A â€˜garbage canâ€™ for ribosomes: how eukaryotes degrade their ribosomes

A ‘garbage can’ for ribosomes: how

eukaryotes degrade their ribosomes

Denis L.J. Lafontaine

1,2

Fonds de la Recherche Scientiﬁque (FRS-F.N.R.S.), Institut de Biologie et de Me´ decine Mole´ culaire (IBMM), Universite´ Libre de

Bruxelles (ULB), Charleroi-Gosselies, Belgium

Center for Microscopy and Molecular Imaging (CMMI), Acade´ mie Wallonie – Bruxelles, Charleroi-Gosselies, Belgium

Ribosome synthesis is a major metabolic activity that

involves hundreds of individual reactions, each of which

is error-prone. Ribosomal insults occur in cis (alteration

in rRNA sequences) and in trans (failure to bind to, or

loss of, an assembly factor or ribosomal protein). In

addition, speciﬁc growth conditions, such as starvation,

require that excess ribosomes are turned over efﬁciently.

Recent work indicates that cells evolved multiple strat-

egies to recognize speciﬁcally, and target for clearance,

ribosomes that are structurally and/or functionally

deﬁcient, as well as in excess. This surveillance is active

at every step of the ribosome synthesis pathway and on

mature ribosomes, involves nearly entirely different

mechanisms for the small and large subunits, and

requires specialized subcellular organelles.

Ribosome synthesis is a multi-step, error-prone process

Ribosomes comprise two subunits of unequal size that

carry out specialized functions in translation: mRNA

decoding and peptidyl-transfer reaction for the small

and large subunits, respectively [1]. Each eukaryotic ribo-

some consists of 4 rRNAs and !80 ribosomal proteins. The

synthesis, maturation and transport of individual riboso-

mal components and their assembly into ribosomal sub-

units requires the intervention of !200 protein trans-

acting factors, and numerous small nucleolar RNAs (snoR-

NAs) that are involved in hundreds of individual, error-

prone, reactions [2,3] (Figure 1). Many ribosomal proteins

perform additional non ribosomal functions [4]. Likewise,

functions in processes not connected directly to ribosome

biogenesis are currently being assigned to ribosome syn-

thesis factors: connections to cell cycle progression, pre-

mRNA splicing, the DNA damage response, nuclear organ-

ization and telomere maintenance are emerging. With so

many reactions in the ribosome assembly pathway, the

possibility to introduce mistakes with potential deleterious

consequences for cell viability and human health is

immense. As such, failure to bind, or the loss of, a synthesis

factor could lead to the production of ribosomes that are

structurally defective (e.g., lacking individual or subsets of

ribosomal proteins or carrying misfolded rRNA) with func-

tional consequences in translation. To circumvent such

problems, cells have evolved multiple quality control mech-

anisms. For example, synthesis factors involved in late

cytoplasmic ribosome assembly steps can bind pre-rRNAs

at early nucleolar stages, thereby committing pre-ribo-

somes to productive synthesis pathways [5]. In other cases,

trans-acting factors with partial homology to ribosomal

proteins or translation factors might bind pre-ribosomes

to monitor and tether the structural integrity of ribosomal

protein-binding sites [6,7]. In wild-type cells, ribosome

assembly defects can result from the delayed binding of

a trans-acting factor. In these circumstances, and provid-

ing that the proper assembly reaction occurs within a

deﬁned timeframe, faithful assembly presumably resumes;

otherwise, pre-ribosomes are identiﬁed as defective and

targeted for rapid degradation by active surveillance mech-

anisms. In addition to alterations in trans, mutations can

occur in cis either during RNA synthesis or, more fre-

quently, as a consequence of exposure to genotoxic stress.

The importance of ‘clearing the system’ of such mutations

is that they could otherwise accumulate to pathological

levels owing to the high stability (half-life of several days)

and abundance (up to 80% of total cellular RNA) of the

rRNAs. Indeed, human diseases, and in particular, neu-

rodegenerative disorders, have recently been linked to the

accumulation of such defective ribosomes [8]. Finally, there

are situations where excess ribosomes must be degraded to

recycle essential cellular building blocks; quite unexpect-

edly, this pathway also involves their speciﬁc recognition.

Recent research indicates that surveillance exerts its

action at every step of the ribosome assembly line, as well

as on the ﬁnal product, that it involves dedicated subcel-

lular structures, and that distinctive pathways prevail for

monitoring the assembly and function of the small and

large ribosomal subunits. This multi-faceted ribosome

surveillance is the focus of this review (Figure 2).

Mutations in cis: non-functional ribosomal RNA decay

Several surveillance pathways have been described that

monitor the structural and functional integrity of mature

RNA molecules [9]. One such pathway that monitors

mRNAs is the ‘no-go’ decay pathway (NGD), wherein

mRNAs that induce stalled ribosomes are degraded [10].

To test whether mutations in functionally relevant and

conserved ribosomal sites affect rRNA stability, LaRivie

et al. introduced substitutions in the decoding site (DCS; in

18S rRNA) and peptidyl transferase centre (PTC; in 25S

rRNA) at positions that are essential for ribosome function

in bacteria [11] (Table S1). This led to the identiﬁcation of

‘non functional rRNA decay’ or NRD [11], a pathway that

detects and eliminates functionally defective components

of mature ribosomes (Box 1). Indeed, in each case, these

Review

Corresponding author: Lafontaine, D.L.J. ([email protected]).

267

changes resulted in a reduced accumulation (!5 to 10 fold)

of the mature rRNA containing the mutation.

Small ribosomal subunit NRD: a process reminiscent of

mRNA no-go decay

During mRNA NGD, a stalled ribosome triggers initiating

endonucleolytic cleavage events on the defective mRNA at

the vicinity of the pause site, and this is followed by the

exoribonucleolytic digestion of the 5’- and 3’-cleaved mRNA

products by the RNA exosome and the 5’!3’ exoRNase

Xrn1, respectively [10]. The RNA exosome is a conserved

multiprotein 3’!5’ exoRNase complex active in the syn-

thesis, degradation and surveillance of most classes of

cellular RNAs, as well as some viral RNAs [12] (Box 2).

Figure 1. Eukaryotic ribosome synthesis is a multi-step, error-prone process. Ribosome synthesis starts in the nucleolus where the core, and catalytically active, pieces of

the machinery, the rRNAs, are synthesized. Three out of the four rRNAs (18S, 5.8S and 25S rRNAs) are produced by RNA polymerase I (Pol I) from a single transcription unit.

The fourth rRNA (5S) is synthesized by RNA Pol III with a 3’-extension (inset). The 18S, 5.8S and 25S rRNAs are interspersed with non-coding sequences: the 5’- and 3’-

external transcribed spacers (5’-ETS and 3’-ETS) and internal transcribed spacers 1 and 2 (ITS1 and ITS2). An actively transcribed rDNA unit is depicted as a ‘Christmas tree’,

reminiscent of its visualization by Miller chromatin spread, with the ‘trunk’ of the ‘tree’ (brown) representing the rDNA locus and the ‘branches’ (green) corresponding to

nascent pre-rRNA transcripts. Pre-rRNA processing is initiated either post-transcriptionally (left of the ‘tree’) or co-transcriptionally (to the right). In fast-growing yeast cells,

up to 50-70% of nascent transcripts are cleaved co-transcriptionally at site A

within ITS1, generating pre-40S and pre-60S ribosomes [57] (other pre-rRNA cleavage sites are

described in Figure 3). In the remaining cases, a full-length transcript (35S) is generated and is assembled into a 90S pre-ribosome (not represented) that is cleaved post-

transcriptionally at site A

. Many ribosome synthesis factors escort pre-ribosomes from their initial site of synthesis in the nucleolus to their cytoplasmic site of function.

