Eur. J . Biochem. f52, 259-266 (1985) ( FFBS 1985

Plant small nuclear RNAs Nucleolar U3 snRNA is present in plants: partial characterization

Tamis KISS, Mhrta TOTH and Ferciic SOLYMOSY

Institute of Plant Physiology, Biological Research Center, Hungarian Academy or Sciences, Szcged

(Received June 3. 1985) - EJR 85 0594

Nuclei, isolated from a number of plant species by either of two independent, newly developed methods, regularly contained a common set of low-molecular-mass RNAs. Partial characterization of these RNAs, based on cell fractionation, polyacrylamide gel electrophoretic and chemical sequencing techniques, as well as comparison with literature data, revealed that, in addition to tRNA, 5s RNA and 5.8s RNA, plant nuclei contain two families of low-molecular-mass RNAs, that are counterparts of vertebrate U1 and U5 RNAs respectively, and three individual low-molecular-mass RNA species. One of these may be related to vertebrate U6 RNA. The two others are true eukaryotic U2 and U3 RNAs, respectively, on the basis of the following lines of evidence obtained from analyses of broad bean nuclear RNAs.

1. The 3‘-end portion (121 nucleotidcs sequenced) of broad bean U2 RNA shows a nearly perfect sequence homology with that of authentic pea U2 RNA.

2. Broad bean U3 RNA is localized in the nucleolus and its 3’-end portion (164 nucleotides sequenced) ( a ) shows sequence homology with that of both rat U3 RNA (48%)) and Dictyostc~liun7 D2 RNA (39%), (b) has a secondary structure which fits perfectly that proposed for both rat U3 RNA and Dictyostrlium D2 RNA, and (c) contains the specific sequence which, in a model based on the primary structure of rat U 3 RNA, is supposed to be involved in the processing of eukaryotic 32s pre-ribosomal RNA. This is the first report on the occurrence in plants of nucleolar U3 RNA.

Eukaryotic cells are known to contain a special class of uridine-rich small nuclear RNAs (snRNAs). There are six well-defined molecular species (U1 to U6) in this group, of. which U3 snRNA is localized in the nucleolus [I , 21, while the remainder are in the nucleoplasm. Thus far U-rich snRNAs have been isolated and characterized with respect to their size and in many cases to both their primary and second- ary structures from all eukaryotes examined (cf. [3]) .

These structural studies have shown (cf. [3]) that U-rich snRNAs (a) are linear single-stranded molelcules containing 106 ~ 21 6 nucleotide residues, depending on the molecular species, (b) are capped a t their 5’-end (m:*2*7G for U1 to U5 snRNAs), (c) contain modified nucleosides, (d) occur in snRNP particles [4] often associated with hnRNP particles [5 - 71, (e) are highly conserved in nature with respect to their size as well as their primary and secondary structure and (i] are metabolically stable [I].

In contrast to this detailed knowledge of their structure, very little is known about their function. U1 snRNA seems to be involved in the processing of pre-mRNA and U3 snRNA in that of pre-rRNA [9 ~ 121.

One approach to obtaining a deeper insight into the functional implications of the structural features of U-rich snRNAs is the compilation of a catalog of primary and sec- ondary structures of the different U-rich snRNA specks from phylogenetically remote eukaryotic organisms, and searching for common sequences with special emphasis on their location

in the secondary structure of the respective U-rich snRNA. Such an approach has already led to a remarkable discovery. Branlant et al. [I 31 have found in U1, U2, U4 and U5 snRNAs of evolutionary distant species a consensus sequence (domain A) that might well be a protein binding site [14].

Strangely enough, apart from the determination of partial sequences of variants of an RNA (WI, possibly related to vertebrate U2 snRNA) of wheat [15] as well as that of the sequences of five variants of U5 snRNA and of the 3‘-terminal region of U1 and U2 snRNAs of pea [16], no work has been published so far on the structure of U-rich snRNAs from plants. One reason for this shortcoming may lie in the difficulties involved in different methods of isolation of plant nuclei and nucleoli in suitable amounts and purity for bio- chemical analysis. The only procedure which has proven useful thus far involved use of anti-cap antibodies for the sclective isolation of some U-rich snRNAs from a total nucleic acid extract of pea nuclei [16]. The use of this approach, however, did not reveal the presence of either a U3-like snRNA, or that of a U6-like snRNA (because of the absence in U6 snRNA of the canonical cap structure [17]).

In this paper we not only report a procedure suitable for the investigation of the entire spectrum of snRNAs from plant material but show, on the basis of sequencing data, that at least two of the molecular species obtained in this way are characteristic eukaryotic U-rich snRNAs.

