Visualization of a group II intron in the 23S rRNA of a stable ribosome

  1. Jacoba G. Slagter-Jäger*,†,
  2. Gregory S. Allen†,‡,
  3. Dorie Smith*,
  4. Ingrid A. Hahn‡,
  5. Joachim Frank‡, and
  6. Marlene Belfort*,§
  1. *Center for Medical Science, Wadsworth Center, New York State Department of Health, 150 New Scotland Avenue, Albany, NY 12208; and
  2. ‡Health Research, Inc. at the Wadsworth Center, Howard Hughes Medical Institute, Empire State Plaza, Albany, NY 12201

  1. Contributed by Marlene Belfort, May 16, 2006

  2. ↵ †J.G.S.-J. and G.S.A. contributed equally to this work.

 
Next Section

Abstract

Thousands of introns have been localized to rRNA genes throughout the three domains of life. The consequences of the presence of either a spliced or an unspliced intron in a rRNA for ribosome assembly and packaging are largely unknown. To help address these questions, and to begin an intron imaging study, we selected a member of the self-splicing group II intron family, which is hypothesized to be the progenitor not only of spliceosomal introns but also of non-LTR retrotransposons. We cloned the self-splicing group II Ll.LtrB intron from Lactococcus lactis into L. lactis 23S rRNA. The 2,492-nt Ll.LtrB intron comprises a catalytic core and an ORF, which encodes a protein, LtrA. LtrA forms a ribonucleoprotein (RNP) complex with the intron RNA to mediate splicing and mobility. The chimeric 23S–intron RNA was shown to be splicing proficient in its native host in the presence of LtrA. Furthermore, a low-resolution cryo-EM reconstruction of the L. lactis ribosome fused to the intron–LtrA RNP of a splicing-defective Ll.LtrB intron was obtained. The image revealed the intron as a large, well defined structure. The activity and structural integrity of the intron indicate not only that it can coexist with the ribosome but also that its presence permits the assembly of a stable ribosome. Additionally, we view our results as a proof of principle that ribosome chimeras may be generally useful for studying a wide variety of structured RNAs and RNP complexes that are not amenable to NMR, crystallographic, or single-particle cryo-EM methodologies.

Introns are common residents of rRNA genes (rDNAs). The small and large subunit rDNAs can harbor the autocatalytic group I and group II introns, as well as the non-self-splicing, archaeal, and spliceosomal introns (1). The location of introns in the 16S and 23S rRNAs is not random; rather, many introns of different types cluster near tRNA-binding sites. Some of these sites are located at the interface between the small and large subunits of the ribosome (2), in dynamic regions that undergo conformational changes during translation (3). The distribution of these introns suggests that their location could be related to their mode of entry into rDNA or rRNA or to their retention at specific sites. Intron maintenance could be a function of the splicing efficiency at a particular position or of the functional preservation of a ribosome harboring an intron.

Besides homologous recombination among rDNA genes, there are two different mechanisms of integration of mobile introns that could explain how introns spread into specific ribosomal sequences: homing into sites on DNA and reverse splicing into sites on RNA (4). For this study, we used a member of the self-splicing group II intron family, which is hypothesized to be the progenitor not only of spliceosomal introns (5) but also of non-LTR retrotransposons (6). Our model intron is the Lactococcus lactis group II intron, Ll.LtrB. In a study of the dispersal of Ll.LtrB by retrotransposition, the intron was found to be able to insert at two different positions in 23S rRNA (7). In a later study, this intron was shown to retain splicing and RNA-based retromobility activities from both positions in the 23S rRNA (8). Also, a previous study showed that the precursor of the Neurospora crassa mitochondrial large subunit rRNA containing a group I intron could be assembled into ribonucleoprotein (RNP) particles containing nearly all of the ribosomal proteins (9). Although the insertion of an intron into a rDNA may be consistent with, or even advantageous to, the intron’s splicing function, intron presence at some sites in the pre-rRNA could arrest ribosome assembly, either by altering local secondary structure or by sterically blocking tertiary contacts.

We therefore wished to test intron function at what we expected to be a tolerant site in the ribosome and to then study the ability of the ribosome to assemble in the presence of a group II intron. Additionally, the integrity of an intron–ribosome chimera would suggest a unique means of examining intron structure. To make the fusion constructs, we used the L. lactis Ll.LtrB intron. The Ll.LtrB intron RNA is 2,492 nt in length, comprising a catalytic core of 695 nt and an ORF of 1,797 nt. The ORF encodes a protein, LtrA, that dimerizes and forms a RNP complex with the RNA to mediate intron splicing and mobility (reviewed in ref. 10). Splicing is initiated by the 2′ OH of an adenosine in domain VI of the intron in a process guided by intron–exon [exon-binding site (EBS)–intron-binding site (IBS)] pairings (Fig. 1 A) (reviewed in ref. 11). To form a stable ribosome–intron chimera, we deleted the “branch-point” adenosine to inhibit the intron from splicing out of the rRNA while providing EBS-IBS pairings for the intron to maintain its native conformation.

Fig. 1.
View larger version:
Fig. 1.