Different ribosome synthesis factors display distinctive patterns of association with pre-ribosomes, resulting in a progressive reduction in protein complexity. The nuclear

dwelling time for pre-40S and pre-60S subunits is distinctly different, with pre-40S subunits reaching the cytoplasm much faster than pre-60S. Translocation through the

various nucleolar domains is facilitated by specific trans-acting factors. Nuclear export involves redundant pathways. About 2000 ribosomes are exported every minute in

exponentially growing yeast cells, and control has been suggested to occur at the nuclear pore in response to nutrient availability [58,59]. There appears to be good timing

between nucleo-cytoplasmic partitioning and the acquisition of prominent ribosomal structural features. For example, final shaping of the ‘beak’, a protruding structure of

the small subunit, is thought to only occur once pre-40S has reached the cytoplasm [60]. Ribosome synthesis factors are available in limited quantities; therefore their

recycling is essential. Recent work identified a mechanochemical process powered by a AAA-type ATPase distantly related to the motor protein dynein involved in this

recycling [61]. Late sets of cytoplasmic trans-acting factors are released sequentially following further structural rearrangements fueled by a cascade of GTPase-mediated

reactions and recycled to the nucleus and nucleolus, ready to engage further rounds of assembly. The stable incorporation of each resident ribosomal protein is then

finalized. Major classes of trans-acting factors are listed. To illustrate the dynamics of ribosome synthesis and the plasticity of pre-ribosomes, several ribosome synthesis

factors (arbitrarily color-coded) are represented.

Review

Trends in Biochemical Sciences Vol.35 No.5

268

The possibility that stalling induced by defective ribosomes

might elicit a similar pathway was quite appealing and,

indeed, 18S NRD substrates are stabilized 2-fold in the

absence of the translation termination factor-like proteins

Dom34 and Hbs1 that are key components of NGD [13].

Strikingly, this is not the case for 25S NRD substrates,

thus indicating that mechanistically, NRD comprises at

least two distinct pathways.

During 18S NRD, Dom34 and Hbs1 act together in the

same pathway (their simultaneous inactivation is not

synergistic), and, consistently, the two proteins interact

functionally in vitro [14] and in vivo [15]. Small molecule

inhibitors of translation (cycloheximide and hygromycin B)

speciﬁcally stabilized 18S but not 25S NRD substrates [13],

thus providing further evidence that 18S NRD activation

requires elongating ribosomes, and that 18S and 25S

NRD are mechanistically different. The cytoplasmic

5’!3’ exoRNase Xrn1 and Ski7 (a cytoplasmic exosome

recruitment cofactor) have both been linked to NGD, and

also contribute to 18S NRD. Hbs1 and Ski7 are GTPases

homologous to the translation release factor eRF3. Delet-

ing both HBS1 and SKI7 enhanced the stabilization of 18S

Figure 2. The major pathways of eukaryotic ribosomal RNA decay. The five pathways of rRNA decay described to date. (i) Nucleolar and nuclear pre-40S and pre-60S

ribosomes are monitored actively by the ‘TRAMP–exosome’ pathway during which misfolded pre-ribosomes are identified by TRAMP binding followed by polyadenylation

of the 3’-ends of defective precursor rRNAs in a step that stimulates both the recruitment and the decay activity of the exosome. It is not yet known how TRAMP detects

defective ribosomes. Polyadenylation occurs both at normal and cryptic pre-rRNA processing sites. Alterations in RNA Pol I activity, as well as in pre-rRNA processing

kinetics, can activate cryptic cleavage sites. Cytoplasmic mature subunits carrying cis mutations in functionally relevant ribosomal sites (and perhaps late cytoplasmic

precursors, Box 4) are monitored by NRD, a process that involves different mechanisms for the small and large subunits. (ii) During 18S NRD, small subunits deficient in

normal progression along the mRNA (‘stalling subunits’) are identified by the translation termination factor-related proteins Dom34 and Hbs1. They are probably cleaved

endonucleolytically by an unknown activity (by analogy to mRNA NGD), the released products are digested by Xrn1 and the RNA exosome, assisted by its cofactor Ski7 (Box

2). (iii) During 25S NRD, defective 60S subunits are targeted for proteasomal degradation by Rtt101–Mms1-mediated ubiquitylation of unidentified, associated ribosomal

component(s). During conditions of nutrient deficiency and stress, excess ribosomes and pre-ribosomes are turned over by ribophagy and PMN, respectively. (iv)

Ribophagy is a specific form of macroautophagy that involves the engulfment and targeting of bulk cytosolic fractions to the vacuole by the de novo formation of an

‘isolation membrane.’ This membrane extends into an autophagosomal structure that fuses to the vacuolar membrane, and triggers the release of an autophagic body and

its subsequent degradation by resident vacuolar hydrolases. For 60S subunits, ribophagy involves the specific deubiquitylation of unidentified ribosomal components by

the Ubp3–Bre5 complex. The effector(s) for 40S ribophagy remain unknown. (v) In PMN, a specific form of microautophagy, a portion of the nuclear envelope is ‘pinched

off’ by the vacuole, generating a specialized organelle termed a NVJ that comprises protein ‘velcro-like’ patches formed by the interaction between Nvj1 (small red circles)

and Vac8 (small green squares). The NVJ matures sequentially into a bulge, a tethered bleb, and finally a vesicle that is degraded by resident hydrolases. The late steps

(vesicle release) of PMN require several ATG genes also involved in macroautophagy [62]. Ribosome surveillance occurs at every step along the assembly pathway, starting

in the nucleolus, where a specialized locale coined the ‘No-body’ concentrates polyadenylated pre-rRNA species and exosomal components, continuing in the nucleoplasm

and ending in the cytoplasm. Cytoplasmic degradation involves several subcellular organelles, including P-bodies (where 18S NRD substrates, as well as mRNA NGD

substrates, are localized), the autophagosome, the proteasome and a cytosolic perinuclear compartment enriched for 25S NRD substrates. Key trans-acting factors involved

in each pathway are represented.

Review

Trends in Biochemical Sciences Vol.35 No.5

269

NRD substrates, indicating that the products of these two

genes probably operate in parallel pathways, possibly

competing for binding an empty-A site on a ribosome

stalled at a sense codon [13]. The core exosome (tested

for Rrp44, also called Dis3; Box 2) contributes, but is

apparently not rate-limiting, for 18S and 25S NRD [13].

18S NRD and mRNA NGD substrates both accumulate

in P-bodies that are conserved RNA–protein cytoplasmic

granules containing untranslated mRNAs complexed with

a set of translational repressors, the mRNA decapping

machinery and Xrn1 [16] . Strikingly, 25S NRD substrates

do not co-localize to P-bodies. Other differences between

18S and 25S NRD include the contribution of Xrn1 and

Ski7 to 18S, but not 25S, NRD [13,17].

It is quite striking that 18S rRNA NRD and mRNA NGD

operate with distinctively different kinetics (half-lives of up

!100 min and !10 min, respectively [10,13]). This differ-

ence might reﬂect the higher complexity and compaction of

mature ribosomal ribonucleoprotein particles (rRNP) ver-

sus mRNP substrates.

