Arabidopsis small nucleolar RNA monitors the efficientpre-rRNA processing during ribosome biogenesisPan Zhua,1, Yuqiu Wanga,1, Nanxun Qina, Feng Wanga, Jia Wanga, Xing Wang Denga,2, and Danmeng Zhua,2

aState Key Laboratory of Protein and Plant Gene Research, Peking-Tsinghua Center for Life Sciences, School of Advanced Agriculture Sciences and School ofLife Sciences, Peking University, Beijing 100871, China

Contributed by Xing Wang Deng, September 10, 2016 (sent for review August 13, 2016; reviewed by James A. Birchler and Jiming Jiang)

Ribosome production in eukaryotes requires the complex and precisecoordination of several hundred assembly factors, including manysmall nucleolar RNAs (snoRNAs). However, at present, the distinct roleof key snoRNAs in ribosome biogenesis remains poorly understoodin higher plants. Here we report that a previously uncharacterizedC (RUGAUGA)/D (CUGA) type snoRNA,HIDDEN TREASURE 2 (HID2), actsas an important regulator of ribosome biogenesis through a snoRNA–rRNA interaction. Nucleolus-localized HID2 is actively expressed inArabidopsis proliferative tissues, whereas defects in HID2 cause aseries of developmental defects reminiscent of ribosomal protein mu-tants. HID2 associates with the precursor 45S rRNA and promotes theefficiency and accuracy of pre-rRNA processing. Intriguingly, disrupt-ing HID2 in Arabidopsis appears to impair the integrity of 27SB, akey pre-rRNA intermediate that generates 25S and 5.8S rRNA and isknown to be vital for the synthesis of the 60S large ribosomal subunitand also produces an imbalanced ribosome profile. Finally, we dem-onstrate that the antisense-box of HID2 is both functionally essentialand highly conserved in eukaryotes. Overall, our study reveals thevital and possibly conserved role of a snoRNA in monitoring the effi-ciency of pre-rRNA processing during ribosome biogenesis.

C/D box small nucleolar RNA | pre-rRNA processing | ribosome biogenesis

The eukaryotic ribosome responsible for protein synthesis iscomposed of a 60S large ribosomal subunit and a 40S small

ribosomal subunit. To date, ribosome biogenesis has been stud-ied most extensively in yeast. The generally conserved process,which encompasses a number of highly orchestrated steps, ispivotal to all organisms and requires the generation of maturerRNAs, ribosomal proteins with hundreds of processing and as-sembly factors, and many small nucleolar RNAs (snoRNAs) (1, 2).In plants, mutants in genes encoding essential ribosomal biogenesisfactors display broad abnormal growth and developmental pheno-types and altered responses to stress stimuli (3). Likewise, a varietyof human diseases have been shown to be caused by defects in ri-bosome assembly (4).In plants, ribosome biogenesis begins in the nucleolus when 45S

rRNA transcription is initiated by RNA polymerase I (5). Theprimary long transcript is comprised of three ribosomal RNAs, 18S,5.8S, and 25S rRNA, which are separated by two internal tran-scribed spacers (ITS1 and ITS2) and flanked by external transcribedspacers (5′- and 3′-ETS) (6). After the initial splicing steps, theresulting 35S rRNA primarily follows the “ITS1-first” pathway,which is common among metazoans, to produce mature rRNAs.This pathway starts with an early cleavage of ITS1 at the A3 site,followed by complete elimination of the 5′-ETS (7, 8). Recently aminor pathway, known as “5′ ETS-first pathway,” also has beenreported to coexist in plants and has been shown to resemble thestrict processing pathway in yeast (9, 10). Each of the processes ofrRNA transcription, processing, and modification has been shownto require snoRNAs (11).Plants possess an abundance of snoRNAs, which are primarily

transcribed in polycistronic gene clusters (12). snoRNAs rangingfrom 70 to 250 nt have been shown to work in concert with ribo-nucleoproteins (RNPs) during the RNA-modification process (13).C (RUGAUGA)/D (CUGA) box and H (ANANNA)/ACA box

snoRNAs constitute the two main classes of snoRNAs, with anumber of the former group directing the 2′-O-methylation ofthe ribose and a number of the latter group guiding pseu-douridination (14, 15). Both types of small nucleolar RNP(snoRNP) recognize their target modification site via specificsequence base-paring.At present, few snoRNAs involved in rRNA processing and/or

rRNA folding have been explored. The U3 C/D box snoRNPshown to be universally conserved across eukaryotes is the corecomponent of the small subunit processome essential for 18SrRNA processing (16, 17). In plants, it has been reported that theU3 snoRNP forms a stable complex with nucleolin protein 1(NUC1), which binds nascent pre-RNA at the 5′ ETS and spe-cifically cleaves pre-rRNA at the P site (18–20). Moreover,NUC1-U3 snoRNP binds rDNA before its interaction with pre-rRNA (20, 21). A few other snoRNAs also have been reported tobe involved in pre-rRNA processing. For instance, in yeast U14,snR30, and snR10 are required for 18S rRNA processing, and inmetazoans U8 and U24 have been demonstrated to be involvedin large-subunit processing (22–25). Overall, however, with theexception of U3 snoRNA, little is known regarding the roles ofplant snoRNAs in the regulation of rRNA processing.Our previous work has annotated dozens of snoRNA gene

clusters, as well as single snoRNAs, in both rice and Arabi-dopsis (26, 27). Here, taking advantage of our establishedcollection of sequenced Arabidopsis snoRNAs, we identify andcharacterize HIDDEN TREATURE 2 (HID2), an Arabidopsis C/D

