INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In eukaryotic cells, most RNA molecules are extensively processed during and following transcription from their corresponding genes. Modifications in RNAs include 5' capping, 3' polyadenylation, splicing, and editing as well as modifications on the base, sugar, and/or phosphate residues. The CCA addition on the 3'-end of transfer RNAs and polyadenylation on the 3'-end of eukaryotic mRNAs are two important 3'-end modifications that have been extensively studied.

Addition of CCA on the 3'-end of tRNAs is ubiquitous, occurring in many evolutionarily distant species including bacteria, yeast, plants, and animals (1, 2). The terminal adenylic acid residue in tRNAs serves as an attachment site for amino acids, and the CCA sequence also participates in the aminoacyl reaction on the ribosome (reviewed in Ref. 1). In some cases the CCA sequence is encoded within the gene; however, in most cases the CCA sequence is not encoded in the corresponding tRNA genes. The CCA sequence turns over rapidly in all tRNAs, and in most cases, tRNA nucleotidyltransferase, the enzyme responsible for CCA addition/turnover, is an essential enzyme (1). In addition to tRNAs, RNAs synthesized by Q (phage Q RNA polymerase) replicase possess a 3'-terminal adenylic acid that is not encoded in the gene (3). The function of this 3'-terminal adenylic acid in the case of Q is not known.

Poly(A) sequences, first discovered in eukaryotic mRNAs (4-7), are also present in some bacterial RNAs (8). The eukaryotic pre-mRNAs are cleaved 10-30 nucleotides downstream of a conserved sequence AAUAAA, which also serves as a binding site for cleavage and polyadenylation specificity factor. The polyadenylation is a complex reaction requiring at least five factors, several of which have multiple protein subunits (5-7). The poly(A) tail has important functions in translation, mRNA degradation, and possibly in transport across the nuclear membrane and in intracellular localization (5, 9).

The human small RNAs in the 75-400 nucleotides size range are synthesized by either RNA polymerase I, II, or III. In human cells, ribosomal 5S, 7SL, 7SK, mitochondrial RNA processing (MRP)/7-2, RNase P, and U6 RNAs are synthesized by RNA polymerase III and terminate with a sequence of UUUU-OH on the 3'-ends (10-14). It is well established that there is trimming of 3'-terminal uridylic acid residues and post-transcriptional uridylation in the case of Xenopus ribosomal 5S RNA (15) and human U6 snRNA¹ (16-19). The human U1-U5 snRNAs are synthesized by RNA polymerase II and terminate 10-15 nucleotides downstream of the mature RNA 3'-ends; this termination is dependent on a termination signal termed 3' box (Ref. 20; for reviews, see Refs. 14 and 21). Many of the recently identified small nucleolar RNAs are derived from intervening sequences of pre-mRNAs, which are synthesized by RNA polymerase II (22-24).

In this study, we characterized the 3'-terminal nucleotide of several small RNA species and found that in every case examined, a fraction of the RNA contained a post-transcriptionally added adenylic acid residue that is not present in the corresponding gene. In the case of 7SL and 7SK RNA, this post-transcriptional adenylation was found in 70% of the RNA molecules. These data indicate that in many human small RNA molecules, a deletion of one or more 3'-end nucleotides that had been incorporated during transcription is followed by the addition of one adenylic acid residue. In addition, an in vitro system capable of efficient and accurate adenylation has been developed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals and Isotopes-- All radioisotopic nucleotides [-³²P]ATP, [-³²P]CTP, [-³²P]GTP, [-³²P]UTP, and [5'-³²P]pCp were purchased from Amersham Pharmacia Biotech. RNA ligase was obtained from New England Biolabs. All other fine chemicals were from Sigma. Nuclease P1, T1 RNase, T2 RNase, and monoclonal anti-trimethylguanosine antibodies were obtained from Calbiochem.