Large ribosomal subunit NRD: a role for ubiquitylation in

ribosome turnover

Fujii et al. conducted a genome-wide screen on the yeast

knock-out collection aimed at the identiﬁcation of

mutations that stabilize 25S NRD substrates. This work

led to the discovery of Mms1, a component of an E3

ubiquitin ligase that had previously been characterized

as being involved in DNA repair [17]. There are two main

intracellular protein degradation systems in eukaryotes

[18]: the ubiquitin-proteasome system (UPS) that often

targets short-lived proteins and involves ubiquitylation

of the substrate and its targeting to the proteasome [18],

and autophagy, a process that leads to the degradation of

long-lived proteins and excess or aberrant organelles.

Originally identiﬁed as a protein involved in the repair

of DNA damage induced by the alkylating agent methyl

methanesulfonate (MMS), Mms1 belongs to the Mms22

module that comprises Mms1, Mms22, Rtt101 and Rtt107.

Rtt101, but not the other members of the complex, is also

required for 25S NRD [17]. Mutations in MMS1 and

RTT101 are not synergistic, indicating that their gene

products act together in the same pathway. Consistently,

Mms1 and Rtt101 interact in vitro and in vivo [19,20].

Strikingly, neither Mms1 nor Rtt101 is required for 18S

NRD, lending further credence to the idea that, mechan-

istically, NRD comprises distinct pathways.

Fujii et al. further demonstrated that the level of ubi-

quitylation observed in large ribosomal subunit fractions

increased strikingly in strains that expressed 25S NRD

rRNA, and that this modiﬁcation required speciﬁcally

Mms1 and Rtt101. Although the identity of the ubiquity-

lated components remains unknown, the overexpression of

a ubiquitin variant that inhibits proteasomal function

demonstrated clearly the requirement of ribosomal com-

ponent ubiquitylation for 25S NRD.

25S NRD substrates accumulate in the cytoplasm in a

previously undescribed perinuclear compartment that, the

Box 1. When and where does NRD occur?

NRD does not depend upon rRNA synthesis, because mutations that

affect the accumulation of 18S rRNA do not impact the stability of

25S rRNA, and vice versa (both RNAs are co-expressed; Figure 1

inset). NRD is believed to occur primarily in the cytoplasm on

mature ribosomal subunits, and, perhaps, on late cytoplasmic

precursors (Box 4), for the following reasons. First, neither of the

mutations tested in the DCS or the PTC g rossly affect the

accumulation of pre-rRNA precursors. Second, NRD substrates

colocalize on velocity gradients with mature ribosomal subunit-

sized and/or polysomal fractions. Third, trans-acting factors in-

volved in NRD (Dom34, Hbs1, Xrn1 and Ski7) are all involved in

cytoplasmic processes, and neither the nuclear-specific exosomal

cofactor Rrp6 (Box 2) nor the nuclear 5’!3’ exoRNase Rat1 impact

18S or 25S NRD. Fourth, NRD substrates accumulate strikingly in

cytoplasmic structures (P-bodies for 18S NRD and cytosolic peri-

nuclear foci for 25S NRD). Fifth, NRD-decay occurs with much

slower kineti cs (with half-l ives ranging b etween !50-100 min

[11,13]) than nuclear decay (for example, in the absence of the

export factor Nmd3, nuclear 25S rRNA decays with a half-life of

!4 min [65]). Finally, 18S NRD requires ongoing translation, and

implies that the 18S rRNA is made, i.e. that the last and cytoplasmic

pre-rRNA cleavage has occurred.

Box 2. The exosome and TRAMP

The eukaryotic RNA exosome consists of a core of nine subunits, and

one or two associated RNases (Rrp44, also called Dis3, and Rrp6)

[12,26]. Six subunits of the core (homologs to bacterial phosphorolytic

RNase PH and PNPase, as well as the archaeal exosome) form a

structural inactive ‘ring’. The remaining three core subunits are S1/KH-

RNA binding proteins that provide a cap that bridges the ring subunits

together. Rrp44 is homologous to bacterial RNase II/R and is endowed

with both 3’!5’ exoRNase and endoRNase activities. Rrp44 endoRNase

activity might contribute to the progression of the exosome on its

substrates by resolving complex RNP structures. The exosome

functions in concert with many cofactors that provide substrate

specificity and modulate its functions. In the nucleus, the exosome

functions together with Rrp6 (homologous to bacterial RNase T/D),

another catalytically active 3’!5’ exoRNase subunit of the complex.

Rrp44 (Dis3) and Rrp6 are hydrolytic enzymes that share complemen-

tary biochemical properties: Rrp44 is a processive enzyme, whereas

Rrp6 is distributive. Rrp44 favors both structured and unstructured

substrates and poorly degrades poly(A) and AU-rich sequences in vitro,

whereas Rrp6 shows a marked preference for unstructured substrates

and for poly(A) and AU-rich sequences in vitro. Trf4 and Trf5 are

characterized by distinctive subcellular distribution (Trf5 is mostly

nucleolar; Trf4 is nuclear), cellular abundance (Trf4 is >3x more

abundant than Trf5), and requirement for normal growth (TRF5 deletion

has no associated phenotype whereas TRF4 loss confers a slow-growth

and cryo-sensitive phenotype). These differences probably reflect the

subtle substrate preferences of the two proteins. TRAMP-mediated

RNA polyadenylation stimulates nuclear RNA degradation; this con-

trasts with the role of canonical mRNA poly(A) tails that stimulate

mRNA export, stability and translation. The role of short poly(A) tails as

RNA ‘degradation tags’ has been compared to the role of RNA

oligoadenylation in bacteria and proposed to reflect its ancestral origin

[25,26]. The difference in function of the two types of poly(A) tails could

reflect the difference in processivity of the two RNA polymerases: Pap1

is highly processive, adding long tails of !250 adenines residues in

human cells and !60-90 in yeast, thus probably ensuring that no free 3’-

end is available until a long tail has been generated and covered by

poly(A) binding proteins, such as Pab1. By comparison, Trf4 and Trf5

display low processivity and are distributive (adding short tails of !15-

30 nucleotides) indicating that the 3’-ends of tails added by TRAMP

could be available frequently as exosome substrates.

Review

Trends in Biochemical Sciences Vol.35 No.5

270

authors suggest, could be related to the recently described

juxtanuclear quality control domain (JUNQ) that is

enriched for both misfolded proteins and proteasomes

[21]. Strikingly, localization of 25S NRD substrates to

perinuclear foci relies upon Mms1, indicating that it might

be the site of active degradation.

Mutations in trans: the TRAMP-exosome pathway and

nucleolar surveillance

It has long been appreciated that mutations in ribosome

synthesis factors that speciﬁcally inhibit nuclear pre-rRNA

processing reactions do not systematically lead to the

expected accumulation of pre-rRNA precursors. This ﬁnd-

ing indicates that aberrant nucleolar and nuclear pre-

rRNAs are degraded actively by surveillance mechanisms.

This nuclear surveillance involves, at least in part, the

addition of unstructured oligoadenylate tails at the 3’-end

of ﬂawed pre-rRNAs by a poly(A) polymerase activity

residing in TRAMP complexes, followed by their degra-

dation by the RNA exosome [22]. TRAMP complexes con-

sist of a poly(A) polymerase (either of the paralogous

proteins Trf4 or Trf5 in TRAMP4 and TRAMP5 complexes,

respectively), a zinc-knuckle-containing and putative

RNA-binding protein (either Air1 or Air2) and the DEVH

helicase Mtr4 (also called Dob1) [23–25] (Box 2). In this

pathway, the addition of short poly(A) tails and/or the

actual binding of the TRAMP complex to the RNA commit

aberrant molecules to degradation by both discriminating

them from ‘normal’ RNA and by stimulating exosomal

activity [26].

