Structures of the TNF-specific RNA Lassos ATR1 (a) and ALR 562-2 (b). The top structures correspond to unprocessed lasso transcripts (pre-Lasso), while the bottom structures depict fully processed lassos circularized around the murine TNF-α target RNA. Antisense sequences within the Lassos are shown in red, hairpin ribozyme sequences are in black, linker sequences are in blue, the U25→C25 substitution is in green, and target RNA is purple. The ribozyme cleavage/ligation sites are marked by arrows. Potential base pairing of a triplex strand is indicated by black dots.

Inhibition of in vitro translation of a TNF-luciferase reporter mRNA by Lasso ATR1. (a) Schematic of TNF22-luc. (b) Products of translation of TNF22-Luc mRNA and pro-IL-1 mRNA (negative control) in rabbit reticulocyte lysate in the presence increasing amounts of Lasso ATR1 and ³⁵S-labeled methionine. Lane 1 contains only the TNF22-Luc mRNA RNA, while lanes 2–6 contain both TNF22-Luc and pro-IL-1 mRNAs. Lanes 3–6 additionally contain 20-, 40-, 80- and 160-fold molar excess of ATR1 Lasso with respect to the targets. The translation products were analyzed by 12% SDS–PAGE.

Binding of circular and linear ATR1 forms to TNF RNA target depends on presence of Mg²⁺. (a) Internally ³²P-labeled ATR1 transcripts incubated at 37°C for 2 h in TM buffer containing 10 mM MgCl₂ (lanes 1–3) or in buffer containing 10 mM EDTA (lanes 4–6) either without target (lanes 1 and 4) or in the presence of nonradioactive TNF685 target (lanes 2–3 and 5–6). Samples in lanes 3 and 6 were additionally postincubated at 90°C for 2 min to dissociate the Lasso-target complexes. The products were analyzed by 6% d-PAGE. (b) Schematic of the mechanism of formation of complexes between Lasso (with antisense region shown in red) and target RNA (blue) under −Mg (top) and +Mg (bottom) conditions.

Schematic of self-processing (cleavage and ligation) of the RNA Lasso precursor. Self-processing of a primary Lasso transcript (pre-Lasso) having 5′-triphosphate (5′-PPP) and 3′-hydroxyl (3′-OH) groups can proceed by either the right or left pathway. Self-cleavage catalyzed by the HPR generates specific fragments having 2′,3′-cyclic phosphate (2′,3′ > P) and 5′-hydroxyl (5′-OH) termini. The fully 5′- and 3′-processed RNA Lasso exists as an equilibrium between circular and linear forms. The cleavage/ligation sites are marked by magenta arrows with asterisks. Loops 1, 2 or 3 indicated in the bottom structure can be used for insertion of an antisense sequence. Here, the antisense sequence (red) is inserted into loop 2 (blue). Species are identified as unprocessed Lasso (UPL), 5′-half-processed Lasso (5′-HPL), 3′-half-processed Lasso (3′-HPL), 5′- and 3′-processed ‘linear’ Lasso (LL) and circular Lasso (CL).

Transcription, self-processing, and target binding of Lasso ATR1. (a) Lane 1 shows internally ³²P-labeled products of ATR1 transcription and spontaneous self-processing during the transcription reaction. Lane 2 shows gel-purified LL (linear Lasso), and lane 3 its partial conversion to CL (circular Lasso) through self-ligation after incubation in TM buffer (50 mM Tris–HCl, 10 mM MgCl₂) for 1 h at 37°C. Lane 4 shows gel-purified CL, and lane 5 its partial conversion to LL through self-cleavage after the same incubation as for lanes 2 and 3. Lanes 6–8 compare internally ³²P-labeled ATR1 Lasso species (lane 6, the same as lane 1) with 3′-end-labeled (lane 7) and 5′-end labeled ATR1 (lane 8). Because of efficient Lasso self-processing, the unprocessed Lasso (UPL) and the half-processed Lasso (comigrating 5′-HPL and 3′-HPL) forms are present in relatively small amounts (lanes 1 and 6). (b) Time-course of binding of internally ³²P-labeled TNF685 RNA target with an excess of non-radioactive Lasso ATR1 transcription mixture in TM buffer at 37°C for the indicated time periods. (c) Internally ³²P-labeled ATR1 incubated with or without an excess of nonradioactive TNF685 fragment (as indicated) in TM buffer for 2 h at 37°C. Samples in lanes 3–6 were additionally incubated for 2 min at elevated temperatures as indicated and transferred immediately to ice to prevent rehybridization of the complexes. The products were analyzed by electrophoresis on denaturing 6% polyacrylamide gels (d-PAGE).

Lasso ALR562-2 binding to linear TNF685 and a circular target (CMT). (a) Products of self-processing of internally ³²P-labeled ALR562-2 in transcription buffer containing 6 mM Mg²⁺ (lane 1) and in buffer containing 50 mM Tris–HCl, 10 mM MgCl₂, 20% formamide (lane 2). Lane 3: formation of the strong complex between ALR562-2 (from lane 2) and nonradioactive TNF685 target provided in excess. Lane 4: products of dissociation of the complex from Lane 3 upon addition of 50% formamide and 10 mM EDTA and post-incubation for 2 min at 90°C. The products were analyzed by 6% d-PAGE and bands were assigned to Lasso-target complexes as well as unbound circular (CL), unprocessed (UPL), half-processed (HPL) and linear (LL) Lasso forms as indicated. (b) Demonstration of topological linkage between CMT and Lasso ALR562-2. Internally ³²P-labeled Lasso ALR562-2 was incubated in TM–formamide buffer containing 50 mM Tris–HCl (pH 7.5), 10 mM MgCl₂, 20% formamide (lanes 1–3) or 50 mM Tris–HCl (pH 7.5), 10 mM EDTA, 20% formamide (lanes 4–6) for 2 h at 37°C either alone (lanes 1 and 4) or with the CMT target (lanes 2–3 and 5–6). Samples in lanes 3 and 6 were additionally postincubated at 95°C for 5 min followed by quenching on ice. The products were analyzed by denaturing 6% PAGE and assigned to complexes of CMT with various forms of the Lasso (HPL-CMT, LL-CMT and CL-CMT) as well as unbound circular (CL), unprocessed (UPL), half-processed (5′-HPL and 3′-HPL) and linear (LL) forms as indicated.

RNA Lassos can redirect splicing of Fas pre-mRNA. (a) Target sequences (red) of Lassos Fas1, Fas2 and Fas3 that span the intron 5/exon 6 splice junction. (b) Schematic of intron/exon structure of Fas receptor primary transcript (not to scale). The Lasso target site (intron 5/exon 6 junction) chosen to prevent inclusion of exon 6 in the final mRNA product is shown. The PCR primer site locations for analysis are shown as horizontal arrows. (c) RT–PCR amplification (using the exon 4 and exon 9 primers, black arrows) of products of Fas splicing in the presence of Lassos Fas1, Fas2 or Fas3. No shorter DNA fragment corresponding to Fas lacking exon 6 was seen in the absence of Fas Lassos (ct). β-actin serves as a loading control. Stds is a DNA ladder (100–1000 bp). (d) Confirmation that Fas Lassos lead to exon 6 skipping revealed by PCR analysis using a downstream primer spanning the exon 5 and exon 7 junction (red arrow) along with the exon 4 upstream primer, which are unable to amplify a product that includes exon 6. Lanes 2–4 and 5–7 correspond to independent experiments with anti-Fas Lassos 1, 2 and 3. Control lane 8 corresponds to the RT–PCR products obtained with cells that were not transfected by Lassos (ct), and lane 9 contains no PCR template (n/a).