View all hidden authors and organizations

Contributor: Joseph G. Gall, September 10, 2021 (submitted for review on August 12, 2021; reviewed by Brenda L. Bass and Anita K. Hopper)

↵ 1G.JST and SD have made equal contributions to this work.

Small nucleolar (sno) RNA usually guides the modification of ribosomal RNA and small nuclear RNA, which is an important event in the biogenesis and function of ribosomes and spliceosomes. Most are processed in the nucleus from the lasso intron RNA, which is an unstable by-product of splicing. We report here that some snoRNAs are encoded in an abnormally stable lasso. These stable lassos with snoRNA (or slb-snoRNA) can be found in the nucleus and cytoplasm associated with snoRNA binding proteins. They do not function as modified guide RNAs, and their output competes with the maturation of snoRNA. Therefore, slb-snoRNAs provide a new level of regulation of snoRNAs and snoRNA binding proteins.

Small nucleolar (sno) RNA directs post-transcriptional modifications critical to the biogenesis and function of its target. Most snoRNAs in higher eukaryotes are encoded in introns. They are first released from the nascent transcript in the form of a lasso, and are quickly targeted by debranching enzymes and exnucleases for linearization and further trimming. In this study, we reported that some snoRNAs are encoded in abnormally stable intronic RNAs. These intron sequences can escape the branching enzyme and accumulate as a lasso. A stable lasso with snoRNA or slb-snoRNA is related to snoRNA binding protein, but does not guide post-transcriptional modification. Although most slb-snoRNA accumulate in the nucleus, some can be exported to the cytoplasm. We found that this output competes with the maturation of snoRNA. Slb-snoRNA provides a previously unknown regulatory layer for snoRNA and snoRNA binding proteins.

Small nucleolar RNA (snoRNA) represents a very abundant type of nuclear RNA. Most snoRNAs direct the site-specific uridine isomerization (H/ACA box snoRNA) and 2'-O-methylation of ribosomal RNA (rRNA), small nuclear RNA (snRNA), and transfer RNA (tRNA) C/D box snoRNA), and possibly other RNAs (1, 2). In vertebrates, almost all snoRNAs are encoded by introns. SnoRNAs are released in the form of a lasso and must be linearized by the debranching enzyme DBR1, and then further trimmed to their mature size, ~60 to 300 nt by ribonuclease. In addition, the lasso can be linearized by endonucleases (such as RNT1 in yeast) to promote maturation (3 ⇓ ⇓ –6).

In a recent study, we analyzed RNA extracted from the oocyte nucleus or germinal vesicle (GV) of the tropical Xenopus laevis. This analysis allowed us to classify more than 400 snoRNAs in this species (7). When we checked the intron readings in the Integrative Genomics Viewer (IGV) browser, we noticed that many Xenopus snoRNAs overlap with other non-coding (nc) RNAs, that is, stable intron sequence (sis) RNAs.

sisRNA refers to a type of intron-derived RNA first described by our laboratory (8). sisRNA lasts for several hours in cultured cell lines and even several days in frog oocytes (8, 9). In many species, including frogs, humans, mice, and chickens, the most abundant sisRNA exists in the form of a lasso (10). Unlike "classic" intron RNA, sisRNA avoids linearization of DBR1 and degradation by nucleases. These lassos are unusually stable, partly due to their non-canonical C branch point (10), a poor substrate for DBR1, which is most effective against A branch RNA (11). In addition, the lasso can be exported into the cytoplasm, which largely depletes DBR1 (10).

Some reports indicate that a stable nuclear lasso plays a role in the regulation of cis or trans transcription; these sisRNAs tend to accumulate at the transcription site (9, 12 ⇓ –14). Alternatively, circular sisRNA can be recruited to nucleosomes and chelating proteins, such as Dicer (15). It is speculated that viral sisRNA can also regulate host proteins through isolation (16, 17). The function of cytoplasmic sisRNA is still unknown.

In this study, we reported that hundreds of snoRNA-encoding introns produced circular sisRNA in different vertebrates. We call this type of sisRNA a stable lasso with snoRNA or slb-snoRNA. slb-snoRNA does not function as a modified guide RNA, but they are related to the canonical modified snoRNP protein (for example, the snoRNA binding protein pseudouridine synthase stain protein). A part of slb-snoRNA can be exported to the cytoplasm, thereby preventing the maturation and accumulation of snoRNA in the nucleus. We discussed the possible regulatory role of slb-snoRNA in snoRNA biogenesis and segregation of snoRNA binding proteins.

In a previous study, we used the GV transcriptome to annotate almost all snoRNAs encoded in the genomes of tropical pathogens (7). When we checked the sequence alignment of the GV transcriptome in the IGV browser and searched for snoRNA reads (Figure 1A), we noticed that the reads almost covered some snoRNA-encoding introns (306 out of all 405 introns snoRNA). ) The entire length (data set S1). In each case, no reads are mapped to the exon-intron junction. In fact, the coverage of these readings starts near the 5'splice site and ends a little upstream of the 3'splice site (Figure 1A, red overlay). This pattern indicates that the oocyte accumulates a stable partially processed intronic snoRNA in the nucleus.

Typical slb-snoRNA in frog oocytes. (A) IGV browser view of the X.tropicalis eif4a1 locus. In the nucleus, mature snoRA48 and snoRD10 (green) and slb-snoRA48 and slb-snoRD10 (red) were detected. After RNase R treatment, only slb-snoRNA was detected. The bottom track shows a "reverse" reading, indicating that the RNA is circular. (B) IGV browser view of the X.tropicalis gln3 locus. Slb-snoRD69 (red) is easily detected in the control and RNase R-treated samples. The bottom track shows reverse reading. (C) Northern blot analysis of snoRD69 in liver (5 µg) and oocyte RNA (500 GV) samples of Decachroma tropicalis. Two bands were detected in RNA from oocytes: fully processed snoRD69 (lower band) and slb-snoRD69 (higher band). Note that the migration speed of the higher bands is much slower than expected for the corresponding full-length linear intron RNA.