The early steps of nucleolar pre-rRNA processing (clea-

vage at sites A

and A

in yeast, Figure 3), that separate

the primary RNA polymerase I (Pol I) transcript into

precursors destined to the small and large subunits,

respectively, require a large RNP known as the SSU-

processome that consists of the box C+D snoRNA U3

and !40 associated proteins referred to as UTP (for U

Three-associated Proteins) [27,28]. The SSU-processome

comprises autonomous protein building blocks that are

loaded onto nascent pre-rRNAs and assemble into cataly-

tically active pre-rRNA processing complexes in a stepwise

and highly hierarchical process following alternative path-

ways (Figure 4). At least some of these autonomous

modules are evolutionarily conserved [29,30]. It is cur-

rently understood that failure to assemble the SSU-pro-

cessome with proper kinetics activates nucleolar

surveillance, as depletion of individual SSU-processome

components leads to early nucleolar pre-rRNA processing

inhibition (cleavage at sites A

-A

), the concomitant acti-

vation of a cryptic cleavage at site A

(further downstream

in internal transcribed spacer 1; ITS1, Figure 3) by the

endoRNase MRP, and the synthesis of the aberrant 23S

RNA. This 23S RNA is polyadenylated, mostly by

TRAMP5, and targeted for rapid degradation by the exo-

some (Figure 3) [31,32]. As TRAMP is a distributive

enzyme that adds short poly(A) tails, it is thought that

ﬂawed RNAs undergo multiple rounds of TRAMP-

mediated polyadenylation, followed by exosomal digestion

to achieve greater degradation efﬁcacy. Trf5 co-localizes

with the SSU-processome at the rDNA in living yeast cells,

indicating that this surveillance starts co-transcriptionally

on nascent pre-ribosomes [31]. Interestingly, low levels of

normal pre-rRNA intermediates are polyadenylated and

can be detected in strains defective for the exosome com-

ponent Rrp6 (Box 2) [33,34]. This indicates that every

available 3’-end is probably polyadenylated by TRAMP,

and that physiological pre-rRNA processing sites are also

used as entry points for rRNA surveillance. Moreover,

stabilized pre-rRNA fragments that terminate at multiple

positions within the coding sequence of the 18S rRNA can

be detected in surveillance-defective strains [31,32]. Thus,

several cryptic pre-rRNA processing sites appear to be

activated under conditions of defective ribosome assembly.

The accumulation of such fragments is also observed in

RNA polymerase I mutants defective for transcription

elongation, pointing to a possible regulatory role for

RNA polymerase elongation rates and pausing in ribosome

assembly [35].

The assembly of the large ribosomal subunit includes

many more nuclear steps than that of the 40S subunit, and

involves many more trans-acting factors (Figure 1). Similar

to UTP sub-complexes (Figure 4), several autonomous

structural neighborhoods have been characterized within

pre-60S ribosomes, and emerging evidence indicates that

at least some of them correspond to functional modules

(reviewed in [2]). Precursor rRNAs for large ribosomal

subunits are also monitored by the TRAMP-exosome

surveillance pathway; however, for the large ribosomal

subunit, it is mostly TRAMP4 that tags the RNAs. In

some instances, surveillance has been suggested to

occur in a specialized nucleolar domain termed the ‘No-

body’ that is enriched in both TRAMP and exosomal com-

ponents [36].

Ribosome surveillance exerted by TRAMP–exosome

complexes initiates 3’ to 5’ RNA degradation at multiple

entry points; however, pre-rRNAs are long and highly

structured molecules, and it is expected that 5’!3’ nuclear

and nucleolar exoRNAse activities also contribute to pre-

rRNA decay.

Bulk ribosome and pre-ribosome decay by ribophagy

and PMN

Under limiting growth conditions, ribosome synthesis is

shut down at multiple levels. De novo synthesis is blocked

and nascent pre-ribosomes and mature ribosomes are

targeted to bulk degradative pathways. Pre-ribosomes

and mature ribosomes are turned over to recycle essential

cellular building blocks to cope with stress, and adapt to a

novel environment. It is quite striking that ubiquitin,

which plays an important role in 25S NRD, is also involved

in bulk ribosome degradation.

Ribophagy: a role for deubiquitylation in speciﬁc 60S

macroautophagy

Under starvation conditions, bulk portions of the cytosol,

including protein aggregates and entire organelles, such as

mitochondria and ribosomes, are recycled via two forms of

autophagy (micro- and macroautophagy). Macroautophagy

entails the formation of a double-membrane-bound vesicle

in the cytoplasm: the ‘autophagosome’ that sequesters

cytoplasmic material and delivers it for breakdown and

eventual recycling to the lysosome, or the vacuole in yeast

Review

Trends in Biochemical Sciences Vol.35 No.5

271

and plants (Figure 2). Autophagy can degrade nearly any

biomolecule, as the lysosome/vacuole contains non-speciﬁc

hydrolases, including proteases, nucleases, lipases, and

glycosylases. By contrast, microautophagy entails the

direct uptake of cytosolic components through invagina-

tions of the lysosome’s limiting membrane. Both macro-

and microautophagy contribute to cell survival during

starvation, and the two processes involve common and

speciﬁc sets of trans-acting factors [18,37]. Moreover,

although both micro- and macroautophagy are primarily

bulk degradative processes, selective forms of both path-

ways have been uncovered.

Figure 3. Nucleolar surveillance, the TRAMP-exosome pathway. Failure to assemble pre-ribosomes with proper kinetics leads to activation of nucleolar surveillance and

subsequent rRNA degradation. (a) In wild-type yeast cells, an early step in SSU-processome assembly is the binding of the UTP-A subcomplex (gray) to nascent pre-rRNA.

This is followed by the recruitment of other subcomplexes (UTP-B: orange; and UTP-C: green), additional individual ribosome synthesis factors (blue), the box C+D snoRNA

U3 (gold) and other snoRNAs, leading to SSU-processome catalytic activation in RNA cleavage. The nascent pre-rRNA is cleaved at sites A

and A

(red) leading to the

production of the 20S pre-rRNA (the immediate precursor to the 18S rRNA) that is exported to the cytoplasm, dimethylated and cleaved at site D to produce mature 18S

rRNA. For simplicity, only co-transcriptional cleavage in ITS1, that occurs in !50-70% of the cases in fast-growing yeast cells, is illustrated here. Co-transcriptional cleavage

in ITS1 at site A

occurs when RNA polymerase I has reached the 5’-end of the 25S gene [57]. At this position, pre-40S compaction has brought the SSU-processome at its

closest position relative to the rDNA; in living yeast cells, this correlates with a strong RNA-dependent ChIP interaction of SSU-processome components [31]. In the

remaining cases, cleavage in ITS1 is delayed until transcription has reached the 3’-end of the gene, and a full-length pre-rRNA precursor (35S) has been generated (not

represented). (b) Mutations in any components of the SSU-processome (illustrated here as a red dot on UTP-B) inhibits SSU-processome assembly and/or function, thereby

resulting in inhibition of cleavage events at sites A

-A

, and activation of a downstream cleavage in ITS1 at site A

(purple) by endoRNase MRP. Cleavage of the nascent

transcript at site A

generates an aberrant 23S RNA that is rapidly polyadenylated by TRAMP5, followed by its digestion by the nuclear exosome; no mature 18S rRNA is

made. Upon surveillance inactivation (deletion of Trf5, Rrp6 or both), the 23S RNA is stabilized and not further matured (not shown). (c) The UTP-A subcomplex of the SSU-

processome is dispensable for cleavage at sites A

-A

. As described in panel B for UTP-B, a mutation in a component of UTP-A leads to alterations in pre-rRNA processing

kinetics (A

occurs prior to A

-A

). However, for UTP-A and by ensuring that the newly generated 23S RNA is stabilized by surveillance inactivation, this RNA is further

matured into 20S pre-rRNA and 18S rRNA. Despite this pre-rRNA processing restoration, cells are unable to survive, thus providing a strong indication that surveillance is

essential [31].