The group II intron and 23S rRNA insertion site. (A) Secondary structure map of the Ll.LtrB intron (gray). EBS–IBS and δ–δ′ interactions are shown by dotted black lines. The exon sequences, including IBS1, IBS2, and δ′, that are present in the chimeric construct are shown. The branch-point adenosine (A) is circled. (B) Secondary structure map of a segment of the L. lactis 23S rRNA. The intron insertion site (IS) in helix 98 is enlarged, as are the nucleotides CCUAGG that create the BamHI site engineered for cloning into DNA.


For this study, the L. lactis ribosome was favored over the Escherichia coli ribosome, because the Ll.LtrB group II intron is ≈100-fold more efficient in retrohoming in L. lactis, its natural host (12). Intron variants were therefore inserted into loop 98 of L. lactis 23S rRNA (Fig. 1 B), which is located on the surface of the 50S subunit and which had been shown to be hospitable to a foreign tRNA sequence (13). This approach was selected over the use of retrotransposed introns in L. lactis rRNA that are splicing proficient, to facilitate cloning and because the tRNA previously inserted into loop 98 of 23S E. coli rRNA had been imaged by cryo-EM (13). We reasoned that if the intron was active at this site, then image reconstruction of the group II intron would be facilitated in the chimera. We show not only that the intron is catalytically active at this site but also that its presence is consistent with assembly of an intact ribosome. Furthermore, the images indicate that the intron forms a large, defined structure, suggesting ribosome chimeras as a general means of imaging RNA-based complexes from a variety of sources.

Previous SectionNext Section

Results and Discussion

Construction of Ribosome–Intron Chimeras.

To construct the ribosome–intron chimeras, the Ll.LtrB intron was inserted into vector pLNRK (14), into which we had cloned the 23S rDNA from L. lactis. This insertion placed expression of the chimeric RNA under the control of the nisin-inducible promoter PnisA (15) (Fig. 2 A). Then, four different intron variants were cloned into rDNA at loop 98 of L. lactis 23S rRNA (Figs. 1 B and 2 A). First, 23S-LtrB contains a WT intron. Second, 23S-LtrBΔA contains an intron in which the branch-point adenosine (Fig. 1 A) had been deleted (16), thereby rendering the intron splicing deficient while maintaining the catalytically active conformation of the RNA. Third, 23S-LtrBΔORF contains an intron in which 1,590 nt of the ORF encoding LtrA had been deleted (17), resulting in a catalytically active intron that cannot splice because of the absence of LtrA. Finally, 23S-LtrBΔORFΔA is a double mutant, which combines the defects of the two aforementioned constructs. In all four cases, base pairings between the intron (EBS1, EBS2, and δ) and the flanking 5′ exon (IBS1 and IBS2) and 3′ exon (δ′) (18) were maintained by preserving 15 nt of natural exon sequence on the 5′ side of the intron and 3 nt on the 3′ side (Figs. 1 A and 2 A). These pairings are required for proper folding of the intron RNA and for accurate splice-site recognition (19).

Fig. 2.
View larger version:
Fig. 2.

Construction and expression of a 23S rRNA–group II intron chimera. (A) Plasmid maps of chimeric constructs. The exon sequences are boxed, and the intron is shaded gray. The plasmid shown for 23S-LtrB (WT) consists of rRNA sequences (thick black lines) and the pLNRK backbone with the nisin promoter. B, BamHI site. (B) Induction of 23S-LtrBΔORFΔΑ with varying nisin concentration. (Upper) RNA was separated on a 1% agarose gel and visualized with ethidium bromide. (Lower) Subsequent Northern blot hybridization was with an intron- and then a 5S-RNA-specific probe. Nisin concentrations are as follows: lane 1, 0 ng/ml; lane 2, 1 ng/ml; lane 3, 2.5 ng/ml; lane 4, 5 ng/ml; lane 5, 7.5 ng/ml; lane 6, 10 ng/ml; lane 7, 50 ng/ml; lane 8, plasmid pLNRK23S-L98 treated with 10 ng/ml nisin; lane 9, IL1403 cells treated with 10 ng/ml nisin. Bands marked “UIB” are unidentified intron-containing bands. The positions of the 23S and 16S rRNA on the Northern blot are indicated by arrowheads.


To test inducibility of the chimeric RNAs, we investigated intron RNA production from the 23S-LtrBΔORFΔA construct by induction with different concentrations of nisin by using agarose gel electrophoresis (Fig. 2 B). In addition to the bands corresponding to the 23S, 16S, and 5S rRNAs, a band is faintly visible at increasing intensity above the 23S rRNA band at nisin concentrations >2.5 ng/ml. This band, labeled 23S-GII, was confirmed to carry intron sequences by Northern blot hybridization with an intron-specific probe. The chimeric 23S-GII band was most intense at 10 ng/ml nisin, the concentration used for subsequent experiments to monitor splicing activity and overproduction of the chimeric 23S-LtrBΔA for cryo-EM.

The Ll.LtrB Intron Is Active in the Ribosome–Intron Chimera.

The activity of the Ll.LtrB intron in the 23S rRNA was investigated by monitoring the presence of the splicing precursor and product by RT-PCR (Fig. 3 A). The WT 23S-LtrB (lane 6) and the 23S-LtrBΔA variant (lane 12) both show a cDNA of ≈2.6 kb corresponding to the precursors of 2,631 and 2,630 bp, respectively. In addition, for the WT, a small, ≈0.15-kb fragment could be detected (lane 6). By sequencing, this small PCR fragment was confirmed to represent the 139-bp splicing product. In the splicing-deficient variants, no or very low amounts of the splicing product were apparent. The ≈1-kb bands representing the precursors in the 23S-LtrBΔORF and the 23S-LtrBΔORFΔA variants (lanes 8 and 10) are 1,041 and 1,040 bp, respectively, consistent with the ORF deletion. Also, in these cases, a small quantity of splicing product is apparent. These bands can likely be attributed to self-splicing, where the intron uses hydrolysis at the splice sites, instead of transesterification reactions, to generate ligated exons (20).

Fig. 3.
View larger version:
Fig. 3.

Intron activity assays. (A) RT-PCR assay (Upper). A schematic of the RT-PCR templates is shown in Lower. Lane M has DNA markers. Lanes 1–12 show RT-PCR products made with forward (F) and reverse (R) primers (see schematic). Odd-numbered (−) and even-numbered (+) lanes are without and with reverse transcriptase, respectively. The RNA templates used were extracted from strain IL1403 containing the following plasmids: lanes 1 and 2, no plasmid; lanes 3 and 4, pLNRK23S-L98 with no intron; lanes 5 and 6, pLNRK23S-L98 with WT intron; lanes 7–12, pLNRK23S-L98 with intron mutants indicated above lanes. Primer extension products corresponding to precursors are marked with a filled arrowhead, and those corresponding to the ligated exon product are marked with an open arrowhead. The sizes of RT-PCR products are shown by solid lines in the schematic. B, BamHI site. (B) Primer extension termination assay (Upper). A schematic of the different products that are predicted is shown in Lower. Lanes in the gel are labeled as in A. Gray arrowhead points to lariat-3′ exon cDNA; black and white arrowheads are as in A. The band in lane 4 corresponds to a G in the 23S RNA at residue 12.


The splicing reaction was further investigated by primer extension termination assays (Fig. 3 B). In this assay, cDNA is made by reverse transcription starting from a specific primer complementary to the 3′ exon. The primer was designed such that, in the presence of 2,3-dideoxycytidine triphosphate, the cDNA will terminate at a different guanosine, depending on whether the precursor or the product serves as a template. Furthermore, the branched nucleotide of the lariat will cause chain termination. Bands corresponding to splicing precursor (primer plus 20 nt), product (primer plus 18 nt), and lariat intermediate (primer plus 9 nt) could be detected for the WT intron (lane 6). For the ΔORF intron in the 23S-LtrB chimera, only cDNA that corresponds to the precursor was observed (lane 8). Furthermore, for the ΔA and ΔORFΔA introns (lanes 12 and 10, respectively), both precursor bands, each of which is one residue shorter because of the ΔA template, as well as several lower bands, are apparent. These smaller bands likely represent cDNAs from lariats formed by ectopic branching in the absence of the bulged adenosine. Together, these results demonstrate the activity of the intron embedded within the 23S rRNA sequence.

Separation and Imaging of Ribosome–Intron Chimeras.

We selected the 23S-LtrBΔA chimera to determine whether a ribosome that harbors a group II intron could be correctly assembled and to see whether the intron formed a defined structure. In this construct, the ORF-containing intron RNA would be associated with LtrA protein and would be in an active conformation, but it could not splice because of deletion of the branch-point adenosine, which initiates splicing. We reasoned that such an intron would fold properly but would remain attached to the 23S rRNA exons. First, we separated these complexes on a sucrose density gradient on which the chimeric particles were present in a peak preceding the 70S ribosome peak (Fig. 4). To verify the identity of the leading peak, as well as to ensure that LtrA had copurified with the ribosome–intron chimera, we tested pooled fractions for the presence of LtrA protein by Western blot analysis (Fig. 4 Inset). For this analysis, 2 pmol of purified ribosome–intron chimera was loaded on a SDS/polyacrylamide gel. The amount of LtrA in the sample was deduced to be between 4 and 5 pmol, from comparison to the standard curve on the same gel, suggesting a 1:2 ratio of chimera to LtrA. This value is in agreement with the fact that LtrA binds the intron as a dimer (17).

Fig. 4.
View larger version:
Fig. 4.

Ribosome–intron chimera purification by sucrose gradients. The plots show WT L. lactis 70S ribosomes (red) and the chimera containing the LtrBΔA intron (blue) separated on a 15–35% sucrose gradient [labeled Top and Bot (bottom)]. The ribosome–group II intron chimeras are contained in a peak (shaded) that precedes the 70S ribosome peak. The fractions in this leading peak were pooled for cryo-EM sample preparation. (Inset) Western blot analysis of sucrose gradient-purified ribosomes and the ribosome–intron chimera. For the LtrA standard curve, 1, 2, 4, and 10 pmol of purified LtrA were loaded. Lane 0, IL1403 ribosomes; lane ΔORF, 23S-LtrBΔORF chimera; lane ΔA, 23S-LtrBΔA chimera.