Corre.spon~~cnc.e to F. Solymosy, Novi-nyelettani Intezet, Magyar Tudomhnyos Akademia, Szegedi Biologiai Kozpont, Postafiok 521, AND METHODS

H-6701 Szeged, Hungary Clirnz icrr Is

sulfatc; SII, small nuclear. Dithiothreitol was purchased from BDH (Broom), T4 polynucleotide kinase from Boehringer (Mannheim), Dul-

(6.5.1.3). becco’s modified Earl’s medium from Flow Laboratories

Ahhwvimtions. hn, heterogeneous nuclear; SDS, sodium dodecyl

Enz,pe .r . T4 polynucleotide kinasc (2.7.1.78); T4 R N A ligase

260

(Irvine), guanidine thiocyanate from Fluka (Buchs), Macerozyme R-10 and cellulase 'Onozuka' R-10 from Kinki Yakult (Nishinomiya), hydrazine from Eastman Kodak (Rochester), aniline, dimethylsulfate, dimethylsulfoxide and VOS04 from Merck (Darmstadt), Percoll from Pharmacia (Uppsala), T4 RNA ligase from P. L. Biochemicals (Mil- waukee, WI), EDTA, 2-mercaptoethanol, phenylmethyl- sulfonyl fluoride and sodium dodecyl sulfate (SDS) from Serva (Heidelberg), and bovine serum albumin, cytidine 3'- monophosphate, diethylpyrocarbonate, ethidium bromide, Mes, N,N-methylcnebisacrylamide, and Nonidet P-40 from Sigma (St Louis). Vanadyl ribonucleoside complex was pre- pared from VOS04 according to Berger and Birkenmeier [I81 with the modification reported by Lichtler et al. [19]. Aniline was used after distillation and acrylamide after recrystalliza- tion from ethyl acetate. All other chemicals were of analytical grade and received through Reanal (Budapest).

Filters

The nylon filters were obtained from Zuricher Beutel- tuchfabrik AG (Ruschlikon).

Plan/ muteriul

Fully expanded leaves of 3-week-old broad bean (Viciu fuha L.), 6- 8-week-old tomato (Lycopersicon esculentum Mill cv. K-700), 5 - 6-week-old cucumber (Cucumis sativus cv. Belair) and 6 - 7-week-old tobacco (Nicotianu tahucum cv. Samsun) plants, grown under ordinary greenhouse conditions at 25 C and 16 h light period, were used. Lyropersicon peruvianum (Mill) cell suspension cultures were maintained according to Nover et al. [20] and used 3 days after passage.

Animals

Male rats (PVGIC) weighing 200 - 250 g were used.

Crll culture

Baby hamster kidney BHK-21 cells were grown in mono- layer cultures in Dulbecco's modified Earl's medium sup- plemented with 10% new-born calf serum, 50 IU/ml penicillin and 50 pg/l streptomycin.

Isolation af'nuclei,fi.om lcwf material

Nuclei were isolated by either of two methods. With both methods all operations were done at 4"C, all centrifugations were carried out in swinging-bucket rotors and all solutions, except those containing Percoll, Tris and Mes, were treated with diethylpyrocarbonate according to Solymosy et al. [21].

Method I was a combination of the citric acid method, originally described for use with mammalian tissues [22], and certain steps specially suited for plant tissues, as reported by Willmitzer and Wagner [23]. It was carried out as follows. Leaves (about 100 g) were homogenized in 6-8 volumes of isolation buffer containing 25 mM citric acid, 250 mM su- crose and 0.7?4 Triton X-100 by using an Ultra-Turrax homogenizer at full speed for 4 x 20 s. The homogenate was filtered through nylon filters of decreasing pore size (67 pm, 20 pm and 10 pm). With broad bean and tobacco, filtering through the 10-pm nylon filter was omitted. The filtrate was centrifuged at 800 x g for I 0 inin, the resulting pellet washed

with 0.1 volume of isolation buffer and centrifuged at 600 x g for 5min, then suspended in 20ml Percoll adjusted to pH 6.2--6.5 by a few drops of 5 x concentrated isolation buffer. This suspension was then centrifuged at 4000 x g for 5 min. During this step starch grains sedimented to the bottom of the tube and nuclei floated to the top of the Percoll cushion. Finally the nuclei were washed three times with the isolation buffer. An aliquot of the nuclear preparation was stained with carmine and viewed in the light microscope. The usual yield was about 2 x 108 nuclei per 100 g leaf material.

Method 11 was based on the method described by Willmitzer and Wagner [23] with some modifications of buffer compositions according to M. Van Montagu (personal com- munication), Hamilton et al. [24], Rizzo et al. [25), Mennes et al. [26] and Luthe and Quatrano [27]. It was carried out as follows. Leaves (about 100 g) were rinsed once with cold diethyl ether, once with the isolation buffer, consisting of 5 mM Mes, pH 6.5, 150 mM sucrose, 2.5 mM CaC12, 40 pM phenylmethylsulfonyl fluoride, 10 mM vanadyl ribonucleo- side complex, 0.7% Triton X-100, and 50% glycerol, then homogenized in 6-8 volumes of the isolation buffer under the conditions described in method I. Filtration through cheesecloth and nylon filters as well as centrifugation of the filtrate was also done as described for method I. The pellet, washed twice with 60- 80 ml isolation buffer and centrifuged at 800 x g for 5 min, was suspended in 20 ml of a 95% Percoll solution containing 25 mM Mes, pH 6.5, 250 mM mannitol, 1 mM dithiothreitol, 10 mM KC1, 5 mM MgClz and 0.6% Nonidet P-40. The suspension was then filtered through a nylon filter (pore size of 20 pm). The purified nuclei from the filtrate were obtained by flotation in Percoll as described in method I and washed twice with a buffer containing 10 mM Mes, pH 6.5, 300 mM sucrose, 5 mM CaCI,, 5 mM 2-mercaptoethanol and 0.6% Nonidet P-40. The usual yield was about 3 x 10' nuclei per 100 g leaf material.

