Experimental RNomics: Identification of 140 Candidates for Small Non-Messenger RNAs in the Plant Arabidopsis thaliana

Transcription

1 Current Biology, Vol. 12, , December 10, 2002, 2002 Elsevier Science Ltd. All rights reserved. PII S (02) Experimental RNomics: Identification of 140 Candidates for Small Non-Messenger RNAs in the Plant Arabidopsis thaliana Claudia Marker, 1 Anja Zemann, 1 Tanja Terhörst, 1 Martin Kiefmann, 1 James P. Kastenmayer, 2 Pamela Green, 3 Jean-Pierre Bachellerie, 4 Jürgen Brosius, 1 and Alexander Hüttenhofer 1,5 1 Institute of Experimental Pathology ZMBE Von-Esmarch-Str. 56 Introduction We initiated an experimental approach, termed Experimental RNomics, with the goal of identifying novel nonmessenger RNAs in model organisms [1 4]. Sizes of non-messenger RNAs range from very large, e.g., up to 17 kb in the case of Xist RNA [5], down to extremely Münster small (21 23 nt), as observed for micro RNAs (mirnas) Germany (for a review, see [6]). In general, the sizes of the majority 2 Department of Energy Plant Research Laboratory of non-coding RNAs vary from 20 nt to 500 nt, well Michigan State University below the size of most mrnas, and are therefore termed Lansing, Michigan snmrnas (small non-messenger RNAs). 3 Delaware Biotechnology Institute In this study, we present the experimental identifica- University of Delaware tion of novel snmrna species in Arabidopsis thaliana Newark, Delaware by generating a specialized cdna library encoding 4 Laboratoire de Biologie Moléculaire Eucaryote snmrnas from this plant. Recently, the 125 Mb genome du C.N.R.S. of A. thaliana [7] has been sequenced and predicted to Université Paul-Sabatier contain 25,498 genes encoding proteins. In the same Toulouse study, from the class of snmrnas, 589 trnas, 27 organ- France elle-derived trnas, and 13 trna pseudogenes could be identified. Also, all spliceosomal snrnas have been experimentally identified in A. thaliana [7]. Summary A large fraction of snmrnas identified so far in Eukarya corresponds to members of the two expanding Background: Genomes from all organisms known to subclasses of small nucleolar RNAs (snornas), termed date express two types of RNA molecules: messenger antisense box C/D and H/ACA snornas, that guide RNAs (mrnas), which are translated into proteins, and modification of rrnas and snrnas. Homologs of the non-messenger RNAs, which function at the RNA level two subclasses of snornas have been recently identified and do not serve as templates for translation. in Archaea, in which they appear to perform the Results: We have generated a specialized cdna library same guide function for rrna modifications [4, 8 10]. from Arabidopsis thaliana to investigate the population The presence in antisense box C/D snornas of hallmark of small non-messenger RNAs (snmrnas) sized box motifs and nt-long complementarity to the nt in a plant. From this library, we identified 140 candidates site of ribose methylation has provided the basis for for novel snmrnas and investigated their expres- their identification by a computer genomic search in sion, abundance, and developmental regulation. Based several eukaryal organisms, including A. thaliana [11 on conserved sequence and structure motifs, ]. In contrast, no effective genomic search of H/ACA snmrna species can be assigned to novel members of snornas has been reported so far, mostly due to shorter known classes of RNAs (designated Class I snmrnas), box motif sequences. As a result, only two different namely, small nucleolar RNAs (snornas), 7SL RNA, U members of this large snorna family have been identi- snrnas, as well as a trna-like RNA. For the first time, fied in a serendipitous way in A. thaliana [12]. 39 novel members of H/ACA box snornas could be Quite recently, two independent studies have focused identified in a plant species. Of the remaining 36 on the presence of mirnas in A. thaliana, thereby identifying snmrna candidates (designated Class II snmrnas), no numerous candidates for these RNAs species in sequence or structure motifs were present that would plants [14, 15]. In this study, we present the identification enable an assignment to a known class of RNAs. These of 140 candidates for novel, small non-messenger RNAs RNAs were classified based on their location on the from the genome of A. thaliana. In addition, we have A. thaliana genome. From these, 29 snmrna species analyzed size, abundance, developmental regulation, located to intergenic regions, 3 located to intronic se- and tissue-specific expression of these RNA species by quences of protein coding genes, and 4 snmrna candidates Northern blot analysis. This is the first survey of the were derived from annotated open reading nuclear and organellar small RNA species population in frames. Surprisingly, 15 of the Class II snmrna candidates a plant cell. were shown to be tissue-specifically expressed, while 12 are encoded by the mitochondrial or chloroplast Results and Discussion genome. Conclusions: Our study has identified 140 novel candi- cdna Library Construction and Analysis dates for small non-messenger RNA species in the plant To investigate the population of small non-messenger A. thaliana and thereby sets the stage for their functional RNAs in a plant species, we have generated a specialanalysis. ized cdna library from A. thaliana that encodes novel candidates for snmrnas. In order to include developmentally 5 Correspondence: huttenh@uni-muenster.de regulated snmrnas, we used total RNA from