Labeling of HeLa Whole Cell and Nuclear Extract-- Preparation of HeLa cell nuclear extract from cultured HeLa cells was carried out by the procedure of Dignam et al. (25); the whole HeLa cell extract was prepared by the method of Weil et al. (26). The final protein concentration of the extract was 4 mg/ml. For in vitro labeling of RNAs, 5 µl of 10× in vitro transcription buffer (6 mM each GTP, UTP, and CTP, 250 µM ATP, 10 mM dithiothreitol, 200 mM KCl, 60 mM creatine phosphate, and 100 mM Tris-HCl, pH 8.0), 40 µl of nuclear extract, and 50 µCi of [-³²P]ATP were mixed in a total reaction volume of 50 µl and incubated at 30 °C for various time periods. Labeled RNA was extracted using the phenol-chloroform procedure, purified, and fractionated on a 12% polyacrylamide, 7 M urea gel. Whenever necessary, the labeled RNAs were excised from the gel, and the RNA was eluted and purified. These RNAs were then digested with various enzymes and analyzed by chromatography and/or size fractionation on 20% polyacrylamide gels. HeLa cell 4-8S RNA was isolated as described earlier (27) and used for ligation with [5'-³²P]pCp and RNA ligase according to England et al. (28).

Hybrid Selection and Immunoprecipitation-- DNA dot hybridizations were carried out as described by Kafatos et al. (29). Immunoprecipitation of RNA was carried out as described by Lerner and Steitz (30). The hybrid-selected RNAs or the RNAs from the immunoprecipitates were further purified by electrophoresis on a polyacrylamide gel, and individual RNAs were subjected to further analyses.

Digestion of RNAs with Nuclease P1, T1, and T2 RNase-- The RNAs were digested with nuclease P1 at a 1:1500 (w/w) enzyme to substrate ratio at 37 °C for 30 min. Complete digestion with T1 or T2 RNase were carried out as described by Brownlee et al. (31).

Fractionation of Nucleotides-- Electrophoresis of T2 RNase digestion products was carried out on Whatman 3MM paper in a Savant voltage electrophoresis unit using 5% acetic acid, ammonium hydroxide buffer, pH 3.5). Chromatography on cellulose plates was performed in a solvent isobutyric acid/water/ammonium hydroxide (66:33:1, v/v/v) as described by Silberklang et al. (32). The radioactivity in each nucleotide was quantitated on a Betagen scanner.