Significance

Box C (RUGAUGA)/D (CUGA) and H (ANANNA)/ACA small nu-cleolar RNAs (snoRNAs) are important for the modification andprocessing of rRNA during ribosome biogenesis in eukaryotes.However, the molecular role of snoRNAs throughout the mul-tiple steps of pre-rRNA processing remains poorly understood.This study shows that an uncharacterized C/D box snoRNA,HIDDEN TREASURE 2 (HID2), functions as a prominent player inthe monitoring of efficient pre-rRNA processing, which, in turn,is essential for accurate ribosome assembly in Arabidopsis. Ourdata explore a link between a spatially regulated snoRNA andthe complexity and the precise control of ribosome biogenesis.Further, the conservation of HID2’s signature motif and func-tion highlights its importance in multicellular organisms. hid2appears to be a representative snoRNA mutant exhibiting pleio-tropic developmental defects in plants.

Author contributions: P.Z., Y.W., F.W., X.W.D., and D.Z. designed research; P.Z., Y.W., N.Q.,F.W., J.W., and D.Z. performed research; P.Z., Y.W., N.Q., F.W., J.W., X.W.D., and D.Z.analyzed data; and P.Z., Y.W., X.W.D., and D.Z. wrote the paper.

Reviewers: J.A.B., University of Missouri-Columbia; and J.J., University of Wisconsin-Madison.

The authors declare no conflict of interest.1P.Z. and Y.W. contributed equally to this work.2To whom correspondence may be addressed. Email: zhudanmeng@pku.edu.cn or deng@pku.edu.cn.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1614852113/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1614852113 PNAS | October 18, 2016 | vol. 113 | no. 42 | 11967–11972

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 22

, 202

0

box snoRNA that we show to be enriched in proliferating tissuesknown to be active in rRNA processing. Knocking down HID2results in an increase of both 45S rRNA and pre-rRNA processingintermediates, overaccumulation of 27SB rRNA (5.8S-ITS2-25S)aberrant products, which most likely impede ribosomal assembly,and the occurrence of pleiotropic developmental defects. Thus, byelucidating the regulation of pre-rRNA processing pathways bysnoRNAs, our study of HID2 provides insights into ribosomesynthesis.

ResultsReduction in Arabidopsis HID2 Expression Impairs Normal Growth andDevelopment. To identify snoRNAs that serve as prominent playersin Arabidopsis growth and development, we screened a collection ofAgrobacterium transferred DNA (T-DNA) insertion mutants (27).The hid2 mutant identified in this screen defectively expressed fivesnoRNAs organized in a polycistronic gene cluster and exhibitedretarded and pointed leaf-growth phenotypes (Fig. S1).To determine whether the leaf phenotypes observed in the hid2

mutant were a direct result of the decrease in snoRNAs, wetransformed a genomic DNA fragment encoding all five snoRNAs(nc0216–nc0220) under the control of their endogenous promoterinto an hid2 mutant background (Fig. 1A). The resulting trans-genic plants (pHID2: ALL/hid2) expressed all five snoRNAs atlevels similar to or higher than their levels in WT plants andshowed a complete rescue of the hid2 phenotype (Fig. 1B), in-dicating that the absence of these specific snoRNAs was re-sponsible for the hid2 phenotypes observed at the seedling stage.Next, to determine which of these snoRNAs was the predominantplayer, we expressed the first snoRNA in the cluster, nc0216,under regulation of its endogenous promoter (pHID2:HID2) inhid2 mutants (Fig. 1A). Interestingly, the expression of nc0216alone in hid2 mutants resulted in a dosage-dependent recovery ofthe mutant phenotype in our two independent transgenic lines(Fig. 1B). Furthermore, we found that the pleiotropic phenotypesobserved in the hid2 mutant, including delayed seed germination,retarded root growth, and narrow, pointed leaves at the adultstage, were all complemented to WT in the pHID2:HID2/hid2transgenes (Fig. S2). In contrast, transgenic lines expressing theother four snoRNAs (pHID2:ΔHID2/hid2) at WT or higher levels

were observed to exhibit the hid2 mutant phenotype (Fig. 1B).These results suggest that nc0216 is physiologically important inhid2 mutants. We also analyzed the expression of genes located ateither end of the nc0216-coding sequence in WT, hid2, andpHID2:HID2/hid2 plants using qRT-PCR (Fig. S3). Our resultsrevealed that the expression levels of the neighboring genes wereeither comparable in all three genotypes or were not correlatedwith the levels of nc0216 expression. Together, our data indicatethat nc0216 is a major driver of hid2, and we therefore refer tonc0216 as “HID2” hereafter. hid2 is a representative snoRNAmutant exhibiting developmental defects in plants. HID2 waspreviously annotated as a 90-nt C/D box snoRNA (27).