The purified 23S-LtrBΔA ribosome–group II intron complexes (Fig. 4) were next applied to Quantifoil grids and were flash-frozen for cryo-EM data collection. In the image analysis, we made use of a previously determined cryo-EM structure of the E. coli ribosome tagged with a tRNA inserted into the same L98 loop of the 23S rRNA (13) for reference-based reconstruction (Fig. 5 A). We used spider (21) to modify this map as follows. A sphere of constant density and radius of 110 Å was added at a point 130 Å from the tRNA on a line drawn from the center of the ribosome to the tRNA (Fig. 5 B). We used the modified map as a reference to automatically pick particles depicting the chimera from the micrographs. Reconstruction, based on ≈10,000 such particles, revealed that the 23S-LtrBΔA ribosome–group II intron complex, at 34-Å resolution, is composed of two globular structures, one of which is clearly a 70S ribosome (Fig. 5 C). Because the 30S ribosomal subunit appears associated with the 50S subunit in the reconstruction, we conclude that the insertion of the intron does not disrupt subunit–subunit interactions. The intron is connected to the 70S ribosome by a thread of density that we infer to be the RNA carrying the IBS1 and IBS2 sequences. Surprisingly, the intron–LtrA complex appears much larger than expected, considering that its calculated mass is less than one-half that of the ribosome (980 kDa for the WT intron RNP vs. 2.5 MDa for the ribosome). However, cuts through the interior of the intron reveal a large cavity and several tunnels (data not shown) that account for the large volume (Fig. 5 D).

Fig. 5.
View larger version:
Fig. 5.

Cryo-EM density map of ribosome–group II intron chimera. (A) The ribosome–tRNA chimera (13). (B) Reference map. (C) Cryo-EM reconstruction of the ribosome–group II intron chimera at ≈34-Å resolution. (D) Cross section of the group II intron displays internal density of the reconstruction. The large internal cavity present in the intron portion of the reconstructed chimera is visible in the cross section. The line drawn through the intron in C indicates the cutting plane used for the cross-sectional view shown here, as well as the axis with respect to which the cross section is rotated. All density below the plane has been removed for ease of viewing. The 50S ribosomal subunit is colored blue, the 30S ribosomal subunit is yellow, the tRNA is green, and the group II intron is red. Landmarks on the ribosomes are as follows: CP, central protuberance; L1, L1 stalk; h, head; sp, spur.


These findings are in contrast to other biophysical studies of homogeneously folded and functional introns, in which the introns appear as tightly packed structures (22, 23). The loose packing shown here may reflect a difference in intron constructs; the biophysical studies were performed on domains of the aI5γ intron, whereas the cryo-EM reflects the intact chimeric LtrB intron. Alternatively, different conformational and/or activity states of the intron or variations in experimental conditions may account for the differences. Regardless, improvement of the resolution of the intron RNP in various structural configurations, which correspond to different conformational states of the ribozyme, will undoubtedly provide greater insight into group II intron structure and function.

Evolutionary and Practical Implications of Ribosome–Intron Chimeras.

To what extent ribosome assembly is affected by natural insertion events remains to be investigated, but the question is now amenable to analysis. Also, whether the considerable size of the group II intron is an accurate reflection of the packing of the RNA or, instead, is a condition-dependent phenomenon awaits further study. Certainly, the large volume of the intron RNP suggests that sites that are buried in the ribosome may not be permissive for ribosome packing unless splicing occurs before ribosome assembly. In contrast, a surface position would place lower demands on splicing for ribosome function. Therefore, localization and maintenance of introns could represent a balance between intron position in the ribosome on the one hand and RNA splicing efficiency on the other.

In addition to evolutionary implications, our results show that the ribosome can be used as a scaffold for the facile cryo-imaging of structured RNAs or RNPs. More generally, not only does the physical connection of the molecule of interest to the ribosome allow rapid purification by using established protocols, but orientation relative to the ribosome is fixed, thereby greatly facilitating subsequent image reconstruction and interpretation. RNAs and RNPs in the 104- to 106-Da range may be too small to image by using conventional single-particle cryo-EM, whereas these particles are cumbersome for current NMR methods. Our chimeric cryo-EM approach offers a potential solution to obtaining low-resolution structures of RNAs and RNPs in this size range without the time commitment of x-ray crystallography. With minor changes in strategy, this chimeric method is applicable to structured RNAs and RNPs from both prokaryotes and eukaryotes, requiring only that the RNA of interest be cloned into a specific surface loop of the 23S rRNA and that, where needed, the RNA-binding protein(s) is coexpressed with the chimeric RNA. Finally, the use of the ribosome as a molecular anchor may provide a viable approach for the study of various functional states of other structured RNAs and RNPs.

Previous SectionNext Section

Materials and Methods

Construction of Plasmids.

First, L. lactis 23S rRNA sequences that were engineered to contain a BamHI site in loop 98 were cloned into an L. lactis/E. coli shuttle vector, pLNRK, a pLE1-based plasmid (14) containing a nisin promoter (15), and the nisR and nisK genes cloned into the ApaLI site. Two PCR fragments were made by using genomic DNA from L. lactis IL1403 as a template and primers carrying XbaI or BamHI cleavage sites. Primers used to amplify the 2,813-bp 5′ end fragment of 23S rRNA were as follows: 23S-XbaI-forward, 5′-GCTCTAGAGGCAAAGTTAATAAGGGCGCAC-3′; L98-BamHI-reverse, 5′-CGGGATCCGAATGGGTAATCTCATCTTGAG-3′. Primers for the 114-bp 3′ end fragment of the 23S rRNA were as follows: L98-BamHI-forward, 5′-CGGGATCCATTAAGAGCCCAGAGAGATGAT-3′; 23S-XbaI-reverse, 5′-GCTCTAGATTGTAAAGTCCTCGAGCGATT-3′. (Cleavage sites are italicized.) The two PCR fragments were cleaved with BamHI and ligated. The resulting 2,917-bp fragment was cleaved at the ends with XbaI, ligated into plasmid pLNRK, which had been linearized by XbaI cleavage, and introduced into E. coli DH5α by transformation. The resulting plasmid, called pLNRK23S-L98, was used for the cloning of four different intron variants.

PCR fragments carrying intron sequences were obtained by using plasmid pET-LtrA as a source for WT ltrB sequences (24), pET-LtrAΔA as the template for the ΔA variant, pGM-ΔORF for the ΔORF variant, and pGM-ΔORFΔA for the double mutant (16). Primers had BamHI restriction sites (shown in boldface in the following) to facilitate ligation of the PCR fragment into BamHI-linearized pLNRK23S-L98. In addition, primer IDT0058 carried 15 bases of the 5′ exon corresponding to IBS1 and IBS2 (shown in italics): 5′-CGCGGATCCGAACACATCCATAACGTGCGCCCAGATAGGGTGTTA-3′. Primer IDT0059 carried three bases of the 3′ exon (shown in italics): 5′-CGCGGATCCATGGTGAAGTAGGGAGGTACCGCCTTG-3′. For amplification of the ΔA variant, primer IDT0110 (5′-CGCGGATCCATGGTGAAGAGGGAGGTACCGCCTTG-3′) was used. This primer is similar to primer IDT0059, except for one thymine, which was deleted. The resulting plasmids with PCR fragments in pLNRK23S-L98 were called pLNRK23S-LtrB, pLNRK23S-LtrBΔORF, pLNRK23S-LtrBΔA, and pLNRK23S-LtrBΔORFΔA. The intron sizes in these plasmids are 2,492, 902, 2,491, and 901 bp, respectively.

RNA Isolation.

Total RNA was isolated from L. lactis IL1403 cells carrying the different plasmids according to methods described by Mills et al. (25), with some changes. Briefly, 11-ml cultures grown overnight at 30°C in M17 medium supplemented with 0.5% glucose and 10 μg/ml chloramphenicol were diluted at a ratio of 1:1 with medium, induced with 10 ng/ml nisin, and allowed to grow for an additional 3 h at 30°C. The culture was mixed at a ratio of 1:1 with ice slurry (0.02 M Tris·HCl, pH 7.3/0.005 M MgCl2/0.02 M NaN3/400 μg/ml chloramphenicol), and cells were pelleted by centrifugation at 4,350 × g for 5 min at 4°C. The pellet was resuspended in 4 ml of 25% sucrose/50 mM Tris·HCl (pH 8). One hundred microliters of a mixture of lysozyme (100 mg/ml) and mutanolysin (1 μg/ml) was added, and the sample was incubated for 5 min on ice. Cells were pelleted by centrifugation at 3,030 × g for 5 min at 4°C. The pellet was resuspended in 600 μl of lysis buffer (20 mM Tris·HCl, pH 8/3 mM EDTA/0.2 M NaCl) and divided into two tubes. To each tube, 350 μl of prewarmed (95°C) lysis buffer containing 1% SDS was added, and the samples were incubated at 95°C. Subsequently, the samples were mixed, and hot (65°C) phenol was added. The samples were incubated for 3 min at 65°C and mixed occasionally. Samples were cooled on ice for 5 min and centrifuged at maximum speed in a Hermle Z 230MR microcentrifuge for 10 min at 4°C. The phenol extraction was repeated twice. Next, a phenol-chloroform (1:1) extraction was performed, which was repeated twice, followed by a chloroform extraction. Finally, the RNA was precipitated with 3 volumes of ethanol (95%), 1/10 of the volume of 5 M NH4Cl, and 20 μg of glycogen.

Northern Blot Analysis.