Because the most abundant sisRNA accumulates as a lasso (18), we want to know whether these partially processed snoRNAs also accumulate as a lasso. We analyzed the circular transcriptome dataset generated by processing nuclear RNA with RNase R, a 3'-5' exonuclease that can degrade linear RNA and keep circular RNA intact (18). As expected, we can no longer detect mature linear snoRNAs in this dataset. However, we detected reads mapped to the entire intron, which is consistent with the detection of circular intron RNA (Figure 1A, red overlay). In addition, we detected many reads, which can only be mapped to the linear genome when they are segmented and rearranged (SI appendix, Figure S1A). These so-called "reverse" readings correspond to the junction of circular RNA (Figure 1A, black overlay). Therefore, these data indicate that a small portion of some snoRNA can be stabilized in the form of a lasso.

In order to confirm the existence of a stable lasso encoding snoRNA, we analyzed the GV RNA of tropical decacolor by Northern blotting. Using a probe specific to snoRD69 (Figure 1B), we detected a band of about 70 nt in the liver and oocyte RNA, corresponding to the mature form. Only higher bands can be easily detected in the RNA lanes of oocytes (Figure 1C). This higher molecular weight band runs slower than the expected linear intron RNA, consistent with it being a round molecule. When we used probes against mature snoRA48, we observed a similar pattern, although the lasso was detected at lower levels than mature snoRNA (SI appendix, Figure S2A). We concluded that in frog oocytes, in addition to the mature linear form, a small fraction of certain snoRNAs can also accumulate in the form of a lasso. We call these circular transcripts a stable lasso with snoRNA or slb-snoRNA.

Slb-snoRNAs tend to be more abundant than other nuclear circular sisRNAs (SI Appendix, Figure S3). Nevertheless, some abundant lassos do not encode known snoRNAs. This observation raises the question of whether these abundant sisRNAs exhibit similarities to slb-snoRNA. SnoRNA forms a stable stereotyped structure and contains basic short sequences called "boxes": C/C' (RUGAUGA) and D/D' (CUGA) in box C/D snoRNA, or H in box C/D snoRNA (AANNA) and ACA (AYA) box H/ACA snoRNA. We use the prediction software snoReport 2.0 (19) to search for such motifs among the abundant circular sisRNA sequences (mapping reads per million [FPKM] ≥ 100, ~1,300 introns in the RNase R experiment). We found snoRNA motifs (score ≥ 0.9) in 75 sisRNAs, including 33 of 63 known snoRNAs that overlap with rich lasso. All in all, 8% (105) of abundant lasso contain snoRNA or snoRNA motifs (data set S1). About 90% of these rich lassos are shorter than 500 nt. Therefore, we investigated annotated introns that are 100 to 500 nt in length and do not encode a stable lasso (~7,500 introns); only about 2% of these introns contain snoRNA motifs. In summary, we show that frog oocytes accumulate hundreds of stable lassos with snoRNA or snoRNA-like sequences. In addition, compared with the unstable lasso of the same size, the snoRNA motifs in the stable lasso are more abundant.

Is slb-snoRNA restricted to Xenopus oocytes? Previously, we performed RNA high-throughput sequencing studies on circular RNA purified from cells treated with the transcription inhibitor α-manitin, and identified hundreds of cells in human, mouse, chicken, and frog cell lines. A stable lasso. This analysis resulted in the annotation of hundreds of sisRNAs (10). We re-examined these sisRNAs and found that 8 encode snoRNA in human HeLa cells, 6 in mouse 3T3 cells, 3 in chicken DF1 cells, and 11 in frog Xenopus XTC cells. There is almost no overlap between snoRNA orthologs (Figure 2A and data set S1). In addition, in all four analyzed species, the snoReport 2.0 algorithm predicted snoRNA motifs in 391 sisRNAs. Of the 391, 24 corresponded to the snoRNA annotated above, 188 showed high predictability scores (≥0.9), of which 14 had perfect scores (=1) (Figure 2B and data set S1).

Identification of slb-snoRNA in cell lines from different vertebrate species. (A) IGV browser view of a typical gene encoding a circular sisRNA that overlaps with snoRNA from humans, mice, chickens, and frogs. Unlike exon reads (yellow), intron reads were detected in RNA samples treated with RNase R, displayed as purple (human), red (mouse), orange (chicken), and green (frog). Black is the coverage of the intron inverted reading. These unconventional readings indicate that the intron transcript is round. (B) RT-PCR analysis of slb-snoRNA in four research species. Use Episcript-RT or AMV-RT for reverse transcription. Intron cDNA was amplified by PCR using inward and outward primers; the latter only detects circular molecules (see schematic diagram in the SI appendix, Figure S1B). The predicted snoRNA motifs of the tested slb-snoRD-like transcripts are shown below the corresponding gel image. The end stem (red) and the predicted C and D boxes (black boxes) are highlighted.

From the RNA-sequencing (RNA-seq) data generated by our previous samples processed for RNase R, there are three pieces of evidence that indicate the cyclic nature of these slb-snoRNAs (10). First, we observed that sisRNA with snoRNA persisted after RNase R treatment (Figure 2A). Second, intron reads are not limited to snoRNA sequences; instead, they extend from the 5'end to the branch point. Finally, for some sisRNAs with snoRNA, we detected unconventional reads that contained the branch point region connected to the 5'end of the intron. These readings correspond to the connection of the lasso (Figure 2A, reverse reading track).