Preparation of cDNA for 7SL and 7SK RNA-- The 3'-terminal sequences of 7SL and 7SK RNAs were determined by a novel T4 RNA ligase/PCR-based approach (33). HeLa cell 4-8S RNA was fractionated on a 10% polyacrylamide gel containing 7 M urea. 7SL RNA and 7SK RNAs were recovered and ligated to a 5'-phosphorylated deoxyoligonucleotide that was blocked at its 3'-end with cordecypin (oligo 1, 5'-pGATAGTGTCACCTAAATGAATTCC(3'-dA)-3') with T4 RNA ligase. The ligation product was then used as a template for synthesis of cDNA and subsequent PCR amplification using the Titan^TM One Tube reverse transcription-PCR System (Boehringer Mannheim). The two primers used for reverse transcription-PCR were an oligonucleotide complementary to oligo 1 (oligo 2, 5'-GGAATTCATTTAGGTGACACTATC-3') and an internal primer corresponding to positions 220-239 of 7SL RNA (oligo 3, 5'-ACTCCCGTGCTGATCAGTAG-3') or an internal primer (oligo 4, 5'-TGCTAGAACCTCCAAACAAGC-3') corresponding to positions 211-232 of human 7SK RNA. PCR amplification products were gel-purified, ligated into the TA cloning vector PCR^TM 2.1 (Invitrogen), and transformed into DH5 cells. Colonies were randomly picked and sequenced using oligo 3 or oligo 4 as primers.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Adenylic Acid Residue Is Added to the 3'-end of Several Small RNAs-- HeLa whole cell extract prepared by the method of Weil et al. (26) was incubated with [-³²P]ATP, and the RNAs were fractionated on a polyacrylamide gel. Since tRNAs are known to undergo terminal CCA turnover (1), the tRNAs were the predominant labeled RNAs (Fig. 1, lane 1). Instead of whole cell extract, HeLa nuclear extract prepared by the method of Dignam et al. (25) was used, and when [-³²P]ATP was used as the precursor in addition to tRNA, several RNAs were labeled; two of these labeled RNAs were prominent, and their mobility corresponded to 7SL RNA (Fig. 1, lane 5). There was no detectable labeling in 7SL RNA when [-³²P]UTP, [-³²P]GTP, or [-³²P]CTP was used as the labeled precursor (Fig. 1, lanes 6-8).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Adenylation of 7SL RNA in vitro. A, HeLa whole cell extract prepared according to Weil et al. (26) was incubated (lanes 1-4) in the presence of the [-32P]-ribonucleoside triphosphates indicated above the lanes. B, HeLa cell nuclear extract prepared according to Dignam et al. (25) was used (lanes 5-8). The labeling conditions and other details are as described under "Materials and Methods" section. After incubation, the labeled RNAs were extracted, purified, and fractionated on a 12% polyacrylamide, 7 M urea gel and subjected to autoradiography. C, hybrid selection of the labeled RNAs by plasmid DNAs containing sequences corresponding to human 7SK, 7SL, RNase P, and mitochondrial RNA processing (MRP) RNA (lanes 10-13). The labeled RNAs obtained with nuclear extracts in the presence of [-32P]ATP (see lane 5) were used for hybrid selection. The labeled RNAs hybridized to the DNAs were eluted, purified, and fractionated on a 12% polyacrylamide, 7 M urea gel and subjected to autoradiography.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   A, time course of adenylation of 7SL RNAs and tRNAs. Top panel, HeLa nuclear extract was labeled in the presence of [-32P]ATP for differing periods of time. The numbers above the figure refer to the period of incubation. The reaction conditions are given under "Materials and Methods." Bottom panel, quantitation of radioactivity in 7SL RNA and tRNAs. The radioactivity in the RNAs was quantitated using a Betagen scanner. B, time course of uridylation of U6 snRNA. Top panel, the HeLa nuclear extract was incubated as described under "Materials and Methods" in the presence of [-32P]UTP. At the end of incubation period, the RNAs were isolated, fractionated on a polyacrylamide gel, and subjected to autoradiography. Bottom panel, quantitation of radioactivity in U6 snRNA; the radioactivity in the RNA was quantitated using a Betagen scanner. C, turnover of adenylated 7SL RNA and tRNAs in vitro. Top panel, the HeLa nuclear extract was labeled or 2 h, and a 200-fold excess of unlabeled ATP was added. The concentration of ATP was 10 µM before unlabeled ATP was added. The incubation was continued for different periods of time, indicated above the lanes. The RNAs were isolated, fractionated on a polyacrylamide gel, and subjected to autoradiography. Bottom panel, quantitation of radioactivity in 7SL RNA and tRNAs. The radioactivity in the RNAs was quantitated using a Betagen scanner. D, effect of preincubation with and without unlabeled ATP. Top panel, the nuclear extract was incubated in the absence or presence of 10 µM ATP for 2 h at 30 °C, and then 30 µCi of [-32P]ATP was added and incubated for various time periods indicated above the lanes. For samples preincubated without unlabeled ATP, 10 µM unlabeled ATP was added along with labeled ATP. At the end of incubation period, the RNAs were isolated, fractionated on a polyacrylamide gel, and subjected to autoradiography. Bottom panel, quantitation of radioactivity in 7SL RNA; the radioactivity in the RNAs was quantitated using a Betagen scanner.

Kinetics of Adenylation of 7SL RNA-- The time course of adenylation of tRNAs and 7SL RNA was studied. The labeling of transfer RNA reached a plateau within 1 h of incubation and remained nearly constant for up to 10 h (Fig. 2A). In contrast, the adenylation of 7SL RNA was continuing even after 10 h of incubation, the longest time period tested. In addition, there appears to be a lag period of 2 h before significant labeling in 7SL RNA was detectable (Fig. 2A, bottom panel). These data indicate that the kinetics of adenylation of tRNAs and 7SL RNA are distinct. In addition to the adenylation of 7SL RNA, there was also labeling of 5S RNA (Fig. 2A), and the kinetics of labeling of 5S RNA was similar to that observed for 7SL RNA. However, the kinetics of 5.8S RNA labeling was different. There was rapid labeling within 1 h, and there was no radioactivity detectable after 3 h of incubation (Fig. 2A, top panel).

A Single Adenylic Acid Residue Is Added to the 3'-end of 7SL RNA-- To determine the number of adenylic acid residues added to the 3'-end of 7SL RNA, the labeled 7SL RNAs (see Fig. 1, lane 5) were isolated and digested with various enzymes, and the products were analyzed. The human 7SL RNA from HeLa cells fractionated into two distinct bands, designated 7SL-1 and 7SL-2 RNAs. The 3'-end sequences of both 7SL-1 and 7SL-2 RNAs are identical (34, 35), and therefore, in many instances data obtained with only 7SL-1 RNA are presented; identical results were obtained when 7SL-2 RNA was analyzed.

