mobile phase B The list of transcripts and proteins was annotated using the was ACN containing 0.1% v/v FA; 0.5 µg of each sample were functional annotation [selecting GO Molecular Function (MF), loaded for each run. Biological Process (BP) Cellular Component (CC), KEGG, and REACTOME] using DAVID 6.82 (Huang da et al., 2009) and Mass Spectrometry later assigning them manually to different groups of interest Peptides were eluted directly into the LTQ Orbitrap XL according to keywords shown in Tables 1, 2, and avoiding mass spectrometer through a nanoelectrospray capillary source redundancy. Afterward, the list of gene/proteins was loaded (Proxeon Biosystems), at 300 nl/min and using a 120-min linear into Cytoscape v.3.8.03 (Shannon et al., 2003), and molecule gradient of 3–40% ACN, followed up by an increase to 40% interactions were retrieved from STRING App v.1.5.1 (Doncheva ACN for the next 30 min. The mass spectrometer automatically et al., 2019) with the confidence cutoff value of 0.4 and no switched between MS and MS/MS acquisition in DDA mode. Full addition of interacting molecules, against STRING experimental MS scan survey spectra (m/z 400–2,000) were acquired in the database, text-mining, and co-expression. Those nodes without Orbitrap with mass resolution of 30,000 at m/z 400. After each edges were discharged, and the rest were clustered using the survey scan, the six most intense ions above 1000 counts were MCODE algorithms provided with clusterMaker2 v.1.3.1 (Morris sequentially subjected to collision-induced dissociation (CID) in et al., 2011). Finally, the clusters were manually annotated based the linear ion trap. Precursors with charge states of 2 and 3 were 2 https://david.ncifcrf.gov/ 1 3 http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ https://cytoscape.org/ Frontiers in Cell and Developmental Biology | www.frontiersin.org 3 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity TABLE 1 | Transcripts regulated in 3T3-L1 cells after EV treatment, classified by TABLE 2 | Proteins regulated in 3T3-L1 cells after EV treatment, classified their involvement in different cellular processes, according to DAVID 6.8 annotation according to their functional annotation in cellular processes, employing DAVID of each transcript. 6.8 database for the annotation of each protein. Process Transcript Process Proteins Transcription DNA replication DNA binding Dbp, Gm14288, Gm14420, Ctcfl, Gm14440 Histone H4C1 Translation Chromosomal segregation ARL8A Proteasome Shfm1, Ube2c Transcription Ribosome Gm5148, Gm12191, Gm8430, Gm8430, Gm13826, Bc1 RNA splicing DCPS, DDX39A, EDC4, TARDBP, EFTUD2 Mitochondria Ppm1k Ribonuclease inhibitor ABCE1 Signaling Cytoskeleton MAPK Dusp6 Anchoring of the basal body PPP2R1A, RAB11A EGF Srf to the plasma membrane Chemokines Cxcl12 Cadherin binding involved in ANXA1, CNN2, ENO1, HSP90AB1, EIF5, cell–cell adhesion RACK1, SEPTIN2, SNX9, SPTBN1, SPTAN1 ITFN Ifit3, I830012O16Rik, Ifi27 Nuclear pore complex NUP155 HIF Serpine1 Protein transport for nuclear signal IPO7, KPNB1 NFKb Ptgs2 Actin filament formation ACTA2, CCT7 TGFb Peg10 Focal adhesion FLNC, CFL1, DSTN, TPM4 Cell adhesion Zyx, Tnc, Prc1 Plasmatic membrane repair MYOF, TNS3 Cell division Bub1, Foxm1,Spry2, Ccnb1 Remodeling cytoskeleton DPYSL3 Cytoskeleton Tubb6, Mical2 Trafficking AP2M1, COPB2 In red, upregulated transcripts; in black, downregulated genes. Phagosome RAB7A, TUBB4B, TUBB5, ATP6V0D1, ATP6V1B2, ATP6V1E1 on the enrichment groups obtained by functional enrichment Apoptosis PDIA3, LMNA analysis with STRING App v.1.5.1 (Doncheva et al., 2019) Calcium ion binding ANXA4, ANXA6, LRP1, RCN1 for each cluster (against GO Function, GO Components, GO Ion channel CLIC4 Processes, Reactome pathways, and UniProt keywords) as shown Protein metabolism in Figure 2. The functional pathway enrichment of Table 3 was Protein biosynthesis EIF3M, ETF1, EIF5A, GARS1, HARS1, LARS1 performed using the Statistical Enrichment Test performed by Proteasome PLAA, PSMC6, PSMC1, PSMA2, PSMA4, PSMB3, PSMD12, PSMD1 Panther v.15.04 (Mi et al., 2019). A Fisher exact test was used in Protein degradation PREP, UBA1, UBA3, PPIA, IPO7, CAND1, order to determine whether the proportion of genes considered ISG15, NEDD4 into the GO term or categories differed significantly between the Ribosomal complex HNRNPC, HNRNPM, EIF4A3, ILF3, PCBP2, dataset and the background [p adjusted (FDR method) < 