Importantly, we confirmed by RT-PCR that sisRNA with circular snoRNA is indeed a lasso. We use Episcript RT (a reverse transcriptase that is active across branch points) or AMV-RT (a reverse transcriptase that terminates at branch points) to generate cDNA from RNA extracted from cells treated with α-amanita (18). Intron cDNA generated using any enzyme can be amplified using inward-facing primers. However, for the outward primers, only the cDNA generated with Episcript RT was amplified (Figure 2B and SI appendix, Figure S1B). The specificity of the PCR product was confirmed by sequencing. Therefore, we demonstrated that many snoRNA and snoRNA-like sequences are stabilized in the form of a lasso in cell lines from various vertebrate species.

Can slb-snoRNAs guide 2'-O-methylation and pseudouridine? In yeast, the dbr1Δ strain is a mutant of the debranching enzyme DBR1 that can survive and accumulate lasso (3). Although most yeast snoRNAs are expressed by independent mono- or polycistronic genes, eight are encoded in introns. In dbr1Δ, these intronic guide RNAs cannot be processed correctly, so they mainly accumulate in the form of lasso and partially processed linear transcripts (4). It is worth noting that even if mature snoRNA molecules are almost undetectable, as in the case of snR24, the positions targeted by the guide RNA (C1437, C1449 and C1450 of 25S rRNA) are still correctly modified (4). Although the author of the earlier study thoughtfully pointed out that "the unprocessed intron U24 snoRNA seems to play a role in 2'-O-methylation", it is tempting to "implement that the former U24 in the form of a lasso is functional" Powerful (4).

When we were unable to detect the modification mediated by snoRA29 in human 18S and 28S rRNA, we first began to question the possibility of the function of snoRNA in the form of a lasso (7). SnoRA29 is highly conserved among vertebrate species, but in humans it only accumulates in a partially processed form, most likely a lasso (Figure 3A). In fact, after RNase R treatment, high-molecular-weight bands are still present in human RNA samples, which indicates its cyclic nature. One might argue that the accumulation level of human slb-snoRA29 is very low, which may not be sufficient for modification activities. In addition, the 3'end of the human snoRA29 sequence has a point mutation in the canonical ACA box, which may affect the function of the guide RNA (Figure 3B).

Test snoRA29 modification activity in mouse and human cells. (A) Northern blot analysis of snoRA29 in mouse (3T3) and human (HeLa) RNA samples treated with RNase R. Only fully processed snoRA29 was detected in the control mouse RNA. In human RNA, the snoRA29 sequence is only present in the high molecular weight band; this band is resistant to RNase R, indicating its circular nature. No snoRA29 band persisted in mouse RNA treated with RNase R, indicating that snoRA29 and higher molecular weight bands are linear molecules in mice; the latter represents partially processed mouse snoRA29 encoding introns. 5S rRNA was used as a control for loading and RNase R treatment efficiency. (B) Hypothetical base pairing of snoRA29 with 18S and 28S rRNA. (C) Diagram of the construct showing ectopic snoRA29 expression in HeLa cells. (D) Northern blot analysis of human and mouse snoRA29 expression in HeLa cells transfected with the construct described in C. The amount of total RNA loaded on the gel and the relative expression levels of ectopic slb-snoRA29 and mature snoRA29 are shown at the bottom. (E) Draw pseudouridine from the 5'end region of 28S rRNA from mouse 3T3 cells and human HeLa cells, and control and transfect with the 5 snoRA29 expression constructs shown in C and D. Mouse 28S rRNA is usually pseudouridineylated at position 45 (red trace at the top). This modification does not exist in the human 28S (black trace at the bottom). Expression of fully processed snoRA29, whether it is mouse (construct #5, gray trace) or human (construct #4, green trace), induces 28S-Ψ45 (stars) in HeLa cells. The expression of snoRA29 in the form of a lasso (construct #1, blue locus and construct #2, dark green locus) cannot induce pseudouridine of U45 in human 28S rRNA. To make all traces comparable, the y-axis of each sample is scaled relative to the total arbitrary fluorescence intensity of all detected peaks in the sample, which corresponds to the number of RNA molecules analyzed.

We made various constructs to express slb-snoRA29 ectopic or fully processed snoRA29, with or without restoring the ACA box (Figure 3C). Only fully treated snoRA29, whether it is a human with a recovered ACA box or a mouse as a positive control (Figure 3D), can induce 28S rRNA at position 45 (Figure 3E) and 18S rRNA at position 220 (SI Appendix, Figure 3D). 3) Pseudouridine. S4) In HeLa cells. Importantly, the expression level of mature human snoRA29 is very low, but it can mediate rRNA modification. At the same time, regardless of the point mutation in the ACA box (Figure 3D), the double overexpression of human slb-snoRA29 is not sufficient to modify 28S-45 and 18S-220 (Figure 3E and SI appendix, Figure 3). S4).

X. laevis snoRA75 provides another hint that slb-snoRNA may not function as a modified guide RNA. We previously found snoRA75 in two frogs of the same genus, namely tropical horned toad and clawed frog (7). The orthologs have almost the same structure (Figure 4A and B), but the mature snoRA75 is expressed in tropical hydatid, but not in the spiny spider (Figure 4 C and D). Therefore, X. laevis 18S rRNA lacks U93 pseudouridine (Figure 4E). In oocytes, we detected snoRA75 expression in Staphylococcus tropicalis and toads by RNA-seq and Northern blotting. However, X. laevis snoRA75 is only expressed as a lasso (Figure 4 C and D). Although slb-snoRA75 is very abundant and concentrated in GV where RNA modification usually occurs, 18S-U93 is not modified in X. laevis oocytes (Figure 4E). These results indicate that slb-snoRNA and slb-snoRNA-like RNA do not support post-transcriptional modification.

