Nucleolar Organization in Response to Transcriptional Stress

ABSTRACT

The nucleolus, a prominent membrane‐less nuclear compartment, is organized around ribosomal RNA (rRNA) gene (rDNA) clusters, known as nucleolar organizing regions (NORs), located on the short arms of acrocentric chromosomes. It serves as the primary site for ribosome biogenesis, an energy‐intensive process crucial for cell growth and proliferation. This involves RNA polymerase I (Pol I)‐mediated transcription of 47S precursor rRNA (pre‐rRNA), pre‐rRNA processing, and ribosomal subunit assembly, reflected in its tripartite structure maintained by liquid–liquid phase separation. Recent evidence indicates that only about 30% of nucleolar proteins are exclusively involved in ribosome production. The remaining proteome participates in diverse cellular functions, establishing the nucleolus as a multifunctional organelle. It functions as a critical stress sensor and signaling hub, responding to various intracellular insults such as nutrient starvation, DNA damage, and viral infection. Many chemotherapeutic agents also induce the response called nucleolar stress via disruption of the nucleolar structure or function, potentially leading to rDNA instability. Nucleolar stress frequently leads to dynamic transition of nucleolar proteins, inducing nucleolar reorganization. Of these, the stress induced by transcriptional changes leads to the unique nucleolar structures termed nucleolar caps and nucleolar necklaces. In this review, we summarize the recent findings about the molecular mechanism of nucleolar changes upon stresses and discuss the possible relationship between rDNA instability and cancer.

In this review, we provide an overview of the research field on nucleolar regulation, mainly focusing on nucleolar organization under transcriptional stress. We also highlight the importance of ribosomal DNA stability, particularly in disease prevention, including cancer.