Isolation of nuclei froni cell mspensiori cultures

This was done by method 1. The starting material consisted of cells (20-30 g wet weight) which had been subjected to plasmolysis and partial enzymatic digestion basically accord- ing to Willmitzer and Wagner [23] except that the plasmolysis buffer consisted of 0.7 M mannitol and 10 mM Mes, pH 5.8 (L. Willmitzer, personal communication) and that partial enzymatic digestion was carried out at room temperature for 1 h in the above plasmolysis buffer supplemented with 0.05% Macerozyme, 0.1% cellulase and 10 mM vanadyl ribonucleoside complex. The cclls treated in this way were rinsed once with the isolation buffer (see method I).

Prepurcrton of nuclei from rut liver

This was done according to Reddy et al. [28].

Isolution of nuclroli,from nuclei of' V. faba

Nucleoli were isolated according to Schumacher et al. [29] froin lo8 nuclei obtained by the use of method 11. The supernatant above the sucrose cushion was termed the extra- nucleolar fraction.

Crll fructionation from leaf nzuteriul

Leaves (100 g) were homogenized according to method I, except that the isolation buffer contained no detergent. The

261

nuclei were centrifuged at 250 x g for 2 - 3 min, and the nu- clear pellet was washed with the isolation buffer supplemented with 0.7% Triton X-100. From this stage on, method I was followed. The supernatant was subjected to centrifugation at 800 x g for 5 min to sediment the chloroplasts. The super- natant of this second centrifugation step was considered as crude cytoplasmic fraction, an aliquot of which served as a source for the isolation of ribosomes. To this end a 10-ml portion of the crude cytoplasmic fraction was clarified by centrifugation at 13,000 x g for 15 min and the ribosomes from the clarified sap were sedimented through 20 ml 1.75 M sucrose in 0.5% citric acid by centrifugation at 40,000 rpm for 45 min in a type 42.1 rotor in the Beckman ultracentrifuge.

Extraction and~fractionation qf R N A

RNA was extracted essentially by the hot phenol/SDS method of Steele et al. [30] from rat liver nuclei, baby hamster kidney cells, plant nuclei, plant nucleoli, chloroplasts and ribosomes, as well as from ethanolic precipitates of the cytoplasmic fraction from tobacco and of the extranucleolar fraction from V.,faba. RNA (1/30 to 1/40 part of an extract from 2 x 10' to 3 x 10' nuclei) was fractionated on 10% polyacrylamide gel made up in 50 mM Tris, 50 mM boratc, 1 mM EDTA, pH 8.3, containing 8 M urea. The bands were visualized by ethidium bromide staining and excised. The RNA was eluted at 37°C for 4-6 h using a solution that contained 0.5 M ammonium acetate. 10 mM magnesium ace- tate, 0.1 mM EDTA and 0.1% SDS. The gel pieces were removed by centrifugation and the volume of the supernatant was reduced by consecutive extractions with 2-butanol before ethanol precipitation.

Preparation oj'3'-end-luheled R N A

RNAs was achieved according to England et al. [31]. Synthesis of [5'-32P]pCp and 3'-end labeling of purified

Nucleotide sequence analysis

Chemical sequencing of 3'-end-labeled RNA was per- formed according to Peattie [32] with the following modifi- cations. The guanosine, adenosine and the strand scission aniline reactions were run for 1.5 min, 4 min and 10 min re- spectively. The digestion products were fractionated on 20% and 10% polyacrylamide gels according to Sanger and Coulson [33].

RESULTS AND DISCUSSION

Isolation of nuclei

Since U-rich snRNAs from plant material cannot be isolated by anti-Sm antibodies [8] and the use of anti-cap antibodies apparcntly did not reveal the entire spectrum of the anticipated plant U-rich snRNAs [16], we applied conventional biochemical methods for the extraction and characterization of low-molecular-mass RNAs present in plant nuclei. A suitable method had to yield plant nuclear preparations with a minimum degree of cytoplasmic con- tamination and endogenous nucleolytic degradation. After a number of initial attempts to use published procedures, which met with little success owing to substantial degradation of high-molecular-mass KNAs that interfered with thc detection

of low-molecular-mass nuclear RNAs, we developed two methods for the isolation of nuclei from plant material.