View larger version (48K): [in this window] [in a new window]   Fig. 3.   Determination of 3'-end nucleotides of 7SL RNA adenylated in vitro. The tRNA and 7SL RNAs (see lane 5 of Fig. 1) were isolated, purified, digested with nuclease P1 or T2 RNase, and subjected to chromatography on a cellulose plate. The cellulose plate was dried and subjected to autoradiography. The position of the unlabeled mononucleotides used as standards are indicated by broken circles.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Analysis of 3'-end sequence heterogeneity of 7SL and 7SK RNAs. A, 7SL RNAs labeled in vitro using HeLa nuclear extract were isolated, purified, digested with T1 RNase, and separated on a 20% polyacrylamide gel containing 7 M urea (lanes 2 and 3). M indicates oligonucleotide size markers. The hexanucleotide UCUCUA-OH migrated slower than the octanucleotide marker because the markers contained 3' phosphate. B, small RNAs from HeLa cells were ligated to [5'-32P]pCp using T4 RNA ligase and purified by hybrid selection and fractionation on a polyacrylamide gel. The RNAs were digested with T1 RNase, separated on 20% polyacrylamide gel, and then subjected to autoradiography. C, the 7SL RNA fragments A and B from lane 5 and 7SK RNA fragments C, D, and E from lane 7 were isolated and digested with T2 RNase. The digestion products were separated by electrophoresis on Whatman 3MM paper and then subjected to autoradiography. The radioactivity in the RNAs was quantitated using a Betagen scanner.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Identification of post-transcriptionally added A residue in 7SL and 7SK cDNA preparations. The 7SL (Fig. 5A) and 7SK cDNA (Fig. 5B) clones prepared as described under "Materials and Methods" were subjected to sequencing by the method of Sanger et al. (49) and subjected to autoradiography. The post-transcriptionally added adenylic acid is shown with an asterisk.

Adenylation on the 3'-end of 7SL RNA Occurs in Vivo-- The small RNAs isolated from HeLa cells were ligated to [5'-³²P]pCp using RNA ligase, and the 7SL RNAs were hybrid-selected and then purified by fractionation on a polyacrylamide gel. The 3'-end-labeled 7SL-1 and 7SL-2 RNAs were isolated, digested with T1 RNase, and fractionated on a 20% polyacrylamide gel. Two major bands were obtained in the case of 7SL-1 RNA (Fig. 4, lane 5) and 7SL-2 RNA (Fig. 4, lane 6). These two fragments, designated A and B, were isolated, and the 3'-end nucleotide was determined by digesting these fragments with T2 RNase and analyzing the products. Fragment A yielded only Up (Fig. 4, lane 8), indicating that this fragment corresponds to UCUCU-OH in the 7SL RNA. Fragment B yielded mostly Ap (88%) and some Up (12%), (Fig. 4, lane 9), showing that these fragments correspond to UCUCUA-OH (88%) and UCUCUU-OH (12%) of 7SL RNA. These data also show that adenylic acid residues are present on the 3'-end of 7SL RNA isolated from cells, and the adenylation observed in vitro does reflect adenylation that is occurring in HeLa cells. These results are also consistent with our earlier observation (36) where rat 7SL RNA was shown to contain both A-OH and U-OH on its 3'-end.

Identification of Post-transcriptionally Added Adenylic Acid in cDNA Preparations-- In addition to ligation of pCp to determine the 3'-end sequence, a deoxyoligonucleotide with a 5'-phosphate was ligated to 7SL RNA, and cDNA clones were prepared as described under "Materials and Methods." Eighteen independent cDNA clones were sequenced, and 10 out of 18 clones contained adenylic acid corresponding to position 300 of 7SL RNA; five of the clones contained U at this position. Two representative sequences corresponding to A at position 300 (Fig. 5A, left panel) and U at position 300 (Fig. 5A, right panel) are shown. These data provide additional evidence for the presence of post-transcriptional addition of adenylic acid residue on the 3'-end of 7SL RNA in HeLa cells. A similar strategy was used to characterize cDNA preparations prepared from 7SK RNA, and clones containing U as well as A on the 3'-end were obtained. Two cDNA representative clones are shown in Fig. 5B. These data are consistent and support the notion that in many small RNAs, 3'-trimming of these RNAs is followed by adenylation.

Adenylic Acid Residues Are Also Present on the 3'-end of Other Small RNAs-- To investigate the possibility that addition of adenylic acid residue is a common phenomenon, several small RNAs were isolated, and their 3'-end nucleotides were determined. In every case that was analyzed, there was some proportion of each RNA that contained adenylic acid residue on the 3'-end (Fig. 6). In the case of U6 snRNA, it is clear that adenylic acid present on the 3'-end is due to the post-transcriptional addition, since there is no adenylic acid in the corresponding position in U6 snRNA gene. However, since the U1, U4, and U5 snRNA genes contain A in the corresponding position, one cannot conclude that post-transcriptional adenylation is occurring in these three snRNAs.

View larger version (51K): [in this window] [in a new window]   Fig. 6.   The 3'-end nucleotide analysis of other small RNAs. Small RNAs from HeLa cells were ligated to [5'-32P]pCp and hybrid-selected with corresponding DNAs immobilized on nitrocellulose (7SL and 5S RNA) or obtained by immunoprecipitation with anti-trimethylguanosine antibodies (U1, U2, U4, and U5 RNAs) or anti-methylphosphate cap antibodies (U6 and 7SK RNAs). The RNAs were further purified by electrophoresis on a polyacrylamide gel, isolated, and digested with T2 RNase. The resulting mononucleotides were fractionated by electrophoresis on a 3MM Whatman paper and subjected to autoradiography. The radioactivity in the RNAs was quantitated using a Betagen scanner.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The main observations made in this study are that human 7SL RNA as well as human 5S, U1, U2, U6, and 7SK RNAs have a post-transcriptionally added adenylic acid residue on their 3'-ends in a subpopulation of these RNAs. In addition, we developed a cell-free system where post-transcriptional adenylation is occurring accurately. Data obtained in this study indicate that adenylation may be a common event occurring in many cellular RNAs.

The adenylation of 7SL RNA observed in vitro was also found to be occurring in the HeLa cells. Adenylic acid residue was reported on the 3'-end of rat (36) and human 7SL RNA (37). Ullu and Weiner (38) characterized the human 7SL RNA and the corresponding genes and concluded that adenylic acid on the 3'-end is post-transcriptionally added. Since the 7SL RNA adenylated in vitro contained UCUCUA-OH (Fig. 4), which corresponds to the 3'-end sequence in 70% of the 7SL RNA molecules isolated from HeLa cells (Table I and 36-38), the adenylation occurring in the nuclear extracts faithfully reflects the in vivo situation. It is interesting to note that only a single adenylic acid residue is added to the 3'-end of 7SL RNA. Although 70% of the 7SL RNA 3'-ends contain A-OH, only 30% of the 7SL RNA molecules contain U-OH as the 3'-end. However, virtually all of the adenylation occurs on the U-OH containing 7SL RNAs. Only 3% of the molecules have two adenylic acid residues (see Fig. 3). Consistent with these data, all 10 3' A-containing clones (Fig. 5A) contained only one adenylic acid residue. Therefore, in logarithmically growing HeLa cells, in vitro adenylation of 7SL RNA is limited to the addition of one adenylic acid residue. The 7SL RNA molecules that are already adenylated are poor substrates for further adenylation. Analysis of the 3'-ends of 5S RNA, 7SK RNA, and U2 snRNA also showed that only one adenylic acid residue to be present on their 3'-ends. However, these data do not rule out a small percentage of 7SL, 7SK, 5S, or U2 snRNAs that contain multiple adenylic acid residues. O'Brien and Wolin (39) characterized 66 Xenopus 5S RNA clones and found that 13 of the sequenced clones contain a single adenylic acid residue on their 3'-end. These data show that only a single A residue is being added to the 3'-end of 5S RNA, and adenylation of small RNAs is also occurring in amphibians.

Although adenylation was found to be the major post-transcriptional addition found in these small RNAs, small RNAs also contained a small percentage of C and G on their 3'-ends (see Fig. 6). A fraction of the ribosomal 5S RNA stored in the oocyte was found to contain post-transcriptionally added C, G, A, or U on its 3'-end (15). These results indicate that 3'-ends of small RNAs are extended in many different ways.

The kinetics of adenylation of 7SL RNA were very different from the adenylation of tRNAs (Fig. 2). Although adenylation of tRNA reached steady state within 1 h, the adenylation of 7SL RNA continued even at 10 h, the longest time period tested. The kinetics of adenylation of 5S RNA was very similar to that of 7SL RNA (Fig. 2) in that there was continued labeling at 10 h. In addition, the turnover of terminal adenylic acid in 7SL RNA was very slow when compared with the turnover of tRNA. The slow labeling and slow turnover of 7SL RNA indicates that the enzyme(s) involved in the adenylation of 7SL and other RNAs may be a complex, and transfer of the labeled ATP to these enzyme(s)/cofactors may be the rate-limiting step in this adenylation reaction.

