-
Inducible transcriptional regulation system enables precise spatiotemporal control over transcriptional activities [1, 2]. Clustered regularly interspaced short palindromic repeats (CRISPR)-associated protein 9 (Cas9) has been an important tool in the system [3-5]. CRISPR-TRs rely on interaction between a catalytically deficient mutant Cas9 (dCas9) and target DNA under the guidance of sgRNA [6, 7]. CRISPR-TRs mostly focus on dCas9 protein engineering and blocking design of the spacer region on the sgRNA. For dCas9 protein engineering, the main method is to control the connection of the fusion protein on dCas9 and fusion protein on effectors (VP64, VPR, P65, and so on) using light [8], small molecules [9, 10] and dimerizing proteins [11]; for blocking design of the spacer region on the sgRNA, the most common approach is to form a hairpin structure in the spacer region of the sgRNA, which utilizes RNA strands [12], mRNA strands [13], aptamers [14], and proteins with exonuclease activity [15] to achieve unshielding of sgRNA activity. Spacer-sequence design of sgRNA gives CRISPR-TRs more flexibility and utility relative to protein engineering. However, blocking the design of the spacer suffers from sequence dependence and does not apply well to endogenous molecules. Therefore, it is worth exploring how to modify sgRNA to overcome sequence dependence for achieving responses to more endogeous molecules.
To solve the above problem, a recent effort has designed a microRNA (miRNA) inducible CRISPR-TR for cell-specific genome regulation, using the RNA-induced silencing complexes (RISC) for sgRNA activation [16]. This design can target miRNAs of arbitrary sequence, but cannot be used in response to other small molecules. In addition, TMSD’s high specificity and programmable Watson-Crick base pairing have been used to design conditional response systems [17]. Based on the riboswitch and RNA interference mechanism in Escherichia coli, a series of synthetic RNA regulatory elements that can control the transcriptional and translational reactions of homologous RNAs have been developed through TMSD reaction [18-26]. Besides, recently, several studies have also allowed conditional activation of CRISPR function through the direct strand displacement of rationally designed gRNA [27-30]. The assembly of two-component gRNA can be induced by RNA triggering [31] and RNA replacement [32], enabling the control of CRISPR/Cas9 in Escherichia coli. But these elements are rarely applied to mammalian cells.
Here, we reported two approaches for constructing sequence-independent CRISPR-TRs responding to endogenous and exogenous molecules in mammalian cells. The first method was a split method, where the main effector unit was designed separately from dCas9, and bRNA was utilized to achieve the connection between the effector unit and dCas9. By modifying the complementary regions and corresponding structures of bRNA and sgRNA, we could realize TMSD reaction in response to miRNA and mRNA. In second method, we designed the complementary sequence of stem loop 2 on the 3' end or 5' end of sgRNA to block the activity of sgRNA. This inactivated sgRNA prevented dCas9 protein from targeting a designed tandem mini-CMV promoter, thereby interrupting the transcriptional circuit responsible for reporter RNA expression. The sgRNA activity was released by exonuclease cleavage (Cas6A), allowing amplification of the signal. These results indicated that the sgRNA-based design for implementing sequence-independent CRISPR-TRs enables both endogenous and exogenous responses. This design endows Cas9 with the ability to generate responsive behaviors to a wide range of signals, thereby expanding the applicability of flexible and controllable genetic engineering tools.
-
Enzymes T5 exonuclease, Phusion DNA polymerase, Taq DNA ligase, T4 DNA ligase, T4 polynucleotide kinase, NheI-HF, BamHI-HF were obtained from New England Biolabs; PrimeSTAR MAX DNA polymerase was purchased from Takara Bio; GeneJET Gel Extraction Kit and GeneJET Plasmid Miniprep Kit were purchased from Thermo Fisher Scientific. DNA oligonucleotides were obtained from Sangon Biotech. The RNA sequences were first optimized and designed by NUPACK software to ensure that no secondary structure exists.
