• In Silico Optimization of gRNA Design for CCND1 Gene Silencing in Breast Cancer Stem Cells Using CRISPRoff and Structural Bioinformatics
  • Mohammad Mehdi Sadehsani,1,* Nasim Nasirinejad,2 Arefeh Kalantari,3 Farimah Haji Habib Yazdi,4 Farima Hosseini,5 Yasamin Shabani,6
    1. Department of Cellular and Molecular Biology, Faculty of Basic Science, Sari Branch, Islamic Azad University, Sari, Iran
    2. Department of Medical Research, Alimohammadi Research Institute, Sari, Iran
    3. Department of Cellular and Molecular Biology, Faculty of Basic Science, Sari Branch, Islamic Azad University, Sari, Iran
    4. Department of Genetics , Faculty of Advanced Science and Technology, Tehran Medical Sciences, Islamic Azad University, Tehran, Iran
    5. Department of Cellular and Molecular Biology, Faculty of Basic Science, Sari Branch, Islamic Azad University, Sari, Iran
    6. Department of Genetics , Faculty of Advanced Science and Technology, Tehran Medical Sciences, Islamic Azad University, Tehran, Iran


  • Introduction: Cyclin D1 (CCND1) is the major contributor to malfunction of the cell cycle and is typically co-upregulated in breast cancer stem cells (BCSCs) that have later been involved in tumor progression, therapy resistance, and relapse. Therefore, targeted repression of CCND1 is a powerful strategy to stop these cells from reproducing. The study here is about the development of a completely new bioinformatics pipeline to designing and evaluating guide RNAs (gRNAs) with high clonogenicity for CRISPRoff-mediated epigenetic silencing of CCND1 activity. By utilizing the CHOPCHOP and CRISPRoff web tools, we extracted the top-score gRNAs in the human GRCh38 genome and recognized one lead (GGTTGGCATCGGGTACGCGCGG) with a very high predicted target efficiency (CRISPRon = 73.144) and specificity (CRISPRspec = 9.673). Primers that border the region were designed and checked in silico to guarantee the right amplification whilst the CRISPRoff v2.1 plasmid was chosen for easily targeted delivery. A genome-wide off-target study found 15 nearest possible sites, all with lower risk scores than the target site. These outcomes prove that the method of computationally guided design of highly selective and efficient repression of CCND1 at the locus without occurrence of double-strand DNA breaks is feasible, which paves a way for epigenetic therapy of breast cancer stem cells.
  • Methods: 2. Materials and Methods 2.1 Selection of Target Gene CCND1 (Cyclin D1) was chosen as the target gene because of its critical role in G1–S cell cycle progression and its frequent overexpression in breast cancer and presumably BCSC populations. Cyclin D1 dysregulation results in uncontrolled proliferation, and its modulation is therefore a sensible therapeutic target for controlling BCSC-driven tumor initiation and progression. We examined the CCND1 locus, genomic coordinates and transcript isoforms in GRCh38 reference assembly prior to gRNA design, to ensure targeting of the promoter and regulatory regions. 2.2 gRNA Design Process Guide RNAs were designed by utilizing the web server CHOPCHOP (GRCh38, Homo sapiens) as a target library with the focus on CRISPRoff targeting and thus, proper parameters related to CRISPRoff targeting. The CHOPCHOP server outputs candidate spacer sequence options ranked by predicted on-target activity and predicted likelihood of off-target activity across the genome (Li et al., 2018). From the complete CHOPCHOP output, we selected the top ten ranked gRNA candidates of interest for further in-silico examination based on predicted activity, compatibility with PAM sequences, proximity to important promoter/enhancer elements of CCND1, and to avoid SNPs or other repetitive sequence components that may reduce specificity. Candidate sequences were also examined for GC content, predicted secondary structure, and potential regulatory-associated motif overlap. 2.3 Final gRNA Selection and CRISPRoff Analysis Shortlisted gRNAs underwent all of the proposed analyses using the CRISPRoff web server, which provides prediction of on-target silencing potential and genome-wide off-target distribution for epigenetic editing. The guide with the highest rank from the CRISPRoff pipeline was GGTTGGCATCGGGTACGCGCGG, based on its favorable measures: a CRISPRon (predicted silencing) score of 73.144 and a CRISPRspec specificity score of 9.673 (High). The CRISPRoff analysis showed considerable on-target binding to chr11:69641354–69641377 (CCND1 locus), with few predicted high-risk off-target sites, based on chosen design thresholds. The in-silico data provided rationale for primer design and downstream experimental design considerations. 2.4 Primer Design and Validation To permit PCR amplification and downstream validation of on-target activity, flanking primers were designed around the binding site of the selected gRNA. In designing primers, following primer design best practices: primer length of 18–24 nt, balanced GC content (40-60%), minimal self-complementarity, and estimated melting temperature (Tm) of 60-62 °C. Five primer pairs producing amplicons of ~282–284 bp were downloaded from the GRCh38 whole genome assembly, and were tested in silico to confirm uniqueness, and to exclude predicted off-target amplification. Secondary structure and dimer potential were assessed from standard primer design tools; all primer pairs were found to have reasonable thermodynamic stability levels, and little off-target binding in silico. 2.5 Plasmid Selection and Construct Strategy After reviewing numerous available vectors, the CRISPRoff v2.1 plasmid (Addgene) selection provided the ideal components for the delivery of CRISPRoff components into the target cells. The CRISPRoff v2.1 plasmid has dCas9-KRAB-Dnmt fusion components or modular cassettes for programmable DNA methylation and histone modifications, but no nuclease activity. Also to be more compatible with mammalian expression systems used to transfect the CRISPRoff effectors, the plasmid gut was chosen primarily for compatibility with mammalian expression systems because of its selectable markers that could be followed up in assays. In the gRNA expression cassette, if required, it is planned to clone into the suitable guide scaffold with this back to allow for the co-expression of the targeting sgRNA and epigenetic effector complex.