Ten micrograms of total RNA was loaded on a 1% agarose gel and electrophoresed in 1× TBE buffer (2.5 mM EDTA/5 mM boric acid/9 mM Trizma-base). Next, the RNA was transferred to a Hybond N+ membrane (Amersham Pharmacia Biosciences) by capillary blotting according to the protocols of the manufacturer. The membrane was hybridized overnight at 42°C with 32P end-labeled deoxyoligonucleotide intron probe IDT0059. After image capture, the membrane was hybridized with the 5S RNA-specific oligonucleotide IDT0198 (5′-ACATGGGAACAGGTGTATCT-3′).

RT-PCR.

For reverse transcription, ≈2 μg of total RNA was mixed with 4 μl of dNTPs (2.5 mM), 1 μl of 23S-XbaI-reverse oligonucleotide (10 μM), and 10.5 μl of water. The RNA was denatured at 80°C for 5 min and placed on ice. Subsequently, 2 μl of 10× Moloney murine leukemia virus (MMLV) reverse transcriptase buffer, 1 μl of RNasin (40 units/μl; Promega), and 1 μl of MMLV reverse transcriptase (100 units/μl; Ambion, Austin, TX) were added and incubated at 42°C for 1 h. Next, 1 μl of this reaction mixture was mixed with 1 μl of 23S-XbaI-reverse oligonucleotide (10 μM), 1 μl of RT-PCR oligonucleotide (5′-CGCGGATCCGAACACATCCATAAC-3′) (10 μM), 2 μl of dNTPs (2.5 mM), 2 μl of 10× ThermoPol buffer, 1.2 μl of MgCl2 (25 mM), 11.8 μl of water, and 0.5 μl of Vent DNA polymerase (2 units/μl; New England Biolabs). The samples were incubated in a Peltier Thermal Cycler (MJ Research, Cambridge, MA) at 94°C for 30 s, at 55°C for 30 s, and at 72°C for 3 min. This program was repeated 35 times. Finally, the samples were incubated at 72°C for 10 min. PCR products were analyzed on 1% agarose gels in 0.5× TBE buffer (1.25 mM EDTA/2.5 mM boric acid/4.5 mM Trizma-base).

Primer Extension Termination Assay.

Reverse transcription reactions were set up as described above, with a few modifications. Deoxyoligonucleotide IDT0126 (5′-CTCTGGGCTCTTAATGGATCC-3′) was radioactively labeled at the 5′ end, and dCTP was exchanged for 2,3-dideoxycytidine triphosphate. Reactions were analyzed on a denaturing (7 M urea) 20% polyacrylamide gel in 1× TBE buffer. The gel was run at 60 W and exposed to a phosphorimaging screen. Images were analyzed with a Typhoon 9400 scanner (Amersham Pharmacia Biosciences).

Purification of 23S-LtrBΔA Ribosome.

L. lactis IL1403 containing plasmids were grown at 30°C in GM17 medium to an OD600 of 0.5. Nisin was added at 10 ng/ml, and cells were grown for an additional 3 h. The culture was then centrifuged, and the pellet was resuspended in buffer containing 20 mM Tris (pH 7.5), 10 mM NH4Cl, 10.5 mM MgOAc, and 0.5 mM EDTA. Resuspended cells were prepared for lysis by the addition of the following: 1 μg/ml DNase (Rnase-free), 80 μl/ml protease inhibitor mixture (Sigma), and 1 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP). Cells were stirred to homogeneity at 4°C. The cell suspension was then passed through a French press (Thermo Electron, San Jose, CA) three times at 14,000 psi. Cell debris was removed twice by centrifugation in a Ti70.1 rotor (Beckman Instruments) for 20 min at 20,000 rpm. Each 5 ml of the clarified lysate was passed through a 5-ml sucrose cushion (composed of 20 mM Tris·HCl, pH 7.5/500 mM NH4Cl/10.5 mM MgCl2/1.1 M sucrose) and centrifuged for 19 h in a Ti70.1 rotor at 28,000 rpm. The resulting pellets were washed with 1 ml of 20 mM Tris·HCl, pH 7.5/500 mM NH4Cl/10.5 mM MgOAc/0.5 mM EDTA/1 mM TCEP and then resuspended in 20 mM Tris·HCl, pH 7.5/100 mM NH4Cl/10 mM MgCl2/1 mM TCEP. The pellets obtained from the sucrose cushion step were loaded on a 10–35% sucrose density gradient in 20 mM Tris·HCl, pH 7.5/100 mM NH4Cl/10 mM MgCl2 and were run at 16,000 rpm for 20 h in an SW28 (Beckman Instruments) rotor. Fractions (750 μl) of the sucrose density gradient were collected and analyzed on a UV-VIS spectrophotometer. Samples of interest were pooled and pelleted at 35,000 rpm for 20 h in a Ti70.1 rotor. Pellets were resuspended in Polymix buffer (95 mM KCl/5 mM NH4Cl/5 mM MgOAc/0.5 mM CaCl2/8 mM putrescine/1 mM spermidine), flash-frozen in liquid nitrogen, and stored at −80°C.

Western Blot Analysis.