2 Non-Coding RNAs in the Plant A. thaliana 2003 Figure 1. Expression Analysis of Novel snmrnas A selection of expressed candidates for novel snmrnas (Class I and II) in A. thaliana, as deduced by Northern blot analysis including two developmental growth phases of A. thaliana, the seedling (I) or the adult plant (IV). Sizes of RNAs, as estimated by comparison with an internal RNA marker, are indicated on the right. Exposure times of autoradiograms are shown below each blot (d: days; h: hours; min: minutes). four developmental stages for cdna library construc- using oligonucleotides complementary to the respective tion: the seedling, the adult plant, and two intermediate RNA sequence (for selected examples, see Figure 1). We growth states. Equal amounts of total RNA prepared determined abundance and size of snmrna candidates from the four different growth states were size selected from A. thaliana. In most cases, as observed for other (from nt) on denaturing PAGE and reverse tran- libraries, sizes of RNAs as determined by Northern blot scribed for cdna library construction (see the Experimental analysis were larger compared to the sizes of the reable Procedures in the Supplementary Material avail- spective cdnas. This is due to the library construction with this article online). strategy (see the Experimental Procedures in the Sup- Subsequently, we spotted PCR-amplified DNA frag- plementary Material), which interferes with cloning of ments from 30,000 cdna clones in high-density arrays the very 5 ends of novel snmrna species, or to the fact on filters. Filters were hybridized with radiolabeled oligo- that RNA structure or modification impeded a complete nucleotides directed against the most abundant known conversion of the snmrna into cdna. Therefore, our snmrnas. By this approach, we avoided redundant se- sequences resemble ESTs (expressed sequence tags) quencing of known snmrnas or rrna fragments. A total of mrnas, and we therefore designated them as ERNS of 5000 clones exhibiting the lowest hybridization scores (expressed RNA sequence) in our tables (Tables 1 and 2). were sequenced. The sequences that did not map to a In order to identify candidates for novel snmrnas that previously annotated gene were considered as potential are specifically expressed at certain stages of developcandidates for novel snmrnas. A total of 140 clones ment, we have constructed the cdna library from four were assigned to this group. developmental growth phases of A. thaliana (see above). We did not detect a strictly developmentally specific Expression, Abundance, and Developmental expression of the RNA species for any of the snmrna Regulation of Novel snmrnas candidates. In some cases, however, snmrnas were To analyze the expression of the 140 candidates for somewhat more stongly expressed at stage I (the seednovel snmrnas, we performed Northern blot analyses ling) compared to stage IV (the adult plant) or vice versa

3 Current Biology 2004 Table 1. Class I: Candidates for snmrnas Exhibiting Known Sequence or Structure Motifs cdna Northern ERNS snorna Nr. (nt) Blot (nt) Abund. Location Comments Acc. Nr. Group 1: C/D Box snornas Target: rrnas Modification Guide Box Ath-66 snor , IR, cluster Um1483/18S and D (11 nt) AJ Um2873/25S D (11 nt) Ath-109 snor , IR Gm398/25S D (12 nt) AJ Ath-119 snor , 62, 170 3, IR, cluster Cm38/18S D (12 nt) AJ presumptive C/D snorna, Ath-119b, 60 nt downstream (Cm2187/25S) Ath-301 snor , IR Um1261/18S D (12 nt) AJ presumptive C/D snorna, Ath-301b, 840 nt downstream (Gm75 and Gm84/U6 snrna) Ath-671 snor , IR Um2726/25S D (12 nt) AJ Target: snrnas Ath-151 snor , IT (AAF80615) Cm29/U2 D (12 nt) AJ Ath-313 snor , 3 UTR Gm41/U5 D (10 nt) AJ (AAK76602) Target: Unknown Ath-282 snor , IR related to Ath-350, snrna promotor AJ Ath-303 snor , 340 5, IR, cluster variant gene copy 65 bp downstream AJ Ath-349a snor , IR AJ Ath-350 snor , IR related to Ath-282; snrna promotor AJ Computationally Found Before Ath-55 snor8b , IR see [12] AJ Ath-140 snor n.d. n.d. 4, IR see [11] AJ Ath-221 snor n.d. n.d. 5, IR see [12] AJ Ath-224 U n.d. n.d. 3, IR see [12] AJ Ath-268 snor n.d. n.d. 1, IR see [12] AJ Ath-339 snor n.d. n.d. 1, IR see [12] AJ Ath-436 snor58y 1 54 n.d. n.d. 1, IR see [12] AJ Ath-611 snor41y 1 56 n.d. n.d. 3, IR see [11] AJ Group 2: H/ACA Box snornas Target: rrnas Modification Ath-99 snor , IT (AAB63817) P1: 634 a /18S; P2: 1520/18S AJ Ath-125 snor , IR P2: 2913 a /25S, conserved in A. fulgidus AJ same target as Ath-163 and Ath-182 Ath-127b snor , IR P2: 918/18S and 214/25S AJ Ath-136 snor , IR P1: 826 b /25S; P2: 752/18S AJ Ath-157 snor , IR, cluster P1: 2620/25S ; P2: 1523/18S; related to Ath-331, same target AJ Ath-162 snor , IT (BAA32529) P1: 2248 a /25S, conserved in E. coli AJ continued