Review

Trends in Biochemical Sciences Vol.35 No.5

272

In yeast, prolonged nitrogen starvation leads to prefer-

ential targeting of both small and large ribosomal subunits

to the vacuole by selective macroautophagy [38]. This

process, speciﬁc to ribosomal subunits, is termed ‘ribo-

phagy’. For the large subunits, the increased turnover

kinetics rely upon the Ubp3 ubiquitin protease and its

activator Bre5. Mutations in UBE3 or BRE5 do not affect

detectably the increased turnover rates of small subunit

reporter constructs, indicating that, much as for 18S and

25S NRD, small and large ribosomal subunit ribophagy

relies on distinctly different pathways; notably, however,

the effector(s) for small ribosomal subunit ribophagy

remain unknown. Ubp3 deubiquitylation activity is

required for 60S ribophagy, and the ubiquitylation level

of several, still unidentiﬁed, ribosome-associated proteins

is increased in cells lacking Ubp3 [38]. It is thought that

deubiquitylation of Ubp3–Bre5 target(s) assists in the

packaging of ribosomes in the autophagosome, or enables

its maturation and/or fusion to the vacuole [38]. Recent

data indicate that the ubiquitin ligase Rsp5 might help to

regulate 60S ribophagy as cells harboring mutant versions

of both Rsp5 and Ubp3 show enhanced synthetic sickness

and reduced ribosome turnover [39].

PMN: the involvement of a specialized nucleo-vacuolar

junction

Piecemeal microautophagy of the nucleus (PMN) is a se-

lective autophagic pathway in which the vacuole ‘pinches

off’ and degrades non-essential portions of the nucleus.

This process involves the formation of a speciﬁc inter-

organelle contact dubbed the nucleo-vacuole junction

(NVJ) which acts as true ‘Velcro-like’ patches formed by

interactions between the vacuolar membrane protein Vac8

and the outer nuclear membrane protein Nvj1 [40]. PMN is

a constitutive process that is induced to high levels upon

nutrient starvation. It bears relevance to ribosome degra-

dation, because there are cases where the site of vacuole

docking to the nucleus map to a position juxtaposing the

nucleolus [41] (Figure 2). In these captures, granular

material of nucleolar origin transfers directly from the

nucleolus to the vacuolar lumen; this event is indicative

of bulk destruction of pre-ribosomes.

Ribosome surveillance in human health

RNA damage occurs under normal cell growth as well as

during stress and in disease situations. Chemical modiﬁ-

cations to nucleobases, as well as RNA–RNA and RNA–

Figure 4. Small Subunit (SSU)-processome assembly involves concurrent parallel pathways. The SSU-processome comprises individual autonomous preassembled

building blocks that are sequentially loaded onto nascent pre-rRNA following multiple concurrent pathways, rather than via a single route [28,63]. This strategy, which

minimizes complexity, is likely to be the rule rather than the exception in ribosome assembly, and probably prevails in LSU-processome (pre-60S) formation. The currently

described SSU-processome subcomplexes include the UTP-A (gray), UTP-B (orange) and UTP-C (green), each comprising !6-7 proteins, as well as additional ribosome

synthesis factors (blue), including the Mpp10–Imp3–Imp4 and the Rcl1–Bms1 subcomplexes (reviewed in [2]). Components of UTP-A (also referred to as t-UTPs) influence

efficient Pol I activity in yeast and humans but are not strictly required for pre-rRNA synthesis in yeast [28,30–32]. UTP-C (also known as the CURI complex) integrates pre-

rRNA processing and ribosomal protein production [64]. Specific functions have not yet been assigned to UTP-B. (i) Binding of UTP-A to nascent transcripts is an early step

that initiates assembly. (ii) In one of two currently described alternative pathways, binding of UTP-A is followed by the sequential recruitment of UTP-B, U3 snoRNA and

additional ribosome synthesis factors. (iii) In an alternative route, UTP-A binding initiates the sequential recruitment of Rrp5 and UTP-C.

Review

Trends in Biochemical Sciences Vol.35 No.5

273

protein crosslinks, are introduced into RNA and RNPs as a

result of exposure to ultraviolet light, oxidation, chlori-

nation, nitration and alkylation [42]. These alterations all

constitute potential physiological triggers to ribosome

degradation pathways. Some cases of ribosomal RNA

damage induced by UV irradiation and oxidation have

been documented, notably in conjunction with human

neurodegenerative diseases such as Alzheimer’s and Par-

kinson’s, as well as atherosclerotic plaques [8] (Box 3 ). It

has been suggested that RNA oxidation could be involved

in disease progression and that RNA susceptibility to

oxidative damage is probably inﬂuenced by various factors,

including the degree of association with proteins (protec-

tion) or iron (sensitization). Mature 5.8S and 25S rRNA are

fragmented heavily in yeast cells exposed to elevated levels

of reactive oxygen species (ROS) generated by oxidative

stress (including exposure to hydrogen peroxide and mena-

dione), chronological aging, and other apoptotic cues [43–

45]. Yeast cells treated with the anti-metabolite and che-

motherapeutic agent 5-ﬂuorouracil (5-FU) accumulate

polyadenylated pre-rRNAs, and this accumulation is

exacerbated in exosome mutants, that are hypersensitive

to the drug. These ﬁndings suggest that the degradation of

5-FU containing RNAs occurs via nucleolar surveillance

[34]. Nucleolar dysfunction has been associated with

numerous human diseases, including cancer [46]. Various

‘nucleolar stresses’, such as drug-mediated interference of

rRNA synthesis, pre-rRNA processing (e.g. actinomycin D,

5-FU), or inhibition of ribosome synthesis factor function,

lead to nucleolar disruption, p53 stabilization and, ulti-

mately, cell cycle progression defects and/or apoptosis [47–

50]. Of interest to the general public is the observation of

an increased abundance of polyadenylated rRNA frag-

ments in the gut of western honey bees originating from

populations infected with colony collapse disorder (CCD)

[51]. The insect gut serves as a primary interface with the

environment, as it is the principal site of pesticide detox-

iﬁcation and an integral component in the immune defense

against pathogens. CCD has been linked to picorna-like

viral infections (known to ‘hijack’ cellular ribosomes) and

the extensive use of pesticides; both stimuli could poten-

tially trigger a ribosome degradation response.

Concluding remarks

Ribosome synthesis is a major cellular metabolic activity

that can enforce a rapid ‘energy drain’ in the absence of tight

regulation [52,53]. Control is exerted at the level of the

synthesis and assembly of the pieces, as well as in the

function of the ﬁnal product. Ribosome synthesis has

evolved to be fully integrated with complex nutrient-sensing

cascades, such that defective precursor ribosomes, as well as

damaged and excess mature ribosomes, are targeted speciﬁ-

cally for rapid breakdown and recycling. Several surveil-

lance pathways that either select defective or excess pre-

ribosomes or mature particles have been identiﬁed; they

display strong speciﬁcity towards small and large ribosomal

subunits, involve specialized subcellular locales, and,

altogether, are active at each step of a ribosome’s life. The

question is now to know whether and how these pathways

interconnect (Box 4).

Box 4. Outstanding questions

1. How tightly is pre-RNA synthesis coupled to pre-rRNA proces-

sing and pre-rRNA surveillance? Do polymerase elongation

rates and/or pausing influence nucleolar surveillance?