Protein samples were mixed with loading dye (100 mM Tris·HCL, pH 7/200 mM dithiotreitol/4% SDS/0.2% bromophenol blue/20% glycerol) and loaded onto a 5% polyacrylamide stacking gel (70 mM Tris, pH 6.8/0.1% SDS) and 10% SDS/polyacrylamide separating gel (375 mM Tris, pH 8.8/0.1% SDS). The gel was run at 40 mA in TGS buffer (25 mM Tris, pH 8.3/192 mM glycine/0.1% SDS) and subsequently incubated in transfer buffer (48 mM Tris, pH 9.2/39 mM glycine/20% MeOH/0.0375% SDS) for 30 min. The proteins were transferred to a poly(vinylidene difluoride) membrane (Millipore) by using a TransBlot Electrophoretic Transfer Cell (Bio-Rad), according to the manufacturer’s instructions. LtrA protein was immunolocalized by using LtrA-specific antibody (kindly provided by Gary Dunny, University of Minnesota, Minneapolis) and the ECL-plus Western Blotting Detection System (Amersham Pharmacia Biosciences).

Cryo-EM and Image Processing.

Thin carbon was floated onto Quantifoil grids (Quantifoil Micro Tools, Jena, Germany). A 5-μl aliquot of ribosome–LtrBΔA (diluted to 30 nM in 20 mM Tris·HCl, pH 7.5/100 mM NH4Cl/10 mM MgCl2) was placed on each grid. The grids were then blotted, plunge-frozen in liquid ethane (Vitrobot; FEI, Hillsboro, OR), and stored in liquid nitrogen until data collection. Data were collected on a Tecnai F30 (FEI) at 300 kV and a magnification of ×39,000 (±2%) under low-dose conditions. Micrographs were scanned at 14-μm resolution, giving 3.72 Å per pixel. Ten micrographs were each assigned to one of 10 defocus groups, ranging from 1.54 to 3.45 μm underfocus. Particles (9,804) could easily be identified on the micrographs because they appeared as doublets. Eighty-three equispaced reference projections were created to classify and align the particle images. The aligned images were used to produce a 3D reconstruction of the 70S-LtrBΔA particle. The spider/web suite of programs (21) was used for all image processing. The resolution of the reconstruction was estimated as 34 Å by using Fourier shell correlation with a cutoff of 0.5.

Previous SectionNext Section

Acknowledgments

We thank Maryellen Carl for help with the manuscript; John Dansereau and Michael Watters for figure preparation; Robert A. Grassucci and the staff of the Molecular Genetics Core (Wadsworth Center) for help with cryo-EM and DNA sequencing, respectively; our Wadsworth Center colleagues for discussions and comments on the manuscript; Dr. Gary Dunny for anti-LtrA antibody; and Alan Lambowitz and Anna-Marie Pyle for useful discussions. This work was supported by National Institutes of Health Grants GM39422 and GM44844 (to M.B.), the Howard Hughes Medical Institute, and the Swedish Research Council (J.G.S.-J.).

Previous SectionNext Section

Footnotes

  • §To whom correspondence should be addressed. E-mail: belfort{at}wadsworth.org

  • Author contributions: J.G.S.-J., G.S.A., J.F., and M.B. designed research; J.G.S.-J., G.S.A., D.S., and I.A.H. performed research; J.G.S.-J. and G.S.A. analyzed data; and J.G.S.-J., G.S.A., and M.B. wrote the paper.

  • Conflict of interest statement: No conflicts declared.

  • Abbreviations:

    Abbreviations:

    rDNA,
    rRNA gene;
    RNP,
    ribonucleoprotein;
    EBS,
    exon-binding site;
    IBS,
    intron-binding site.
Previous Section
 