4 Non-Coding RNAs in the Plant A. thaliana 2005 Table 1. Continued cdna Northern ERNS snorna Nr. (nt) Blot (nt) Abund. Location Comments Acc. Nr. Group 2: H/ACA Box snornas Target: rrnas Modification Ath-163 snor , IT (AAC18792) P2: 2913 a /25S, related to Ath-182,same target as Ath-125 AJ Ath-166 snor , IR, cluster P1: 1130/25S; P2: 999/25S AJ Ath-176 snor , IT (AAD25605) P2: 1013 a /25S, same target as Ath-438 AJ Ath-182 snor , IT (AAF99734) P2: 2913 a /25S, related to Ath-163 with same target as Ath-125, in AJ intron/rpl17 gene Ath-204 snor , IR P2: 783/25S AJ Ath-220 snor , IR P1: 2304 a /25S; P2: 2339/25S AJ Ath-223 snor , IR P2: 827/18S AJ Ath-226 snor , IR P2: 78/5.8S AJ Ath-235 snor , IR P1: 2555/25S; P2: 360 b /18S AJ Ath-236 snor , IT (AAG51445) P1: 715/18S and 2126 c /25S AJ Ath-245 snor , 220 5, IR, cluster P1: 1302/18S; strongly related to Ath-291a, same target AJ Ath-258 U , IT (BAB09703) P1: 2250 a /25S, one additional stem AJ Ath-263 snor , IT (BAB10811) P1: 1000 a /18S; P2: 1118 b /18S AJ Ath-271 snor , 500 ( ) 3, IR P1: 220/18S; P2: 35/25S AJ Ath-277 snor , IT (AAB80636) P1: 948 b /18S AJ Ath-280 snor , 162 3, IT (CAB43405) P1: 604/18S; P2: 761 b and 1090/18S AJ Ath-291a snor , IR, cluster P1: 1302/18S; P2: 1742/18S, strongly related to Ath-245, same target AJ Ath-300 snor , IR P2: 2244 c /25S AJ Ath-317 snor n.d. 3, IT (AAF23276) P2: 2773 c /25S AJ Ath-318 snor , IR P2: 132/25S AJ Ath-331 snor , IR, cluster P1: 2620/25S; related to Ath-157, same target AJ Ath-334 snor , IR, cluster P1: 2703/25S; presumptive C/D snorna, Ath-334b, 105 nt downstream AJ (Cm375/25S and Cm1471/25S) Ath-335 snor , 220 5, IR P1: 1246/25S AJ Ath-363 snor , IR, cluster P2: 2707 b /25S; variant of C/D AtsnoR nt upstream AJ Ath-438 snor , IT (CAC43286) P2: 1013 a /25S, same target as Ath-176 AJ Ath-648 snor , 190 4, IR P1: 803/18S; P2: 2201 a /25S AJ Ath-678 snor , 260 5, IT (CAB93719) P2: 111 b /18S AJ Target: snrnas Ath-402 snor , IT (BAB08857) P1: 50 d /U5; P2: 47 d /U5 AJ Ath-422 snor , IT (AAG35792) P1: 42 a /U2 AJ Target: Unknown Ath-210 snor n.d. 4, IR AJ Ath-458 snor , IT (AAG12705) AJ Ath-488 snor , IR, cluster AJ Ath-637 snor , IT (CAB81918) AJ continued

5 Current Biology 2006 Table 1. Continued cdna Northern ERNS snorna Nr. (nt) Blot (nt) Abund. Location Comments Acc. Nr. Group 3: snrna Homologs Ath-16a , IR similarity to U1a (X53175) 87% AJ Ath-16b , IR similarity to U1a (X53175) 89% AJ Ath-16c , IR similarity to U1a (X53175) 92% AJ Ath-16d , IR similarity to U1a (X53175) 91% AJ Ath-191a n.d. n.d. 2, IR similarity to U2 (X06477) 95% AJ Ath-191d 1 39 n.d. n.d. 1, IR similarity to U2 (X06477) 94% AJ Ath-227c n.d. n.d. 2, IR similarity to U2 (X06474) 98% AJ Ath-188a n.d. n.d. 1, IR similarity to U4.1 (X67145) 96% AJ Ath-188b n.d. n.d. 1, IR similarity to U4.1 (X67145) 98% AJ Ath-134a , IR similarity to U5 (X13012) 92% AJ Ath-134b , IR similarity to U5 (X13012) 93% AJ Ath-134c , IR similarity to U5 (X13012) 89% AJ Ath , IR similarity to U5 (X13012) 92% AJ Group 4: 7SL Homologs Ath n.d. n.d. 1, IR similarity to 7SL gene (X55111) 84% AJ Ath , IR similarity to 7SL-1gene (X72228) 84% AJ Group 5: trna-like snmrnas Ath ,IR similar to trna Phe AJ Compilation of novel expressed RNA sequences from Class I from an A. thaliana cdna library derived from RNAs sized nts. ERNS: expressed RNA sequences (Ath-); snorna: nomenclature for snorna in A. thaliana; Nr.: number of independent cdna clones identified from each RNA species; cdna (nt): length of cdna encoding a snmrna candidate, as assessed by sequencing; Northern blot (nt): length of RNA, as assessed by Northern blot analysis; Abund.: relative abundance of snmrna, as estimated by Northern blot analysis; Location: chromosomal location of the snmrna gene (IR: intergenic region; IT: intron; cluster: indicates that the snorna gene is part of a larger snorna cluster; for all snorna genes located in introns, the accession number of the respective host gene is given in brackets); Comments: comments related to proposed function of snmrna; in the case of Class I/Groups 1 and 2 snmrnas, modification refers to predicted modified nucleotides within rrnas or snrnas; for C/D snobox RNAs, guide box refers to the length of the antisense element (indicated in nt), preceded by its location in the 5 domain (D ) or3 domain (D) of the snorna; for H/ACA box snornas, P1 refers to the pseudouridylation pocket in stem 1 of the snorna, P2 refers to the pseudouridylation pocket in stem 2; evolutionary conservation of modifications among species are indicated as: a conserved in vertebrates and yeast, b conserved in vertebrates, c conserved in vertebrates and fruitfly, d conserved in plants and vertebrates; Acc. nr.: accession number of snmrna sequence in the DDBJ/EMBL/GenBank database.