0.05]. RPL23A, RPLP2, RPS19, ATP5F1A To determine the status of the pathway, fold change values Protein folding/viral secretion PPIA were provided, and the category set employed was the GO Signaling Biological Processes. Inhibitor Wnt G3BP1 Insulin ACACA, PRKAR2B, PRKACB, MAPK1, PCK2 RESULTS Negative regulation of NF-kappaB ERBIN Interferon IFI204 Transcriptomics Transcription inhibitor BASP1 The treatment of 3T3-L1 cells with MLP29-derived EVs induced Carbon metabolism ENO3, FH, PSAT1, SUCLA2, TKT a change in the expression of 45 transcripts above the Carbohydrate metabolism GAA, UGP2 significance threshold (see Figure 1A). Most of those changes Purine metabolism NME2, PFAS were downregulating the expression of genes involved in many Lipid metabolism ACSF2 active processes of the cell. In Table 1, we show the main cellular Mitochondrion VDAC1 processes affected by EV treatment, while in Supplementary Oxidation–reduction process AKR1B1, GLUD1, ETFB, GSTO1, MDH1, MDH2, NSDHL, PCYOX1, UGDH Table 1, we show the differential gene expression fold changes and p values. In addition, the data were deposited in the GEO In red, upregulated proteins; in black, downregulated ones. repository under the access number GSE157022. A total of 45 transcripts were found significantly altered in EV-treated fibroblasts, and among them, 9 transcripts were upregulated. and Gm19820 were the most downregulated. Furthermore, the We found that Cxcl12 and Ifit3 were the most upregulated most downregulated protein coding gene was Ptgs2, a gene transcripts after treatment, while the predicted genes Gm19592 involved in inflammation response. Some of the transcripts were assayed by qPCR, and we confirm the upregulation of Cxcl12 and 4 http://www.pantherdb.org/ Ifit3, as well as the downregulation of Tubb6 and the same trend Frontiers in Cell and Developmental Biology | www.frontiersin.org 4 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity TABLE 3 | Most significant enriched pathways according to Panther v.15.0 Enrichment Analysis (GO Biological Processes). Process GO Term Sign Molecules Negative regulation of transcription, GO:0045892 Up UBA3; Ifi27; Foxm1; Srf; BASP1; ILF3; NEDD4 DNA-templated Regulation of actin filament-based process GO:0032970 Down CNN2; DSTN; ACTA2; CFL1; LRP1; SPTAN1; SPTBN1; SNX9 Ribosome biogenesis GO:0042254 Up RPL23A; RPS19; ABCE1 Positive regulation of hydrolase activity GO:0051345 Down PLAA; EIF5; RACK1; SNX9; HSP90AB1 Cellular amino acid catabolic process GO:0009063 Down GLUD1; PCYOX1 Regulation of phospholipase A2 activity GO:0032429 Down ANXA1; PLAA; LRP1 Positive regulation of cell motility GO:2000147 Down ANXA1; Cxcl12; TKT; ACTA2; PLAA; CFL1; Spry2; LRP1 Positive regulation of innate immune response GO:0045089 Up RPS19; IFI204 Cellular catabolic process GO:0044248 Down ISG15; PCK2; GLUD1; EDC4; PSMD12; PLAA; EIF4A3; PSMC6; LRP1; Ube2c; ETFB; Dbp; PSMA4; GAA; PSMA2; PSMB3; RAB7A; ETF1; PSMD1; NEDD4; PCBP2; UBA1; GSTO1; PCYOX1; DCPS Regulation of phagocytosis GO:0050764 Down CNN2; LRP1; RACK1 Regulation of protein secretion GO:0050708 Down ANXA1; PCK2; GLUD1; PPIA; LRP1; ANXA4; SPTBN1 G protein-coupled receptor signaling pathway GO:0007186 Down ANXA1; LRP1; NME2; PRKACB Muscle structure development GO:0061061 Up Srf; MYOF; FLNC; LMNA Inflammatory response GO:0006954 Down ANXA1; PLAA; LRP1; Ptgs2 Regulation of focal adhesion assembly GO:0051893 Down CFL1; LRP1 The software defines that each pathway was upregulated (red) or downregulated (black) according to the fold change values of each gene/protein group and their function. For the molecules’ annotation, red color indicates overexpression in fibroblast incubated with EVs, and black denotes downregulation with respect to untreated fibroblasts. FIGURE 1 | Volcano plot representing the changes in 3T3-L1 cells after MLP29-derived EV treatment. The graphs show fold changes (log2 transformed) versus p values (-log10 transformed). Significant regulated transcripts (A) and proteins (B) are represented as black dots in red circles. A gene or protein was considered regulated when its DE resulted in an adjusted p value < 0.05 (1.3 after −log10 transformation). Each experiment was performed in triplicate. for Peg10, although in this case, the result was not significant expression with fold changes and p values. The expression of (Supplementary Figure 3A). 