-
Plasmid dCas9 fusions and effector fusions were cloned into plasmid SP-dCas9-VPR (Addgene #63798) digested with SacI and BsrGI (NEB) using the Gibson assembly method. Effector domains VP64 and p65-HSF1 were amplified from pJZC34 (Addgene #62331) and lenti MS2-p65-HSF1_Hygro (Addgene #61426), respectively. All guide RNAs and bRNAs were cloned into M-SP-sgRNA (Addgene #48671) with two BsaI restriction sites digested with BsaI (NEB) using the golden gate assembly method. Sequences for the sgRNAs and bRNAs are listed in Table S1 and Table S2 (Supplementary materials, SM), respectively and sequences for the BRS-sgRNAs and SRB-bRNAs are listed in Table S3 (SM). The reporter plasmid with a dCas9 binding site upstream of a minimal CMV promoter was purchased from addgene (#47320). The binding sequence is AGAATTCGCCCTTGCGACGT. The sequences for Cas6A protein are listed in Table S4 (SM).
-
HEK293T cells (ATCC) were maintained in high-glucose Dulbecco’ s modified Eagle medium (Gibco) supplemented with 10% FBS (Gibco). The cells were maintained at 37 °C and 5% CO2 and not routinely tested for mycoplasma contamination. 120000 HEK293T cells were seeded into each well of 24-well plates 24 h before transfection. If not specially stated otherwise, in dCas9-VPR system, 600 ng dCas9-VPR, 80 ng bRNA, 80 ng sgRNA, and 50 ng reporter plasmid were delivered to each well by Lipofectamine 2000 (Thermo Fisher Scientific) according to the manuals. In dCas9-VP64 system, 355 ng dCas9-VP64, 65 ng MS2-p65-HSF1, 80 ng bRNA, 80 ng sgRNA, and 50 ng reporter plasmid were delivered to each well by Lipofectamine 2000 according to the manuals. 70 ng Cas6A plasmid or control plasmid was for responding to Cas6A protein. 80 ng BRS-sgRNA or SRB-sgRNA was used for the blocking of sgRNA. The cells were assayed for 48–72 h after transfection using flow cytometry (BD Accuri C6, BD Biosciences). For detection of endogeous molecues and exodogeous molecues, the transfection reagent is jetPRIME, transfection reagent volume to plasmid mass ratio is 2:1.
-
Inverted microscope IX71 (Olympus) was used to assess cell growth and dTomato expression.
-
We used synergistic activation mediator (SAM) [33] structure to amplify the output signal. The proteins p65-HSF1 and VP64 were split, with VP64 being fused to dCas9, and p65-HSF1 being fused MCP protein to bind bRNA through MCP protein-RNA aptamer interaction. The crystal structure of Cas9/sgRNA/target DNA revealed that the tetraloop, stem loop 2 and 3' end of sgRNA protrude outside of the Cas9-sgRNA ribonucleoprotein complex [34]. We set the corresponding complementary sequences (s') of bRNA on the sgRNA, which were put on tetra loop (sgRNA 1.1), stem loop 2 (sgRNA 1.2), tetra loop and stem loop 2 (sgRNA 2.0), and 3' end (sgRNA 1.3) respectively, whose positions can set arbitrary sequence without interfering with the function of sgRNA (FIG. 1(a)).
We attempted bRNA on three systems, dCas9-VP64, dCas9-VPR and dCas9. bRNA was the bridge of combining three systems and MCP-p65-HSF1 unit. For the three systems without bRNA, adding the s' to the tetra loop or 3' end of sgRNA resulted in a little signal suppression, while adding the s' to stem loop 2 of sgRNA resulted in a little signal increase. Meanwhile, for the three systems with bRNA, when we added the s' to stem loop 2 of sgRNA, the enhancements of output signals were 15.56-fold in dCas9-VP64 system, 40.04-fold in dCas9-VPR system and 1.75-fold in dCas9 system, respectively (FIG. 1(b–d)). These results indicated when the s' of the bRNA was put on stem-loop 2 of sgRNA, signal amplification was the optimal with the dTomato as output signal.
We ligated MCP-p65-HSF1 to sgRNA 1.2 via bRNA, and at this point we replaced dCas9-VP64 with dCas9, named the dCas9+p65-HSF1 system. We then compared dCas9+p65-HSF1 with the dCas9-VP64 system and the dCas9-VP64+p65-HSF1 system, and these three systems were relatively more effective than the dCas9 system by 5-, 1.56- and 35.8-fold, respectively, suggesting a synergistic transcriptional increase of VP64 and p65-HSF1 (FIG. 2).