Abbreviations

53BP1

p53‐binding protein 1

AMD

Actinomycin D

ATM

ataxia telangiectasia mutated

ATR

activating ataxia telangiectasia and Rad3‐related protein

BRCA1

breast cancer gene 1

CHK1

check point kinase 1

CHK2

check point kinase 2

CITI

Condensates Induced by Transcription Inhibition

CtIP

carboxy‐terminal binding protein (CtBP) interacting protein

DDR

DNA damage response

DFC

dense fibrillar component

DJ

distal junction

DRB

5,6‐Dichlorobenzimidazole 1‐β‐D‐Ribofuranoside

DSB

double‐strand break

DYRK1B

dual specificity tyrosine phosphorylation regulated kinase 1B

ETS

external transcribed spacer

FBL

fibrillarin

FC

fibrillar center

GC

granular component

HR

homologous recombination

IGS

intergenic spacer

ITS

internal transcribed spacer

LLPS

liquid–liquid phase separation

MRE11

meiotic recombination 11 homolog 1

MST2

mammalian sterile‐20‐like kinase 2

NBS1

Nijmegen breakage syndrome protein 1

NCL

nucleolin

ncRNA

non‐coding RNA

n‐DDR

nucleolar‐DNA damage response

NEAT1

nuclear paraspeckle assembly transcript 1

NHEJ

non‐homologous end joining

NONO

non‐POU domain containing octamer binding

NOR

nucleolar organizing region

NPM1

nucleophosmin

PH

perinucleolar heterochromatin

PJ

proximal junction

Pol I

RNA polymerase I

Pol II

RNA polymerase II

RAD51

radiation sensitive protein 51

rDNA

ribosomal DNA

RPA2

Replication protein A2

rRNA

ribosomal RNA

SETX

senataxin

SFPQ

splicing factor proline and glutamine rich

snoRNA

small nucleolar RNA

snoRNP

small nucleolar ribonucleoprotein

TAF15

TATA‐box binding protein associated factor 15

TCOF1

treacle biogenesis factor1

TOPBP1

DNA topoisomerase II binding protein 1

UBF

upstream binding factor

UFL1

UFM1 specific ligase 1

XRCC4

X‐ray repair cross complementing 4

1. Introduction: rDNA and Nucleolar Organization

The foundation of the nucleolus lies in the nucleolar organizing regions (NORs), which are clusters of tandemly repeated rRNA genes [1, 2, 3]. In human cells, each NOR comprises multiple copies of the rDNA repeat unit, arranged predominantly in a head‐to‐tail orientation and located on the short (p) arms of the five acrocentric chromosomes: 13, 14, 15, 21, and 22 (Figure 1) [2, 3, 4, 5]. The human diploid genome is estimated to contain an average of approximately 300 to 400 copies of rDNA repeats [6, 7], with each acrocentric chromosome harboring between 16 and 76 copies [7]. A typical human rDNA repeat unit also spans approximately 43–45 kilobases (kb), consisting of a ~13 kb transcribed region and a ~31 kb intergenic spacer (IGS) [3]. The transcribed region by Pol I encodes the large 47S (or 45S) precursor rRNA (pre‐rRNA), which contains the sequences for the mature 18S, 5.8S, and 28S rRNAs, flanked by external transcribed spacers (ETS) and separated by internal transcribed spacers (ITS1 and ITS2) (Figure 1) [3, 8]. The IGS contains crucial regulatory elements, including enhancer repeats and promoters for both the 47S pre‐rRNA and spacer transcripts (Figure 1) [3, 4]. The entire rDNA array on each acrocentric chromosome is localized between conserved proximal (centromere‐facing) and distal (telomere‐facing) junction sequences (PJ and DJ) (Figure 1) [2, 5, 8].

FIGURE 1.

Schematic representation of ribosomal RNA (rRNA) genes in human. The rRNA gene array is located at the nucleolar organizer regions (NORs). It is positioned between telomeres and centromeres on the short arms (p‐arms) of the acrocentric chromosomes 13, 14, 15, 21, and 22 in humans. They are flanked by heterochromatic distal and proximal junctions (DJ and PJ, respectively). Each ribosomal DNA (rDNA) repeat unit consists of an intergenic spacer (IGS) and a transcribed region encoding the 47S pre‐rRNA, which is further processed into the mature 18S, 5.8S, and 28S rRNAs. The coding region is flanked by external transcribed spacers (5′ ETS and 3′ ETS), and rDNA genes are separated by internal transcribed spacers (ITS1 and ITS2). The IGS includes the promoter for rDNA transcription.

The nucleolus is considered a biomolecular condensate formed through liquid–liquid phase separation (LLPS), creating distinct, immiscible sub‐compartments without a membrane [9, 10]. In mammalian cells, the structure of the nucleolus typically exhibits a tripartite internal organization comprising the fibrillar center (FC), dense fibrillar component (DFC), and granular component (GC), often surrounded by a layer of perinucleolar heterochromatin (PH) (Figure 2) [1, 5, 11, 12]. The FCs store the RNA Polymerase I (Pol I) and transcription regulators, such as the upstream binding factor (UBF) and treacle biogenesis factor1 (TCOF1) [1, 5, 12]. TCOF1 potentially acts as a scaffold for the FCs (Figure 2) [13]. Transcription of the 47S pre‐rRNA occurs at the interface between the FC and the surrounding DFC [14, 15]. The DFC is the site for early pre‐rRNA processing and modification, containing factors such as fibrillarin (FBL), Nucleolin (NCL), and small nucleolar ribonucleoproteins (snoRNPs) (Figure 2) [1, 11]. FBL plays a pivotal role in sorting nascent pre‐rRNAs into the DFC, driven by its RNA‐binding and self‐assembly properties [10, 16]. On the other hand, NCL is likely to serve as a histone chaperone that facilitates rDNA transcription by promoting a euchromatic state [17]. The outermost layer, the GC, is where the late rRNA processing and assembly of pre‐ribosomal subunits (pre‐40S and pre‐60S) occur (Figure 2) [11, 18]. Key GC protein is NPM1 (also as known as B23), which functions in ribosome assembly and export [19, 20, 21]. NPM1 also plays a crucial role in phase separation of GC through multivalent interactions [22, 23]. Although some NORs are silent [24, 25], the active rDNA transcription by Pol I facilitates LLPS via the interaction with rRNA and the nucleolar proteins FBL and NPM1, underpinning the complex and vital structure of the nucleolus [26, 27].

