Evaluating target silencing by short hairpin RNA mediated by the group I intron in cultured mammalian cells

Abstract

Background: The group I intron, a ribozyme that catalyzes its own splicing reactions in the absence of proteins in vitro, is a potential target for rational engineering and attracted our interest due to its potential utility in gene repair using trans-splicing. However, the ribozyme activity of a group I intron appears to be facilitated by RNA chaperones in vivo; therefore, the efficiency of self-splicing could be dependent on the structure around the insert site or the length of the sequence to be inserted. To better understand how ribozyme activity could be modulated in cultured mammalian cells, a group I intron was inserted into a short hairpin RNA (shRNA), and silencing of a reporter gene by the shRNA was estimated to reflect self-splicing activity in vivo. In addition, we appended a theophylline-binding aptamer to the ribozyme to investigate any potential effects caused by a trans-effector.

Results: shRNA-expression vectors in which the loop region of the shRNA was interrupted by an intron were constructed to target firefly luciferase mRNA. There was no remarkable toxicity of the shRNA-expression vectors in Cos cells, and the decrease in luciferase activity was measured as an index of the ribozyme splicing activity. In contrast, the expression of the shRNA through intron splicing was completely abolished in 293T cells, although the silencing induced by the shRNA-expressing vector alone was no different from that in the Cos cells. The splicing efficiency of the aptamer-appended intron also had implications for the potential of trans-factors to differentially promote self-splicing among cultured mammalian cells.

Conclusions: Silencing by shRNAs interrupted by a group I intron could be used to monitor self-splicing activity in cultured mammalian cells, and the efficiency of self-splicing appears to be affected by cell-type specific factors, demonstrating the potential effectiveness of a trans-effector.

Figure 1

shRNA directed against firefly luciferase and interruption of shRNA by an intron. (A) Schematic representation of the formation of the stem-loop structure through the self-splicing of a group I intron inserted into the loop site. (B) Expression vectors were constructed for shRNAs targeting the "155-173" or "851-875" region of firefly luciferase. The shRNA cassettes were placed under control of the CMV promoter of the pRNA-CMV3.1-Neo vector. The loop portion of the shRNA was from the loop region of hsa-mir-371 with a slight modification. The loop sequences are underlined. The inserted site of the group I intron is illustrated, and the modified nucleotides are shown in red. (C) Schematic representation of the interactions involved in self-splicing including the pairing between the intron and the 5' and 3' exon sequences. These pairings, namely the P1 helix (5'exon-intron pairing) and P10 helix (intron-3'exon pairing), are shown in bold, and the modified nucleotides are shown in red.

Figure 2

Characterization of shRNA and shRNA interrupted by an intron in cultured cells. (A) The efficiency of the shRNA was analyzed by transiently transfecting Cos cells with vectors expressing firefly luciferase and Renilla luciferase. Firefly and Renilla luciferase activity was analyzed 48 h and 60 h after transfection using the Dual-luciferase Reporter Assay System in which firefly luciferase activity is normalized to Renilla luciferase activity. The pRNA-CMV3.1-Neo empty vector was used as a control, and the results are expressed as the mean ± S.D. of the percentage of control. (B) Silencing activity of the shRNA1-intron and shRNA2-intron in Cos cells 48 h and 60 h after transfection. (C) Cell viability was determined microscopically by trypan blue exclusion 48 h and 60 h after transfection with the pRNA-CMV3.1-Neo, shRNA1, shRNA1-intron, shRNA2 or shRNA2-intron vector. The total cell number and the viability were normalized by the values for cells transfected with the control vector and are expressed as the mean ± S.D. of the percent of control. (D) The efficiency of the shRNA and shRNA interrupted by the intron was also analyzed by transiently transfecting 293T cells and using the Dual-luciferase Reporter Assay System.

Figure 3

Characterization of the processing of the shRNA-intron in Cos and 293T cells. (A) The predicted processing of the shRNA-intron in vivo as well as the primers and a stem-loop primer are schematically illustrated. Primer 1 detected the unspliced mRNA and a spliced out intron, and primer 2 detected the siRNA. (B) The presence of siRNA and the intron was examined by end-point RT-PCR 48 h after the transfection of Cos cells (C) and 293T cells (T) with vectors expressing shRNA1 or shRNA1-intron. The empty pRNA-CMV3.1-Neo vector was used as a control. (C) The siRNA levels detected in cells transfected with the shRNA2 and shRNA2-intron vectors. (D) The levels of siRNA compared to those of the intron produced from the shRNA1-intron or shRNA2-intron vector in Cos cells and 293T cells were analyzed by RT-qPCR (each normalized to G3PDH mRNA). The normalized values of the siRNA/intron levels were set to 100 in Cos cells.

Figure 4

Sequence-specific silencing effect of the shRNA-intron. (A) The firefly luciferase sequences in the pGL3-Control vector that correspond to the targeting regions of 155-173 and 851-875 have a slightly different sequence composition than those in the psiCHECK-2 vector. The sequences from the pGL3-Control plasmid were synthesized and cloned downstream from the stop codon of the Renilla luciferase gene in the psiCHECK-2 vector. (B) The sequence-specific repression by shRNA interrupted by an intron was analyzed by transient cotransfection into Cos cells with the GL155 or GL851 vector. The empty pRNA-CMV3.1-Neo vector was used as a control. Luciferase activity was assessed using the Dual-luciferase Reporter Assay System, in which Renilla luciferase activity was normalized to firefly luciferase activity, and the data are shown as the mean ± S.D. of the percentage of the value for the empty psiCHECK-2 vector.

Figure 5

Characterization of the shRNA-intron fused to a theophylline-binding aptamer. The theophylline-binding aptamer is shown on the left. A schematic representation of the shRNA1-intron construct is shown on the right. The shRNA1-intron-Apt was constructed by substituting the region of the P6 stem, highlighted by a gray-shaded ellipse with a broken line, with the theophylline-binding aptamer in an analogous fashion to the construct "Th2P6" of Thompson, et al. (2002). In the diagram, straight lines are used to connect the three structural domains (labeled P4-P6, P1-P2 and P3-P9). P stands for the paired region. The arrowheads on the lines indicate 5' to 3' polarity. The dotted lines indicate tertiary interactions to help in correct folding. (B) Cotransfection of the shRNA1-intron-Apt into Cos cells along with vectors expressing firefly luciferase and Renilla luciferase was performed. The medium was removed and exchanged for fresh medium with or without 10 mM theophylline 24 h after transfection. Thereafter, the plates were incubated for 6 h and medium was replaced with normal, fresh medium. Luciferase activity was analyzed 36 h after transfection. (C) Caffeine treatment at a concentration of 10 mM was performed in the same way as theophylline. The results are expressed as the mean ± S.E. of the percentage of the value for the control vector, with significance determined by t tests shown by *P < 0.05.