To achieve TMSD, we divided the bRNA into three parts: toehold region, hairpin region and MCP aptamer region (MS2), and we put the toehold region in the middle region or the end of the bRNA for comparison (the secondary structures were shown in FIG. S1 (SM)). When we placed the toehold region into the middle region of the bRNA and the lengths of the toehold were set to 0, 6, 8, and 10 nt, we found that the signal did not increase, and at this point all signal amplifications were lower than 1.1-fold; when we placed the toehold region into the end of the bRNA and the lengths of the toehold region were 8 and 10 nt, the signal amplifications were 1.51- and 2.61-fold, respectively. We extended the lengths of the toehold region at the end of the bRNA to 12 and 14 nt, and found that the signal amplification increased significantly as the length of the toehold region increased, with a 10.46-fold signal amplification for 12-nt toehold region and a 19.30-fold signal amplification for 14-nt toehold region (FIG. 3(a)). These results indicated that the signal could be amplified when the toehold region was located at the end of the bRNA and the length of the toehold region was greater than 8 nt, whereas the signal was suppressed when the toehold region was located in the middle region of bRNA. So we have initially succeeded in setting up the CRISPR-TRs with TMSD system.
-
MiRNA typically inhibits the expression of mRNA target gene at the post-transcriptional level via partially Watson-Crick base pairing with the target gene mRNA in mammalian cells [35]. Many miRNAs are expressed in specific tissues and cell types, particular cancer cells, thereby making them critical biomarkers in numerous diseases [36]. MiR-20a can serve as a marker for colorectal cancer [37], with high expression in HEK293T cells [38]. Using the above TMSD system, we detected miR-20a on HEK293T cells, so we established two types of bRNAs. The first one was to set the toehold at the end of bRNA with 6 and 8 nt, and the signal amplifications were 3.55- and 2.37-fold, respectively. The second one was to set the toehold on the ring of the stem loop structure with 6, 8, and 10 nt, and signal amplifications were 4.28-, 1.95- and 1.73-fold, respectively (FIG. S2 in SM and FIG. 3(b)). Two structural designs showed the best signal amplification when the length of the toehold was 6 nt, which may be related to the seed region of miRNA, which is generally 7 or 8 consecutive bases in the 5' end of miRNA [39]. We could provide a platform to detect miRNA by fluorescence values alone.
Next, in response to endogenous mRNA, we utilized highly expressed endogenous TK1 (thymidne kinase 1) mRNA as a response molecule for signal amplification. We modified the sequence of bRNA based on the 30-nt region which was stably present in TK1 mRNA, and the lengths of the toehold were set to 12 and 14 nt to achieve 30-bp complementary pairing, whose control groups had a 1-nt mismatch in the hairpin region of the bRNA (the secondary structures are shown in FIG. S3 in SM). According to FIG. 3(b), when the lengths of the toehold were set to 12 and 14 nt, the signal amplifications of the experimental groups were 2.33- and 2.12-fold relative to the control groups, respectively. In the fluorescence images, the responses of different bRNAs to miR-20a and TK1 mRNA could be clearly shown, producing fluorescence of different brightness (FIG. S4 and FIG. S5 in SM). Therefore, we could detect small molecules and visualize the endogenous environment through TMSD system.
-
To achieve the response to exogenous molecules, we first explored the effect of the complementary sequence of stem loop 2 on transcriptional activation, whether we reduced the bases (from 22 nt to 19 nt) or placed the complementary sequence of the bRNA at the 3' end of bRNA, which had high transcriptional effect on dCas9-VP64+p65-HSF1 system and dCas9-VPR+p65-HSF1 system (FIG. 4(a)). Then, we chose two random 23-nt sequences as the response sequences, and added the bRNA sequences and the response sequences on the 3' or 5' end of the sgRNA to form an interlocking structure by utilizing the bRNA’s complementarity with the sequences of stem loop 2 on the sgRNA 1.2, called SRB-sgRNA and BRS-sgRNA, respectively (the secondary structures were shown in FIG. S6 and FIG. S7 (SM)). We found that the interlocking of the 3' end of sgRNA had no inhibition of transcriptional activation, and the interlocking of the 5' end of sgRNA had a great inhibitory effect on transcriptional activation, we obtained similar results in dCas9-VP64 system and dCas9-VPR system (FIG. 4(b)). We thought that the interlocking of 5' end of sgRNA could achieve effective blocking activity of sgRNA, and we added the corresponding sequence to the 5' end of sgRNA to achieve unblocking of sgRNA by shearing the sequence, called SUI. To compare with SBH, SBH structure is realized by adding the response sequence and the complementary sequence of the spacer region of the sgRNA at the 5' end of the sgRNA (FIG. 4(c), the secondary structures are shown in FIG. S8 in SM). So we added the response sequence which is sheared by exogenous molecule (Cas6A protein) to the two inducible structures (SUI and SBH). We found SUI structure shielded the sgRNA activity more strongly and resulted in a low transcriptional signal relative to SBH for responding to the Cas6A protein (48.24-fold for SUI system vs. 64.09-fold for SBH system, FIG. S9 in SM). However, the overall fold transcriptional activation ratio between response and non-response to Cas6A protein in SUI system was higher than that in SBH system (49.00-fold vs. 21.50-fold) (FIG. 4(d)).