The Expression of HID2, Driven by a Site II Cis Element, Is Enriched inActively Proliferating Tissues. To gain initial insight into the de-velopmental control of HID2, we first characterized the promoteractivity of HID2 by introducing a pHID2:GUS reporter gene intoWT plants. In shoots, β-glucuronidase (GUS) activity was signifi-cantly enriched in the shoot apex but not in the cotyledon. Thisresult was also confirmed by Northern blot analysis using a HID2-specific antisense probe (Fig. 2 A and B). We further examined theexpression pattern of HID2 at the shoot apical meristem using insitu hybridization. We detected high HID2 expression in the leafprimordia and growing young leaves and weak expression in thevascular tissues (Fig. 2 C and D). In addition, our FISH analysisvalidated the localization ofHID2 in the nucleolus (Fig. 2E). Thus,our results indicate that nucleolus-localized HID2 is preferentiallyexpressed in the rapidly proliferating tissues of Arabidopsis.To understand better the spatial control of HID2’s expression

pattern, we analyzed the HID2 promoter sequence using MEMEsoftware that enabled us to identify the specific motifs drivingHID2 expression in vivo. We found two site II (TGGGC) motifsand one site II-related (GCCCA) motif in the HID2 promoters(Fig. 2F). Site II motifs have been shown to be present in pro-moters of some, but not all, snoRNA genes. They also have beenshown to be present in ribosomal genes (28–30). To examine therelation between this cis element and the expression of HID2 inplanta, we constructed pHID2WT:GUS and pHID2mut:GUS and

Fig. 1. Down-regulation of HID2 results in developmental defects. (A) Sche-matic illustration of constructs containing all the members of the snoRNA genecluster (pHID2:ALL), nc0216 only (pHID2:HID2), and nc0217–nc0220 (pHID2:ΔHID2)driven by the endogenous HID2 promoter. (B, Upper) Phenotypes of WT, hid2,and the indicated transgenic seedlings. (Scale bar: 5 mm.) (Lower) Northernblot analysis showing the expression levels of nc0216–nc0220, with 5.8S rRNAas the loading control.

Fig. 2. Nucleolus-localized HID2 is highly expressed in cell proliferating re-gions. (A) Histochemical staining showing the expression pattern of pHID2:GUSin 8-d-old WT seedlings. The apex and cotyledon (CE) are indicated. (B) Ex-pression of HID2 in the cotyledon and apex, with 5S rRNA as the loadingcontrol. (C and D) In situ hybridizations of the apex in WT seedlings with theHID2 antisense (C) and sense (D) probes. (Scale bars: 100 μm.) (E) Nucleolarlocalization of HID2. FISH using the HID2 antisense (AS) and sense (SE) probescompared with nuclear staining with DAPI. (Scale bars: 1 μm.) (F) Schematicillustration showing the positions of site II and related motifs in the HID2promoter. (G–J) GUS staining of pHID2WT:GUS (G) and pHID2mut:HID2 (H–J)transgenic seedlings. (Scale bars: 2 mm.)

11968 | www.pnas.org/cgi/doi/10.1073/pnas.1614852113 Zhu et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 22

, 202

0

compared their activity. We observed that mutation of either the siteII (m12) or site II-related (m3) motifs reduced the GUS reporteractivity (Fig. 2 H and I), and mutation of all three motifs (m123)abolished GUS activity (Fig. 2J). Taken together, these results in-dicate that site II cis elements are essential drivers of HID2 ex-pression in vivo. Given that the site II element can be recognized byTEOSINTE BRANCHED1/CYCLOIDEA/PROLIFERATINGCELL FACTOR1 (TCP), a family of key transcriptional factorsknown to regulate cell proliferation in planta (31, 32), we next testedthe binding of selected TCPs to the WT and mutated HID2 pro-moter by gel-shift analysis. Both TCP20 and TCP24 were found tobind the HID2 WT promoter, but not the mutant HID2 promoterm123, directly in vitro (Fig. S4). This finding suggests that TCPproteins may be important determinants of HID2 transcription.

2′-O-Methylation on the 25S rRNA G2620 Apparently Is Not Affectedin hid2. To gain insight into HID2 function, we first searched forcanonical targets of HID2 in Arabidopsis. We found that the HID2antisense-box element uniquely matched a conserved fragment in25S rRNA. Within this fragment, the G2620 site is predicted toundergo 2′-O-methylation (Fig. S5 A and B). To test the connec-tion between the 2′-O-methylation of 25S rRNAG2620 andHID2,we performed reverse transcription at low deoxy-ribonucleotidetriphosphate concentrations (RTL-P) followed by PCR inWT andhid2 mutants (33). Our data support the presence of a 2′-O-methylated site within the tested region. However, no obvious dif-ference in band intensity was observed in either the cotyledon orapex of the WT and hid2 samples (Fig. S5C), and this result wasfurther supported using primer extension analysis (Fig. S5D).