In method I the low pH of the isolation buffer was meant to minimize endonucleolytic degradation of RNAs during nuclear isolation, and flotation of the crude nuclear prepara- tion in Percoll led to the separation of plant nuclei from starch grains. This is an essential step because any starch remaining in the RNA extracts will cause smears in the gel patterns.

Method I1 was used for the isolation of plant nuclei under non-acidic conditions. In this case inclusion of vanadyl ribonucleoside complex in the isolation buffer was absolutely essential to prevent endogenous nucleolytic activity.

Fig. 1 shows light microscopy photographs of nuclei pre- pared from broad bean leaves by methods I (Fig. 1 A) and I1 (Fig. 1B). One can see that these nuclear preparations contained little contamination except for some filamentous material in that obtained by method I. This contamination apparently represents acid-precipitated chromatin fragments as revealed by their intensive fluorescence in ultraviolet light after staining with ethidium bromide (not shown).

Gel electrophoretic putterns oflou~-molecular-mass RNAs extracted f rom broad bean nuclei

When preparations of total RNA, isolated from broad bean nuclei obtained by methods I and 11, were subjected to polyacrylamide gel electrophoresis in parallel with RNA extracted from rat liver nuclei isolated according to Reddy et al. [28], and used as a standard throughout, the patterns shown in Fig. 2 were obtained. The overall patterns were quite similar with all three samples. A comparison of the RNA patterns obtained from nuclei isolated from rat (lane A) and broad bean (lanes B and C) cells reveals that the number of RNA regions was the same, but the complexities of the individual regions exhibited notable differences. The most conspicuous agreement in mobility was observed (in addition to the bands in the tRNA region) between band 3 (lanes B and C) and U3 (lane A) on the one hand, and between band 2 (lanes B and C) and U2 (lane A) on the other.

The broad bean low-molecular-mass RNAs in these bands and regions (a) have a non-phosphorylated 3' end, because they could be labeled by an RNA-specific enzyme, T4 RNA ligase, without prior dephosphorylation (not shown) and (b) are neither artifacts of nuclear isolation nor nucleolytic breakdown products, because they could be also detected in total plant nucleic acid extracts obtained by three independent methods, namely the guanidine thiocyanate method [34], the diethyl pyrocarbonate method [35,36] and the phenol method, in all of which RNase activity is known to be inhibited from the very beginning of the extraction procedure (not shown).

Comparison of the gel electropizoretic palterns of low~-molecular-ma.ss R N A s extractedjrom nuclei q f ' d i t f t ~ n t plant species

Fig. 3 shows that the general pattern of low-molecular- mass RNAs, extracted from nuclei isolated by method I from a number of plant species, was quite similar, implying that these RNAs represent a class of molecules characteristic of plants in general. The individual components of this class of RNA, however, do not seem to be identical in the different plant species examined, since both the staining intensities and thc clectrophoretic mobilities of the individual RNA bands and regions varied among the plant species. It is of particular interest to note that in the RNA sample from tomato nuclei

262

Fig. 1 , Light microscopy photogruph of carmine-stained nuclei prepured.fiorn hroud heun leuves b.v rnc4iod I ( A ) ond mrtliod I1 i B ) . The bar represents 20 pm. For more details, see Materials and Methods

(lane C), band 3 was a double one while in that from cucumber (lane E) instead of band 3 two faster migrating major bands appeared. This suggests that the low-molecular-mass RNA in band 3 consists of two subspecies in tomato and cucumber.

CIiararIerization ?f'lowmolecular-mass RNAs cxtracled from plant nuclei: Subcellular localization and clectrophorrtic mobility

For these experiments subcellular fractions were prepared from tobacco rather than broad bean leaves, since in broad bean an excellent chloroplast marker, 4.5s rRNA, has been shown to migrate to the same position as tRNA and thus be easily mistaken for it (cf. [37]). The results in Fig. 4 first of all aid in the unequivocal identification of the 5.8S, 5s and tRNA bands in lane A (rat liver nuclei) since the molecular species designated 5.8S, 5s (both mostly ribosome-derived RNA species) and tRNA (mostly cytoplasm-derived RNA species) were less abundant in lane A than in lane B, on which low- molecular-mass RNAs from whole baby hamster kidney cells, rather than from nuclei, werc loaded.