X.tropicalis and X. laevis snoRA75 and slb-snoRA75. (A) Alignment of two snoRA75 orthologs. The antisense element used to locate 18S-Ψ93 is highlighted in blue; the H and ACA boxes are framed by black lines. (B) Putative base pairing between snoRA75 and 18S rRNA. (C) IGV browser view of snoRA75 and slb-snoRA75 in oocytes and somatic tissues of tropical diseases and toads. (D) Northern blot analysis of somatic cells, XTC cell lines and liver (10 µg total RNA per lane) and oocytes from toads (300 GV) and liver (7 µg total RNA) and snoRA75 in oocytes ( 100 GVs) from Pseudomonas tropicalis. The fully processed snoRA75 was only detected in tropical pathogen liver and oocyte RNA samples. High-molecular-weight bands were detected in the oocytes of both frogs; the migration speed of these bands was much slower than expected for linear full-length intronic RNA molecules, which is characteristic of the intron lasso. (E) Draw pseudouridine on the 5'end region of Xenopus 18S rRNA. The 18S rRNA from the liver (green trace) and oocytes (red trace) of Decachroma tropicalis is pseudouridineized at positions 34 and 93 (star). X. laevis 18S rRNA is modified at position 34 instead of position 93 (arrow) in any tissue, including oocytes (pink trace), liver (gray trace) and cultured cell line XTC (brown trace at bottom) String). The trajectory is scaled as shown in Figure 3E.

To further explore the modification activity of snoRNA in the form of a lasso, we returned to the yeast cell system. Since part of the snoRNA processing in dbr1Δ mutant strains may be driven by RNT1, which is an endonuclease involved in snoRNA processing (6), we naively hope to use the double mutant dbr1Δrnt1Δ to distinguish lasso and partially processed snoRNA activity. However, in the dbr1Δrnt1Δ mutant, all 8 intronic snoRNAs still accumulate in the lasso, in partially processed and mature forms, although the latter two are at low levels (SI appendix, Figure S5A), and normal rRNA modifications are detected Mode (SI Appendix, Figure S5B). It is worth noting that in dbr1Δrnt1Δ (SI Appendix, Figure S5C), almost no independently transcribed snoRNAs were detected, but we still detected modifications at their target locations (SI Appendix, Figure S5B). Yeast's endogenous snoRNA processing mechanism seems very powerful. It has multiple backup mechanisms to ensure proper modification of functionally important RNA molecules.

We propose that, unlike endogenous yeast intron snoRNA, vertebrate slb-snoRNA may be resistant to alternative processing, and in dbr1Δ and dbr1Δrnt1Δ yeast mutant strains, exogenous slb-snoRNA will exclusively accumulate in the form of a lasso. We chose two relatively abundant slb-snoRNAs of Staphylococcus tropicalis: slb-snoRA28 and slb-snoRD41. Advantageously, the unmodified position (18S-U808) of vertebrate snoRA28 in yeast mediates pseudouridine of 18S rRNA. In addition, we have previously shown that this modification can be induced in yeast by expressing the corresponding guide RNA (7). Unfortunately, the target position of snoRD41 was modified in yeast (25S-Um2729). In order to minimize the changes in the already sensitive mutant strains, we chose to replace the antisense element in snoRD41 to target U2 snRNA at the C41 position. Yeast U2-C41 is usually not modified, but this modification is easily induced (20).

We inserted the Xenopus slb-snoRNA into the yeast EFB1 gene to replace the intron encoding snR18, but retained the typical yeast splicing motif (Figure 5A). As a control, we constructed a construct expressing snoRA28 and snoRD41 as independent genes (Figure 5A). When we expressed snoRA28 from independent gene constructs in the dbr1Δ strain, low levels of mature snoRA28 were detected by Northern blot (Figure 5B) and 18S-U808 was modified in these cells (Figure 5C, green trace). Similar to endogenous yeast snoRNA, the minimum amount of fully processed snoRA28 is sufficient for modification. At the same time, when we expressed the intron snoRA28 in dbr1Δ, the mature snoRA28 failed to accumulate. Instead, we detected slb-snoRA28 and linearized slb-snoRA28 (Figure 5B). Even though lasso and "partially processed" linear transcripts are abundant, they cannot induce pseudouridineization of 18S-U808 (Figure 5C, red trace). These results indicate that neither partially processed box H/ACA snoRNA nor its lasso can mediate RNA modification.

Test slb-snoRNA modification guidance activity in a yeast cell system. (A) Schematic diagram of DNA constructs used to express fully processed linear snoRNA (independent) and lasso-form snoRNA (intron) in yeast. A28 is tropical decacolor snoRA28; c-D41 is a chimeric snoRD41 with U2-C41 antisense element, not 28S-U4276 (yeast 25S-U2729). The RNT1 cleavage site and snR13 terminator are described. The intron construct is made by replacing the intron in the yeast EFB1 gene. The specific intron regions of Xenopus are shown in red. (B) Northern blot analysis of xtsnoRA28 expressed from independent and intronic constructs in the yeast dbr1Δ strain. The mature snoRA28 is expressed only from independent constructs. Expression from the intron construct yielded snoRA28 in the form of a lasso and a partially processed linear lasso. (C) Plot the pseudouridine in the 18S rRNA from the dbr1Δ yeast strain. Usually, pseudouridine is detected at positions 759 and 766 in yeast 18S (black trace at the bottom). When the fully processed xtsnoRA28 was expressed from the independent construct (green trace), additional pseudouridine was detected at position 808 (asterisk), but when xtsnoRA28 was lassoed from the intronic construct (red trace) It was not detected in the form of expression. (D) Northern blot analysis of xtsnoRD41 expression of independent and intron constructs in yeast dbr1Δ and dbr1Δrnt1Δ strains. The mature snoRD41 is expressed only from independent constructs. The expression of the intron construct produces a lasso and its linearized form in both mutants (which may be destroyed during the RNA preparation process). Note that in addition to the lasso, partially processed transcripts can also be detected in the dbr1Δ single mutant. (E) Analysis of 2'-O-methylation in yeast U2 snRNA. Generally, yeast snRNAs do not have 2'-O-methylated residues (black trace at the bottom). The expression of fully processed (top green trace) or partially processed snoRD41 (red trace) induces 2'-O-methylation of C41 in yeast U2 snRNA (star). In the dbr1Δrnt1Δ double mutant, snoRD41 is only expressed as a lasso from the intron construct and cannot induce 2'-O-methylation of U2 snRNA (blue trace).