6 Non-Coding RNAs in the Plant A. thaliana 2007 Table 2. Class II: Candidates for snmrnas Lacking Known Sequence or Structural Motifs cdna Northern ERNS Nr. (nt) Blot (nt) Abund. Location Tiss.-Spec. Comments/Homology Acc. Nr. Group 1: snmrnas in Intergenic Regions Nucleus Ath , AJ Ath , AJ Ath AJ Ath homolog in B. oleracea (96%) AJ Ath homolog in B. oleracea (86%), O. sativa (87%), and G. max (94%) AJ Ath AJ Ath ( ) 1 AJ Ath AJ Ath AJ Ath AJ Ath ( ) 1 AJ Ath AJ Ath AJ Ath AJ Ath homolog in B. oleracea (100%) AJ Ath AJ Ath AJ Ath AJ Ath ( ) 2 ribosomal operon/its AJ Ath ( ) 5 AJ Mitochondrion Ath , 79, 170 M,2 repeat unit AJ Ath M AJ Ath M,2 homolog in B. oleracea (100%) AJ Ath ( ) M,2 repeat unit, homolog in B. oleracea (98%) AJ Chloroplast Ath C homolog in P. micrantha (93%) and P. fremontii (93%) AJ Ath ( ) C homolog in B. napus (99%) AJ Ath C homolog in B. oleracea (97%) AJ Ath C AJ Ath C AJ Group 2: snmrnas in Introns Ath ( ) 1 (AAC ) AJ Ath ( ) 1 (AAG ) AJ Ath C (BAA ) homolog in B. oleracea (91%) and S. oleracea (86%) AJ continued

7 Current Biology 2008 Table 2. Continued cdna Northern ERNS Nr. (nt) Blot (nt) Abund. Location Tiss.-Spec. Comments/Homology Acc. Nr. Group 3: snmrnas from ORFs Nucleus Ath (NP ) 5 UTR/exon AJ Ath (AAM ) exon/3 UTR AJ Mitochondrion Ath , 162 M,2 exon/3 UTR AJ (NP ) Chloroplast Ath C (AAF ) exon/3 UTR AJ Compilation of novel expressed RNA sequences from Class II from an A. thaliana cdna library derived from RNAs sized nts. ERNS: expressed RNA sequences (Ath-); Nr.: number of independent cdna clones identified from each RNA species; cdna (nt): length of cdna encoding a snmrna candidate, as assessed by sequencing; Northern blot (nt): length of RNA, as assessed by Northern blot analysis; Abund.: relative abundance of snmrna, as estimated by Northern blot analysis; Location: chromosomal location of the snmrna gene (IR: intergenic region; IT: intron; M: mitochondrium; C: chloroplast; for all snmrna genes located in introns or derived from mrnas, the accession number of the host gene is given in brackets); Comments/Homology: comments related to proposed function of snmrna or homologs of snmrnas found in other plant species (percent of homology is indicated in brackets); Acc. nr.: accession number of snmrna sequence in the DDBJ/EMBL/GenBank database. (see Figure 1). Subsequent to Northern blot analysis, we grouped candidates for novel snmrnas based on known sequence or structure motifs (Class I snmrnas, Table 1) or the lack of these (Class II snmrnas, Table 2). Class I snmrnas Exhibiting Known Sequence or Structure Motifs From our library, 104 cdna sequences could be assigned to Class I RNAs with known sequence or structure motifs (Table 1 and Figure 2). The 104 snmrna candidates were divided into 5 subgroups, whereby 88 belong to members of C/D box or H/ACA box snornas (Groups 1 and 2), 13 represent homologs to U snrnas (Group 3), 2 represent homologs to 7SL RNA (Group 4), and 1 contains trna sequence motifs (Group 4). For C/D and H/ACA box snornas, a nomenclature (see Table 1) that is consistent with the one applied to previously identified snornas in A. thaliana (nomenclature according to John Brown, Scottish Crop Research Institute, Dundee, UK) was used. Group 1: C/D snornas C/D box snornas guide the 2 -O-methylation of riboses within ribosomal or spliceosomal RNAs [16]. The presence of hallmark box motifs in antisense box C/D snornas has provided the basis for their identification by computer genomic search in several eukaryal organisms. In A. thaliana, a total of 131 presumptive gene copies of 69 different antisense box C/D snornas have been detected by this approach, but only a fraction of these have been experimentally verified [11 13]. In our screen, we have identified 49 snmrnas belonging to the class of C/D box methylation guide snornas (Table 1, Group 1). From these, 30 have previously been identified by computational analyses and have been confirmed by experiments [11 13]. Hence, we did not list them in Table 1. Of the remaining 19 snmrnas in that class, 11 represent entirely novel C/D box snornas, while the remaining 8 snmrnas of the C/D type were previously predicted by computer searches; however, their expression had not been experimentally verified as of yet. In the mammalian system, methylation guide snornas are exclusively encoded in pre-mrna introns ([3], for reviews, see [17, 18]). However, the vast majority of C/D snornas detected by genomic search have been found in intergenic regions of the A. thaliana genome [11 13]. This is also the case for all but two of the novel C/D snornas identified in our screen. Interestingly, the two exceptions, Ath-151 and Ath- 313, which are located in an intron or the 3 untranslated region of a protein-coding gene, respectively, are not predicted to target rrna but are predicted to target spliceosomal RNAs U2 and U5, respectively (see below). This is the first report of snornas in plants that are able to modify RNA species other than ribosomal RNA. Importantly, the predicted methylations, Cm29 in U2 and Gm41 in U5, have also been experimentally verified in plants [19]. Four snmrnas displaying all the canonical motifs of bona fide C/D box snornas (including a 4- to 7-bp terminal stem) do not have a cognate target within A. thaliana rrnas or snrnas (Table 1, Group 1). These novel snornas species might target other non-proteincoding RNAs or even mrnas [3].

8 Non-Coding RNAs in the Plant A. thaliana 2009 Figure 2. Novel snmrnas in A. thaliana A schematic overview and classification of 140 candidates for small non-messenger RNAs in A. thaliana. Group 2: H/ACA Box snornas As observed for C/D box snornas, two novel H/ACA H/ACA snornas direct the formation of the snorna candidates that are able to target U2 or U5 pseudouridines generally found in eukaryal rrnas [16, snrnas for pseudouridylation were identified in our 20 22]. So far, this large family of snornas has been screen (see Figure S1). Ath-402 exhibits two pseudouribest characterized in vertebrates, in which about half dylation pockets, which target U47 and U50 of U5 of its presumptive members have been identified [3, 20], snrna, respectively, while Ath-422 targets U42 in U2 and to a lesser extent in the yeast S. cerevisiae [20, 23]. snrna. Modifications of U47 and U50 in U5 snrna have In plants, only two H/ACA box snornas, designated been shown to be conserved between plants and verteas snor2 and snor5, have been identified in A. thali- brates, while pseudouridylation of U42 in U2 snrna has ana [12]. been determined in vertebrates and yeast [19]. In our A thaliana screen, a total of 39 novel members of Four snmrnas, designated as Ath-210, Ath-458, Ath- H/ACA box snornas have been found. They are mostly 488, and Ath-637, exhibiting all the canonical structure located in intergenic regions of the genome (Table 1, and sequence motifs of bona fide H/ACA box snornas Group 2), in contrast to vertebrate H/ACA box snornas, lack complementarity to rrnas or snrnas within their which are exclusively located in pre-mrna introns [17, pseudouridylation pockets. In analogy to three C/D box 18]. Intriguingly, however, the proportion of A. thaliana snornas identified in our screen, which also lack the H/ACA snornas that are intron derived (16 out of 39) expected RNA targets (see above), this could imply that is substantially higher than for the C/D class in this other RNA targets such as mrnas exist. For analysis organism [11 13]. of genomic organization and promotor elements of C/D Of the 39 novel H/ACA snornas, 33 are unambiguously and H/ACA snorna genes, see the Supplementary Reuridines able to direct the isomerization of a total of 47 sults and Discussion in the Supplementary Material. within 5.8S, 18S, or 25S ribosomal RNAs from Group 3: snrna Homologs A. thaliana; this is based on their ability to form the In our screen, we have identified 12 snmrna candidates canonical guide duplex in which the target uridine is with similarity to previously reported small nuclear RNAs always positioned nt from the box H or ACA motif (snrnas, Table 1, Group 3). From these, four snmrna (see Figure S1 in the Supplementary Material). Although species represent homologs to U1 RNA, three represent pseudouridines in nuclear-encoded rrnas have not homology to U2 snrna, two represent homology to U4.1 been experimentally mapped in plants so far, their identification snrna, and four represent homology to U5 snrna (Ta- in humans, mouse, the fruit fly D. melanogaster, ble 1, Group 3). The similarity indicated to respective and the yeast S. cerevisiae shows that a substantial annotated snrna (Table 1, Group 3) always refers to fraction of them are conserved in all or a subset of these the closest match of the novel snmrna to all previously organisms [24]. This provides a useful background for reported U snrna isoforms. The degree of similarity an indirect assessment of A. thaliana predicted sites. ranges from 87% (Ath-16a) to 98% (Ath-227c). As detailed in Table 1, a large fraction of them correspond Group 4: 7SL Homologs to phylogenetically conserved pseudouridines, In analogy to U snrna homologs, we could also identify experimentally determined in other model organisms, two novel snmrna species that are related to 7SL RNA which supports the predicted modification in A. thaliana. (Table 1, Group 4). Four 7SL RNA species have been A substantial fraction of the novel H/ACA snornas previously reported in A. thaliana, 7SL RNA, 7SL-1, targeting rrna (12 out of 33) appear to have pseudouri- 7SL-2, and 7SL-3 RNA. The two novel snmrna species dylation guide activity in both hairpin domains, like several identified in our screen, Ath-29 and Ath-383, show 84% of their vertebrate or yeast homologs. Ath-280 is similarity to either the 7SL RNA gene or the 7SL-1 RNA even able to target three pseudouridylation sites, since gene (Table 1). The base changes within Ath-29 and its 3 pseudouridylation pocket exhibits complementar- Ath-383 snmrnas, compared to previously found 7SL ity to two different sites in 18S rrna (the 5 pseudouridylation RNA species, either preserve the canonical secondary pocket targets still another site in 18S rrna). structure of 7SL RNA (data not shown) or are located

9 Current Biology 2010 Figure 3. Tissue-Specific Expression of snmrnas from A. thaliana Northern blot analysis showing tissue-specific expression of Class II snmrna candidates. The clone number is indicated on the left of each Northern blot, and the estimated size of the snmrna is indicated on the right. The four tissues used in Northern blot analysis are R (root), S (stem), L (leaf), and F (flower). For clone Ath-241, which exhibits two bands on a Northern blot (with sizes of 162 nt and 70 nt), only the 162 nt band shows tissue-specific expression. As an internal control, ubiquitous expression of 7SL RNA is indicated at the top of each panel. sequence or structure motifs (Table 2 and Figure 2). We cannot rule out the possibility that these RNAs encode small proteins, although we were unable to identify ex- tended open reading frames (ORFs) within RNA se- quences: on average, most Class II RNAs lacked reading frames longer than 12 aa. In cases where longer ORFs were detected (up to 44 aa), these ORFs completely lacked ribosome binding sites (a Kozak sequence, or, in the case of organellar snmrna candidates, a Shine- Dalgarno-like sequence) or a termination codon at their 3 end, which would be indicative of translated mrnas. The 36 snmrnas were assigned to three groups ac- cording to