107 proteins was affected by the EV treatment, and among them, 16 were upregulated, and the rest were downregulated Proteomics after the treatment. The protein most over-represented was The treatment of 3T3-L1 cells with MLP29-derived EVs ERBIN, a protein that binds to the unphosphorylated form of the induced a change in the expression of different proteins ERBB2 protein and regulates ERBB2 function and localization above the threshold of significance (see Figure 1B). Most of (Tao et al., 2014). The second most upregulated protein was those changes were a protein expression downregulation. In ILF3, a protein with RNA binding capability involved in the Table 2, we show the annotation of proteins modified, while formation of the ribosomal complex (Higuchi et al., 2012). in Supplementary Table 2, we show the differential protein Among the most downregulated proteins we found was RACK1, Frontiers in Cell and Developmental Biology | www.frontiersin.org 5 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity a scaffolding protein with participation in multiple functions To further explore the relationship between regulated such phagocytosis and cell-to-cell adhesion processes (Thorslund proteins/genes, we loaded all regulated molecules in the et al., 2011). The second most downregulated protein was Cytoscape v.3.8.0 software (Shannon et al., 2003). We retrieved MDH2, a mitochondrial malate dehydrogenase involved in ox– protein interactions from STRING App v.1.5.1 (Doncheva et al., redox processes (Minarik et al., 2002). We should mention that 2019) as described in the section “Materials and Methods,” and regarding fold change value, the most downregulated protein we discovered clusters of interacting molecules employing the was GLUD1, a mitochondrial glutamate dehydrogenase also MCODE algorithm from the clusterMaker2 App v.1.3.1 (Morris involved in catabolism and TCA cycle (Tzimagiorgis et al., et al., 2011). Non-connected proteins were eliminated, and the 1993;Coloff et al., 2016). annotations of the different clusters were decided according to At a molecular level, we did not observe any overlapping the STRING functional enrichment annotations retrieved for between regulated genes and proteins. However, regarding the each group of molecules, as well as protein individual annotation. process affected, we encountered some similarities between the Figure 2 illustrates the resulting interaction clusters, where routes affected according to transcriptomic and proteomic data. the ribosome complex and the proteasome represent two well- In both cases, we observed downregulation for the machinery defined groups. The machinery for mRNA splicing and different related to control of cytoskeleton and cell motility. We also building blocks of the cytoskeleton appear as sub-groups within observed downregulation of major signaling cascades (i.e., the largest cluster. Analysis in this interaction cluster showed MAPK1, NFKb, TGFb, and EGF) in both types of analyses, that Cxcl2-Ptgs2 were related to regulation of immune activity while some interferon-related molecules were upregulated. We and connected to other molecules related to cell migration. had performed a qPCR to see if the regulation of the observed Other important groups were connected to cell cycling and the proteins may reflect on the transcriptome, and we have observed metabolism of carboxyl groups. the increase of Myof transcript, as well as the downregulation of Regarding the functional meaning of the observed changes, Plaa and Mdh1 transcripts, according to what was observed in the the network presented in Figure 2 showed an increase in the proteomics. However, this agreement was not clearly observed presence of ribosome machinery, metabolism of fatty acids, and for Ccnb1 or Mapk1 (Supplementary Figure 3A). Interestingly, molecules related to interferon signaling. However, most of the when 3T3 cells were treated with the vesicular fraction of the cellular processes seemed to be inhibited, such as mRNA splicing, SEC-purified preparation of EVs (see Supplementary Materials carbon metabolism, cell migration, the proteasome, and some and Methods), the downregulation of these last transcripts was molecules of the cytoskeleton. There was also a shortage in observed (Supplementary Figure 3B), suggesting that some molecules involved in ATP hydrolysis. According to the results responses depend on the purity of the EV preparation, and to shown on the individual enrichment analysis, the interaction clearly visualize, one needs to remove protein aggregates that network suggested a slowdown of major cell activity, despite an could unmask the response. increase of translation. We have observed as well a decrease in the metabolic activity measured by the MTT assay (Supplementary Enrichment Analysis and Protein Materials and Methods and Supplementary Figure 2A) that Networks could be explained by the downregulation described by the To get a better insight of the general effect achieved by regulations enrichment analysis. of genes and proteins, in this section, the corresponding analysis was performed considering both types of molecules as a single set. As described in the “Materials and Methods” section, the DISCUSSION enrichment analysis performed with the Panther v.15.0 platform5 (Mi et al., 2019) takes into account both the gene/protein In this work, we characterized the main effects produced by regulated, as well as the direction of such regulation, and EVs derived from MLP29 liver progenitor cells in 3T3-L1 accordingly, it calculates the probability for the pathways stored fibroblasts. The MLP29 cell line is an epithelial homogeneous in the Gene Ontology Biological Processes database to be up- cellular clone obtained by limiting dilution from a mouse or downregulated. We performed this analysis with regulated embryonic liver cell line (Medico et al., 1996; Muller et al., genes/proteins, and the result is displayed in Table 3. The 2002). This cell line responds to HGF with the whole array list includes a selection of pathways significantly regulated (p- of its biological effects: scattering, survival, proliferation, and adjusted <0.05), manually curated for redundancy and for lack of tubular morphogenesis (Medico et al., 1996). Moreover, their cellular context. The results indicate a reduction in genes related EVs carry a transcriptome enriched in signals regulating cell to transcriptional activity in fibroblasts treated with MLP29- fate, proliferation, and cell organization (Royo et al., 2013). derived EVs and upregulation of genes related to ribosome Accordingly, we analyzed the main changes in transcriptome and assembly, when compared to untreated cells. Interestingly, we proteome in 3T3-L1 fibroblasts treated with MLP29-derived EVs. also observed the upregulation in genes related to innate We should remark that purification of EVs through differential immunity. Some of the pathways that are downregulated centrifugation could carry over other non-vesicular elements, included genes involved in migration, catabolism, inflammation, which may also be responsible for the observed response, or phagocytosis. the reason why we have compared and further purified the preparation by SEC and observed a similar response to the 5 http://www.pantherdb.org/ fraction corresponding to vesicles. These results suggest that the Frontiers in Cell and Developmental Biology | www.frontiersin.org 6 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity FIGURE 2 | Clusters of protein interaction networks generated within the Cytoscape v.3.8.0 software environment. Red and blue color of the molecules denotes up- and downregulation, respectively. Annotation of the clusters was done manually according to functional enrichment generated with STRING App v.1.5. observed effects are mainly due the vesicular component of our that disrupts the association between PAK2 and GTP-bound preparations (Supplementary Figures 2B, 3B). CDC42/RAC1. In fibroblasts, TGF-β activates p21-activated We observed that treatment with MLP29-derived EVs kinase 2 (PACK2), which was activated in response to TGF-β in promoted a decrease in pathways related to catabolic activity, different fibroblast cell lines, but not in epithelial cell lines (Wilkes cell proliferation, and cytoskeleton remodeling. It should be et al., 2003; Wilkes et al., 2009). ERBIN also leads to deactivation highlighted that the protein most differentially expressed in of the ERK signal pathway, which is one of the main survival treated cells is ERBIN, an ERBB2 interaction protein blocking signaling pathways downstream HER2 (Yu et al., 2013). ERK activation (Huang et al., 2003). ERBIN is an intracellular We have also observed anti-apoptotic responses, such as the protein of the LAP family (LRR and PDZ), containing an LRR increase of the cytoskeleton protein myoferlin (MYOF), whose domain and a PDZ domain (Borg et al., 2000; Kolch, 2003), and it loss in fibroblasts is associated with a decrease in the expression is believed to play a role in basolateral targeting in epithelial cells of tight junction molecules and an increment in the number (Dillon et al., 2002). The mechanism for inhibition of the route of of cells positive for apoptotic markers (Leung et al., 2012). TGF-β involves the formation of the ERBIN–MERLIN complex Taking into account that cells did not decrease their viability, Frontiers in Cell and Developmental Biology | www.frontiersin.org 7 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity the inhibitory responses observed do not lead to cell death. This DATA AVAILABILITY STATEMENT agrees with previous studies of stem cell-derived EVs, showing the inhibition of fibrosis progression in models of kidney injury The datasets generated for this study can be found the NCBI GEO (Grange et al., 2019). Together, our results indicate that general acession GSE157022. cellular activity was slowed down in EV-treated compared to untreated fibroblasts, since transcription, cell motility, protein catabolism, and carbon metabolism were downregulated. Some AUTHOR CONTRIBUTIONS enzymes with ATP hydrolytic activity were also downregulated. Consistently, we also detect reduction in metabolic activity HAP, MY-M, AM, HP, IG, FB, MV-G, and JF-P originally measured by an MTT assay. conceived and designed the study. FR, MA, MG-L, and At the same time, there was an increase of signals related LB performed the experiments. JL and AM performed the with immune system activation, centered in the activation of the transcriptomics analysis. MA and FE performed the proteomics transcription of the cytokine Cxcl12, a potent chemoattractant for analysis. FR, MA, JL, MC-G, AC, UC, and TS performed hematopoietic cells. This molecule has been associated to fibrosis the enrichment analysis and data visualization. All authors after radiation injury (Cao et al., 2019) and also to pro-metastatic participated in data interpretation, discussion, and manuscript stimulus of colon cancer cells through the PI3K/AKT/mTOR drafting and editing. signaling pathway (Ma et al., 2017). However, in massive liver injury models where oval cell repair is involved, there is an upregulation of Cxcl12 that interacts with oval cells because this FUNDING cell type expresses CXCR4, the only known receptor for CXCL12 (Hatch et al., 2002). This work was supported by GEIVEX, Network of Excellence The idea of a damage response is also reinforced by in the Research and Innovation on Exosomes REDIEX, the observation of the increase of pathways related to SAF2015-71231-REDT, Network of Excellence in Translational ribosome biogenesis. These phenomena have been described NeTwork for the CLinical application of Extracellular VesicleS to be associated to dsDNA sensing, a response to restrict (TeNTaCLES), and RED2018-102411-T. The review is supported virus reproduction and inflammatory response regulation by Spanish Ministry of Science and Innovation, within the (Bianco and Mohr, 2019). In this context, it is important to national Plan RTI2018-094969-B-I00, and Excellence Severo mention that MLP29-derived EVs contain dsDNA compatible Ochoa grant Innovative Research Grant (SEV-2016-0644). with apoptotic features (Kruglik et al., 2019). Ribosome accumulation may also coordinate IFNB1 production, a pathway increased in fibroblasts that received MLP29-derived EVs. The ACKNOWLEDGMENTS upregulated molecules associated with the interferon pathway We would like to acknowledge all the participants in the were Ifit3 and IFI204 (Bianco and Mohr, 2019). To conform Practical Workshop Bioinformatics Tools to study exosomes’ to the hypothesis of ribosome accumulation, we measure effects, organized by GEIVEX and REDIEX at CIC bioGUNE the changes of transcriptional activity of 45S pre-ribosomal in November 2017. RNA, observing a non-significant increase in cells treated by EVs by qPCR, and almost significant (p value 0.07) in cells treated with the vesicular fraction of our preparations SUPPLEMENTARY MATERIAL (Supplementary Figure 3). All the changes observed suggests that MLP29-derived EVs The Supplementary Material for this article can be found promoted somehow signals related to tissue damage in 3T3- online at: https://www.frontiersin.org/articles/10.3389/fcell.2020. L1 fibroblasts that may be partially explained by the presence 