Interestingly, when we transformed the dbr1Δ mutant with the snoRD41 construct, U2-C41 was modified by independent transcription and intron chimeric guide RNA (Figure 5E, green and red traces, respectively), even though the intron was constructed In the case of the body, the mature snoRD41 is not obvious (Figure 5D). When snoRD41 was expressed from an intron construct, we detected a partially processed form of slb-snoRD41 and snoRD41; the guide RNA activity may come from one or two forms of this box C/D snoRNA. We next transformed the dbr1Δrnt1Δ double mutant. In this mutant, the intron chimeric snoRNA mainly accumulates in the form of a lasso (Figure 5D), while U2-C41 is not modified (Figure 5E, blue trace). Because the level of slb-snoRD41 in the dbr1Δrnt1Δ strain was three to five times higher than the level of linear snoRD41 expressed as an independent transcript (Figure 5D), we concluded that the lasso form of snoRNA cannot function as a modified guide RNA. Based on all these experiments, we ruled out the possibility that slb-snoRNA can guide post-transcriptional modification.

In order to act as a modified guide RNA, snoRNA forms a snoRNA-protein complex (snoRNP) with four different proteins, including a catalytic enzyme: pseudouridine synthase in H/ACA box snoRNP, chromophorin or C/D box snoRNP The methyltransferase in fibrillin. snoRNP assembly usually occurs in a co-transcriptional manner. Since we found that the lasso form of snoRNA cannot guide post-transcriptional modification, we want to know whether the classic snoRNP protein binds slb-snoRNA stably. Initially, the human snoRA29 sequence (Figure 3A), which only exists in the form of a lasso, was identified in the RNA fraction co-precipitated with the GAR1 protein, which is one of the four core proteins in box H/ACA snoRNP (21). This finding indicates that snoRNP protein can bind to slb-snoRNA. We decided to focus on dyskerin, which is the only snoRNA binding protein with catalytic activity and RNA binding domain, and test whether it is related to slb-snoRNA. As the RNA counterpart of this assay, we chose the abundant slb-snoRNA, Xenopus slb-snoRA75, because this lasso can be expressed ectopic at a high level and is easy to detect by Northern blot and RT-PCR.

We injected X. laevis oocytes with a DNA construct encoding HA-tagged chromophorin and a construct optimized to express a stable lasso (10). The latter construct includes the Staphylococcus tropicalis (xt) slb-snoRA75 sequence inserted in a part of the X. laevis ncl gene downstream of mCherry (Figure 6A). Two days after injection, the nucleus of mCherry oocytes was isolated and lysed, and RNA co-immunoprecipitation was performed with anti-HA tag antibody. Using Northern blotting, we detected ectopic expression of xtslb-snoRA75 in the RNA fraction co-immunoprecipitated with HA-labeled dyskerin, indicating that dyskerin binds to the lasso (Figure 6B). Then we tested whether the endogenous toad (xl) slb-snoRA75 also interacts with dyskerin. We use outbound primers for RT-PCR analysis to detect xlslb-snoRA75. It should be pointed out that newly synthesized endogenous RNA is almost undetectable in this analysis. Surprisingly, we found that endogenous xlslb-snoRA75 also co-precipitated with HA-labeled stain protein (Figure 6C). These results indicate that dyskerin and slb-snoRNA form a dynamic RNP complex, and snoRNP can bind to slb-snoRNA after transcription.

The ectopic expression of slb-snoRNA, its co-immunoprecipitation with stain protein and its active export to the cytoplasm. (A) Diagram of the expression construct used for ectopic expression of slb-snoRNA in toad oocytes. (B) Northern blot analysis of co-precipitation of Staphylococcus tropicalis snoRA75 and slb-snoRA75 with stain protein. Ectopic xtslb-snoRA75 was detected in both input samples: co-injected HA-labeled chromophorin and simulated co-injected oocytes. xtslb-snoRA75 only co-precipitates in the presence of HA-labeled stain protein. (C) RT-PCR analysis of RNA co-immunoprecipitated with dyskerin. The reverse primer (the blue arrow in the schematic diagram of the gene model) is only used to detect slb-snoRNA. The ectopic expressed xtslb-snoRA75 and endogenous xlslb-snoRA75 co-precipitate with keratin in the nuclear and cytoplasmic parts. (D) Northern blot analysis of xtslb-snoRA75 accumulation in toad oocytes co-injected with a construct expressing GFP mRNA as a competitor. When the xtsnoRA75 construct was injected alone, ectopic slb-snoRA75 (higher band) was detected in the nucleus and cytoplasm of toad oocytes; the expression of mature xtsnoRA75 was only detectable in the nucleus exposed for longer periods of time (Not shown). Each lane is loaded with RNA from a dissected nucleus (GV) and the cytoplasm of the injected oocyte. After co-injection with 750 pg GFP construct, slb-snoRA75 was mainly detected in the nucleus, and fully processed snoRA75 accumulated at high levels in the nucleus.