-
In summary, CRISPR-TRs have transformed the current synthetic biology landscape by allowing specific activation or repression of any target gene. Sequence-independent CRISPR-TRs give them more flexibility and utility in response to endogenous and exogenous molecule.
In the study, our strategy relied on the sequence-independent design of stem loop 2 on sgRNA. With dTomato fluorescent protein as the output signal, we added bRNA and achieved a synergistic transcriptional increase of VP64 and P65-HSF1 by utilizing endogenous small molecules (miR-20a and TK1 mRNA) to achieve TMSD, and thus detecting the endogenous environment. When the toehold region was located at the end of the bRNA and the length of the toehold region was greater than 8 nt, the signal could be amplified. Two structural designs in response to miR-20a showed the best signal amplification when the length of the toehold was 6 nt. For responding to TK1 mRNA, when the lengths of the toehold were set to 12 and 14 nt, the signal amplifications of the experimental groups were 2.33- and 2.12-fold relative to the control groups, respectively. Then, we designed s' of stem loop 2 on the 3' end or 5' end of sgRNA to block activity of sgRNA. We found the interlocking of the 5' end of sgRNA tremendously inhibited transcriptional activation, and adding the response sequence between the sequence of bRNA and the sequence of sgRNA to respond a exogenous molecule (Cas6A protein), the transcriptional activation ratio between response to and non-response to Cas6A protein was higher than SBH (49.00-fold vs. 21.50-fold). Therefore, we offered a modular and versatile framework for enabling endogenous and exogenous molecule responding CRISPR-TRs in mammalian cells without sequence limitation.
Supplementary materials: FIG. S1–FIG. S9 show the secondary structures of all RNAs, representative images in CRISPR-TRs and flow cytometry scatter plots in iSBH and iSUI systems. All sgRNAs, bRNAs, BRS-gRNAs, SRB-gRNAs and Cas6A sequences are shown.
Sequence-Independent CRISPR-Based Transcription Regulators Responding Endogenous and Exogenous Molecules
- Received Date: 14/12/2023
- Available Online: 27/08/2025
-
Key words:
- Inducible transcriptional regulation /
- dCas9 /
- Endogenous molecule /
- Exogenous molecule
Abstract: CRISPR-based transcription regulators (CRISPR-TRs) have revolutionized the field of synthetic biology by enabling targeted activation or repression of any desired gene. However, the majority of existing inducible CRISPR-TRs are limited by their dependence on specific sequences, which restricts their flexibility and controllability in genetic engineering applications. In this study, we proposed a novel strategy to construct sequence-independent inducible CRISPR-TRs, which is achieved by the design of stem loop 2 in the single guide RNA (sgRNA). Under this strategy, by utilizing toehold-mediated strand displacement (TMSD) reactions between small endogenous molecules (miR-20a and TK1 mRNA) and bridge RNA (bRNA) to link bRNA with sgRNA, we achieved synergistic transcriptional activation of VP64 and p65-HSF1 in response to endogenous molecules. To enable response to exogenous molecules, we added response sequences and bRNA sequences to the 5' end of sgRNA to block sgRNA activity, and achieved activation of sgRNA by shearing the response sequence, called sequential unlimited interlocking (SUI). Compared with conventional sequence-restricted interlocking (spacer-blocking hairpin (SBH)), the transcriptional activation ratio between response and non-response to the Cas6A protein using our approach was increased by 2.28-fold. Our work presents a modular and versatile framework for endogenous and exogenous molecule-responsive CRISPR-TRs in mammalian cells, without limitations imposed by sequence dependence.
首页
登录
注册


DownLoad: