The six mycotoxins AFB1, OTA, CIT, DON, NIV, and Pen A were applied to the human hepatoblastoma cell line HepG2 and CIT, OTA and a combination of both on the human kidney epithelial cell line IHKE in sub-cytotoxic concentrations. The results are described based on two evaluation methods both assisted by STRING DB.

First, bubble enrichment plots display the significant enrichment of protein groups with a common biological function to the down-, upregulated or both sides of a whole proteome (comparable to gene set enrichment analysis, for detailed information see chapter Interaction analysis). In this case, the biological function is described as terms of Gene Ontology (GO) annotations, which contain biological processes (BPs), cellular components (CCs) and molecular functions (MFs) as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) terms, which contain important signaling and metabolic pathways. The bubble plots presented here show only the strongest effects (highest enrichment scores, ESs) of each experiment on the cellular proteomes. The ES describes how distant the specific, term-associated proteins are from the middle of the proteome. In other words, the ES characterizes the intensity of a deregulation of a certain biological function, expressed in the form of up- or downregulated proteins. It should be noted that ESs for bubble plots are calculated individually for each experiment (see chapter Interaction analysis) and for this reason, the scale of the x-axis is not comparable. In addition, FDR scale and size of bubbles, representing the identified percentage of a term, are individual per plot. However, all shown effects have an FDR < 0.01.

Second, functional enrichments within the group of DAPs from either side were identified and taken into account for interpretation. In this case, not the whole proteome data were used, but only the interactions between significantly altered proteins from either side were analyzed—regardless of their fold change. These results are shown in the Online Resources 3 (HepG2) and 4 (IHKE).

All proteomic effects of the investigated mycotoxins are provided in Online Resources 3 and 4—individually per protein and as enriched terms by STRING analyses. Additional to enriched terms, changes in the abundance of individual proteins were considered. From all these effects, the main cellular targets of the mycotoxins’ toxicity were derived and potential MoA were illustrated.

The overall number of altered proteins in the HepG2 proteome varied according to the incubated mycotoxin. In terms of amount of DAPs, DON had the strongest effect, as it induced significant changes for 17% of all proteins (see Online Resource 1, Figure S2), according to the Perseus analysis (see chapter Statistical analysis). OTA and NIV had comparable effects, with 6.7% and 5.3%, respectively. In comparison, Pen A, CIT and AFB1 showed fewer significant alterations, with 1.2%, 0.80% and 0.53%, respectively, but the overall effect of AFB1 was increased by the pretreatment of HepG2 cells with β-NF to 3.2%. In IHKE cells, OTA also had a stronger overall effect on the proteome, with 14% DAPs, than CIT, with 0.62%. Their combination led to 2.3% of DAPs in the IHKE proteome. The following sections provide an overview of the effects caused by the individual mycotoxins.

HepG2 cells were treated with 200 nM OTA for 24 h (Fig. 1, left). Strongest effects were observed in the upregulation of the “MCMFootnote 1 complex”, which is a heterohexamer controlling DNA replication in the late M to early G1 phase of the cell cycle (Lei 2005). The “CMGFootnote 2 complex” showed the same values as the “MCM complex” of ES 2.43 and an FDR of 0.0074. These two terms include the same six modulated proteins (MCM2–MCM7), but three CMG-specific proteins were not identified. These results were supported by enriched terms associated with DNA replication and the MCM complex itself in the group of upregulated DAPs. The deregulation of the MCM complex is associated with the development of hepatocellular carcinoma and genomic instability (Lei et al. 2021). MCM plays a key role in DNA replication—replicative stress is discussed as a potential cause for genotoxic properties of OTA. However, the underlying mechanism remains unclear (EFSA Panel on Contaminants in the Food Chain (CONTAM) et al. 2020; Klotz et al. 2022). Our results support the mechanism of DNA replication as a relevant target of OTA toxicity, as the upregulation of the MCM complex was the strongest enriched term. Potentially, cells counter-regulate the inhibitory effect of OTA on DNA replication by upregulation of the MCM complex.

Enrichment bubble plots of the top 15 enriched terms in the whole proteome of HepG2 cells treated with ochratoxin A (OTA, 200 nM) and citrinin (CIT, 20 µM) for 24 h. Effects on cellular components (CC), biological processes (BP), molecular functions (MF, all from Gene Ontology) and terms of the Kyoto Encyclopedia of Genes and Genomes (KEGG) are shown. The color scale describes the significance as the false discovery rate (FDR) and the bubble size scales with the percentage of identified proteins affiliated to the described term. Terms are annotated by + for upregulation and * for downregulation. Enrichment analysis was performed by STRING DB

The BPs on ribonucleoside and nucleoside monophosphate biosynthesis were upregulated with ESs of 1.80 and 1.72, respectively, and FDRs of 0.0016, both, likewise supported by the DAP results (see Online Resource 3). The upregulation of proteins involved in nucleotide biosynthesis could be an indirect effect of the disturbed DNA replication, for which a balanced nucleotide pool is required. The induction of respective genes might be regulated via the transcription factor c-Myc, as there is a high overlap between its regulated genes, the proteins upregulated by OTA and those affiliated to nucleoside monophosphate biosynthesis (Liu et al. 2008). Liang et al. (2015) also described the deregulation of nucleotide metabolism and cell cycle as well as DNA repair mechanisms and the blockade of RNA synthesis by OTA in human embryonic kidney cells (HEK293), which mostly supports our results. They concluded that OTA activated the apoptosis signal-regulating kinase 1 (ASK1) via oxidative stress, which in turn led to apoptosis initiation by the mentioned mechanisms.

Several terms of the whole proteome data containing upregulated proteasomal proteins were shown to be enriched with ESs from 1.75 for the CC “proteasome regulatory particle” (FDR < 0.0001) to 1.05 for the CC “proteasome complex” (FDR < 0.0001). The proteasome is a multisubunit protein complex that is responsible for the intracellular degradation of proteins. Its upregulation might be caused by direct effects of OTA or might represent an unspecific stress response. On the one hand, the Nrf2 pathway is described to induce the proteasome as a reaction to oxidative stress (Pickering et al. 2012), which is reported to be a key mechanism of OTA toxicity (Frangiamone et al. 2024). On the other hand, OTA could directly bind to proteins, like described for human and murine serum albumin (Sueck et al. 2018; Kuhn et al. 2024). Malfunctioning modified proteins need to be degraded, which could induce the upregulation of the proteasome. Furthermore, Perugino et al. (2024) recently proposed the inhibition of a prolyl 3-hydroxylase involved in protein synthesis, by 3-dimensional modeling. This effect might induce protein damage resulting in an increased requirement for degradation. Akpinar et al. (2019) also observed a time-dependent deregulation of proteasomal proteins in human kidney proximal tubule cells (HK-2) caused by 10 µM OTA. Furthermore, within the upregulated DAPs, different extracellular components were identified. Effects with ESs < 1 are not further discussed.

In IHKE cells, OTA induced a strong downregulation of several histones and high mobility group nucleosome-binding proteins, which resulted in the enriched MFs “nucleosomal DNA binding” and “structural constituent of chromatin” in the whole proteome data and within the downregulated DAPs (see Online Resource 4). Histone downregulation can be caused by the G1 checkpoint pathway, which in turn can be activated by DNA damage (Su et al. 2004). On the upregulated side, mainly proteins of extracellular components and various metabolic processes were enriched (see Online Resource 4).

Taking results from both cell lines together, OTA seems to affect the cell cycle, which is already described (Kőszegi and Poór 2016). Different responses of the two cell lines could be explained by very different methods used to obtain the cells (López-Terrada et al. 2009; Tveito et al. 1989) and by eventually high differences in abundances of proteins like the tumor suppressor p53 in cancer cells (Zhou and Elledge 2000).

HepG2 cells were treated with 20 µM CIT for 24 h (Fig. 1, right). Similar to the effects of OTA, CIT strongly induced the six MCM subunits MCM2–MCM7 resulting in an ES of 3.06 and an FDR of 0.0022 for the CC terms “MCM complex” and “CMG complex”. This effect is supported by respective terms enriched in the DAPs and several terms regarding DNA replication (see Online Resource 3). Remarkably, the same terms of upregulated proteasomal proteins and nucleotide biosynthesis-related proteins as for OTA were observed to be enriched by CIT as well. For proteasomal proteins, ESs ranged from 1.75 for the CC “proteasome accessory complex” (FDR < 0.0001) to 1.43 for the CC “proteasome complex” (FDR < 0.0001, not part of top 15 terms). Concerning nucleotide biosynthesis, the BP “ribonucleoside monophosphate biosynthesis” showed an ES of 1.88 with an FDR of 0.0014 and the respective nucleoside term showed an ES of 1.61 with an FDR of 0.0022. Supporting these results, terms on the MF “nucleotide binding” were enriched in the analysis of upregulated DAPs (see Online Resource 3). Concerning the MCM upregulation and induction of the proteasome and the nucleotide biosynthesis, CIT showed effects on the proteome of HepG2 cells comparable to OTA (see chapter Ochratoxin A). We assume that this observation indicates a similarity in their toxicity pathways in terms of replication and oxidative stress. A comparison of the log2 FC values of the six upregulated MCM proteins between the OTA and the CIT experiment by Student’s t test revealed a p-value of 0.931, demonstrating a resemblance between the mentioned effects of OTA and CIT. Oxidative stress is a comprehensively analyzed mechanism of CIT and OTA toxicity (Rašić et al. 2019), but replication stress is only discussed for OTA so far (EFSA Panel on Contaminants in the Food Chain (CONTAM) et al. 2020). Thus, the described results reveal a new potential mechanism of CIT toxicity and concurrently suggest the polyketide-derived coumarin part of the molecules (Geisen et al. 2018) as responsible for this mechanism.

The second strongest enriched term was the “catenin complex” (ES 2.85, FDR 0.0045), caused by the downregulation of cadherins (CDH) 1 and 2, catenins α−1 and δ−1 and junction plakoglobin, all of which are junction proteins. As this term was headed by CDH1 (log2 FC = − 2.45, − log10p value = 0), which was identified in only one out of six replicates of CIT treatment, this effect is not discussed in more detail.

Proteins of the KEGG terms “fructose and mannose metabolism” (ES 2.29, FDR 0.00046) were upregulated, which could affect energy production from glycolysis, but also ascorbate metabolism and N-glycan biosynthesis are associated with this pathway (KEGG: hsa00051). On the other hand, some proteins affiliated to “complement and coagulation cascades” (ES 1.91, FDR 0.00052) were downregulated, which were mainly complement factors, serine protease inhibitors and fibrinogens (see Online Resource 3). This could affect the blood coagulation in vivo (Amara et al. 2008). However, no such effects for CIT have been described in the literature.

Enzymes of the “folate biosynthesis” were upregulated (ES 1.93, FDR 0.0087), which could explain the strong enrichment of dihydrobiopterin observed in HepG2 cells (Gerdemann et al. 2022). This effect might be related to nucleotide synthesis—and thereby probably to replication stress—as the output of folate metabolism also includes nucleotide precursors (Zheng and Cantley 2019). Gerdemann et al. (2022) also postulated the inhibition of the enzymes pyruvate carboxylase (PYC) and succinyl-CoA ligase (SUCL), which are part of the citrate cycle. Our results might also explain this result, as PYC (log2 FC − 0.743, − log10p value 1.29) and SUCL (log2 FC − 0.646, − log10p value 1.33 for subunit G2) were the strongest downregulated proteins of the enriched KEGG term “citrate cycle” (ES 1.43, FDR 0.0034) that was not listed within the top 15 terms.

In IHKE cells, only the CC “mitochondrial protein-containing complex” and the KEGG term “systemic lupus erythematosus” were enriched in the whole proteome data (see Online resource 4). The latter term was mainly driven by downregulation of histones, which was comparable to OTA. However, the histone downregulation still indicates a similarity in their MoA, which is supported by observations in both HepG2 and IHKE cells.

The experiment with a combination of CIT and OTA did not show any effects that point towards specific combinatory effects on the proteome in the used concentrations of 15 µM and 20 nM, respectively, beyond the addition of the single compound effects—based on the enriched terms in the whole proteome data (see Online Resource 4). In contrast, the combination showed a smaller number of DAPs than OTA alone (see Online Resource 1, Figure S2). This could be caused by the inhibitory effect of CIT on the uptake of OTA described by Knecht et al. (2005). They described, that 15 µM CIT reduced the uptake of OTA by more than 60% in IHKE cells.

HepG2 cells were treated with 10 µM AFB1 for 24 h (Fig. 2, left). Since the metabolic activation of AFB1 was shown to be critical for certain toxic mechanisms (Gerdemann et al. 2023; van Vleet et al. 2002), the same experiment was additionally conducted in β-NF pretreated HepG2 cells (10 µM, 16 h) to induce the metabolic activity especially of CYP1A variants (Westerink and Schoonen 2007; Gerets et al. 2012). In this study, only CYP1A1 was observed as induced (see Online Resource 3, sheet “β-NF proteins”), but CYP1A2 was shown to be induced by β-NF in HepG2 cells in former investigations (data not shown) and presumably lacks abundance to meet the limit of detection. The latter experiment included its own corresponding control pretreated with β-NF and afterwards treated with solvent control (see chapter Cell treatment). The results are shown in the right-hand part of Fig. 2.

Enrichment bubble plots of the top 15 enriched terms in the whole proteome of HepG2 cells treated with aflatoxin B1 (AFB1, 10 µM) for 24 h without (left) or with previous metabolic induction by 10 µM β-naphthoflavone (β-NF, right). Effects on cellular components (CC), biological processes (BP), molecular functions (MF, all from Gene Ontology) and terms of the Kyoto Encyclopedia of Genes and Genomes (KEGG) are shown. The color scale describes the significance as the false discovery rate (FDR) and the bubble size scales with the percentage of identified proteins affiliated to the described term. Terms are annotated by + for upregulation, * for downregulation and ~ for enrichment on both ends. Tricistronic rRNA transcript describes the SSU-rRNA, 5.8S rRNA, LSU-rRNA variant. Enrichment analysis was performed by STRING DB

For AFB1, by far the strongest enriched term was the “cytokine–cytokine receptor interaction” (ES 2.79, FDR 0.0057), mainly caused by the upregulated cytokine “growth and differentiation factor 15” (GDF15, log2 FC 1.72, − log10p value 4.87), but also by tumor necrosis factor receptors and interleukin receptors (see Online Resource 3). In β-NF pretreated cells, again the KEGG term “cytokine–cytokine receptor interaction” was found enriched (ES 2.29, FDR 0.0026) with the same proteins involved, but in this case, enrichment on both ends (up- and downregulation) was identified. Like for AFB1 without pretreatment, GDF15 was the main driver for this term (log2 FC 3.72, − log10p value 4.27), but also affected the enriched BP term “regulation of pathway-restricted SMAD protein phosphorylation” (ES 2.79, ES 0.0046). The strong upregulation of GDF15, even enhanced by pretreatment with β-NF to increase metabolic activation of AFB1, probably indicates an inflammatory response or other cellular dysfunctions (Wang et al. 2021). GDF15 is used as a biomarker for cardiovascular disease, cancer and other diseases in humans (Luan et al. 2019). Therefore, the effect of AFB1 on GDF15 abundance in vivo should be analyzed to prevent false-positive diagnosis of these diseases. Our results indicate that the metabolic activation of AFB1 through phase I metabolism is required for the inflammatory response in HepG2 cells or at least induces it. The hypothesis of inflammatory processes is supported by respective enriched terms that include GDF15 as well as different cytokine receptors. Except for GDF15, no cytokine was detected, which could be related to their low molecular weight and subsequent loss during sample preparation. During inflammatory processes, interleukins are excreted to cell culture media which was removed prior to sample preparation and, therefore, not analyzed in this study. Eventually, the abundance of cytokine receptors could have been upregulated because of a high, but not detectable cytokine concentration. Iori et al. (2022) observed the same KEGG term “cytokine–cytokine receptor interaction” as strongest enriched in a transcriptomic approach in bovine liver cells. They proposed an activation of the toll-like receptor 2 linked to inflammatory response and oxidative stress. Other transcriptomic studies using the chicken hepatocellular carcinoma cell line LMH and the bovine fetal hepatocyte cell line BFH12 found impaired genes associated with inflammation as well (Choi et al. 2020; Pauletto et al. 2020).

Within the upregulated DAPs caused by AFB1 in non-pretreated cells, several terms related to the mitotic cell cycle were identified as enriched, although only 13 proteins (0.43% of all quantified) were identified as significantly upregulated (see Online Resource 3). Within these terms, the aurora kinases A and B (AURKA, log2 FC 0.63, − log10p value 3.28; AURKB, log2 FC 0.75, − log10p value 3.04) and the inner centromere protein (INCENP, log2 FC 0.76, − log10p value 3.53) were the key proteins. Aurora kinases and INCENP are parts of the chromosomal passenger complex and play a central regulatory role in mitosis and cytokinesis. The deregulation of these processes could lead to the described general cytotoxicity (Cimbalo et al. 2022), but could also induce chromosomal defects (Ruchaud et al. 2007) and thereby contribute to the carcinogenicity of AFB1.

Further enriched terms in the whole proteome dataset of only AFB1 describe the downregulation of enzymes involved in glycolysis or nucleotide phosphorylation, with a high overlap within the proteins of these terms: all proteins of the BP “glycolytic process” were also found in the BP “nucleotide phosphorylation” (see Online Resource 3). The downregulation of these enzymes matches the decreased concentration of nucleoside derivatives and several metabolites of glycolysis found after AFB1 treatment in HepG2 cells (Gerdemann et al. 2022).

All other top terms of AFB1 treatment in metabolically induced HepG2 cells describe downregulated proteins of processes or components of the maturation of rRNA or ribosomes. For instance, the BP “endonucleolytic cleavage in 5-ETS of tricistronic rRNA transcript (SSU-rRNA, 5.8S rRNA, LSU-rRNA)” was enriched by ES 2.73 with an FDR of 0.0052. None of these terms occurred in the top 15 enriched terms in non-pretreated cells. The ribosome biogenesis and its related terms were very much represented in the enriched terms within the group of downregulated DAPs as well (see Online Resource 3) and precursors of the ribosomal LSU. The CC “preribosome, large subunit precursor” was already found in the experiment with non-pretreated cells. However, the high abundances and ESs of terms related to ribosome biogenesis or, more specifically, rRNA maturation indicate higher effects after metabolic induction. This suggests that the activation of AFB1 through phase I metabolism enhances its effect on these terms. However, no such effect on ribosomes or ribosomal activity is described for AFB1 so far, except for the aforementioned transcriptomics approach by Iori et al. (2022), who found the term “ribosome biogenesis in eukaryotes” enriched. These results point towards a new potential cellular target of AFB1 that could contribute to its hepatotoxicity. Whether the downregulation of ribosomal proteins actually impairs their activity in the form of protein synthesis needs to be investigated in further experiments. Effects with ESs < 1 are not discussed in detail.

HepG2 cells were treated with 10 µM Pen A for 24 h. The enrichment analysis of the few DAPs (1.2%) revealed no enriched GO or KEGG terms in these groups. However, the analysis of the whole proteome dataset revealed significant enrichments of certain biological functions (Fig. 3). The strongest effect was observed in the downregulation of proteins involved in the “cholesterol biosynthetic process” (ES 1.54, FDR 0.0030), while several other terms containing “sterol”, “steroid” or “secondary alcohol” showed a high overlap with this term. The diterpene part of the chemical structure of Pen A might cause the downregulation of proteins involved in these processes, as it comprises a similarity to the polycyclic backbone of cholesterol. Comparably, the exogenous steroid hypocholamide regulates the expression of genes involved in cholesterol and fatty acid homeostasis via the liver X receptor (Song and Liao 2001). Potentially, Pen A activates negative regulatory feedback pathways of cholesterol synthesis and metabolism by binding to this receptor. Inhibited synthesis of cholesterol in the liver in vivo can affect the uptake, metabolism and transport of lipids, but also more severe effects on the entire organism are described, especially during developmental stages (Peeples et al. 2024).

Enrichment bubble plots of the top 15 enriched terms in the whole proteome of HepG2 cells treated with penitrem A (Pen A, 10 µM) for 24 h. Effects on cellular components (CC), biological processes (BP), molecular functions (MF, all from Gene Ontology) and terms of the Kyoto Encyclopedia of Genes and Genomes (KEGG) are shown. The color scale describes the significance as the false discovery rate (FDR) and the bubble size scales with the percentage of identified proteins affiliated to the described term. Terms are annotated by + for upregulation and * for downregulation. Enrichment analysis was performed by STRING DB

Another term contains downregulated proteins of the “proton-transporting two-sector ATPase complex” (ES 1.09, FDR 0.0056), of which 24 different ATPases or subunits were identified. These also led to the enriched terms concerning proton motive force-driven ATP synthesis. The effect on mitochondrial ATP synthesis was highly specific, as almost all ATPases or subunits were downregulated. Its relevance for cellular energy levels is apparent and could explain cytotoxic effects of Pen A in higher concentrations (Gerdemann et al. 2022; Kalinina et al. 2018). A third effect was the downregulation of valine, leucine and isoleucine degrading enzymes (ES 1.02, FDR < 0.0001). The deregulation of branched-chain amino acid degradation via transamination and oxidative decarboxylation is associated with obesity, insulin resistance and diabetes (Choi et al. 2024).

Few studies analyzed cellular toxicity pathways of Pen A so far and these focused on cytotoxicity (Kalinina et al. 2018), metabolic (Gerdemann et al. 2022) or tremorgenic effects in the central nervous system (Berntsen et al. 2013). Our study suggests three new potential deregulated mechanisms by Pen A, which are sterol biosynthesis and metabolism, mitochondrial energy production and branched-chain amino acid degradation. These results can contribute to further understand the detailed mechanisms behind the (cyto)toxicity of Pen A. Future studies should include the investigation of proteomic alterations in cells of the central nervous system, the main site of action of Pen A toxicity in vivo. For this purpose, e.g., CCF-STTG1 could be used, in which Pen A has shown stronger cytotoxicity than in HepG2 cells (Kalinina et al. 2018).

For the two type B trichothecenes DON and NIV, a deviating presentation of the enrichment analysis results was chosen. Due to their well-described ribotoxicity, effects on the abundance of ribosomal proteins were expected. These were observed in the form of upregulated proteins involved in ribosomal biogenesis. However, the corresponding enriched terms were the strongest ones on the upregulated side, but not within the overall top 15 enriched terms, since the terms on the downregulated side showed much higher ESs. As we still intended to demonstrate the ribotoxicity-related effects of trichothecenes, the following bubble plots are divided in up- and downregulated terms.

HepG2 cells were treated with 1 µM DON for 24 h. The presented results are divided in enriched terms caused by upregulated (Fig. 4, left) and downregulated proteins (Fig. 4, right). On the upregulated side, the strongest enrichments were observed for the CCs “box C/D RNPFootnote 3 complex” (ES 2.32, FDR 0.0026) and “preribosome, large subunit precursor” (ES 2.30, FDR < 0.0001) and the BPs “maturation of LSU-rRNAFootnote 4” (ES 2.26, FDR < 0.0001), specifically from tricistronic (SSU-rRNA, 5.8S rRNA, LSU-rRNA) rRNA transcript (ES 2.30, FDR < 0.0001). Several further terms described the biogenesis of ribosomes directly or indirectly (MF: “rRNA methyltransferase activity”; BPs: “ribosomal large subunit biogenesis”; “rRNA methylation”). In addition, the CC term “nucleolar exosome (RNAse complex)” and two BP terms on the processing of small nucleolar (sno(s)) RNA were enriched. All these terms were also found enriched in the group of upregulated DAPs (see Online Resource 3) as well as several other terms regarding ribosomes or ribosomal biogenesis.

Enrichment bubble plots of the top ten enriched terms for both upregulation (left) and downregulation (right) in the whole proteome of HepG2 cells treated with deoxynivalenol (DON, 1 µM) for 24 h. Effects on cellular components (CC), biological processes (BP), molecular functions (MF, all from Gene Ontology) and terms of the Kyoto Encyclopedia of Genes and Genomes (KEG, in this figure) are shown. The color scale describes the significance as the false discovery rate (FDR) and the bubble size scales with the percentage of identified proteins affiliated to the described term. Terms are annotated by + for upregulation and * for downregulation. Tricistronic rRNA transcript describes the SSU-rRNA, 5.8S rRNA, LSU-rRNA variant. Enrichment analysis was performed by STRING DB

DON is a well-known ribotoxin and thereby impairs the protein synthesis (McCormick et al. 2011). HepG2 cells seem to counter-regulate the inhibited ribosomal activity by upregulating proteins required to generate new ribosomes. Remarkably, the binding to the LSU becomes apparent in the respective enriched terms, like the “maturation of LSU-rRNA”. Besides the terms directly related to ribosomes, the terms concerning the box C/D RNP complex, the nucleolar exosome and sno(s) RNA are associated with the biogenesis of ribosomes as well, since all these are essential for the maturation of rRNA (Kilchert et al. 2016; Maden and Hughes