We have previously reported that the most abundant lasso is exported into the cytoplasm. Naturally, we questioned whether slb-snoRNA was also exported. As an initial experiment, we searched for RNA-seq datasets from cytoplasmic samples processed by RNase R. Because snoRNA is usually found in the nucleus, we expect that slb-snoRNA will be confined in the nucleus. It is worth noting that we detected 29 slb-snoRNAs in the cytoplasm of Staphylococcus tropicalis oocytes (SI appendix, Figure S2B, blue overlay and data set S1). In addition, about 2% of the most abundant cytoplasmic sisRNA (FPKM ≥100 in the RNase R experiment, about 2,000 introns analyzed) have a high probability of snoRNA motifs. Compared with unstable introns of the same size, the cytoplasmic lasso did not show an enrichment of such motifs.

Typical sisRNAs are exported to the cytoplasm through the NXF1/NXT1 mechanism (10). In order to test whether slb-snoRNA is also actively exported, we conducted an export competition analysis on slb-snoRNA. We injected the xtslb-snoRA75 construct, isolated nuclear and cytoplasmic RNA fractions from mCherry oocytes, and analyzed these RNAs by Northern blot. XtsnoRA75 and xtslb-snoRA75 were detected in the nuclear part, but only xtslb-snoRA75 accumulated in the cytoplasm (Figure 6B and D). When the xtslb-snoRA75 construct was co-injected with a large number of GFP expression constructs (competitors of the NXF1/NXT1 export mechanism), the accumulation of xtslb-snoRA75 in the cytoplasm was impaired, indicating that slb-snoRNAs were actively exported to the cytoplasm NXF1/NXT1 Export machinery. We also noticed that export damage is accompanied by an increase in the accumulation of mature snoRA75 in the nucleus (Figure 6D). These results are highly repeatable. In other words, the export of slb-snoRNA to the cytoplasm seems to compete with the maturation of snoRNA.

Finally, we tested whether the cytoplasmic slb-snoRNA is also related to the classic snoRNA binding protein. We analyzed cytoplasmic RNA co-precipitated with HA-labeled stain protein. Ectopic xtslb-snoRA75 and endogenous xlslb-snoRA75 were detected in the cytoplasmic RNA co-immunoprecipated with dyskerin (Figure 6C). The most reasonable explanation for these results is that slb-snoRNA accumulates in the cytoplasm as RNP particles, but still contains stain protein.

SnoRNA is one of the most thoroughly studied non-coding RNAs, but there are still many questions about their biogenesis and function. The traditional model of snoRNA expression is that they are processed from Pol II transcribed gene excision and linearization of introns. However, the expression levels of many intronic snoRNAs cannot be explained by the expression levels of their host genes. Two mechanisms have been proposed to explain the uncoupling of host gene and snoRNA expression levels: nonsense-mediated decay (22) and dual-initiated transcription (23). However, not all snoRNA and host genes are suitable for these models (24). We found that a subset of snoRNA, slb-snoRNA, became stable as a lasso. The accumulation of this form of snoRNA and its further active export to the cytoplasm prevent their processing into mature snoRNA molecules. This accumulation of slb-snoRNA occurs in tissue-specific (Figure 1C and 4D) and species-specific (data set S1). These findings reveal ways to fine-tune snoRNA expression in cells.

The composite ncRNA containing snoRNA has been described previously (25 ⇓ –27). However, in the previous case, the snoRNA domain was found at the end of the linear RNA molecule. The stability and protection of the ends are their main functions (26 ⇓ –28). In slb-snoRNA, the lasso structure itself provides the snoRNA sequence with additional protection against exonuclease activity. Importantly, even if the lasso form of snoRNA is related to the core snoRNP protein, including the modification enzyme Dykerin (Figure 6), this structure prevents snoRNA from acting as a guide RNA to post-transcriptionally modify rRNA and snRNA (Figure 3-5). . This method of exclusion from the functional modification-guided RNP pool may play a role in the negative regulation of the post-transcriptional modification of functionally critical cellular RNA. Recent studies have identified some of the modified positions in rRNA (29 ⇓ ⇓ –32) and snRNA (33); these diverse modification patterns have resulted in ribosomal and spliceosome heterogeneity. In fact, the slb-snoRD RNAs we identified in HeLa cells (data set S1) target very fragile and undermethylated positions in rRNA, or their modifications are very close to the highly sensitive residues identified in humans​​​的位置(34). These correlations support the proposed regulation of slb-snoRNA.

The inability of slb-snoRNAs to mediate the post-transcriptional modifications assigned to their homologous snoRNAs is essential for their proposed regulatory functions. Our experiments clearly show that slb-snoRNA is not a functionally modified guide RNA (Figure 3-5), which is contrary to earlier studies on the function of endogenous yeast snoRNA in dbr1Δ strains (4). These differences can easily be explained by the presence of partially processed snoRNA in mutant yeast. In particular, box C/D snoRNA was found to be functional even if it was not fully processed.

One might argue that slb-snoRNAs show no modification activity because the lasso is quickly exported to the cytoplasm, so they are separated from their substrate in the nucleus. In fact, too much lasso in the nucleus is toxic to the cell, and their export to the cytoplasm is demonstrated in yeast cells lacking DBR1 (35). However, in Xenopus GV, the accumulation level of some slb-snoRNA is comparable to that of fully processed snoRNA (Figure 1C and 4D). Specifically, X. laevis slb-snoRA75 is an example of this, but the lasso form of X. laevis snoRA75 does not support 18S rRNA modification (Figure 4E).

Interestingly, the circularization of snoRNA alone does not eliminate its ability to mediate post-transcriptional modification. Archaeal circular snoRNA-like RNA is a fully functional modified guide RNA (36 ⇓ –38). These circular guide RNAs belong to a class of stable circular RNAs called tRNA intron tric RNAs (39). Why is tricRNA a functionally modified guide RNA instead of slb-snoRNA? Compared to mature snoRNA and tricRNA, slb-snoRNA may be too large. Archaeal ring guides have only 2 to 3 extra nucleotides, and slb-snoRNA is at least 100 nt longer than mature snoRNA. This additional sequence may affect the overall snoRNA folding or stabilize the non-functional alternative conformation. Although the RNA folding software algorithm (SI appendix, Figure S6) is used to predict that the snoRNA domain of most slb-snoRNAs will not change significantly, even minor changes in the snoRNA structure can turn the fully functional modification into non-functional RNP (20) . In addition, slb-snoRNAs have a 5'-2' covalent bond connecting the branch point to the 5'end of the intron, while tricRNAs are composed of regular 5'-3' connections.

Here we should emphasize that the function of snoRNA is not limited to standard guide RNA activity and rRNA processing (40 ⇓ ⇓ ⇓ –44). The dysregulation of snoRNA is related to many diseases, and some snoRNAs have been proposed to have carcinogenic or tumor suppressor functions (45 ⇓ ⇓ ⇓ –49). In some cases, the link between a single snoRNA and tumor progression involves the misregulation of modifications in rRNA (50) or spliceosome snRNA (51), although these are fairly rare examples. The underlying mechanism may be related to smaller RNA processed from snoRNA. These include microRNA (miRNA), PIWI-interacting RNA, and snoRNA-derived RNA (sdRNA), which regulate transcription, translation, and alternative splicing (52 ⇓ ⇓ ⇓ ⇓ ⇓ –58). Therefore, the stability of sdRNA precursor as slb-snoRNA will affect the processing and function of snoRNA and sdRNA. Interestingly, in the human slb-snoRNA (data set S1) we report here, there are four snoRNAs that are up-regulated in cancer: U27, U64, U38B, and U105B.

Different conditions, stress factors and drugs can change the expression and cell distribution of snoRNA (59 ⇓ ⇓ ⇓ ⇓ –64). In response to lipotoxicity and oxidative stress, some snoRNA will accumulate in the cytoplasm (61, 63). The export mechanism needs further study (65). These snoRNAs may be exported in the form of slb-snoRNAs and further processed into mature forms in the cytoplasm under stress conditions. Consistent with this hypothesis, DBR1 has been shown to shuttle between the nucleus and cytoplasm (66).

In addition, slb-snoRNA can regulate the availability of RNA-binding proteins in cells. It has been shown that, driven by the knockdown of debranching enzymes, the artificial accumulation of lasso leads to the segregation of the RNA-binding protein TDP-43 in yeast and mammalian cells (35). It is well known that TDP-43 transiently binds to introns and regulates many cellular processes. In plants, the endonuclease Dicer, which is required for miRNA processing, can be isolated in a similar manner, leading to a general down-regulation of miRNA (15). slb-snoRNA may be necessary to isolate snoRNA binding proteins. In fact, many stable lassos contain snoRNA-like motifs (for example, human slb-snoRA29) (data set S1) instead of the typical snoRNA; the latter forms a fully processed functional snoRNP. Since the dysregulation of snoRNA binding proteins is prominent in certain diseases (67 ⇓ ⇓ –70), their strict regulation is essential.

In this study, we reported hundreds of snoRNAs, which usually accumulate in the form of stable lasso, rather than fully processed snoRNP particles. This type of ncRNA is widely distributed in different species. We propose that the main function of this unusual form of snoRNA is to fine-tune the expression level of mature snoRNA and modify the guided snoRNP. Since more than 15% of snoRNA show a certain degree of tissue specificity (24), and almost no overlap of slb-snoRNA between different species (our data), we predict that when more species and cell types will be More slb-snoRNA was identified and analyzed. As the functions of snoRNA itself become more and more diversified, we predict that the role of slb-snoRNA will also extend beyond the regulation of snoRNA expression.

Culture the human HeLa and mouse 3T3 cell lines according to standard procedures. Transfect HeLa cells with ViaFect reagent (Promega). The SnoRA29 expression construct is shown in Figure 3C. A fragment of the snoRNA encoding gene was amplified from the genomic DNA and cloned into the pCS2 vector under the CMV promoter.

X. laevis females were anesthetized with tricaine methanesulfonate (MS222), pH 7.0, and part of the ovaries were removed. Unless otherwise specified, manually isolated oocytes will be injected with 100 pg of plasmid at the animal pole in a volume of 2.3-nL. After injection, the oocytes were incubated in OR2 for 2 days. The oocytes were dissected in a separation solution (83 mM KCl, 17 mM NaCl, 6.0 mM Na2HPO4, 4.0 mM KH2PO4, 1 mM MgCl2, 1.0 mM DTT) at pH 5.6 to separate the nuclear and cytoplasmic parts.

Amplify human chromophorin cDNA from a construct provided by Mary Armanios of Johns Hopkins University School of Medicine in Baltimore, Maryland, and add a C-terminal HA tag with a reverse primer. The PCR fragment was cloned into the pcDNA3 vector under the CMV promoter. As previously described (10), a construct expressing xtslb-snora75 was generated.

The haploid yeast Saccharomyces cerevisiae strains used in this study are as follows: BY4741 (wild-type control), dbr1Δ, DBR1::KAN mutant strains (courtesy of Tulane University School of Medicine, New Orleans, LA) Jeff Han), and double Knock out dbr1Δrnt1Δ, DBR1::HIS RNT1::TRP strain (courtesy of Sherif Abou Elela, Université de Sherbrooke, Sherbrooke, QC, Canada).