FIGURE 2.

Nucleolar organization and its function. The overall structure of the mammalian nucleolus consists of the internal fibrillar center (FC), dense fibrillar component (DFC), and granular component (GC), all of which are surrounded by condensed perinucleolar heterochromatin (PH). RNA polymerase I (Pol I) transcribes the rRNA genes to form 47S pre‐rRNA at the FC‐DFC border. The transcription factor upstream binding factor (UBF) and Treacle (TCOF1) are required for rDNA transcription. In the DFC, pre‐rRNA is further processed and modified to produce the mature 18S, 5.8S, and 28S rRNAs. The DFC contains small nucleolar ribonucleoproteins (snoRNPs), Fibrillarin (FBL), and Nucleolin (NCL). In the GC, pre‐40S and pre‐60S are assembled and exported to the nucleoplasm. The GC harbors nucleophosmin (NPM1).

2. Nucleolar Reorganization and Dynamics in Response to Transcriptional Stress

The nucleolus has been emerging as a critical sensor and integrator of cellular stress, frequently targeted by various classes of chemotherapeutic agents [5, 8, 28, 29, 30]. Of these, transcriptional stresses elicit characteristic structural reorganization of nucleoli (Figure 3) [5, 26, 27]. In this section, we review the molecular dynamics and mechanisms in response to transcription inhibition of Pol I and Pol II.

FIGURE 3.

Nucleolar reorganization upon transcriptional stress. Upon Pol I inhibition, rDNA together with proteins of the FC and DFC sub‐compartments moves to the nucleolar periphery, forming nucleolar caps. Upon Pol II inhibition, the GC is disrupted and the FC and DFC modules are unraveled, forming CITIs and/or nucleolar necklaces.

2.1. Pol I Inhibition

The response observed upon inhibition of Pol I using compounds like Actinomycin D (AMD), CX‐5461, or BMH‐21 is a profound spatial reorganization categorized as nucleolar segregation [5, 31]. The normally intermingled FC and DFC components undergo a striking phase transition. This indicates that FC proteins like UBF and TCOF1, along with Pol I itself, detach from the nucleolar interior and coalesce into larger, often spherical structures at the periphery of the nucleolus (Figure 3) [5]. These structures are known as nucleolar caps (Figure 3) [32]. The remaining bulk of the nucleolus, primarily consisting of GC material marked by NPM1, typically shrinks and rounds up, forming a residual core structure. Even though the treatment with Pol I inhibitors often causes DNA damages that induce the nucleolar DNA damage response (n‐DDR; reviewed in next section) [33, 34], the treatment with BMH‐21 or AMD at low concentration gives rise to the nucleolar cap formation without inducing DNA damage response [35, 36], suggesting that transcription silencing, not DNA damage, is a trigger of the nucleolar cap formation.

Single molecule tracking experiments showed that RPA194, a component of Pol I, which forms relatively immobile clusters with UBF in the FCs likely through engagement with rDNA during transcription under unperturbed conditions, dissociates from the rDNA chromatin and transitions to rapid, liquid‐like diffusion within the newly formed nucleolar cap droplet upon Pol I inhibition [37]. Furthermore, a very recent study described that, upon Pol I inhibition, the nucleolar cap formation is mediated by TCOF1 through its S/E‐rich central and K‐rich C‐terminal domain that promotes phase separation [38]. These findings provide a mechanism of the Pol I inhibition‐dependent nucleolar cap formation driven by LLPS dynamics. Although some studies suggested that Pol I transcription inhibition leads to p53 activation [39, 40], the biological role of the cap formation in response to Pol I inhibition is still unclear.