613583/full#supplementary-material of dsDNA (Bianco and Mohr, 2019). The effects inferred from changes in proteins and genes indicate a slowdown Supplementary Figure 1 | Characterization of EVs derived from MLP29. in some general cell activities including cell motility and Supplementary Figure 2 | Measure of metabolic activity by MTT assay in 3T3.L1 cytoskeleton remodeling, transcription, and protein degradation cells after EV treatment. machinery, carbon metabolism and TCA processes, and the Supplementary Figure 3 | Changes in the transcription of 3T3.L1 cells induced upregulation of molecules related to innate immune response. by EV treatment measure by qPCR. Moreover, the results offer a new confirmation of the role Supplementary Table 1 | Significantly regulated transcripts. of EVs in the crosstalk between liver stem and stromal cells in the liver. Supplementary Table 2 | Significantly regulated proteins. REFERENCES Borg, J. P., Marchetto, S., Le Bivic, A., Ollendorff, V., Jaulin-Bastard, F., Saito, H., et al. (2000). ERBIN: a basolateral PDZ protein that interacts with the Bianco, C., and Mohr, I. (2019). Ribosome biogenesis restricts innate immune mammalian ERBB2/HER2 receptor. Nat. Cell Biol. 2, 407–414. doi: 10.1038/ responses to virus infection and DNA. eLife 8:e49551. 35017038 Frontiers in Cell and Developmental Biology | www.frontiersin.org 8 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity Brossa, A., Fonsato, V., Grange, C., Tritta, S., Tapparo, M., Calvetti, R., et al. (2020). Katsuda, T., and Ochiya, T. (2015). Molecular signatures of mesenchymal stem Extracellular vesicles from human liver stem cells inhibit renal cancer stem cell-derived extracellular vesicle-mediated tissue repair. Stem Cell Res. Ther. cell-derived tumor growth in vitro and in vivo. Int. J. Cancer 147, 1694–1706. 6:212. doi: 10.1002/ijc.32925 Kolch, W. (2003). Erbin: sorting out ErbB2 receptors or giving Ras a break? Sci Bruno, S., Pasquino, C., Herrera Sanchez, M. B., Tapparo, M., Figliolini, F., Grange, .STKE 2003:e37. C., et al. (2020). HLSC-derived extracellular vesicles attenuate liver fibrosis and Kruglik, S. G., Royo, F., Guigner, J. M., Palomo, L., Seksek, O., Turpin, P. Y., inflammation in a murine model of non-alcoholic steatohepatitis. Mol. Ther. et al. (2019). Raman tweezers microspectroscopy of circa 100 nm extracellular 28, 479–489. doi: 10.1016/j.ymthe.2019.10.016 vesicles. Nanoscale 11, 1661–1679. doi: 10.1039/c8nr04677h Cao, J., Zhu, W., Yu, D., Pan, L., Zhong, L., Xiao, Y., et al. (2019). The involvement Leung, C., Shaheen, F., Bernatchez, P., and Hackett, T. L. (2012). Expression of SDF-1alpha/CXCR4 Axis in radiation-induced acute injury and fibrosis of of myoferlin in human airway epithelium and its role in cell adhesion and skin. Radiat. Res. 192, 410–421. doi: 10.1667/rr15384.1 zonula occludens-1 expression. PLoS One 7:e40478. doi: 10.1371/journal.pone. Coloff, J. L., Murphy, J. P., Braun, C. R., Harris, I. S., Shelton, L. M., Kami, K., 0040478 et al. (2016). Differential glutamate metabolism in proliferating and quiescent Li, X. P., Liu, P., Li, Y. F., Zhang, G. L., Zeng, D. S., and Liu, D. L. (2019). LPS mammary epithelial cells. Cell Metab. 23, 867–880. doi: 10.1016/j.cmet.2016. induces activation of the TLR4 pathway in fibroblasts and promotes skin scar 03.016 formation through collagen I and TGF-beta in skin lesions. Int. J. Clin. Exp. Conde-Vancells, J., Rodriguez-Suarez, E., Embade, N., Gil, D., Matthiesen, R., Valle, Pathol. 12, 2121–2129. M., et al. (2008). Characterization and comprehensive proteome profiling of Liao, Y., Smyth, G. K., and Shi, W. (2013). The Subread aligner: fast, accurate exosomes secreted by hepatocytes. J. Proteome Res. 7, 5157–5166. doi: 10.1021/ and scalable read mapping by seed-and-vote. Nucleic Acids Res. 41:e108. doi: pr8004887 10.1093/nar/gkt214 Damania, A., Jaiman, D., Teotia, A. K., and Kumar, A. (2018). Mesenchymal Love, M. I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change stromal cell-derived exosome-rich fractionated secretome confers a and dispersion for RNA-seq data with DESeq2. Genome Biol. 15:550. hepatoprotective effect in liver injury. Stem Cell Res. Ther. 9:31. Ma, J. C., Sun, X. W., Su, H., Chen, Q., Guo, T. K., Li, Y., et al. (2017). Fibroblast- Dillon, C., Creer, A., Kerr, K., Kumin, A., and Dickson, C. (2002). Basolateral derived CXCL12/SDF-1alpha promotes CXCL6 secretion and co-operatively targeting of ERBB2 is dependent on a novel bipartite juxtamembrane enhances metastatic potential through the PI3K/Akt/mTOR pathway in colon sorting signal but independent of the C-terminal ERBIN-binding domain. cancer. World J. Gastroenterol. 23, 5167–5178. doi: 10.3748/wjg.v23.i28. Mol. Cell Biol. 22, 6553–6563. doi: 10.1128/mcb.22.18.6553-6563. 5167 2002 Medico, E., Mongiovi, A. M., Huff, J., Jelinek, M. A., Follenzi, A., Gaudino, G., Dobin, A., Davis, C. A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., et al. et al. (1996). The tyrosine kinase receptors Ron and Sea control “scattering” (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21. and morphogenesis of liver progenitor cells in vitro. Mol. Biol. Cell 7, 495–504. doi: 10.1093/bioinformatics/bts635 doi: 10.1091/mbc.7.4.495 Doncheva, N. T., Morris, J. H., Gorodkin, J., and Jensen, L. J. (2019). Cytoscape Mi, H., Muruganujan, A., Ebert, D., Huang, X., and Thomas, P. D. stringapp: network analysis and visualization of proteomics data. J. Proteome (2019). PANTHER version 14: more genomes, a new PANTHER GO-slim Res. 18, 623–632. doi: 10.1021/acs.jproteome.8b00702 and improvements in enrichment analysis tools. Nucleic Acids Res. 47, Fatima, F., Ekstrom, K., Nazarenko, I., Maugeri, M., Valadi, H., Hill, A. F., et al. D419–D426. (2017). Non-coding RNAs in mesenchymal stem cell-derived extracellular Minarik, P., Tomaskova, N., Kollarova, M., and Antalik, M. (2002). Malate vesicles: deciphering regulatory roles in stem cell potency. Inflammatory dehydrogenases–structure and function. Gen. Physiol. Biophys. 21, resolve, and tissue regeneration. Front. Genet. 8:161. doi: 10.3389/fgene.2017. 257–265. 00161 Morris, J. H., Apeltsin, L., Newman, A. M., Baumbach, J., Wittkop, T., Su, G., et al. Fatima, F., and Nawaz, M. (2017). Long distance metabolic regulation through (2011). clusterMaker: a multi-algorithm clustering plugin for Cytoscape. BMC adipose-derived circulating exosomal miRNAs: a trail for RNA-Based therapies? Bioinformatics 12:436. doi: 10.1186/1471-2105-12-436 Front. Physiol. 8:545. doi: 10.3389/fphys.2017.00545 Muller, M., Morotti, A., and Ponzetto, C. (2002). Activation of NF-kappaB Grange, C., Tritta, S., Tapparo, M., Cedrino, M., Tetta, C., Camussi, G., et al. (2019). is essential for hepatocyte growth factor-mediated proliferation and Stem cell-derived extracellular vesicles inhibit and revert fibrosis progression in tubulogenesis. Mol. Cell Biol. 22, 1060–1072. doi: 10.1128/mcb.22.4.1060-1072. a mouse model of diabetic nephropathy. Sci. Rep. 9:4468. 2002 Hatch, H. M., Zheng, D., Jorgensen, M. L., and Petersen, B. E. (2002). SDF- Rigo, F., De Stefano, N., Navarro-Tableros, V., David, E., Rizza, G., Catalano, G., 1alpha/CXCR4: a mechanism for hepatic oval cell activation and bone marrow et al. (2018). Extracellular vesicles from human liver stem cells reduce injury in stem cell recruitment to the injured liver of rats. Cloning Stem Cells 4, 339–351. an Ex Vivo normothermic hypoxic rat liver perfusion model. Transplantation doi: 10.1089/153623002321025014 102, e205–e210. Herrera, M. B., Fonsato, V., Gatti, S., Deregibus, M. C., Sordi, A., Cantarella, D., Royo, F., Cossio, U., Ruiz De Angulo, A., Llop, J., and Falcon-Perez, J. M. et al. (2010). Human liver stem cell-derived microvesicles accelerate hepatic (2019a). Modification of the glycosylation of extracellular vesicles alters regeneration in hepatectomized rats. J. Cell Mol. Med. 14, 1605–1618. doi: their biodistribution in mice. Nanoscale 11, 1531–1537. doi: 10.1039/c8nr0 10.1111/j.1582-4934.2009.00860.x 3900c Herrera Sanchez, M. B., Bruno, S., Grange, C., Tapparo, M., Cantaluppi, V., Tetta, Royo, F., Gil-Carton, D., Gonzalez, E., Mleczko, J., Palomo, L., Perez-Cormenzana, C., et al. (2014). Human liver stem cells and derived extracellular vesicles M., et al. (2019b). Differences in the metabolite composition and mechanical improve recovery in a murine model of acute kidney injury. Stem Cell Res. Ther. properties of extracellular vesicles secreted by hepatic cellular models. 5:124. doi: 10.1186/scrt514 J. Extracell Vesicles 8:1575678. doi: 10.1080/20013078.2019.1575678 Higuchi, T., Sakamoto, S., Kakinuma, Y., Kai, S., Yagyu, K., Todaka, H., et al. (2012). Royo, F., Schlangen, K., Palomo, L., Gonzalez, E., Conde-Vancells, J., Berisa, A., High expression of nuclear factor 90 (NF90) leads to mitochondrial degradation et al. (2013). Transcriptome of extracellular vesicles released by hepatocytes. in skeletal and cardiac muscles. PLoS One 7:e43340. doi: 10.1371/journal.pone. PLoS One 8:e68693. doi: 10.1371/journal.pone.0068693 0043340 Shannon, P., Markiel, A., Ozier, O., Baliga, N. S., Wang, J. T., Ramage, D., Huang, Y. Z., Zang, M., Xiong, W. C., Luo, Z., and Mei, L. (2003). Erbin suppresses et al. (2003). Cytoscape: a software environment for integrated models of the MAP kinase pathway. J. Biol. Chem. 278, 1108–1114. doi: 10.1074/jbc. biomolecular interaction networks. Genome Res. 13, 2498–2504. doi: 10.1101/ m205413200 gr.1239303 Huang da, W., Sherman, B. T., and Lempicki, R. A. (2009). Systematic and Tao, Y., Shen, C., Luo, S., Traore, W., Marchetto, S., Santoni, M. J., et al. (2014). integrative analysis of large gene lists using DAVID bioinformatics resources. Role of Erbin in ErbB2-dependent breast tumor growth. Proc. Natl. Acad. Sci. Nat. Protoc. 4, 44–57. doi: 10.1038/nprot.2008.211 U.S.A. 111, E4429–E4438. Frontiers in Cell and Developmental Biology | www.frontiersin.org 9 January 2021 | Volume 8 | Article 613583 Royo et al. Extracellular Vesicles Regulate Metabolic Activity Thery, C., Amigorena, S., Raposo, G., and Clayton, A. (2006). Isolation and Zhao, Y., Sun, X., Cao, W., Ma, J., Sun, L., Qian, H., et al. (2015). Exosomes derived characterization of exosomes from cell culture supernatants and biological from human umbilical cord mesenchymal stem cells relieve acute myocardial fluids. Curr. Protoc. Cell Biol. Chapter 3:Unit 3.22. ischemic injury. Stem Cells Int. 2015:761643. Thorslund, S. E., Edgren, T., Pettersson, J., Nordfelth, R., Sellin, M. E., Ivanova, E., et al. (2011). The RACK1 signaling scaffold protein selectively interacts Conflict of Interest: The authors declare that the research was conducted in the with Yersinia pseudotuberculosis virulence function. PLoS One 6:e16784. doi: absence of any commercial or financial relationships that could be construed as a 10.1371/journal.pone.0016784 potential conflict of interest. Tzimagiorgis, G., Leversha, M. A., Chroniary, K., Goulielmos, G., Sargent, C. A., Ferguson-Smith, M., et al. (1993). Structure and expression analysis of a Citation: Royo F, Azkargorta M, Lavin JL, Clos-Garcia M, Cortazar AR, Gonzalez- member of the human glutamate dehydrogenase (GLUD) gene family mapped Lopez M, Barcena L, del Portillo HA, Yáñez-Mó M, Marcilla A, Borras FE, to chromosome 10p11.2. Hum. Genet. 91, 433–438. Peinado H, Guerrero I, Váles-Gómez M, Cereijo U, Sardon T, Aransay AM, Wilkes, M. C., Murphy, S. J., Garamszegi, N., and Leof, E. B. (2003). Cell-type- Elortza F and Falcon-Perez JM (2021) Extracellular Vesicles From Liver specific activation of PAK2 by transforming growth factor beta independent of Progenitor Cells Downregulates Fibroblast Metabolic Activity and Increase the Smad2 and Smad3. Mol. Cell Biol. 23, 8878–8889. doi: 10.1128/mcb.23.23.8878- Expression of Immune-Response Related Molecules. Front. Cell Dev. Biol. 8:613583. 8889.2003 doi: 10.3389/fcell.2020.613583 Wilkes, M. C., Repellin, C. E., Hong, M., Bracamonte, M., Penheiter, S. G., Borg, J. P., et al. (2009). Erbin and the NF2 tumor suppressor Merlin cooperatively Copyright © 2021 Royo, Azkargorta, Lavin, Clos-Garcia, Cortazar, Gonzalez-Lopez, regulate cell-type-specific activation of PAK2 by TGF-beta. Dev. Cell 16, 433– Barcena, del Portillo, Yáñez-Mó, Marcilla, Borras, Peinado, Guerrero, Váles-Gómez, 444. doi: 10.1016/j.devcel.2009.01.009 Cereijo, Sardon, Aransay, Elortza and Falcon-Perez. This is an open-access article Wisniewski, J. R., Zougman, A., Nagaraj, N., and Mann, M. (2009). Universal distributed under the terms of the Creative Commons Attribution License (CC BY). sample preparation method for proteome analysis. Nat. Methods 6, 359–362. The use, distribution or reproduction in other forums is permitted, provided the Yu, D., Li, M., Ni, B., Kong, J., and Zhang, Z. (2013). Induction of neuronal original author(s) and the copyright owner(s) are credited and that the original mitophagy in acute spinal cord injury in rats. Neurotox Res. 24, 512–522. doi: publication in this journal is cited, in accordance with accepted academic practice. No 10.1007/s12640-013-9397-0 use, distribution or reproduction is permitted which does not comply with these terms. Frontiers in Cell and Developmental Biology | www.frontiersin.org 10 January 2021 | Volume 8 | Article 613583