A Decrease in HID2 Expression Causes an Increase in the TranscriptionRate of rDNA in Arabidopsis. To detect the steady-state levels ofpre-rRNA precursors and intermediates in each processing step,Northern blot analyses were performed, showing an accumulationof 45S and 35S pre-rRNA precursors in hid2mutants (Fig. 3 A andB and Fig. S6A). To examine whether rDNA transcription wasincreased in hid2 mutants, we conducted a nuclear run-on (NRO)transcription analysis using WT and hid2 seedling samples. Ourrun-on results revealed an increased 45S rDNA transcription ratein hid2 (Fig. 3C). Selective inhibition of DNA-dependent RNApolymerase I activity by actinomycin D treatment suggested thatthe degradation rate of 45S rRNA in hid2 is not reduced com-pared with WT (Fig. 3D). Next, to determine whether HID2 di-rectly affects the transcription of rDNA, we used chromatinisolation by RNA purification (ChIRP) to purify HID2 RNA andsubsequently investigated whether it interacts with 45S rDNA.Following retrieval ofHID2-interacting chromatin using anti-senseDNA tilling probes, we could not detect the 45S rDNA locus butclearly detected the HID2 locus, indicating that HID2 may notinteract with 45S rDNA (Fig. S7). Thus, HID2 may modulaterDNA transcription indirectly.

HID2 Is Required for Efficient Processing of Pre-rRNA.Using Northernblot analysis, we determined that P-A3 and 27S intermediate pre-rRNAs were accumulated in hid2 mutants, whereas maturerRNAs were expressed at their normal levels (Fig. 3B and Fig.S6B). This finding suggests that pre-rRNA processing is delayed inhid2mutants. Notably, the appearance of a significantly sized bandaround 1.8 Kb when both 5.8S and ITS2 probes were used (Fig.3B, asterisk) drove us to determine the precise 5′ and 3′ ends ofthe undetermined transcript via circular RT-PCR (Fig. 3B). Mostintriguingly, our sequencing results demonstrated that the over-produced band in hid2 mutants was a series of 3′- truncated 27SBintermediates (27SB* in short) that had been cleaved within the25S region (Fig. 3 E and F and Fig. S8). By analyzing 50 cloneseach for the WT and hid2 mutant circular RT-PCR products, wefound the frequency of cleavage inside the 25S region of 27SB tobe increased in the hid2 mutant (Fig. S8). It is worth noting that

the observed increase in 45S and 35S rRNA and 27SB* accu-mulation is recovered in the pHID2:HID2/hid2 transgenes (Fig.S6C), suggesting that HID2 plays an important role in both theefficiency and accuracy of rRNA processing.To understand the molecular basis ofHID2 function further, we

isolated HID2-interacting RNA by ChIRP. Our results demon-strated that HID2 was efficiently purified (Fig. 3G) and thatamplicons on 45S pre-rRNA were equally enriched (Fig. 3H). Thisfinding, in turn, suggests that HID2 is able to bind 45S pre-rRNA.

hid2 Mutants Confer an Imbalanced Ribosome Profile and AlteredResponses to Specific Antibiotics. To evaluate further the potentiallyimpaired ribosome assembly state caused by defects of pre-rRNAprocessing in the hid2 mutants, we performed a polysome profilingassay. Lysates from WT and hid2 seedlings were fractionatedthrough sucrose gradients to separate the ribosomal subunits andthe 80S monosome (Fig. S9). It appeared that the hid2 mutation

Fig. 3. HID2 interacts with 45S rRNA and promotes efficient pre-rRNAprocessing. (A) Schematic illustration of the rDNA locus encoding 45S pre-rRNA. (B) Northern blots showing the accumulated pre-rRNAs and aberrantprocessing products (asterisk) in WT and hid2 seedlings. The SYBR Green II-stained gel image is shown as a loading control. Probes (black bars) are in-dicated in A. (C) NRO assay showing increased rDNA transcription in hid2seedlings compared with WT. *P < 0.05; **P < 0.01 (t test). Primers are in-dicated by blue arrows in A. (D) Northern blots showing the transcript levelsof 45S pre-rRNA in WT and hid2 seedlings treated with ActinomycinD (50 ng/mL) for the indicated time. 5S rRNA was used as a control.(E) Identification of 27SB* by circular RT-PCR. (F) Schematic illustration of 27SB*.The primers for RT-PCR are indicated by red arrows. (G) Northern blotsshowing that ChIRP enriches HID2. (H) qRT-PCR analysis of HID2-ChIRP RNAshows the retrieval of 45S rRNA. Amplicons are indicated by facing arrows in A.Data represent one of three biological replicates.