2.2. Pol II Inhibition

The inhibition of Pol II transcription by chemicals, such as α‐amanitin or CDK7/9 inhibitors including DRB, THZ1, and AZD4573, induces nucleolar reorganization patterns that differ markedly from the nucleolar caps formed upon Pol I inhibition [5, 41, 42]. Instead of the segregation, Pol II inhibition leads to the translocation of the nucleolar proteins like NPM1 to the nucleoplasm [41, 42, 43], which often end up with the formation of nucleolar necklaces (Figure 3) [5, 42]. This morphology is characterized by the unraveling or peripheral arrangement of the FC/DFC units while Pol I transcription may continue in each spot of the necklace, at least initially [42].

A mechanism of the nucleolar reorganization upon Pol II inhibition was thought to be based on the downregulation of the snoRNAs, which are essential for 3′ rRNA processing [44]. However, recent findings suggest that Pol II is physically present in nucleoli and operates on the rDNA locus directly, primarily within the IGS regions via TATA box‐binding protein‐like 1 (TBPL1) and/or Pol II‐associated factor‐1 (PAF1)‐dependent manner [45, 46]. Here, assisted by factors like the helicase senataxin (SETX), Pol II generates antisense non‐coding RNAs (ncRNAs), which form R‐loop structures (RNA:DNA hybrids) in the IGS regions. These structures appear to function as a regulatory shield, preventing the excessive transcription of sense IGS ncRNAs (sincRNAs) by Pol I and translocation of NPM1 to the nucleoplasm. The accumulation of sincRNAs is potentially associated with the nucleolar reorganization observed in diseases [45].

Another key discovery associated with Pol II inhibition is the formation of novel nucleolar condensates termed CITIs (Condensates Induced by Transcription Inhibition) [47]. CITIs are seeded by the RNA‐binding proteins SFPQ and NONO along with TAF15 and FUS. Under normal conditions, SFPQ interacts with NEAT1, a long ncRNA, forming paraspeckles and also binds to promoter regions of transcriptionally active genes outside of nucleoli; however, when Pol II activity is suppressed, SFPQ binds to pre‐rRNAs, forming CITIs in place of the GC around the FC/DFC. The CITI formation appears to be independent of the accumulation of sincRNAs. Notably, SFPQ/TAF15 recruits transcriptionally active chromatin to the nucleolar periphery through the CITI formation, promoting chromosomal translocations when DSBs are present. CITIs are also induced by stressors including UV irradiation, cold shock, chemotherapeutics like Topoisomerase I inhibitors as well as the depletion of rRNA processing factors. Interestingly, CITIs exhibit a reversible nature in common with nucleolar necklaces [42], implying the crosstalk between them. Given that CITIs are formed in early time points after Pol II inhibition (typically 0.5 to 2 h) and that nucleolar necklaces are observed in later time points (typically 8 to 24 h), CITIs might be an early form of nucleolar necklaces. The functions of CITIs in response to the transcriptional stress are still unclear.

3. The Molecular Mechanism of rDNA Damage Response

In a decade, numerous studies using CRISPR/Cas9 and restriction enzymes I‐PpoI and AsiSI to introduce DSBs within rDNA regions have indicated that the DNA damage response occurring in nucleoli is distinct from the canonical DNA damage response in the nucleoplasm [48, 49]. These nucleolus‐specific responses are collectively referred to as the nucleolar‐DNA damage response (n‐DDR) (Figure 4).

FIGURE 4.