References

  1. ↵
    1. Cannone J. J.,
    2. Subramanian S.,
    3. Schnare M. N.,
    4. Collett J. R.,
    5. D’Souza L. M.,
    6. Du Y.,
    7. Feng B.,
    8. Lin N.,
    9. Madabusi L. V.,
    10. Muller K. M.,
    11. et al.
    (2002) BMC Bioinformatics 3:1–31.
    Medline
  2. ↵
    1. Jackson S.,
    2. Cannone J.,
    3. Lee J.,
    4. Gutell R.,
    5. Woodson S.
    (2002) J. Mol. Biol. 323:35–52.
    CrossRefMedlineISI
  3. ↵
    1. Frank J.,
    2. Agrawal R. K.
    (2000) Nature 406:318–322.
    CrossRefMedline
  4. ↵
    1. Belfort M.,
    2. Derbyshire V.,
    3. Parker M. M.,
    4. Cousineau B.,
    5. Lambowitz A. M.
    1. Craig N.,
    2. Craigie R.,
    3. Gellert M.,
    4. Lambowitz A.
    (2002) in Mobile DNA II, eds Craig N., Craigie R., Gellert M., Lambowitz A. (Am. Soc. Microbiol. Washington, DC), pp 761–783.
  5. ↵
    1. Sharp P. A.
    (1991) Science 254:663.
    FREE Full Text
  6. ↵
    1. Kazazian H. H., Jr.
    (2004) Science 303:1626–1632.
    Abstract/FREE Full Text
  7. ↵
    1. Cousineau B.,
    2. Lawrence S.,
    3. Smith D.,
    4. Belfort M.
    (2000) Nature 404:1018–1021, correction (2001) 414, 84.
    CrossRefMedline
  8. ↵
    1. Conlan L. H.,
    2. Stanger M. J.,
    3. Ichiyanagi K.,
    4. Belfort M.
    (2005) Nucleic Acids Res. 33:5262–5270.
    Abstract/FREE Full Text
  9. ↵
    1. LaPolla R. J.,
    2. Lambowitz A. M.
    (1979) J. Biol. Chem. 254:11746–11750.
    Abstract/FREE Full Text
  10. ↵
    1. Lambowitz A. M.,
    2. Mohr G.,
    3. Zimmerely S.
    1. Belfort M.,
    2. Derbyshire V.,
    3. Stoddard B. L.,
    4. Wood D. W.
    (2005) in Homing Endonucleases and Inteins, eds Belfort M., Derbyshire V., Stoddard B. L., Wood D. W. (Springer, Berlin), pp 121–145.
  11. ↵
    1. Michel F.,
    2. Ferat J.-L.
    (1995) Annu. Rev. Biochem. 64:435–461.
    CrossRefMedlineISI
  12. ↵
    1. Cousineau B.,
    2. Smith D.,
    3. Lawrence-Cavanagh S.,
    4. Mueller J. E.,
    5. Yang J.,
    6. Mills D.,
    7. Manias D.,
    8. Dunny G.,
    9. Lambowitz A. M.,
    10. Belfort M.
    (1998) Cell 94:451–462.
    CrossRefMedlineISI
  13. ↵
    1. Spahn C. M.,
    2. Grassucci R. A.,
    3. Penczek P.,
    4. Frank J.
    (1999) Structure (London) 7:1567–1573.
  14. ↵
    1. Mills D. A.,
    2. Choi C. K.,
    3. Dunny G. M.,
    4. McKay L.
    (1994) Appl. Environ. Microbiol. 60:4413–4420.
    Abstract/FREE Full Text
  15. ↵
    1. de Ruyter P. G.,
    2. Kuipers O. P.,
    3. Beerthuyzen M. M.,
    4. van Alen-Boerrigter I.,
    5. de Vos W. M.
    (1996) J. Bacteriol. 178:3434–3439.
    Abstract/FREE Full Text
  16. ↵
    1. Matsuura M.,
    2. Noah J. W.,
    3. Lambowitz A. M.
    (2001) EMBO J. 20:7259–7270.
    CrossRefMedlineISI
  17. ↵
    1. Saldanha R.,
    2. Chen B.,
    3. Wank H.,
    4. Matsuura M.,
    5. Edwards J.,
    6. Lambowitz A. M.
    (1999) Biochemistry 38:9069–9083.
    CrossRefMedline
  18. ↵
    1. Mohr G.,
    2. Smith D.,
    3. Belfort M.,
    4. Lambowitz A. M.
    (2000) Genes Dev. 14:559–573.
    Abstract/FREE Full Text
  19. ↵
    1. Pyle A. M.,
    2. Lambowitz A. M.
    (2006) RNA World III (Cold Spring Harbor Lab. Press, Woodbury, NY), 469–505.
  20. ↵
    1. Podar M.,
    2. Chu V. T.,
    3. Pyle A. M.,
    4. Perlman P. S.
    (1998) Nature 391:915–918.
    CrossRefMedline
  21. ↵
    1. Frank J.,
    2. Radermacher M.,
    3. Penczek P.,
    4. Zhu J.,
    5. Li Y.,
    6. Ladjadj M.,
    7. Leith A.
    (1996) J. Struct. Biol. 116:190–199.
    CrossRefMedlineISI
  22. ↵
    1. de Lencastre A.,
    2. Hamill S.,
    3. Pyle A. M.
    (2005) Nat. Struct. Mol. Biol. 12:626–627.
    CrossRefMedlineISI
  23. ↵
    1. Swisher J.,
    2. Duarte C. M.,
    3. Su L. J.,
    4. Pyle A. M.
    (2001) EMBO J. 20:2051–2061.
    CrossRefMedlineISI
  24. ↵
    1. Matsuura M.,
    2. Saldanha R.,
    3. Ma H.,
    4. Wank H.,
    5. Yang J.,
    6. Mohr G.,
    7. Cavanagh S.,
    8. Dunny G. M.,
    9. Belfort M.,
    10. Lambowitz A. M.
    (1997) Genes Dev. 11:2910–2924.
    Abstract/FREE Full Text
  25. ↵
    1. Mills D. A.,
    2. McKay L. L.,
    3. Dunny G. M.
    (1996) J. Bacteriol. 178:3531–3538.
    Abstract/FREE Full Text
« Previous | Next Article »Table of Contents

This Article

  1. Published online before print June 19, 2006, doi: 10.1073/pnas.0603956103 PNAS June 27, 2006 vol. 103 no. 26 9838-9843
  1. AbstractFree
  2. Figures Only
  3. » Full Text
  4. Full Text (PDF)

Classifications

Navigate This Article