Skip to main content

Identifying the Capabilities and Limitations of CRISPR in Targeting G-quadruplex Forming Sequences: From Target Recognition to Gene Expression Regulation

Periodic Reporting for period 1 - CRISPR-GQ (Identifying the Capabilities and Limitations of CRISPR in Targeting G-quadruplex Forming Sequences: From Target Recognition to Gene Expression Regulation)

Reporting period: 2018-07-01 to 2019-06-30

Despite being a bacterial immune response against phage attacks, Clustered Regularly Interspaced Palindromic Repeats (CRISPR) and associated proteins (Cas) have provided unprecedented control on genome editing at both cellular and organismal levels. Due to its simple targeting mechanism and high specificity, Cas9 has been widely utilized. Cas9 uses CRISPR-RNA (crRNA) as a guide to target a specific double stranded DNA (dsDNA), which has a complementary sequence to crRNA in one of its strands. When the target sequence is encountered, Cas9 binds to it and a RNA-DNA hybrid (R-loop) between crRNA and target DNA is created. This stable complex formation is followed by cleavage of both strands of the target DNA by Cas9, which creates a double-stranded break and provides a venue to modifying the sequence via homologous recombination.

The main objective of this proposal is to investigate the impact of secondary structure formation in the target DNA on various aspects of Cas9 activity, such as target recognition, binding, stable complex formation, and cleavage efficiency. G-quadruplex (GQ) structures have been used as model secondary structures in this study. GQ structures (GQs) form in guanine-rich segments of the genome and are particularly concentrated in telomeres and promoters. Telomeric GQs stabilize chromosome ends and inhibit telomerase activity, which makes them significant for cancer research. On the other hand, non-telomeric GQs that form in promoters have been shown to modulate transcription level gene expression. GQs can be very stable and often require protein activity to be unfolded. Therefore, targeting GQ forming sequences with CRISPR-Cas9 has implications for genomic stability and gene expression regulation.
The two strands of the target dsDNA differ in terms of their role in this study. The target strand (TS) is complementary to crRNA and together they form the R-loop. Any GQ that might form in TS needs to be destabilized for stable complex formation. On the other hand, the non-target strand (NTS) does not directly interact with crRNA, and should be largely uninhibited for GQ formation after crRNA hybridizes with TS. Formation or destabilization of GQs depends on their thermal stability. Destabilization of a high-stability GQ in TS would be very challenging while such a GQ would readily form in NTS after crRNA hybridizes with TS.
Three different GQ forming sequences with low, medium, and high thermal stability were inserted within either TS or NTS. GQ formed by thrombin binding aptamer (TBA-GQ) was of lowest stability (thermal melting temperature Tm = 51 C) and has the following 15-nt sequence: GGTTGGTGTGGTTGG. GQ formed by human telomeric sequence (hGQ) was of medium stability (Tm = 69 C), and is formed by a 21-nt long sequence: GGGTTAGGGTTAGGGTTAGGG. The most stable GQ studied (called 3L1L-GQ and has Tm > 95 C) was formed by a 15-nt long sequence: GGGTGGGTGGGTGGG. When in TS, TBA-GQ would be expected to minimally disturb the system while 3L1L-GQ is expected to be very challenging to destabilize by Cas9-crRNA. As significant part of the work on hGQ has been conducted by a graduate student (Ms. Viktorija Globyte), this report will focus on the results on TBA-GQ and 3L1L-GQ.
Single molecule Förster resonance energy transfer (smFRET) was used as the primary tool for this work. SmFRET depends on a distance-dependent energy transfer between a donor and an acceptor fluorophore. By properly placing the donor-acceptor fluorophores the relative dynamics between crRNA and dsDNA could be studied, as well as direct monitoring of GQ formation or destabilization. These studies demonstrated that the CRISPR-Cas9 and target dsDNA complex is in general not static and multiple conformations are sampled during these interactions. In the case of TBA-GQ, evidence for GQ formation is observed when GQ is in NTS but not in TS. In the case of 3L1L-GQ, we noticed significantly less binding events between CRISPR-Cas9 and target dsDNA compared to TBA-GQ case. We also observed significantly less Cas9 loading onto crRNA that contains a 3L1L-GQ compared to that containing TBA-GQ. We also observed significantly more complex dynamics for the case of 3L1L-GQ compared to TBA-GQ, suggesting a less stable and more dynamic complex formation with CRISPR-Cas9.
The results of this work will impact the field in several ways. First, the reduced binding efficiency of CRISPR-Cas9 to target DNA with GQ forming sequence suggests that there are limitations about what type of sequences can be targeted by Cas9. Secondly, binding of CRISPR-Cas9 to TS facilitates secondary structure formation in the non-target strand. If these structures are stable enough, they could induce significant distortions in the resulting complex. We have observed significant dynamics in the complex in both of these scenarios, highlighting the underlying competition between secondary structure formation and Watson-Crick pairing. In addition, if NTS can form GQ, the corresponding crRNA could also form GQ. We showed that crRNA loading to Cas9 becomes less efficient if the crRNA contains a GQ forming sequence. These types of issues with crRNA loading to Cas9 might be avoidable by introducing one or two nucleotide mutations in the distal end of crRNA (away from protospacers adjacent motif-PAM). In general, it is possible to eliminate GQ formation by mutating 1-2 guanines. When such a mutation is made, the complementarity of crRNA with the target strand will be lost for the mutated nucleotide(s); however, this might have minimal impact on binding stability and specificity of CRISPR-Cas9 to target dsDNA if a guanine in the distal region is mutated. The impact of such a mutation on Cas9 cleavage activity would require further investigation.
By using a catalytically inactive Cas9 (dCas9), it should be possible to modulate GQ stability without cleaving the DNA. By targeting either the GQ forming G-rich strand or the complementary cytosine-rich (C-rich) strand, it should be possible to inhibit or promote GQ formation, respectively. This is potentially significant as it could be used to target GQ-forming sequences or their complementary C-rich strands in promoters to modulate gene expression. This method would be more specific compared to the alternative method of targeting such sequences with small molecules that have structural but not sequence specificity.
This initial work will be continued in the form of an international collaboration between my lab in USA and Dr. Joo’s lab in Netherlands. The project was expanded from studying one to three GQ forming sequences with different stabilities. This expansion provided a better view about the range of impact that these GQs could have on CRISPR-Cas9 activity. However, certain aspects of the work could not be completed within the 1-year period. The remaining part of the work will be completed by Balci in USA and V. Globyte from the Joo lab in Netherlands. Balci will seek further funding for this work in USA. Finally, Balci was involved with training of undergraduate and graduate students at Delft. In particular, he served as the daily supervisor (along with Mr. Tao Ju Cui) for a master’s student (Rodrigo G. Linares), which served the two-way exchange goal aimed by the Marie Curie Fellowship.
CRISPR-Cas9 in complex with a double stranded DNA that has a GQ in the non-target strand