The nucleolar DNA damage response (n‐DDR). In the left panel, upon the presence of persistent rDNA double‐strand breaks (DSBs) in the nucleolus, the ataxia telangiectasia mutated (ATM) kinase is activated and phosphorylates TCOF1. The MRN complex, consisting of Nijmegen breakage syndrome protein 1 (NBS1), meiotic recombination 11 homolog 1 (MRE11), and the DNA repair protein RAD50, is recruited to the nucleolus through the interaction between NBS1 and phospho‐TCOF1. Phosphorylated TCOF1 recruits DNA topoisomerase II binding protein 1 (TOPBP1), activating the ataxia telangiectasia and Rad3‐related protein (ATR) kinase to elicit transcription silencing. UFL1 involves transcription silencing via TCOF1 modifications. In the right panel, transcription silencing elicits the nucleolar cap formation. The nucleolar caps include the homologous recombination (HR) factors: Breast cancer gene 1 (BRCA1), replication protein A2 (RPA2), carboxy‐terminal binding protein interacting protein (CtIP), and radiation sensitive protein 51 (RAD51). ATM, ATR, the MRN complex, the phosphorylated form of the histone variant H2AX (γH2AX), and p53‐binding protein 1 (53BP1) are clustered at nucleolar caps.

The n‐DDR pathway is initiated by activating the ATM kinase upon DSB induction in rDNA, where ATM phosphorylates TCOF1 [50, 51, 52]. Phosphorylated TCOF1 acts as an adaptor to recruit the MRE11‐RAD50‐NBS1 complex (MRN complex) through the interaction with NBS1 [50, 51, 52, 53]. Furthermore, TCOF1 facilitates the recruitment of the ATR activator TOPBP1 in an ATM‐ or NBS1‐dependent manner [50, 54]. A recent study indicates that TCOF1 is subject to UFMylation mediated by the UFL1 E3 ligase, also contributing to the efficient ATM activation and early response coordination [55]. This ATM‐TCOF1‐MRN‐TOPBP1 axis mediates the rapid inhibition of Pol I transcription under persistent DSB induction in rDNA. The process of silencing is largely dependent on the ATM kinase activity but also requires downstream signaling mediated by ATR for complete silencing [52, 53]. In addition, more sufficient repression is likely to involve additional kinases like CHK1/CHK2, DYRK1B, and MST2 acting through various mechanisms including histone modifications [53, 56, 57]. Although this signaling is relevant to the disruption of the Pol I pre‐initiation complex formation and displacement of the elongating Pol I, the direct mechanism suppressing the Pol I activity upon rDNA damage remains unclear.

Transcriptional silencing by n‐DDR often leads to the nucleolar cap formation similar to the treatment with the Pol I inhibitors. This process also requires the ATM/ATR signaling and is mediated by factors including TCOF1 and TOPBP1 (Figure 4) [52, 53]. Interestingly, the nucleolar caps induced by DSBs in rDNA contain not only the nucleolar proteins but also DDR‐associated factors including ATM, ATR, NBS1, TOPBP1, 53BP1, and γH2AX via an unknown mechanism [52, 53, 54].

As with DSBs in the nucleoplasm in human cells, DSBs in rDNA are also repaired by NHEJ or HR, but the pathway choice is potentially influenced by the rDNA damage persistence and nucleolar structure [48, 49]. A study using I‐PpoI as an inducer of DSBs in rDNA showed that inhibition of DNA‐PK or depletion of XRCC4 or Ku80 increases the rDNA breakages and following cap formation, whereas inhibition of HR factors does not [58]. These data suggest that NHEJ is the predominant mode of immediate DSB repair at rDNA. However, it is still unclear whether the NHEJ in nucleoli is operated in the same way as the NHEJ in the nucleoplasm. On the other hand, persistent DSBs accompanied by the cap formation are likely to be repaired by HR. Several studies using the CRISPR/Cas9 or I‐PpoI system showed that the HR factors, including RPA2, RAD51, CtIP, and BRCA1, accumulate in the cap (Figure 4) [36, 52, 53, 58, 59], reinforcing that the HR repair of DSBs in rDNA occurs exclusively in the caps. Interestingly, this accumulation was observed in not only the S/G2 but also G1 phase [36, 53]. In contrast, under the condition where DSBs were induced in rDNA by AsiSI, the cap formation was restricted to the S/G2 phase [36]. Additionally, the lack of the HR initiating factor CtIP does not abrogate the cap formation but influences rDNA repair within nucleolar caps [52, 53], suggesting the flexibility of the repair within the caps. A recent study showed that RAD51, a key recombination factor, is recruited to nucleolar caps induced by the CRISPR/Cas9‐mediated system, but this recruitment was not dependent on the long‐range resection mediated by BLM [60]. This suggests that the HR pathway in nucleolar caps is different from that in the nucleoplasm. Given that HR repair in rDNA regions was reported to result in a loss or gain of rDNA repeats and reduced cellular viability [59], further investigation will be required to understand how the repair pathway is coordinated at nucleolar caps.

4. rDNA Instability as a Cancer Therapeutics Target

rDNA loci represent genomic instability in oncogenesis, exhibiting a high frequency of structural alterations [61, 62]. Indeed, across various solid tumors and hematological malignancies like Hodgkin's lymphoma, rearrangements such as insertions, translocations, and amplifications within these repetitive sequences are commonly observed [63, 64]. The propensity for rDNA destabilization in cancer is linked to several oncogenic mechanisms that directly impinge upon these regions. For example, the aberrant activity of oncogenes or the deficiency of tumor suppressor functions can trigger substantial replication stress [65], potentially inducing DSBs in rDNA. Additionally, oncogenic signaling pathways and the loss of tumor suppressors frequently culminate in an augmented rate of rRNA synthesis [39, 66]. This excessive transcription of rDNA may facilitate rDNA instability through the accumulation of R‐loops [5, 8].

Furthermore, compromised DNA repair capabilities, a frequent hallmark of neoplastic cells, exert a considerable influence on rDNA integrity. This is explained in monogenic disorders such as Bloom Syndrome and Ataxia‐telangiectasia, where mutations in the BLM and ATM, respectively, lead to a profound cancer predisposition accompanied by the marked rDNA instability [67]. Clinically, many cancers often manifest overt instability within their rDNA arrays with an alternation in the rDNA copy number [68].

These underline that the rDNA instability exhibited in cancer cells provides an attractive target for cancer therapy [8, 30]. The therapeutic potential of Pol I inhibitors, in particular, is underscored by the recent FDA fast‐track designation of CX‐5461 for breast cancer treatment [30]. Furthermore, CDK7/9 inhibitors, which suppress the Pol II activity, are also regarded as promising therapeutic drugs, with several Pol II targeting agents currently undergoing clinical trials [69, 70]. Nevertheless, the antitumor effects of CX‐5461 still remain uncharacterized [8], and it has been reported to induce mutations through an as‐yet‐unelucidated mechanism [71]. Moreover, as delineated in Section 2, CDK7/9 inhibitors may potentially induce chromosomal translocations via CITI formation [47]. Therefore, a comprehensive understanding of the nucleolar reorganization under stress confers novel insights to facilitate the mitigation of adverse effects and enhancement of therapeutic efficacy.

5. Conclusion

The nucleolus and its constituent rDNA are far more than mere ribosome factories; they represent a highly dynamic nexus for sensing and responding to cellular stress, orchestrating intricate reorganizations and specialized DNA damage responses. These responses mediated by nucleoli are crucial for maintaining genomic integrity and preventing diseases. However, this field is still in its infancy. The mechanisms of each nucleolar stress response remain poorly understood. We expect that future research will elucidate these mechanisms and hope that the findings will 1 day contribute to the development of therapies for diseases associated with nucleolar stress and the creation of anticancer drugs targeting the nucleolar stress pathways.

Author Contributions

Rikiya Imamura: writing – original draft, writing – review and editing. Takaaki Yasuhara: supervision, writing – review and editing.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.