their location on the genome: Group 1 con- tains snmrna candidates located in intergenic regions, Group 2 contains snmrna candidates located in in- trons, and Group 3 exhibits snmrna candidates derived from ORFs. To exclude false positives as much as possi- ble, e.g., randomly, at low-level transcribed RNA spe- cies, we omitted all of those RNA species from Class II whose expression could not be confirmed by Northern blot analysis. in single-stranded regions, such as loops or bulges, likely not to interfere with the predicted 7SL RNA secondary structure. These findings are consistent with a preserved functional role of novel identified 7SL RNA isoforms, Ath-29 and Ath-343 in A. thaliana. Group 5: trna-like snmrnas A single clone, designated as Ath-74, can be assigned to Group 5 (Table 1). It is identical to A. thaliana trna Phe from position 33 72, except for a single base substitution, A to U, at position 58 (numbering according to trna Phe ). All 13 cdna sequences from our library differ at the same site from the genomic sequence on chromosome 5, possibly indicating a sequencing error at the reported genomic locus. Alternatively, at this point, we cannot exclude the possibility that Ath-74 is derived from a different locus of the genome, which has not yet been sequenced so far (see above). Ath-74 snmrna is highly abundant in the plant cell, as assessed by Northern blot analysis, and migrates faster than the cognate trna Phe, exhibiting a size of about 53 nt (Figure 1). The sequence similarity of the Ath-74 snmrna to trna Phe from A. thaliana includes the CCA end, the conserved GGTTC sequence in the T loop, and the anticodon stem and loop. Unlike for the canonical trna Phe gene, the presumed genomic locus for Ath-74 encodes the CCA end of the trna. In addition, in contrast to trna Phe, Ath-74 lacks the first 32 nt from the 5 end of trna Phe but contains about 13 nt of non-trna-related sequence instead. Surprisingly, Ath-74 snmrna can be folded into an extended rod-like stem structure rather than the canonical cloverleaf structure observed for trnas (see Figure S2 in the Supplementary Material). Still, the RNA contains, for example, a mature CCA end, indicative that at least parts of the trna-processing machinery is able to recognize the trna-like features of the Ath-74 sequence. Class II snmrnas Lacking Known Sequence or Structure Motifs Class II contains 36 representatives of candidates for snmrnas in A. thaliana, which are devoid of known Tissue-Specific Expression of Class II snmrna Candidates In order to gain further insight into the possible functions of Class II snmrna candidates, we investigated their tissue-specific expression in addition to the analysis of their developmental expression described above. We isolated total RNA from four different tissues of the plant: the root, the stem, the leaf, and the flower. Northern blot analysis was performed for each Class II snmrna for all four different tissues. Surprisingly, 15 members from Class II snmrna can- didates exhibited a tissue-specific expression pattern or were expressed preferentially in certain tissues of the plant (Figure 3). This applies to snmrnas from any of the three subgroups, candidates from intergenic regions, intron-encoded snmrnas, and those derived from an- notated coding regions of the Arabidopsis genome. In many cases, expression could be observed either mainly

10 Non-Coding RNAs in the Plant A. thaliana 2011 Figure 4. Mitochondrion- and Chloroplast-Encoded snmrnas from A. thaliana Location of snmrnas on the mitochondrial (left) and the chloroplast genome (right), not drawn to scale. The snmrna on the left, the genes flanking the snmrna candidate, as well as the distance of the snmrna 5 and 3 ends to flanking genes (in nt) are indicated. Flanking genes are shown by black arrows; the respective snmrna genes are indicated by red arrows (mitochondrial genome) or green arrows (chloroplast genome). For snmrnas that overlap the open reading frame of a gene, the number of overlapping nucleotides is indicated by a negative number. or exclusively in the roots of the plant, or vice versa in the flower (Figure 3). Search for Class II snmrna Homologs Because of the uncertain origin of snmrna candidates from Class II, we also searched for homologs in other plant species to obtain information regarding their evo- lutionary conservation. The sequence similarity search is hampered, however, by the lack of a plant genome closely related to Arabidopsis thaliana. The only com- plete genomic sequence at hand is that from the mono- cot plant rice (Oryza sativa), which is, however, only very distantly related to the dicot Arabidopsis thaliana. The Class II sequences were compared to sequence databases to identify potential homologs. Nine of the Class II sequences showed significant similarity to se- quences from other plant species (sequence similarity ranging from 86% to 100%, Table 2). In several cases, the similar sequences are cdnas providing support for conserved expression. The alignments with potential orthologs are consistent with these Class II sequences corresponding to conserved non-messenger RNAs genes (a representative example shows an alignment of Ath-237 with a genomic sequence from the dicot Brassica oleracea and cdna sequences from Glycine max and Oryza sativa; see Figure S3 in the Supplemen- tary Material). There is also no obvious conservation of ORFs, as there are often single nucleotide insertions or deletions in the orthologs relative to the Class II gene sequences that would result in frame shifts. Similar to the alignment of Ath-237, the overall identity is high for the other alignments as well (Table 2). A lower identity would be expected if protein-coding sequences were conserved, because third position mutations should easily arise. Taken together, these observations support the possibility that several Class II sequences correspond to non-coding RNA genes. As more sequence information becomes available from additional more closely related plant species, we will be able to examine alignments of additional Class II genes. Group 1: snmrnas from Intergenic Regions From the 36 Class II candidates for snmrnas, 29 map to regions between protein-coding genes on the A. thaliana genome (Table 2, Group 1). Of these, 20 are encoded by the nuclear, 4 are encoded by the mitochondrial, and 5 are encoded by the chloroplast genome (Figure 4). Mitochondria and chloroplasts each contain a circular genome with a size of about 376 kb or 154 kb, respec- tively. In mitochondria, three out of four snmrna candi- dates map not only to the mitochondrial but also to the nuclear genome on chromosome 2 (Table 2) due to a gene duplication event involving parts of the mitochon-