Most of the RNAs synthesized by RNA polymerase III, including 7SL, 7SK, 5S, and U6 RNAs, terminate with the UUUU-OH sequence on their 3'-ends. Four small RNAs investigated in this study, namely human 7SL RNA, 7SK RNA, 5S RNA, and U6 snRNA, are synthesized by RNA polymerase III and terminated with UUUU-OH. Therefore, identification of adenylic acid residues corresponding to these uridylic acid residues provides evidence for deletion of uridylic acid residues and for post-transcriptional adenylation. The observation that U2 snRNA synthesized by RNA polymerase II also undergoes adenylation indicates that this adenylation reaction is not confined to RNAs synthesized by one type of RNA polymerase.

It is surprising that in some small RNAs like ribosomal 5S and U6 snRNA, both uridylation and adenylation are occurring on the 3'-end. It is possible that uridylation and adenylation are two competing events, and the variability in the percentage of adenylation on the 3'-end of small RNAs may reflect the product of these two competing events (see Table I). In the case of 5S and U6 snRNA, the uridylation seems to be a dominant event resulting in >80% of uridylic acid on their 3'-ends. In contrast, adenylation appears to be a dominant event in the case of U2, 7SL, and 7SK RNAs, resulting in about 70% adenylic acid on their 3'-ends.

Where does adenylation of RNAs occur in the cells? We do not have data to conclude that adenylation of small RNAs is occurring in any one subcellular compartment. The fact that U6 snRNA, which does not leave the nucleus (40) does contain post-transcriptionally added adenylic acid residues, indicates that the necessary enzymatic machinery must be present in the nucleus. In addition, nuclear extracts adenylated better than whole cell extracts (Fig. 1), suggesting that enzyme/substrate combination is more optimal for adenylation in the nuclear extract. In addition, results from Dahlberg and co-workers (41) previously showed that deletion of the terminal nucleotides in U1 snRNA occurred in the nucleus after its return from the cytoplasm. If U2 snRNA follows the same maturation pathway, it is likely that the adenylation of U2 snRNA is occurring in the nucleus. Further studies are needed to determine where in the cell and at what stage in the maturation of RNAs adenylation of small RNAs occurs.

Finally, the most important question is what is the function of this adenylation? It is unlikely that cells will be carrying out this adenylation reaction unless there is a need for this adenylation. The fact that this adenylation reaction is occurring in RNAs of diverse origin, subcellular localization, and functions indicates that this adenylation may have function(s) related to synthesis and/or turnover of cellular RNAs. In this context it is interesting to note that Piper et al. (42) and Kempers-Veenstra et al. (43) observed that yeast cells deficient in an exonuclease accumulate 5S RNA molecules that have multiple adenylic acid residues on the 3'-end. Xu and Cohen (44) provided evidence for adenylation-mediated degradation of RNAs in bacteria. Following this analogy, it is possible that human small RNAs are also degraded through an intermediate where small RNAs are first polyadenylated.

The results on RNA degradation in the yeast system showed that deadenylation of poly(A) tail is a requirement. Both stable and unstable yeast mRNAs were shown to follow a decay pathway in which deadenylation leads to either internal cleavage or decapping followed by 5'  3' exonucleolytic degradation of the mRNA (45). There are no data on the post-transcriptional adenylation of corresponding small RNAs in the yeast cells. The small RNAs, such as human U2 snRNA, with one adenylic acid residue on the 3'-end, are known to be very stable in the cell (46). Therefore, it is unlikely that the adenylation characterized in this study makes these human small RNAs less stable. However, it is possible that these small RNAs containing a single adenylic acid residue on their 3'-ends are polyadenylated and then targeted for degradation. If true, small RNAs with longer poly(A) tail on the 3'-end should be less stable. This possibility needs to be tested experimentally. In this context, it is interesting to note that two distinct populations of RNAs exist for both mouse B2 RNA and human telomerase RNA; one population contains a poly(A) tail, and another population contains no poly(A) tail on the 3'-end (47, 48). These data show that some small RNAs go through a polyadenylation phase during their metabolic cycle.