2. How does TRAMP discriminate aberrant from normal RNAs?

Are other exosome cofactors involved in nuclear ribosome

surveillance?

3. Do any nucle(ol)ar mechanisms of pre-ribosome surveillance

initiate from the 5’-end of the pre-RNAs? There are suitable

exoRNase activities in the nucle(ol)us.

4. To what exte nt is (pre-)rRNA modification monitored by

surveillance?

5. Does PMN contribute quantitatively to nucleolar and nuclear

ribosome surveillance?

6. How does ubiquitin serve both a stabilizing and destabilizing

function in ribosome turnover? Which ribosome-associated

components are ubiquitylated in the 25S NRD and ribophagy

pathways? A ribosomal protein from each subunit (Rps31 and

Rpl40a/b) is produced as a fusion with ubiquitin [74] and Rpl28

ubiquitylation is required for optimal ribosome function [75].

What is the involvement of ubiquitin in nuclear pre-ribosome

turnover? What are the effector(s) of 40S ribophagy?

7. How are defective large ribosomal subunits functionally mon-

itored by 25S NRD?

8. Which endoRNase(s) initiate the cleavage events in mRNA NGD

and in 18S NRD? What is the fate of associated mRNAs in NRD

and, reciprocally, that of associated ribosomes in NGD? Late

cytoplasmic pre-40S ribosomes might enter translation, and

therefore be actively monitored by NRD [76–78].

9. Do ribosome repair mechanisms exist? A priori, this seems

quite unlikely, owing to the compaction of mature subunits. A

demethylase is, however, known to repair alkylated RNAs in

bacteria and eukaryotes [79,80], suggesting that some rRNA

repair could occur during assembly. Evidence for ribosome

rejuvenation by ribosomal protein replacement is suggested by

in vitro translational reactivation of chemically-damaged bacter-

ial ribosomes [81].

10. How does ribosome surveillance failure contribute to human

disease etiology and progression?

Box 3. RNA damage in human health

Environmentally relevant doses of UVA and UVB are sufficient to

induce lesions in nucleic acids. UV is a source of photoproducts,

crosslinking and oxidative damage. UV crosslinking is dependent on

the geometry (distance, angle) and photoreactivity of the nucleotide

and amino acids, suggesting that certain cellular RNAs are more likely

to be protected from alterations by UV crosslinking than others (e.g.

double stranded RNA region). UV-mediated rRNA damage has been

reported in cultured mammalian cells, consistent with the formation

of photoproducts [66]. The lesions clustered around the active sites of

the large ribosomal subunit (domains V and VI near the PTC), and

were correlated with reduced translation and the activation of a

kinase-mediated stress response. Under oxidizing conditions, ribo-

somes undergo two types of damage: formation of 8-oxo-7,8-

dihydroguanosine (8-oxoG) and cross-linking of rRNA and ribosomal

proteins (shown in yeast cells treated with H

[67]). Most oxidative

damage results from the action of ROS that arise from metabolic

reactions. Increased levels of oxidatively damaged RNAs, as revealed

by the presence of 8-oxoG, have been reported in neurons from

patients with human neurodegenerative diseases, including Alzhei-

mer’s (AD; where it has been detected in nucleoli) and Parkinson’s, as

well as dementia, and in patients with atherosclerosis [68–72]. Studies

in neurons from patients with AD revealed higher levels of oxidized

rRNA [68,69], and ribosomes purified from AD patients showed

elevated levels of associated redox-active iron [73]. Advanced human

atherosclerotic plaques have been linked to severe loss of rRNA

integrity and accumulation of elevated levels of 8-oxoG-containing

RNAs, a probable consequence of oxidative stress, ROS production

and intraplaque hemorrhage that is accompanied by iron deposition

[70,71].

Review

Trends in Biochemical Sciences Vol.35 No.5

274

How are defective (pre-)ribosomes recognized?

The exact molecular basis for recognition is not yet known;

however, the huge diversity of potential ribosome insults

makes it extremely unlikely that each one is monitored by

individual surveillance mechanisms. It is thought that each

reaction in the ribosome assembly pathway is provided with

a time frame for completion and that surveillance is trig-

gered by inefﬁcient kinetics. Work with mature ribosomes

indicates that, in this case, it is the overall dynamics of the

translation process that is monitored. Thus far, this prop-

erty is best demonstrated for small subunits where stalling

ribosomes are identiﬁed by ‘translation-factor like’ proteins

that are also active in mRNA NGD. For 25S NRD, the

kinetics of interaction with translation factors might be

the key parameter that is monitored. It can thus be anticip-

ated that rRNA sequences that have been the most highly

conserved throughout evolution are probably prime targets

for monitoring by surveillance. Likewise, pre-rRNA modi-

ﬁcations that cluster around functional ribosomal sites and

contribute to translation efﬁcacy, are also expected to be

monitored. The specialized functions of the two ribosomal

subunits in translation is reﬂected in their overall 3-D

structure and shape, with a ﬂexible ‘Y-shape’ for the small

subunit and a monolithic block for the large subunit that

might require additional ‘effort’ (i.e., ubiquitylation, and

possible dissociation, of some of the protein components)

to expose naked RNAs to salient cellular endo- and exoR-

Nase activities. The selective cellular compartmentalization

of aberrant pre-ribosomes (No-body), and mature ribosomes

(P-bodies), in speciﬁc organelles might also contribute to the

processes of discrimination and degradation. Functional

compartmentalization could be particularly relevant to

25S NRD substrates that accumulate in cytosolic peri-

nuclear foci and primarily colocalize with 60S subunits in

velocity gradients. These ﬁndings indicate that defective

large subunits might be segregated away from the transla-

tionally active pool of ribosomes. By contrast, 18S NRD

RNAs co-migrate with monosomes, 80S particles and poly-

somes; these ﬁndings are consistent with their monitoring

during translation [11,13,17].

How are defective ribosomes tagged for degradation?

Thus far, two mechanisms have been identiﬁed that involve

either the addition of short poly(A) tails at the 3’-ends of the

RNA in the nucleus or the ubiquitylation/deubiquitylation of

unknown associated proteins in the cytoplasm. Strikingly,

ubiquitin plays both a stabilizing and destabilizing function;

clearly, its role in ribosome turnover is only starting to

unfold. That ubiquitylation also plays a role in nuclear

pre-ribosome degradation is suggested by the severe pre-

rRNA processing inhibition, and production of aberrant and

truncated rRNA fragments that are observed in yeast upon

conditional inactivation of the ubiquitin ligase Rsp5 [54].

Mammalian proteasome inhibition also leads to pre-rRNA

processing defects as well as striking nuclear accumulation

of ribosomal proteins; such ﬁndings are consistent with the

existence of further quality control systems [55,56].

Do we need ribosome surveillance?

The physiological relevance of ribosome surveillance is

illustrated by emerging connections to human diseases

(Box 3), and observations that cells harboring an inacti-

vated nucleolar pathway, 25S NRD or ribophagy, accumu-

late defective or excess ribosomes, and are either unable to

survive or show a much decreased lifespan [17,31,32,38].

Proteomic analyses of puriﬁed nucleoli and pre-ribo-

somes have left us with a myriad of trans-acting factors

to which function(s) and structure must be assigned. The

ﬁeld is entering a new era in which deciphering the post-

translational modiﬁcation of synthesis factors and riboso-

mal proteins will dictate future progress in understanding

the regulated assembly and turnover of ribosomes.

Acknowledgements

Makoto Kitabatake (Kyoto University), Natalia Shcherbik and Dimitri

Pestov (University of Medicine and Dentistry of New Jersey), Martin Kos

and David Tollervey (University of Edinburgh) are acknowledged for

communicating unpublished observat ions prior to publication. I

apologized to my friends and colleagues whose work could not be listed

here owing to space limitation. The laboratory is funded by the FRS-

FNRS, Re

gion Wallonne (Cibles) and Communaute

Franc¸aise de Belgique

(ARC).

Appendix A. Supplementary data

Supplementary data associated with this article can be

found, in the online version, at doi:10.1016/j.tibs.2009.

12.006.

References

1 Steitz, T.A. (2008) A structural understanding of the dynamic ribosome

machine. Nat. Rev. Mol. Cell Biol. 9, 242–253

2 Henras, A.K. et al. (2008) The post-transcriptional steps of eukaryotic

ribosome biogenesis. Cell. Mol. Life Sci. 65, 2334–2359

3 Strunk, B.S. and Karbstein, K. (2009) Powering through ribosome

assembly. RNA 15, 2083–2104

4 Warner, J.R. and McIntosh, K.B. (2009) How common are

extraribosomal functions of ribosomal proteins? Mol. Cell 34, 3–11

5 Lafontaine, D. et al. (1995) The 18S rRNA dimethylase Dim1p is

required for pre-ribosomal RNA processing in yeast. Genes Dev. 9,

2470–2481

6 Senger, B. et al. (2001) The nucle(ol)ar Tif6p and Eﬂ1p are required for

a late cytoplasmic step of ribosome synthesis. Mol. Cell 8, 1363–

1373

7 Gadal, O. et al. (2002) Rlp7p is associated with 60S preribosomes,

restricted to the granular component of the nucleolus, and required for

pre-rRNA processing. J. Cell Biol. 157, 941–951

8 Nunomura, A. et al. (2009) RNA oxidation in Alzheimer disease and

related neurodegenerative disorders. Acta Neuropathol. 118, 151–166

9 Doma, M.K. and Parker, R. (2007) RNA quality control in eukaryotes.

Cell 131, 660–668

10 Doma, M.K. and Parker, R. (2006) Endonucleolytic cleavage of

eukaryotic mRNAs with stalls in translation elongation. Nature 440,

561–564

11 LaRiviere, F.J. et al. (2006) A late-acting quality control process for

mature eukaryotic rRNAs. Mol. Cell 24, 619–626

12 Schmid, M. and Jensen, T.H. (2008) The exosome: a multipurpose

RNA-decay machine. Trends Biochem. Sci. 33, 501–510

13 Cole, S.E. et al. (2009) A convergence of rRNA and mRNA quality

control pathways revealed by mechanistic analysis of nonfunctional

rRNA decay. Mol. Cell 34, 440–450

14 Graille, M. et al. (2008) Structure of yeast Dom34: a protein related to

translation termination factor Erf1 and involved in No-Go decay. J.

Biol. Chem. 283, 7145–

7154

15 Carr-Schmid, A. et al. (2002) Novel G-protein complex whose

requirement is linked to the translational status of the cell. Mol.

Cell. Biol. 22, 2564–2574

16 Balagopal, V. and Parker, R. (2009) Polysomes, P bodies and stress

granules: states and fates of eukaryotic mRNAs. Curr. Opin. Cell Biol.

21, 403–408

17 Fujii, K. et al. (2009) A role for ubiquitin in the clearance of

nonfunctional rRNAs. Genes Dev. 23, 963–974

Review

Trends in Biochemical Sciences Vol.35 No.5

275

18 Beau, I. et al. (2008) Lost to translation: when autophagy targets

mature ribosomes. Trends Cell Biol. 18, 311–314

19 Suter, B. et al. (2007) Examining protein protein interactions using

endogenously tagged yeast arrays: the cross-and-capture system.

Genome Res. 17, 1774–1782

20 Krogan, N.J. et al. (2006) Global landscape of protein complexes in the

yeast Saccharomyces cerevisiae. Nature 440, 637–643

21 Kaganovich, D. et al. (2008) Misfolded proteins partition between two

distinct quality control compartments. Nature 454, 1088–1095

22 Houseley, J. et al. (2006) RNA-quality control by the exosome. Nat. Rev.

Mol. Cell Biol. 7, 529 – 539

23 LaCava, J. et al. (2005) RNA degradation by the exosome is promoted

by a nuclear polyadenylation complex. Cell 121, 713–724

24 Wyers, F. et al. (2005) Cryptic pol II transcripts are degraded by a

nuclear quality control pathway involving a new poly(A) polymerase.

Cell 121, 725–737

25 Vanacova, S. et al. (2005) A new yeast poly(A) polymerase complex

involved in RNA quality control. PLoS Biology 3, e189

26 Andersen, K.R. et al. (2008) Take the ‘‘A’’ tail - quality control of

ribosomal and transfer RNA. Biochim. Biophys. Acta 1779, 532–537

27 Dragon, F. et al. (2002) A large nucleolar U3 ribonucleoprotein required

for 18S ribosomal RNA biogenesis. Nature 417, 967–970

28 Gallagher, J.E. et al. (2004) RNA polymerase I transcription and pre-

rRNA processing are linked by speciﬁc SSU processome components.

Genes Dev. 18, 2506–2517

29 Turner, A.J. et al. (2009) A novel small-subunit processome assembly

intermediate that contains the U3 snoRNP, nucleolin, RRP5, and

DBP4. Mol. Cell. Biol. 29, 3007–3017

30 Prieto, J.L. and McStay, B. (2007) Recruitment of factors linking

transcription and processing of pre-rRNA to NOR chromatin is

UBF-dependent and occurs independent of transcription in human

cells. Genes Dev. 21, 2041–2054

31 Wery, M.

et al. (2009) The nuclear poly(A) polymerase and Exosome

cofactor Trf5 is recruited cotranscriptionally to nucleolar surveillance.

RNA 15, 406–419

32 Dez, C. et al. (2007) Roles of the HEAT repeat proteins Utp10 and

Utp20 in 40S ribosome maturation. RNA 13, 1516–1527

33 Kuai, L. et al. (2004) Polyadenylation of rRNA in Saccharomyces

cerevisiae. Proc. Natl. Acad. Sci. U. S. A. 101, 8581–8586

34 Fang, F. et al. (2004) 5-ﬂuorouracil enhances exosome-dependent

accumulation of polyadenylated rRNAs. Mol. Cell. Biol. 24, 10766–

10776

35 Schneider, D.A. et al. (2007) Transcription elongation by RNA

polymerase I is linked to efﬁcient rRNA processing and ribosome

assembly. Mol. Cell 26, 217–229

36 Dez, C. et al. (2006) Surveillance of nuclear-restricted pre-ribosomes

within a subnucleolar region of Saccharomyces cerevisiae. EMBO J. 25,

1534–1546

37 Nakatogawa, H. et al. (2009) Dynamics and diversity in autophagy

mechanisms: lessons from yeast. Nat. Rev. Mol. Cell Biol. 10, 458–

467

38 Kraft, C. et al. (2008) Mature ribosomes are selectively degraded upon

starvation by an autophagy pathway requiring the Ubp3p/Bre5p

ubiquitin protease. Nat. Cell Biol. 10, 602–610

39 Kraft, C. and Peter, M. (2008) Is the Rsp5 ubiquitin ligase involved in

the regulation of ribophagy? Autophagy 4, 838–840

40 Pan, X. et al. (2000) Nucleus-vacuole junctions in Saccharomyces

cerevisiae are formed through the direct interaction of Vac8p with

Nvj1p. Mol. Biol. Cell 11, 2445–2457

41 Roberts, P. et al. (2003) Piecemeal microautophagy of nucleus in

Saccharomyces cerevisiae. Mol. Biol. Cell 14, 129–141

42 Wurtmann, E.J. and Wolin, S.L. (2009) RNA under attack: cellular

handling of RNA damage. Crit. Rev. Biochem. Mol. Biol. 44, 34– 49

43 Mroczek, S. and Kufel, J. (2008) Apoptotic signals induce speciﬁc

degradation of ribosomal RNA in yeast. Nucleic Acids Res. 36, 2874–

2888

44 Thompson, D.M.

et al. (2008) tRNA cleavage is a conserved response to

oxidative stress in eukaryotes. RNA 14, 2095–2103

45 Thompson, D.M. and Parker, R. (2009) The RNase Rny1p cleaves

tRNAs and promotes cell death during oxidative stress in

Saccharomyces cerevisiae. J. Cell Biol. 185, 43–50

46 Derenzini, M. et al. (2009) What the nucleolus says to a tumour

pathologist. Histopathology 54, 753–762

47 Mayer, C. and Grummt, I. (2005) Cellular stress and nucleolar

function. Cell Cycle 4, 1036–1038

48 Rubbi, C.P. and Milner, J. (2003) Disruption of the nucleolus mediates

stabilization of p53 in response to DNA damage and other stresses.

EMBO J. 22, 6068–6077

49 Azuma, M. et al. (2006) Perturbation of rRNA synthesis in the bap28

mutation leads to apoptosis mediated by p53 in the zebraﬁsh central

nervous system. J. Biol. Chem. 281, 13309–13316

50 Pestov, D.G. et al. (2001) Evidence of p53-dependent cross-talk between

ribosome biogenesis and the cell cycle: effects of nucleolar protein Bop1

on G(1)/S transition. Mol. Cell. Biol. 21, 4246–4255

51 Johnson, R.M. et al. (2009) Changes in transcript abundance relating to

colony collapse disorder in honey bees (Apis mellifera). Proc. Natl.

Acad. Sci. U. S. A. 106, 14790–14795

52 Lempiainen, H. and Shore, D. (2009) Growth control and ribosome

biogenesis. Curr. Opin. Cell Biol. 21, 1–9

53 Jorgensen, P. et al. (2004) Forging the factory: ribosome synthesis and

growth control in budding yeast. In Cell Growth: Control of Cell Size

(Hall, M.N. et al., eds), pp. 329–370, CSHL Press

54 Neumann, S. et al. (2003) Formation and nuclear export of tRNA, rRNA

and mRNA is regulated by the ubiquitin ligase Rsp5p. EMBO Rep. 4,

1156–1162

55 Lam, Y.W. et al. (2007) Analysis of nucleolar protein dynamics reveals

the nuclear degradation of ribosomal proteins. Curr. Biol. 17, 749–760

56 Stavreva, D.A. et al. (2006) Potential roles for ubiquitin and the

proteasome during ribosome biogenesis. Mol. Cell Biol. 26, 5131–5145

57 Osheim, Y.N.

et al. (2004) Pre-18S ribosomal RNA is structurally

compacted into the SSU processome prior to being cleaved from

nascent transcripts in Saccharomyces cerevisiae. Mol. Cell 16, 943–954

58 Honma, Y. et al. (2006) TOR regulates late steps of ribosome

maturation in the nucleoplasm via Nog1 in response to nutrients.

EMBO J. 25, 3832–3842

59 Vanrobays, E. et al. (2008) TOR regulates the subcellular distribution

of DIM2, a KH domain protein required for cotranscriptional ribosome

assembly and pre-40S ribosome export. RNA 14, 2061–2073

60 Schafer, T. et al. (2006) Hrr25-dependent phosphorylation state

regulates organization of the pre-40S subunit. Nature 441, 651–655

61 Ulbrich, C. et al. (2009) Mechanochemical removal of ribosome

biogenesis factors from nascent 60S ribosomal subunits. Cell 138,

911–922

62 Krick, R. et al. (2008) Piecemeal microautophagy of the nucleus

requires the core macroautophagy genes. Mol. Biol. Cell 19, 4492–

4505

63 Perez-Fernandez, J. et al. (2007) The 90S preribosome is a

multimodular structure that is assembled through a hierarchical

mechanism. Mol. Cell. Biol. 27, 5414–5429

64 Rudra, D. et al. (2007) Potential interface between ribosomal protein

production and pre-rRNA processing. Mol. Cell Biol. 27, 4815–4824

65 Ho, J.H. and Johnson, A.W. (1999) NMD3 encodes an essential

cytoplasmic protein required for stable 60S ribosomal subunits in

Saccharomyces cerevisiae. Mol. Cell Biol. 19, 2389–2399

66 Iordanov, M.S. et al. (1998) Ultraviolet radiation triggers the ribotoxic

stress response in mammalian cells. J. Biol. Chem. 273, 15794–15803

67 Mirzaei, H. and Regnier, F. (2006) Protein-RNA cross-linking in the

ribosomes of yeast under oxidative stress. J. Proteome Res. 5, 3249–

3259

68 Ding, Q. et al. (2006) Decreased RNA, and increased RNA oxidation, in

ribosomes from early Alzheimer’s disease. Neurochem. Res. 31, 705–

710

69 Ding, Q. et al. (2005) Ribosome dysfunction is an early event in

Alzheimer’s disease. J. Neurosci.

25, 9171–9175

70 Martinet, W. et al. (2005) RNA damage in human atherosclerosis:

pathophysiological signiﬁcance and implications for gene expression

studies. RNA Biol. 2, 4–7

71 Martinet, W. et al. (2004) React ive oxygen species induce RNA damage

in human atherosclerosis. Eur. J. Clin. Invest. 34, 323–327

72 Nunomura, A. et al. (1999) RNA oxidation is a prominent feature of

vulnerable neurons in Alzheimer’s disease. J. Neurosci. 19, 1959–1964

73 Honda, K. et al. (2005) Ribosomal RNA in Alzheimer disease is oxidized

by bound redox-active iron. J. Biol. Chem. 280, 20978–20986

74 Finley, D. et al. (1989) The tails of ubiquitin precursors are ribosomal

proteins whose fusion to ubiquitin facilitates ribosome biogenesis.

Nature 338, 394–401

Review

Trends in Biochemical Sciences Vol.35 No.5

276

75 Spence, J. et al. (2000) Cell cycle-regulated modiﬁcation of the ribosome

by a variant multiubiquitin chain. Cell 102, 67–76

76 Lacombe, T. et al. (2009) Linear ubiquitin fusion to Rps31 and its

subsequent cleavage are required for the efﬁcient production and

functional integrity of 40S ribosomal subunits. Mol. Microbiol. 72,

69–84

77 Pertschy, B. et al. (2009) RNA helicase Prp43 and its co-factor Pfa1

promote 20S to 18S rRNA processing catalyzed by the endonuclease

Nob1. J. Biol. Chem. 284, 35079–35091

78 Soudet, J. et al. (2010) Immature small ribosomal subunits can engage

in translation initiation in Saccharomyces cerevisiae. EMBO J 29, 80–

79 Aas, P.A. et al. (2003) Human and bacterial oxidative demethylases

repair alkylation damage in both RNA and DNA. Nature 421, 859–863

80 Ougland, R. et al. (2004) AlkB restores the biological function of mRNA

and tRNA inactivated by chemical methylation. Mol. Cell 16, 107–116

81 Pulk, A. et al. (2009) Ribosome reactivation by replacement of damaged

proteins. Mol. Microbiol Epub ahead of print PMID 19968789

Review

Trends in Biochemical Sciences Vol.35 No.5

277