11 Current Biology 2012 drial genome on chromosome 2 [25]. In addition, novel Incorrectly Annotated snmrna Genes snmrna genes are, in most cases, also present in two in the A. thaliana Genome copies on the mitochondrial genome (Figure 4). During the course of our work, we also have identified, In contrast to snmrna candidates mapping to the in addition to known, correctly annotated protein-coding mitochondrial genome of A. thaliana, the five snmrnas or RNA genes, several trna genes, which were found mapping to the chloroplast genome lack counterparts to be wrongly annotated on the genome of A. thaliana. within the nuclear genome. In addition, unlike in mito- The genes for these six trnas are shown in Table S1 chondria, they are single-copy genes. Surprisingly, four (available in the Supplementary Material): all of the trna out of the five snmrnas are also tissue-specifically ex- genes were listed in GenBank as being transcribed opposite pressed (Figure 3). The Ath-230 gene is located between of their proper orientation. two non-coding RNAs, trna Leu and trna Phe. Ath-243 maps to the intergenic region between 5S rrna and 4.5 Conclusions S RNA (Figure 4). The remaining three snmrna candi- Our Experimental RNomics approach applied to the dates, Ath-270, Ath-310, and Ath-587, all precede or plant Arabidopsis thaliana has generated 140 candifollow genes coding for proteins that are integral parts dates for small non-messenger RNAs and has considerof the photosystem I or II complex. ably extended the number of small non-messenger Group 2: snmrnas in Introns RNAs known in plant species to date. From Class I In general, intron sequences from pre-mrnas are rap- snmrnas (snmrnas that exhibit conserved sequence idly degraded upon removal from pre-mrnas. An ex- or structure motifs), we have isolated 39 novel members ception are snornas that are encoded by a portion of of H/ACA box snornas directing pseudouridylation of introns in some eukaryal lineages, such as vertebrates RNAs. In addition, we have found 49 members of C/D [17, 18]. We have found three snmrna candidates in box snornas directing ribose methylations of RNAs. our cdna library encoded by introns (Table 2, Group 2). Among the C/D and H/ACA box snornas, we identified, Ath-260 and Ath-329 snmrnas are located within the for the first time in a plant species, four snornas capanuclear introns of gene F8K4.2 and of a putative glycer- ble of modifying small nuclear RNAs. Another snmrna, inaldehyde-3-phosphate dehydrogenase gene, respec- Ath-74, that is related to trna Phe was found. Because tively. Ath-602 maps to a group II intron of the cyto- of its trna-like features, the snmrna might have some chrome b/f gene on the chloroplast genome. The regulatory function in A. thaliana for example, in propresence of a small, stable RNA species derived from tein synthesis as predicted for other trna-like RNAs [27]. a group II intron might indicate a physiological role for From Class II RNAs (lacking known sequence or structhe RNA species, potentially via binding to (a) protein ture motifs), we have identified 36 candidates for small splicing factor(s). non-messenger RNAs. Surprisingly, 15 of them showed Group 3: snmrna Derived from Annotated a tissue-specific expression. Interestingly, out of 36 Open Reading Frames (ORFs) snmrna candidates, 12 mapped to the mitochondrial Four candidates for snmrnas that were derived from or chloroplast genomes. Especially intriguing was the ORFs or overlapped ORFs were identified in our screen identification of Ath-328 snmrna in chloroplasts, which (Table 2, Group 3). According to Northern blot signals, maps to the 3 UTR of the ribulose 1,5-biphosphate the sizes of the snmrna candidates were estimated carboxylase mrna, encoding the central enzyme for well below the size of the predicted open reading frames CO 2 fixation in plant chloroplasts. Our study sets the from which they are derived. This is consistent with the stage for the functional analysis of all 140 candidates presence of a small, stable RNA species rather than the for snmrnas in A. thaliana. identification of the respective mrna. Still, at this point, we cannot exclude the possibility that these snmrnas Experimental Procedures represent extremely abundant and stable degradation All Experimental Procedures regarding library construction and products of mrnas. analysis have been published previously [3, 4]. More detailed meth- Two snmrna candidates overlapping coding regions ods for library construction and analysis from A. thaliana are deare derived from the mitochondrial or chloroplast ge- scribed in the Supplementary Experimental Procedures in the Supnome (Figure 4). Especially intriguing is clone Ath-328, plementary Material. which maps to the 3 UTR of the ribulose 1,5-biphosphate carboxylase mrna encoding a central enzyme Supplementary Material in chloroplasts that catalyses CO 2 fixation (Figure 4). Supplementary Material including the Experimental Procedures as well as a paragraph on the genomic organization and promoter Interestingly, the 3 portion of the Ath-328 RNA exhibits elements of C/D and H/ACA snorna genes from A. thaliana is availa very stable stem structure, characteristic for many able at small, stable RNAs (see Figure S2). Such stem structures located within the 3 UTRs of mrnas have been pre- Acknowledgments viously implicated in the processing and 3 -end formation of chloroplast mrnas. In addition, they were shown We would like to thank Antje van Schaewen for supplying plant to bind to specific proteins involved in mrna 3 -end material from A. thaliana and Jim Brown and Stefan Binder for helpful formation [26]. It remains to be tested whether Ath-328 discussions. This work was supported by the German Human Ge- nome Project through the Bundesministerium für Bildung und is in some way involved in the regulation of ribulose 1,5- Forschung (BMBF) (#01KW9966) to A.H. and J.-P.B., by an Interdisbiphosphate carboxylase gene expression or represents ziplinäres Zentrum für Klinische (IZKF) grant (Teilprojekt IKF G6, merely a degradation product of the mrna turnover Münster) to A.H., by laboratory funds from the Centre National de pathway in chloroplasts. la Recherche Scientifique and Université Paul Sabatier, Toulouse,