  • Results: Off-target candidates table for the selected CCND1 guide RNA (spacer = GGTTGGGATCGGGGTACGCGCG, on-target = chr11:69641354–69641377). The columns represent (i) TargetSeq – genomic sequence aligned to the guide; (ii) Mismatches – number of nucleotide differences relative to the guide(0-6); (iii) CRISPRoff – predicted Cas9/epigenetic activity score (tool output; the lower the score the lower the predicted activity); (iv) Level - qualitative risk label denoted by the pipeline (ONTARGET vs CRITICAL); (v) Coordinates – GRCh38 genomic position; and (vi) Overlaps – nearest gene or genomic feature overlapping the candidate site. The intended on-target site is shown as having 0 mismatches, and the highest CRISPRoff score (32.48). Candidate off-targets have 3-6 mismatches and lower CRISPRoff scores (13.51-18.72). Off-targets that overlap annotated genes (e.g., ACTL8, TTC6, SGSM1, LINC02109) are highlighted to indicate a higher priority for follow up, because unintended editing at these sites could have functional significance. Overall, computational predictions are meant to help prioritize targeted experimental validation (amplicon deep sequencing, GUIDE-seq/CIRCLE-seq, or targeted PCR), not to be misconstrued as definitive evidence of off-target editing. 3.1 gRNA Candidates with Efficiency and Off-Target Scores From CHOPCHOP, we established a ranked list of candidate gRNAs targeting regulatory regions surrounding CCND1. We conducted CRISPRoff analysis on the top ten candidates to assess predicted silencing efficiency (CRISPRon) and genome-wide specificity (CRISPRspec). Table 1 (to be added) shows these candidates, spacer sequences, genomic coordinates, CRISPRon scores, CRISPRspec values, predicted off-target loci, and notes regarding overlap with regulatory elements. Of these, GGTTGGCATCGGGTACGCGCGG had the best CRISPRon score (73.144) and CRISPRspec (9.673) suggesting the highest likelihood for effective and specific epigenetic silencing at the CCND1 locus. 3.2 Final gRNA Silencing Prediction The locus centric predictions from CRISPRoff suggest appropriate on-target engagement for the selected guide, and no predicted high-risk off-targets in coding or regulatory regions likely to produce deleterious effects were found in the analysis using conservative thresholds. The predicted binding coordinate (chr11:69641354–69641377) coincides with the promoter/enhancer context of CCND1 suggesting a plausible mechanistic pathway for promoter methylation and deposition of repressive histone marks. These computational predictions provided sufficient confidence to move forward with the guide for cloning and cellular testing. 3.3 Primer Validation The five designed primer pairs specifically for amplification of the targeted CCND1 region were checked in silico (all primers had a single predicted product of 282-284 bp when aligned against GRCh38, and there were no off-target matches under standard stringent). Table 2 (to be inserted) contains sequences, Tm, GC content, amplicon size and in-silico specificity metrics. The determination of a positive identification of locus function and assays like bisulfite sequencing (for DNA methylation) or amplicon-based quantification of an expression change will utilize effective targets utilizing these primer pairs. 3.4 Plasmid Integration and Construct Preparedness CRISPRoff v2.1 emerged as an ideal expression backbone containing the necessary epigenetic effector components. The expression plasmid enables expression of the dCas9-KRAB-Dnmt fusion which can also be designed to contain the chosen gRNA cassette. This will reduce the need for extensive cloning of effector modules and will allow for a more rapid transition to cellular assays. The planned downstream validation experiments will involve transient transfection into BCSC-enriched cultures, followed by qPCR, western blot for Cyclin D1, methylation PCR or bisulfite sequencing around the target region and functional readouts like proliferation and sphere formation.
  • Conclusion: Reviewed through the lens of a multi-disciplinary approach, this report attempted to enhance gRNA design to silence the CCND1 gene in breast cancer stem cells with the aid of CRISPRoff and structural bioinformatics. This article presented a deliberately curated library of guide RNAs that were intended to knock down the target gene; a library that was designed to be specific. The best candidate showed high predicted efficiency and specificity using both CRISPRoff and other scoring systems. The proposed pipeline sets forth rational principles that can be used to engineer a fit-for-purpose CRISPRoff approach specific to genomic targets, and demonstrates possible applications for precision medicine. The roadmap put forth in this workshop report is not a destination, but rather a conceptual and practical starting point that can be honed iteratively through experimental validation to inform future therapeutic pursuits and support patients in their therapeutic journey.
  • Keywords: CCND1, BCSCs, CRISPRoff, CHOPCHOP, gRNA Design