The low-molecular-mass RNAs extracted from plant nuclei were free of contamination by chloroplast RNA as indicated by a lack in Lane F of the band representing chloroplast 4.5s rRNA (marked by an arrow in lane E). The individual regions and bands in the gel electrophoretic pattern (lane F) can be characterized as follows. Region 8 represents tRNA, because (a) its mobility in the gels is identical with that of tRNA from rat liver nuclei (compare lanes A and F) and (b) it is equally abundant in all subcellular fractions (lanes C to F). Region 7 is more complex than its counterpart, rat '4.5s' (compare lanes A and F), and seems to contain at least five subfractions (lane F); all of them occur in the cytoplasm as well (lane C). No further characterization of this region was undertaken. Band 6 contains a distinct molecular species (lane F and Fig. 2, lanes B and C) and is purely of nuclear

location. Although it has a higher mobility in this system than rodent U6 snRNA (compare lane F with lanes A and B), it is tempting to suppose that this band contains the plant equiva- lent of vertebrate U6 snRNA [17,38]. Further data are needed to test the validity of this assumption. The material in region 5 (compare lane F with lanes C to E) seems to contain 5s rRNA (faster moving part of region 5) and a family of U5 snRNA variants (slower moving part of region 5 ; [16]). The material in region 4 is the most poorly characterized component of plant snRNA. It is definitely of nuclear origin (compare lane F with lanes C to E), however, the lack of any distinct band in this region a s well as the very low amount of the RNA it contains, made any further characterization of this RNA component impossible. A similar hazy region with the same electrophoretic mobility as that of region 4 seems to be present also in gels onto which nuclear or nucleolar extracts from rat liver and Novikoff hepatoma cells were loaded (see Fig. 2, lane F and Fig. 4 in [28]). In this case, however, a distinct band containing U4 snRNA is also present. It remains to be elucidated whether region 4 rcpresents a family of plant U4 snRNAs or is the result of some endonucleolytic degrada- tion. Region 1 consists of a number of bands. The slowest moving one contains plant 5.8s rRNA (compare lane F with lanes C and D) and the faster moving ones most probably comprise a family of U1 snRNA species. A comparison of lanes A and B with lanes C and D indicates that plant 5.8s rRNA exhibits about the same mobility in 10% urea] polyacrylamide gels as does mammalian U1 snRNA. Although the broad bean 5.8s rRNA could be extracted from our gels in a pure state and sequenced up to a length of 117 nucleotide r,esidues from the 3' end (not shown), thus confirming its identity with authentic broad bean 5.8s rRNA [39], the faster moving components in this region escaped all our attempts at proper fractionation into individual molecular species lending themselves to sequence analysis. Bands 2 and 3 (lane F) contained low-molecular-mass RNAs definitely of

263

Fig. 2. Elwtrophorelic ptrttwns c~f'lo~~~-molc~c~ular-mmss RNAs extracted bj. the I i o t phenollSDS method Jiorn nuclei isolated ,from ral liver uccording 10 Keddjs r t al. /28/ (lane A ) , and.from broad bean Ieuves hi' method I (Iunc B ) or method II (lane C ) . For morc dctails, scc Materials and Mcthods. The designations on the left refer to thc bands and regions in lane A and were established by calculations [niobiliiy versus log (number of nucleotide residucs)] and by comparing the pattern in lane A with that reported by Reddy et al. [28]. The bands and regions in lancs R and C are marked on the right and are numbered to emphasix similarities with the bands and regions in lane A

nuclear location. They are equivalents of the single RNA species obtained from broad bean and shown in bands 2 (Fig. 2, lanes B and C) and 3 (Fig. 2, lanes B and C). These latter are authentic plant U2 and U3 snRNAs, respectively, as shown below.

~ ~ ~ ~ i i t i f ~ ~ ~ t i ~ ) n of broad bean lo~~~-i-mok~~rular-rnass R N A of btIIld 2

Band 2 (Fig. 2, lane B) contained a single species of broad bean U2 snKNA. This was confirmed by our sequencing data (Fig. S) , which revealed a nearly perfect sequence homology between the 3'-end portion of broad bean U2 snRNA and that of pea U2 snRNA [16]. This implies that the secondary structure of these two U2 snRNAs, both derived from legumes, is also identical, including the sequence of Domain A [13], probably a protein-binding site [14], between nucleotides 85 an 93 in broad bean U2 snRNA. It is interesting to note that in addition to domain A, which is present also in wheat W1 low-molecular-mass RNA and rat U2 snKNA, there are a number of additional stretches that are homolo- gous in U2 snRNAs from broad bean, pea and rat. The longest

Fig 3. Elwtrtiphoretic pattcrns of lo,c-molecular-mass R N A s extracted thr hot plrenol/SDS methodfrom nuclei isolated according to Reddy

et ul. [-3;Y],fionz rut liver (lanes A and F ) , and by method Ijirom leaves of'lwoad h e m (lane B ) , tomato (lane C ) and cucumber (lane E) us ~ t v ~ l l as , f rom su.~pen.rion culture cells of' L. peruvianuni (lane D ) . For more details, see Materials and Methods. The designations on the left and on the right refer to lanes A and F, respectively, while those in the insert between lanes B and C refer to lane B

such stretch can be found from nucleotide 60 to nucleotide 67 in broad bean U2 snRNA (Fig. 5) . It is unfortunate that sequence data of W1 low-molecular-mass RNA from wheat [15] are missing for a major part of this particular stretch. It should also be noted that a stretch between nucleotides 97 and 109 of broad bean U2 snRNA exhibits a considerable sequence homology with corresponding stretches of W1 low- molecular-mass RNA from wheat and rat U2 snRNA. Unfor- tunately, sequence data for this particular stretch are not available from pea snRNA for comparison.

Idepitification of broad bean low-molecmlav-mass R N A of hand 3

Band 3 (Fig. 2, lane B) contained a single species of snRNA. The identity of this molecular species with eukaryotic U3 snRNA is supported by four lines of evidence. (a) When RNA was extracted separately from the nucleolar and the extranucleolar fraction of nuclei isolated from broad bean leaves by method I1 and was run on polyacrylamide gel slabs in parallel with appropriate controls, the pattern shown in Fig. 6 was obtained. Although there were some common features (most probably reflecting nuclear contamination) between these two fractions (compare lanes C and D), the

264

Fig. 4. Electrophoretic patterns c~flor~-niolecular-muss R N A s extructed by the hot plictiol/SDS nietliod,frorn nuclei of’rut liver (lane A ) , ,from huhy humster kidncy c d l s (lane B ) , und Jrom cytoplusniic (lane C/ , ribosomal (lane D ) , chloroplast (lane E ) and nuclear (lanil F)Jiactions of rohucLo l ea~es . Nuclci wcrc isolated from rat liver according to Reddy et al. [2S]. For more details, see Materials and Methods. Thc designations on the left refer to lanes A and €3, while those on the right refer to lane F. The arrow in lane E indicates chloroplast 4.5s rKNA (cf. [37])

120

Fig. 6. Electrophoretic patterns c~flow-niole~ular-tiz~is.s R N A s ertracted 1 7 ~ . the hot phcnol/SDS melhodJrom nuclei isolaied according to Ret ldj c’t ul. [28],f iom rut liver (lune A) ,Jkom nuclei isolated by nieihod I 1 fiom broad hrun leuves (lane B ) US i 4 d l us ,from the i~.~trrmucltwlur (lane C ) and izuc,lc.olrrr,fruc~ioii (lune D ) of nuclei isolutcd Iiy tnethod II.from hroudbcutz k m ~ e s . For more details. see Materials and Mcthods. The designations on the left refer to lane A, while those on the right refer to lancs B, C and D

100

40 2P U G U - C ~ C U U U U G C G U U G C A C U A U A G C A A U U G C U G G C G C A C C C C A C O H 20

40

. . . - . l . . . . . . . . . . .=l .=====~========m=========~H 40

U G 6 . C AG..U.....m..8mG U=CUGG..A...........A~~, 43 20

C ~ . - . . A C C . G . U A . . . . . G ~ ~ ~ C U ~ C ~ G G A A C ~ ~ U ~ ~ ~ ~ ~ - - ~ ~

Fig. 5. Comparison o f the sequence ofthe 3’-end region o f t h e R N A in hand 2 f rom broad beun nuclei iviih thosc o f p c u U2 .snRNA 16) . whcut Wl’ R N A [ 151 and rat U 2 s n R N A [3] . The numbering ofnucleotide residues starts from the 3’ ends of the moleculcs. Sequences arc aligned for maximum homology. Dashes indicate deletions. Areas enclosed by parantheses represent unscqucnced stretches and X dcnotcs positions which were countcd but not assigned. Bold lines indicatc known sequences and dashed lincs indicatc unknown sequences up to thc 5’ end of the molecules. Solid squares in thc scqucnccs stand for nuclcotides which are identical i n either pea U2 snRNA or wheat W1’ R N A o r rat U2 s n R N A with thosc in broad k d n U2 snRNA

265 180 14,O

n road bean t i 7 5 ' - - - -A A G G A G A C A G G A A G U G U R A G U C U G G U G G A U G A A U C -

5 ' ~ C U . A C U G U G U A G . . C C C . C . A A A C C A C ~ ~ G ~ ~ C G A - Rat U3B

D 2 5'~~.UC.AUA.UUUUUCCUCUU~.AUAGCU.G..UGAU

160 140

140 160

40 20

G b U U A C - A C U G G C U - - G - G G G U G G U 6 A C A C G G U G A U C U G A - C A G A C o H

C d . . - . G G G G . . G . A A . A ~ . ~ A ~ ~ - G ~ - ~ ~ ~ C A ~ - ~ ~ ~ ~ ~ G U G ~ ~ ~ 40 2?

A.GG.uG.A.u.G.uGuG;.....-G.-uuc..-.-..I-G~u~~

Fig. I. C'onri~arison of [he sequence of the 3' end r q i o n of the R N A in burrd 3 from broud bean nuclei with those of rut U3B snRNA 131 and Dictyostclium 0 2 RNA ( 4 2 / . The numbering of nucleotide residues starts from the 3' ends of the molecules. Scquences arc aligncd for maximum homology. Dashcs indicate dcletions. Bold lines indicate known sequcnces and dashed lines indicate unknown sequences up to the 5' end of the molecules. Solid squares in the sequences stand for nucleotides which are identical in either U3B s n R N A or 0 2 RNA with those in broad bean U3 snRNA

A U- L D s t e m A

J A A

160

--- A A G G AGA c A'

Fig. 8. Prop(i,wd secondury strucfure model of the 3'-end region ofbroad bean US snRNA. Secondary structure was derived by comparison with that reported by Bernstein el al. [40] for both rat U3B snRNA and Dictjmteliurn D2 RNA. Marking of thc stems is identical with that of Bernstein et al. [40]. The dashed line indicates continuation of unknown sequence up to thc 5' end of the molecule

nucleolar fraction was unequivocally characterized as being the actual site of accumulation of the material present in band 3. This band from broad bean nucleoli (lane D) exhibited electrophoretic mobility identical with that of the band containing rat U3 snRNA (compare lanes B and D with lane A) of experimentally verified nucleolar location [2]. (b) When the low-molecular-mass RNA, extracted from band 3 (Fig. 2, lane B), was subjected to chemical sequencing, its 3'-end por- tion (164 nucleotides sequenced) was found to show 48% sequence homology with that of rat U3 snRNA and 39% sequence homology with that of Dictyosteliurn D2 RNA, as shown in Fig. 7. (c) The 3'-end portion (164 nucleotides sequenced) of the low-molecular-mass RNA molecule in band 3 (Fig. 2, lane B) was found to fit perfectly, with all stems (B to F) conserved, the secondary structure model proposed by Bernstein et al. [40] for the corresponding portion of both rat U3B snRNA and Dictynstelium D2 RNA (Fig. 8).

(d) Bachellerie et al. [I01 have proposed that U3 snRNA may be involved in the conversion of eukaryotic 32s pre-rRNA into mature 28s rRNA through base-pairing with precursor- specific sequences at the boundaries of excised domains. In this interaction a distinct role has been attributed to a specific sequence originally described for rat U3B snRNA as GAUGAUCGUC [41] and later modified to GAUGAUCGUUC [3]. This latter sequence was found by us also in broad bean U3 snRNA (nucleotides 48 - 58 in Fig. 7). It is prescnt in a somewhat altered form in D2 snRNA of Dict,yoLstelium as well (Fig. 7).

Concluding remarks

Even though the functions of snRNAs are not yet fully understood, the apparent presence of these particular molec- ular species in plant nuclei can be taken as evidence that they

are functionally impor t an t nuclear componen t s . Our study shows, once again, t h a t particular a t tent ion t o use o f isolation and preparative procedures suited to various peculiarities of plant cells can be of considerable significance in assessing the presence and locale of certain molecules and their functions, hence in furthering our comprehension of comparative bio- chemistry a n d molecular biology of eukaryotes .

We thank Professor J . Schell (Max-I'lanck-Institut fur Zuch- tungsforschung, Kohl) and Dr L. Willmitzer (Max-Planck-lnstitut fur Zuchtungsforschung, Kiilu) for helpful discussions, Professor E. R. Lcadbelter (University of Connecticut, Storrs, Connecticut) for valuable comments on the manuscript, Drs L. Nover (Institut fur Biochemie der Pflanrcn, AdW, Halle) and K. Kariko (Institute of Biophysics, Biol. Res. Center, HAS, Szeged) for supplying 1,. peruriurumi callus tissues and BHK cells, respectively, M r B. Dusha for prcparing the photographs and figures as well as Mrs I . Walsh for typing the manuscript. This work was supported by grants 363/82/1-6 and 811 i400/2/85!1/1 18 of the Hungarian Academy of Sciences.

REFERENCES 1. 2.

3.

4.

5.

6.

7. 8.

9.

10.

12.

12. 13.

14.

15.

Weinberg, R. A. & Penman, S. (2968) .I. Mol. B i d . 38, 289 -304. Ro-Choi, T. S . , Moriyama, Y., Choi, Y. C. & Busch, H. (1970)

Rcddy, R. & Busch. H. (1983) Prog. Nuc1c.i~ Acid. Res. Mol. Bid .

Lerner, M. R. & Steilz, J . A. (1979) Proc. Nut1 Acud. Sci. USA

Sekeris, C. E. & Nicssing, J. (1975) Biochern. Bioiphys. Res.

Guimont-Ducamp, C. , Sri-Widada, J. & Seanteur, P. (1977) Bio-

Gallinaro, H. &Jacob, M. (1979) FEBS Letf. 104, 176-182. Lcrncr, M. R., Boylc, J . A., Mount, S. M., Wolin, S. L. & Steitz,

Prcstayko, A. W., Tonato, M. & Busch, H. (1970) J . Mol. R id .

Bachellerie. J.-P., Michot, B. 62. Raynal, F. (1983) Mol. Biol. Rep.

Crouch. R. J., Kanaya, S. & Earl, P. L. (1983) Mol. Biol. Rep. 9 ,

Tagtie, B. W. & Gerbi, S. A. (1984) J . Mol. Evolut. 20, 362-367. Braiilant, C., Krol, A,, Ebel, J.-P., Lazar, E., Haendler, H. &

Liautard, J . P., Sri-Widada, J., Brunel, C. & Jcanteur, P. (1982)

Skuzeski, J . M. & Jeiidrisak, .I. J. (1985) Pluizt Mol. Biol. 4 , 181 -

J. Biol. Chcnr. 245, 1970 - 1977.

30. 127-162.

76, 5495 - 5499.

C'onirnrin. 62, 642 - 650.

chiinie (Paris) 59, 755 - 758.

S. A. (1980) Nuture (Lond.) 283, 220-224.

47, 505-515.

Y, 79 - 86.

75 - 78.

Jacob, M. (1982) EMBO J. 1, 1259 - 1265.

J . Mol. Biol. 162, 623-643.

193.

16. Krol, A., Ebcl. J.-P., Rinkc, J. & Luhrmann, R. (1983) NucliJic

t 7. Epslein. P., Reddy, R., Heuning, D., & Busch, H. ( I 9S0) J . Biol.

18. Bcrgcr, S . I.,. & Birkenmeier, C. S. (1979) Bioch(wii,srt.y 18, 5143-

19. Lichtler, A. C., Sierra, F., Clark, S., Wells, J . R. E., Stein. J . L. &

20. Nover. L., Kranz, E. & Scharf, K. D. (1982) Bioc.1~~7. Ph>,.siol.

21. Solymosy, F., Fedorcsik, I., Gulyis. A., Farkas. G. I.. &

22. Higashi, K., A d a m , H. R . & Busch, H. (1966) Cancer Rt.s. 1 6 ,

23. Willmitzer, L. & Wagner, K . G. (1981) El-/>. Ccll Rcs. 135.

24. Hamilton. R. H., Kunsch, V. & Temperli, A. (1972) Anul. Bin-

25. Rizzo, P. J . , Pcdcrson, K . & Cherry, J . H . (1978) Pluni Sci. Lrrr.

26. Mennes, A. M., Bouman, H., Van der Burg, M. P. M. & Libbenga.

27. Luthe, D. S. &: Quatrano, R. S. (1980) Pluni PIiysiul 65 , 305-

28. Keddy, K., Li, W.-Y., Henning, D., Choi, Y . C., Nohga, K . &

29. Schuinachcr, J . , Singer, H. L. & Riesner, D. (1983) EMBO J. 2,

30. Steele. W. J . , Okamura, N. & Rusch, H. (1965) .I. B i d . C1wnl.

31. England, T. E., Bruce, A. G. & Uhlenbeck, 0. C. (1980) Mijthods

32. Pcaltie. D. A. (,1979) Proc. Nut1 Acud. Sci. USA, 76, 1760- 1764. 33. Sanger, F. & Coulson, A. R. (1978) FEBS Lett. 87. 107- 110. 34. Chirgwin, J . h4., Prrybyla, A. E., MacDonald, R . J . & Rutter.

35. Solymosy, F., Liz i r , G. & Bagi, G. (1970) Anul. Bioc,/ienz. 38,

36. Ehrenberg, L., Fedorcsiik, I. & Solymosy, F. (1976) Prog. Nuclric Acid Res. Mol. Bid. 16, 1 89 - 261.

37. Leaver, C. J. (1979) in Nucleic acids in pl imts (Hall. T. C. & Davies, .I. W., eds) vol. I , pp. 193-215, CRC, Boca Katon.

38. Harada, F., Kato, N. 62. Nishirnura, S. (1980) Biochern. Biophys. RCS. Conzrnun. Y5, 1332- 1340.

39. Tanaka, Y., Dyer, T. A. & Brownlee, G. G. (1980) Nudric Acids

40. Bernstein, L. B., Mount, S. M. & Weiner, A. M. (1983) Cell 31,

41. Reddy, R., Henning, D. & Busch, H. (1979) J . Bid. Chern. 254,

42. Wise, J . A . & Weiner, A. M. (1980) Ccll22, 109- 118.

Acids Res. 11, 8583-8594.

C'lwn. 155. 8901 - 8906.

5149.

Stein, G. S. (1982) Nuture (Lond.) 298, 195-198.

Pflunz. 177, ,483 -499.

Ehrcnbcrg, L. (1968) Eur. J . Biachern. 5 . 520-527.

2196-2201.

70 - 77.

c.he111. 40, 48 - 57.

12, 133-143.

K. R . (1978) PlmI Sci. Lef t . 13, 329-339.

308.

Busch, H. (1981) J . Biol. Cliern. 256, 8452-8457.

1549- 1555.

240, 1 742 - 11 749.

Eiiz j~nol . 65, 65 - 74.

W. J. (1979) Biochemistry 18, 5294-5299.

40-45.

Rex. 86, 1259 - 1272.

461 -412.

11097-lllO5.