Zhu et al. PNAS | October 18, 2016 | vol. 113 | no. 42 | 11969

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 22

, 202

0

caused a reduction in the 80S/40S ratio and an increased ratio ofpolysomes (Fig. 4A). Further insight regarding the association be-tween each ribosomal particle and rRNAs was gained by usingNorthern blot analyses using a 25S rRNA-specific probe. The re-sults showed an accumulation of prominent bands smaller than themature 25S rRNA that cofractionated with 60S and 80S of hid2mutants (Fig. 4B). Through circular RT-PCR and sequencing, wefound these bands to be 27SB* (Fig. 4 C and D), raising the pos-sibility that the overaccumulation of 27SB* in hid2 mutants mayhave an impact on the lower 80S/40S ratio.To test further whether the ribosome is structurally impaired in

hid2 mutants, we analyzed the mutants’ response to an array ofantibiotics known to target ribosomal locations. Although the hid2mutants showed a response to spectinomycin similar to that ob-served in the WT, they showed mild resistance to erythromycin.The mutants also were found to be resistant to streptomycin andspiramycin, exhibiting green cotyledons (Fig. 4E). Streptomycinbelongs to aminoglycoside class of antibiotics, which inhibit ami-noacyl tRNA binding to position A during elongation. In contrast,spiramycin and erythromycin are macrolide antibiotics and inhibitprotein synthesis by stimulating the dissociation of peptidyl-tRNAfrom ribosomes (34). The specific resistance of hid2 mutants tothese antibiotics suggests that an altered ribosomal structure or anaberrant population of ribosomes in the mutants prohibits theproper binding of these antibiotics by the ribosomes. To assessfurther whether HID2 could affect de novo protein synthesis, wetested the effect of the protein synthesis inhibitor cycloheximideon the hid2 mutants (35). As expected, the hid2 mutant was moresensitive to cycloheximide than the WT, suggesting that the mu-tant has reduced translational activity (Fig. 4E). Taken together,our data reveal that HID2-regulated rRNA biogenesis is essentialfor the proper assembly and activity of the ribosome.

The Contribution of the HID2 Signature Motif to Its Function in Vivo.HID2 has three clearly identifiable signature motifs, the C/D box,the C′/D′ box, and an antisense box that is complementary to 25SrRNA. Our sequence comparison analysis revealed the conser-vation of those signature motifs in yeast, in model metazoansranging from worm to human, and in monocot rice (Fig. 5A). Totest the functional importance of the HID2 antisense box, wedesigned a chimeric construct that included a HID2 antisense-boxsequence and a backbone sequence of snoR117, which is anno-tated to another classic C/D box snoRNA with an antisense-boxsequence upstream of its D′ box (36). The expression of snoR117was found to be comparable in the WT and hid2 mutant (Fig. 5B).We then transformed the chimeric construct into the hid2 mu-tant background under the regulation of the HID2 endogenouspromoter (HID2 AT-117). Intriguingly, the expressed chimeric

snoR117 successfully rescued the hid2 phenotype to WT (Fig. 5C).Moreover, we found the extent of phenotypic recovery to becorrelated with the expression level of HID2 AT-117, confirmingthat the antisense box of HID2 is critical for its function in vivo.To evaluate the functional conservation of HID2 further, we

transformed hid2mutants with OsHID2:OsHID2, anHID2 orthologin rice. Interestingly, the expression of OsHID2 completely rescuedthe phenotype of hid2 mutants without disturbing the expression ofAtHID2 (Fig. 5D), suggesting that the function ofHID2 is conservedin monocots and dicots. Moreover, our findings may imply a generalrole of HID2 throughout the eukaryotic kingdom.

DiscussionIn this study, we show that HID2 encodes a C/D box snoRNAthat is highly expressed in actively proliferative tissues and isessential for the normal growth and development of Arabidopsis.HID2 associates with 45S rRNA and acts as an important reg-ulator of ribosome biogenesis, predominantly by promoting ef-ficient pre-rRNA processing.In plants, a large number of snoRNAs have been identified

and annotated (26, 27, 36). However, to the best of our knowl-edge, no characteristic developmental defects have been pre-viously connected to the function of a single snoRNA in plants,at least in part because of the higher diversity of the snoRNA

Fig. 4. hid2 mutants confer an imbalanced ribosome profile and altered responses to specific antibiotics. (A) Quantitative sucrose density gradient analysisshowing the imbalanced ribosome profile in hid2 mutants. (B) Northern blots showing prominent aberrant products detected by 25S antisense probe that werecosedimented with 60S and 80S ribosomal fractions in hid2. T, total ribosomal extracts. Fraction numbers are indicated above each lane. (C) Identification of 27SB*by circular RT-PCR. The primers are indicated by red arrows in D. (D) Schematic illustration of 27SB*. (E) Phenotypes of WT and hid2 seedlings treated withantibiotics and cycloheximide. CHX, cycloheximide; Ery, erythromycin; Spe, spectinomycin; Spira, spiramycin; Strep, streptomycin. (Scale bars: 10 mm.)

Fig. 5. The conserved antisense box of HID2 is essential for its function inplants. (A) The consensus signature motifs of Arabidopsis HID2 and itsorthologs in Oryza sativa, Saccharomyces cerevisiae, Caenorhabditis elegans,Drosophila melanogaster, and Homo sapiens are presented using theWebLogo software. (B–D, Upper) Phenotypes of WT, hid2, and the indicatedtransgenic seedlings (Scale bars: 5 mm). (Lower) Northern blots showing theexpression levels of snoR117 (B), HID2 AT-117 (C), and AtHID2 and OsHID2(D) in the indicated seedlings, with 5.8S rRNA as a loading control.

11970 | www.pnas.org/cgi/doi/10.1073/pnas.1614852113 Zhu et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 22

, 202

0

genes in plants (12). The sequence redundancy of snoRNA genesalso may make it difficult to obtain a single snoRNA mutantdisplaying apparent physiological defects (12, 26). Thus, theidentification of our hid2 mutant represents a valuable oppor-tunity to dissect and further our understanding of the role ofsnoRNA in regulating ribosome biogenesis.A few snoRNAs have been found to regulate pre-rRNA pro-

cessing in yeast and animals, but little is known about the func-tion of snoRNAs in planta, with the exception of recent studies ofthe U3 snoRNP (18–21). As one of the core components in thesmall subunit processome, U3 snoRNP has been shown to have aconserved function in 18S rRNA biogenesis (8, 17, 23). De-pletion of U3 snoRNA in yeast has been shown to lead to under-accumulation of mature 18S rRNA (37). Because the expressionlevels of mature 18S rRNA were unchanged in hid2 mutants(Fig. 3 and Fig. S6), we assume that HID2 has a function dif-ferent from that of U3 snoRNA in pre-rRNA processing. Giventhat the antisense-box sequence of HID2 is functionally essential(Fig. S5 and Fig. 5) and that HID2 is associated with 45S rRNAin vivo (Fig. 3), we propose that HID2–rRNA base-pairing isrequired for HID2 to regulate pre-rRNA processing. In addition,given the disruptions in pre-rRNA processing documented inribosome biogenesis factor mutants with distinct aberrant pre-rRNAs (3, 7, 9), it is conceivable that HID2 may serve as asafeguard responsible for promoting the processing and mini-mizing the production of aberrant rRNA species in the error-prone processes that require a large set of transacting factors. Acomplete model of HID2’s function has yet to be establishedin vivo, and, consequently, a number of important questionsremain unanswered. For instance, it is still unclear whether thebase-pairing between HID2 and 25S rRNA affects rRNA folding,facilitates the accessibility of HID2–RNP to the target site, or hasa functional relevance to the binding and release of HID2–RNP.Although the general scheme of pre-rRNA processing is con-

served in eukaryotes, differences do occur in different species inthe factors required at specific developmental stages to ensure thesuccessful processing of pre-rRNA (1, 3). We suggest that HID2may represent one of these factors that have plant-specific fea-tures. Our study showed that the site II motif in the promoter of

HID2 is essential for its expression in actively proliferative tissues(Fig. 2), which rely on rapid ribosome biogenesis. However, we didnot find similar cis elements in metazoans and yeast. Thus, regu-lation of the expression of HID2 orthologs in animals and yeastrequires further investigation. In addition, to our knowledge, therehas been no report on the functional analysis of HID2 orthologs inmodel animals, and whether these orthologs have a common or adivergent role in controlling ribosome biogenesis remains of greatinterest for future explorations. Furthermore, changes in snoRNAexpression previously have been shown to affect the physiologicalconditions of cells and to correlate with increases in dysfunctionalribosome activity that, in turn, lead to various human diseases,including various forms of cancer (38–40). This insight into thecontrol of snoRNA expression no doubt will contribute both toour understanding of how growth and development in eukaryotesis controlled and to the development of innovative diagnosticapproaches and therapeutic agents for human disease.

Materials and MethodsThe plant materials, growth conditions, plasmid construction, and transgenicplants generation procedures are described in SI Materials and Methods. Thedetailed procedures of Northern blot analysis, GUS staining, circular RT-PCR, insitu hybridization, FISH, EMSA, primer extension analysis, ChIRP, and the NROassay are provided in SI Materials and Methods. The primers and probes used inthis study are listed in Table S1. The genes mentioned in this article have beengiven the following accession numbers by the Arabidopsis Information Resource(TAIR) database: TCP20, AT3G27010; TCP24, T1G30210; UBQ10, AT4G05320;UBC21, AT5G25760; ACT2, AT3G18780.

ACKNOWLEDGMENTS. We thank Dr. Jian Lu, Dr. Weiqiang Qian, Dr. ShunongBai, Dr. Ligeng Ma, and Dr. Qing Li for their helpful discussions and com-ments; Dr. Yu Zhang, Yue Huang, Yan Hu, Jiawei Pan, and Shengqian Doufor technical assistance; and Dr. Daniel A. Chamovitz, Dr. Sigal Rencus-Lazar,and Abigail Coplin for critical reading of the manuscript. This work was sup-ported by National Key Basic Research Program Grant 2016YFA0500800,Major Program of National Natural Science Foundation of China Grant91540105, General Program of National Natural Science Foundation of ChinaGrant 31471155, and National Basic Research Program (973 Program) Grant2012CB910900, and in part by the State Key Laboratory of Protein and PlantGene Research at Peking University and the Peking-Tsinghua Center for LifeSciences. Y.W is supported in part by a Postdoctoral Fellowship from thePeking-Tsinghua Center for Life Sciences.

1. Henras AK, Plisson-Chastang C, O’Donohue MF, Chakraborty A, Gleizes PE (2015)An overview of pre-ribosomal RNA processing in eukaryotes. Wiley Interdiscip RevRNA 6(2):225–242.

2. Lafontaine DL (2015) Noncoding RNAs in eukaryotic ribosome biogenesis and func-tion. Nat Struct Mol Biol 22(1):11–19.

3. Weis BL, Kovacevic J, Missbach S, Schleiff E (2015) Plant-specific features of ribosomebiogenesis. Trends Plant Sci 20(11):729–740.

4. Teng T, Thomas G, Mercer CA (2013) Growth control and ribosomopathies. Curr OpinGenet Dev 23(1):63–71.

5. Pontvianne F, et al. (2013) Subnuclear partitioning of rRNA genes between thenucleolus and nucleoplasm reflects alternative epiallelic states. Genes Dev 27(14):1545–1550.

6. Layat E, Sáez-Vásquez J, Tourmente S (2012) Regulation of Pol I-transcribed 45S rDNAand Pol III-transcribed 5S rDNA in Arabidopsis. Plant Cell Physiol 53(2):267–276.

7. Zakrzewska-Placzek M, Souret FF, Sobczyk GJ, Green PJ, Kufel J (2010) Arabidopsisthaliana XRN2 is required for primary cleavage in the pre-ribosomal RNA. NucleicAcids Res 38(13):4487–4502.

8. Borovjagin AV, Gerbi SA (2004) Xenopus U3 snoRNA docks on pre-rRNA through anovel base-pairing interaction. RNA 10(6):942–953.

9. Hang R, et al. (2014) Arabidopsis protein arginine methyltransferase 3 is required forribosome biogenesis by affecting precursor ribosomal RNA processing. Proc Natl AcadSci USA 111(45):16190–16195.

10. Fernández-Pevida A, Kressler D, de la Cruz J (2015) Processing of preribosomal RNA inSaccharomyces cerevisiae. Wiley Interdiscip Rev RNA 6(2):191–209.

11. Watkins NJ, Bohnsack MT (2012) The box C/D and H/ACA snoRNPs: Key players in themodification, processing and the dynamic folding of ribosomal RNA.Wiley InterdiscipRev RNA 3(3):397–414.

12. Rodor J, Letelier I, Holuigue L, Echeverria M (2010) Nucleolar RNPs: From genes tofunctional snoRNAs in plants. Biochem Soc Trans 38(2):672–676.

13. Brown JW, Echeverria M, Qu LH (2003) Plant snoRNAs: Functional evolution and newmodes of gene expression. Trends Plant Sci 8(1):42–49.

14. Kiss-László Z, Henry Y, Bachellerie JP, Caizergues-Ferrer M, Kiss T (1996) Site-specificribose methylation of preribosomal RNA: A novel function for small nucleolar RNAs.Cell 85(7):1077–1088.

15. Ganot P, Bortolin ML, Kiss T (1997) Site-specific pseudouridine formation in pre-ribosomal RNA is guided by small nucleolar RNAs. Cell 89(5):799–809.

16. Dutca LM, Gallagher JE, Baserga SJ (2011) The initial U3 snoRNA:pre-rRNA basepairing interaction required for pre-18S rRNA folding revealed by in vivo chemicalprobing. Nucleic Acids Res 39(12):5164–5180.

17. Marmier-Gourrier N, Cléry A, Schlotter F, Senty-Ségault V, Branlant C (2011) A second basepair interaction between U3 small nucleolar RNA and the 5′-ETS region is required for earlycleavage of the yeast pre-ribosomal RNA. Nucleic Acids Res 39(22):9731–9745.

18. Sáez-Vasquez J, Caparros-Ruiz D, Barneche F, Echeverría M (2004) A plant snoRNP com-plex containing snoRNAs, fibrillarin, and nucleolin-like proteins is competent for bothrRNA gene binding and pre-rRNA processing in vitro. Mol Cell Biol 24(16):7284–7297.

19. Samaha H, et al. (2010) Identification of protein factors and U3 snoRNAs from aBrassica oleracea RNP complex involved in the processing of pre-rRNA. Plant J 61(3):383–398.

20. Pontvianne F, et al. (2010) Nucleolin is required for DNA methylation state and theexpression of rRNA gene variants in Arabidopsis thaliana. PLoS Genet 6(11):e1001225.

21. Sáez-Vasquez J, Caparros-Ruiz D, Barneche F, Echeverría M (2004) Characterizationof a crucifer plant pre-rRNA processing complex. Biochem Soc Trans 32(Pt 4):578–580.

22. Dunbar DA, Baserga SJ (1998) The U14 snoRNA is required for 2′-O-methylation of thepre-18S rRNA in Xenopus oocytes. RNA 4(2):195–204.

23. Zhang L, Wu C, Cai G, Chen S, Ye K (2016) Stepwise and dynamic assembly of theearliest precursors of small ribosomal subunits in yeast. Genes Dev 30(6):718–732.

24. Peculis BA (1997) The sequence of the 5′ end of the U8 small nucleolar RNA is criticalfor 5.8S and 28S rRNA maturation. Mol Cell Biol 17(7):3702–3713.

25. Qu LH, et al. (1995) U24, a novel intron-encoded small nucleolar RNA with two 12 ntlong, phylogenetically conserved complementarities to 28S rRNA. Nucleic Acids Res23(14):2669–2676.

26. Liu TT, et al. (2013) A global identification and analysis of small nucleolar RNAs andpossible intermediate-sized non-coding RNAs in Oryza sativa. Mol Plant 6(3):830–846.

27. Wang Y, et al. (2014) Genomic features and regulatory roles of intermediate-sizednon-coding RNAs in Arabidopsis. Mol Plant 7(3):514–527.

28. Trémousaygue D, et al. (2003) Internal telomeric repeats and ‘TCP domain’ protein-binding sites co-operate to regulate gene expression in Arabidopsis thaliana cyclingcells. Plant J 33(6):957–966.

Zhu et al. PNAS | October 18, 2016 | vol. 113 | no. 42 | 11971

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 22

, 202

0

29. Qu G, et al. (2015) Promoter-based identification of novel non-coding RNAs revealsthe presence of dicistronic snoRNA-miRNA genes in Arabidopsis thaliana. BMCGenomics 16:1009.

30. Li C, Potuschak T, Colón-Carmona A, Gutiérrez RA, Doerner P (2005) ArabidopsisTCP20 links regulation of growth and cell division control pathways. Proc Natl AcadSci USA 102(36):12978–12983.

31. Hervé C, et al. (2009) In vivo interference with AtTCP20 function induces severe plantgrowth alterations and deregulates the expression of many genes important fordevelopment. Plant Physiol 149(3):1462–1477.

32. Wang H, Mao Y, Yang J, He Y (2015) TCP24 modulates secondary cell wall thickeningand anther endothecium development. Front Plant Sci 6:436.

33. Dong ZW, et al. (2012) RTL-P: A sensitive approach for detecting sites of 2′-O-methylation in RNA molecules. Nucleic Acids Res 40(20):e157.

34. Wilson DN (2014) Ribosome-targeting antibiotics and mechanisms of bacterialresistance. Nat Rev Microbiol 12(1):35–48.

35. Abbasi N, et al. (2010) APUM23, a nucleolar Puf domain protein, is involved in pre-ribosomalRNA processing and normal growth patterning in Arabidopsis. Plant J 64(6):960–976.

36. Kim SH, et al. (2010) Plant U13 orthologues and orphan snoRNAs identified byRNomics of RNA from Arabidopsis nucleoli. Nucleic Acids Res 38(9):3054–3067.

37. Hughes JM, Ares M, Jr (1991) Depletion of U3 small nucleolar RNA inhibits cleavage inthe 5′ external transcribed spacer of yeast pre-ribosomal RNA and impairs formationof 18S ribosomal RNA. EMBO J 10(13):4231–4239.

38. Valleron W, et al. (2012) Specific small nucleolar RNA expression profiles in acuteleukemia. Leukemia 26(9):2052–2060.

39. Su H, et al. (2014) Elevated snoRNA biogenesis is essential in breast cancer. Oncogene33(11):1348–1358.

40. Thorenoor N, Slaby O (2015) Small nucleolar RNAs functioning and potential roles incancer. Tumour Biol 36(1):41–53.

41. Zhu D, et al. (2008) Biochemical characterization of Arabidopsis complexes con-taining CONSTITUTIVELY PHOTOMORPHOGENIC1 and SUPPRESSOR OF PHYA pro-teins in light control of plant development. Plant Cell 20(9):2307–2323.

42. Wang Y, et al. (2014) Arabidopsis noncoding RNA mediates control of photomor-phogenesis by red light. Proc Natl Acad Sci USA 111(28):10359–10364.

43. Zhang Y, et al. (2008) Arabidopsis DDB1-CUL4 ASSOCIATED FACTOR1 forms a nuclearE3 ubiquitin ligase with DDB1 and CUL4 that is involved inmultiple plant developmentalprocesses. Plant Cell 20(6):1437–1455.

44. Wu MF, Wagner D (2012) RNA in situ hybridization in Arabidopsis. Methods Mol Biol883:75–86.

45. Jing Y, et al. (2016) SUVH2 and SUVH9 couple two essential steps for transcriptionalgene silencing in Arabidopsis. Mol Plant 9(8):1156–1167.

46. Qu LH, Meng Q, Zhou H, Chen YQ (2001) Identification of 10 novel snoRNA geneclusters from Arabidopsis thaliana. Nucleic Acids Res 29(7):1623–1630.

47. Liu F, Bakht S, Dean C (2012) Cotranscriptional role for Arabidopsis DICER-LIKE 4 intranscription termination. Science 335(6076):1621–1623.

48. Sun Q, Csorba T, Skourti-Stathaki K, Proudfoot NJ, Dean C (2013) R-loop stabilizationrepresses antisense transcription at the Arabidopsis FLC locus. Science 340(6132):619–621.

49. Chu C, Quinn J, Chang HY (2012) Chromatin isolation by RNA purification (ChIRP).J Vis Exp (61):3912.

50. Mustroph A, Juntawong P, Bailey-Serres J (2009) Isolation of plant polysomal mRNAby differential centrifugation and ribosome immunopurification methods. MethodsMol Biol 553:109–126.

11972 | www.pnas.org/cgi/doi/10.1073/pnas.1614852113 Zhu et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 22

, 202

0