12 Non-Coding RNAs in the Plant A. thaliana 2013 and by a grant from the Ministère de l Education Nationale, de la cursor snorna to functional snornp. Curr. Opin. Cell Biol. 11, Recherche et de la Technologie (Programme de Recherche Fondamentale en Microbiologie et Maladies Infectieuses et Parasitaires, 18. Filipowicz, W., and Pogacic, V. (2002). Biogenesis of small nu ) to J.-P.B. cleolar ribonucleoproteins. Curr. Opin. Cell Biol. 14, Massenet, S., Mougin, A., and Branlant, C. (1998). Posttranscriptional modifications in the U snrnas. In Modification and Editing Received: August 8, 2002 Revised: September 30, 2002 of RNA: The Alteration of RNA Structure and Function, H. Gros- Accepted: October 1, 2002 jean and R. Benne, eds. (Washington, D.C.: ASM Press), pp. Published: December 10, Ganot, P., Bortolin, M.L., and Kiss, T. (1997). Site-specific pseudouridine formation in preribosomal RNA is guided by small References nucleolar RNAs. Cell 89, Filipowicz, W. (2000). Imprinted expression of small nucleolar 21. Smith, C.M., and Steitz, J.A. (1997). Sno storm in the nucleolus: RNAs in brain: time for RNomics. Proc. Natl. Acad. Sci. USA new roles for myriad small RNPs. Cell 89, , Jady, B.E., and Kiss, T. (2001). A small nucleolar guide RNA 2. Cavaille, J., Buiting, K., Kiefmann, M., Lalande, M., Brannan, functions both in 2 -O-ribose methylation and pseudouridylation C.I., Horsthemke, B., Bachellerie, J.P., Brosius, J., and Huttenof the U5 spliceosomal RNA. EMBO J. 20, hofer, A. (2000). Identification of brain-specific and imprinted 23. Ni, J., Tien, A.L., and Fournier, M.J. (1997). Small nucleolar RNAs small nucleolar RNA genes exhibiting an unusual genomic orgadirect site-specific synthesis of pseudouridine in ribosomal nization. Proc. Natl. Acad. Sci. USA 97, RNA. Cell 89, Hüttenhofer, A., Kiefmann, M., Meier-Ewert, S., O Brien, J., Lehlution of pseudouridine residues in large subunit ribosomal 24. Ofengand, J., and Bakin, A. (1997). Mapping to nucleotide resorach, H., Bachellerie, J.-P., and Brosius, J. (2001). RNomics: an experimental approach that identifies 201 candidates for novel, RNAs from representative eukaryotes, prokaryotes, archaebac- small, non-messenger RNAs in mouse. EMBO J. 20, teria, mitochondria and chloroplasts. J. Mol. Biol. 266, Tang, T.-H., Bachellerie, J.-P., Rozhdestvensky, T., Bortolin, 25. Meinke, D.W., Cherry, J.M., Dean, C., Rounsley, S.D., and M.-L., Huber, H., Drungowski, M., Elge, T., Brosius, J., and Hütgenome analysis. Science 282, Koornneef, M. (1998). Arabidopsis thaliana: a model plant for tenhofer, A. (2002). Identification of 86 candidates for small nonmessenger RNAs from the archaeon Archaeoglobus fulgidus. 26. Hayes, R., Kudla, J., Schuster, G., Gabay, L., Maliga, P., and Proc. Nat. Acad. Sci. USA 99, Gruissem, W. (1996). Chloroplast mrna 3 -end processing by 5. Brown, C.J., Hendrich, B.D., Rupert, J.L., Lafreniere, R.G., Xing, a high molecular weight protein complex is regulated by nuclear Y., Lawrence, J., and Willard, H.F. (1992). The human XIST gene: encoded RNA binding proteins. EMBO J. 15, analysis of a 17 kb inactive X-specific RNA that contains conlar transport and activity-dependent modulation. Results Probl. 27. Brosius, J., and Tiedge, H. (2001). Neuronal BC1 RNA: intracellu- served repeats and is highly localized within the nucleus. Cell 71, Cell Differ. 34, Ambros, V. (2001). micrornas: tiny regulators with great potential. Cell 107, Accession Numbers 7. The Arabidopsis Genome Initiative (2000). Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. The sequence data have been submitted to the DDBJ/EMBL/Gen- Nature 408, Bank databases under accession numbers AJ AJ Gaspin, C., Cavaille, J., Erauso, G., and Bachellerie, J.P. (2000). Archaeal homologs of eukaryotic methylation guide small nucleolar RNAs: lessons from the Pyrococcus genomes. J. Mol. Biol. 297, Omer, A.D., Lowe, T.M., Russell, A.G., Ebhardt, H., Eddy, S.R., and Dennis, P.P. (2000). Homologs of small nucleolar RNAs in Archaea. Science 288, Klein, R.J., Misulovin, Z., and Eddy, S.R. (2002). Noncoding RNA genes identified in AT-rich hyperthermophiles. Proc. Natl. Acad. Sci. USA 99, Barneche, F., Gaspin, C., Guyot, R., and Echeverria, M. (2001). Identification of 66 box C/D snornas in Arabidopsis thaliana: extensive gene duplications generated multiple isoforms predicting new ribosomal RNA 2 -O-methylation sites. J. Mol. Biol. 311, Brown, J.W., Clark, G.P., Leader, D.J., Simpson, C.G., and Lowe, T. (2001). Multiple snorna gene clusters from Arabidopsis. RNA 7, Liang-Hu, Q., Qing, M., Hui, Z., and Yue-Qin, C. (2001). Identification of 10 novel snorna gene clusters from Arabidopsis thaliana. Nucleic Acids Res. 29, Llave, C., Kasschau, K.D., Rector, M.A., and Carrington, J.C. (2002). Endogenous and silencing-associated small RNAs in plants. Plant Cell 14, Reinhart, B.J., Weinstein, E.G., Rhoades, M.W., Bartel, B., and Bartel, D.P. (2002). MicroRNAs in plants. Genes Dev. 16, Bachellerie, J.P., Cavaille, J., and Qu, L.H. (2000). Nucleotide modifications of eukaryotic rrnas: the world of small nucleolar RNAs revisited. In Ribosome: Structure, Function, Antibiotics and Cellular Interactions, R. Garrett, S. Douthwaite, A. Liljas, A. Matheson, P.B. Moore, and H. Noller, eds. (Washington, D.C.: ASM Press), pp Weinstein, L.B., and Steitz, J.A. (1999). Guided tours: from pre-