In order to express foreign RNA in yeast cells, the corresponding coding sequence was amplified from genomic DNA and cloned into p426Gal1 and p416GalS vectors and YEplac195 plasmid containing GPD promoter, RNT1 cleavage site and snR13 terminator ( 71). The RNT1 cleavage site was removed from the construct expressed in the dbr1Δrnt1Δ strain. Overlap extension PCR is used to prepare chimeric fragments. A schematic diagram of these constructs is shown in Figure 5A. Use standard lithium acetate methods to transform yeast cells. Analyze at least two to three independent colonies from each plate. The transformants were grown in SC-URA medium with glucose or galactose as the sugar source, at 30°C or 25°C, when the dbr1Δrnt1Δ mutant strain was used in the experimental setting.

TRIzol reagent (Ambion) is used to extract RNA from vertebrate cell lines and Xenopus oocytes. Hot acid phenol is used for yeast RNA extraction. Purify RNA using Direct-zol RNA MiniPrep kit (Zymo Research). Remove the DNA from the column according to the manufacturer's protocol.

The oocytes were dissected in isolation buffer adjusted to pH 7.0. The nucleus and cytoplasm were collected in separate tubes and homogenized with protease inhibitor mixture (Roche) in lysis buffer (20 mM TrisHCl, pH 7.4, 150 mM NaCl, 5 mM MgCl 2, 1 mM DTT). The lysate was incubated with anti-HA magnetic beads (Pierce) for 1 hour at room temperature. The co-precipitated RNA was extracted with TRIzol.

RNA is separated on an 8% urea polyacrylamide gel, transferred to a nylon membrane (Zeta Probe, Bio-Rad), and probed with a DNA probe labeled with Digoxin (Dig), which is specific for yeast and vertebrate snoRNA sex. Dig is detected with an anti-Dig antibody conjugated with alkaline phosphatase and a chemiluminescent substrate CDP-Star (Roche). The Li-Cor Odyssey Fc imaging system and Image Studio software were used to visualize and analyze the hybridization signal. All Northern blots were repeated twice.

Random hexamers and Episcript-RT (Epicentre) and AMV-RT (New England Biolabs) were used for reverse transcription at 37°C for 1 hour. Use Taq DNA polymerase to amplify cDNA with inward and outward primers (SI Appendix, Table S1). The RT-PCR experiment was performed in two independent replicates. In order to confirm the specificity of the amplified sequence, the PCR fragment was cloned into the pGEM-T Easy vector (Promega) and sequenced.

To map 2'-O-methylated and pseudouridine residues in yeast and vertebrate RNA, we used the non-radioactive modification based on the reverse transcription method described in Deryusheva and Gall (72). Previously designed 6-FAM labeled oligonucleotides dedicated to yeast U2 snRNA and rRNA (7, 73). In short, to detect 2'-O-methylation, we performed primer extension with a low concentration of dNTP (0.004 mM). To detect pseudouridine, first treat the RNA sample with CMC [N-cyclohexyl-N'-(2-morpholinethyl)carbodiimide methyl-p-toluenesulfonate] (Sigma-Aldrich), and then use Treatment with 50 mM sodium carbonate buffer (pH 10.4). Separate fragments on a capillary electrophoresis instrument (Applied Biosystems) using the techniques and parameters recommended by the manufacturer. For each run, the Gene Scan-500 Liz size standard was used to compare fragments from different samples. GeneMapper software (Applied Biosystems) is used to screen data, identify peaks and pinpoint modified nucleotides. All RNA modification analysis experiments were performed in duplicate; each sample was run in three serial dilutions on the capillary column.

The data set used in this study has been deposited in the National Center for Biotechnology Information (NCBI) Sequence Reading Archive (SRA) Biological Project ID PRJNA302326 and PRJNA479418. The stable lasso was previously annotated (10, 18). In short, conventional reading and TopHat (v2.0.7) and Pseudomonas tropicalis genome (v9.1), mouse genome (v10), human genome (v19), chicken genome (v5) or X. laevis genome (v9.0) and intron readings were quantified using Bedtools (v2.15.0). To further verify the circular nature of the lasso, find_lariat.pl (74) was used to map the reverse reading. SnoRNA-like motifs were identified using snoReport (v2) (19).

The RNA sequencing data used in this study can be downloaded from https://www.ncbi.nlm.nih.gov/bioproject/PRJNA302326/ (18) and https://www.ncbi.nlm.nih.gov/bioproject/PRJNA479418/ Obtained (10). All other data is included in the main text and supporting information.

The research reported in this publication was supported by the NIH National Institute of General Medical Sciences, grant R01 GM33397 (grant JGG). JGG is a professor of developmental genetics at the American Cancer Society.

Author contributions: GJST, SD, and JGG design research; GJST and SD conducted research; GJST and SD contributed new reagents/analysis tools; GJST, SD, and JGG analysis data; GJST, SD, and JGG wrote this paper.

Reviewers: BLB, University of Utah School of Medicine; and AKH, Ohio State University.

The author declares no competing interests.

This article contains online support information https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2114156118/-/DCSupplemental.

This open access article is distributed under the Creative Commons Attribution-Non-Commercial-No Derivative License 4.0 (CC BY-NC-ND).

Thank you for your interest in advertising on PNAS.

Note: We only ask you to provide your email address so that the people you recommend the page to know that you want them to see it, and that it is not spam. We do not capture any email addresses.

Feedback privacy/legal

Copyright © 2021 National Academy of Sciences. Online ISSN 1091-6490. PNAS is a partner of CHORUS, COPE, CrossRef, ORCID and Research4Life.