Engineering
Part I: Simulation and Validation of the activity of CRISPR/Cas13a system
Brief Introduction
The CRISPR/Cas13a system is frequently used in detection and cleavage of oligonucleotide due to its
unique collateral cleavage activity. In recent years, the relationship between microRNA (miRNA) and
cancer progression has gained attention, making it an important biomarker. Our project aims to use the
CRISPR/Cas13a system to detect miR-21-5p, which is a biomarker for breast cancer. Usually, optimizing a
CRISPR/Cas13a system that could specifically and efficiently target a miRNA requires a large amount of
wet lab experiments, which needs a lot of experimental materials and time. To save these experimental
materials and time, we utilized a set of computational approaches to simulate the reaction conditions
and validated the activity of CRISPR/Cas13a system by actual experiments. Noteworthy, this set of
approaches can be readily utilized in other iGEM projects that also use engineered CRISPR/Cas13a for
diagnosis or treatment of diseases.
Design
First, We designed four crRNA sequences corresponding to the sequence of miR-21-5p. Previous studies have
shown that the length of the crRNA guide sequence affects the overall activity of the CRISPR/Cas13a
system. The activity of the CRISPR/Cas13a
system was significantly reduced if the length of the guide
sequence was less than 20 nt.
| Name |
Sequence |
| miR-21-5p |
UAGCUUAUCAGACUGAUGUUGA |
Table 1 miRNA candidates' sequences
| crRNA |
Sequence |
| crRNA full length |
GAUUUAGACUACCCCAAAAACGAAGGGGACUAAAAC
UCAACAUCAGUCUGAUAAGCUA |
| crRNA1 |
GAUUUAGACUACCCCAAAAACGAAGGGGACUAAAAC
AACAUCAGUCUGAUAAGCUA
|
| crRNA2 |
GAUUUAGACUACCCCAAAAACGAAGGGGACUAAAAC
CAACAUCAGUCUGAUAAGCU
|
| crRNA3 |
GAUUUAGACUACCCCAAAAACGAAGGGGACUAAAAC
UCAACAUCAGUCUGAUAAGC |
Table 2 crRNA sequence for miR-21-5p
To validate the stability of these crRNA, we evaluated crRNA in three ways: 1) Minimum Free Energy; 2) GC
content; 3) MFE heterodimers.
| crRNA |
MFE (kcal/mol) |
GC content |
MFE heterodimer (kcal/mol) |
| crRNA full length |
-9.6 |
0.40 |
-42.5 |
| crRNA1 |
-9.8 |
0.39 |
-41.6 |
| crRNA2 |
-9.7 |
0.41 |
-40.9 |
| crRNA3 |
-9.6 |
0.41 |
-39.8 |
Table 3 Simulation parameters for crRNA
Figure 1 Secondary structure of crRNA. (A) crRNA1; (B) crRNA2; (C) crRNA3; (D)
crRNA full length(Green: Stems, Yellow: Interior Loops, Blue: Hairpin loops, Orange: 5' and 3' unpaired
region)
The stability of crRNA can be described using two factors: the Minimum Free Energy (MFE) and the GC
content. A smaller MFE value and a larger GC content value indicate a more stable crRNA. In addition,
the MFE Heterodimer is used to describe the stability of the binary complex formed between crRNA and
miRNA. A smaller MFE Heterodimer value indicates a more stable binary complex formed by the combination
of crRNA and miRNA. The stability of both the crRNA structure and binary complex structure are essential
for the proper functioning of the CRISPR/Cas13a system. Our findings indicate that the stability of
crRNA full-length is better than the other crRNA sequences, as is the stability of the complex formed by
crRNA and miRNA.
Build and Test
Computer simulation
Section 1 Molecular Docking
Molecular docking is a method of predicting receptor-ligand interactions. Based on the properties of
receptor and ligand molecules, various possible conformations are docked and scored based on energy and
binding conditions.
The first step of molecular docking is to obtain the PDB structure files of receptors and ligands. Since
PDB file of the LwaCas13a protein (1152AA) was not available in the UniProt, we used a homologous
modeling prediction method to obtain the PDB structure file of LwaCas13a. Due to the 82.56% similarity
between protein sequences of LwaCas13a and LbuCas13a, we used LbuCas13a as the template protein. We
adopted the best result of SWISS-MODEL homology modeling as the receptor PDB file of LwaCas13a for
the
subsequent docking. Results of the homologous modeling is displayed with PyMol software.
Figure 2 Homologous modeling of proteins & Preparation of docking files (A) The
crystal structure of LbuCas13a; (B) A homology modeling structure of LwaCas13a; (C) A superimposed image
of LbuCas13a and LwaCas13a; (D) Structural domain annotation of LwaCas13a.
We used the 3d RNA server to obtain the tertiary structure and PDB file of crRNA from its previously
obtained secondary structure.
Figure 3 Predicted structure of crRNA with 3dRNA. (A) crRNA1; (B) crRNA2; (C)
crRNA3; (D) crRNA full length.
After obtaining the PDB files for LwaCas13a and crRNA required for molecular docking, we performed rigid
docking of the LwaCas13a protein and crRNA using HDOCK server. The server provided scoring and RMSD for
each docking.
Figure 4 Docking scores for crRNA and LwaCas13a proteins calculated by HDOCK
Figure 5 RMSD for crRNA and LwaCas13a proteins calculated by HDOCK
The Docking Score is a widely-used evaluation metric in HDOCK for assessing molecular docking. The metric
measures the capacity of a docking simulation to accurately predict the binding between two molecules. A
lower Docking Score indicates a better docking performance. RMSD (Root Mean Square Deviation) is a
parameter that quantifies the deviation of atoms from their expected positions during molecular docking.
A lower RMSD value corresponds to a more stable docking. By integrating these two parameters, it was
observed that crRNA full length exhibited superior performance in molecular docking.
At HDOCK, we utilize a rigid docking algorithm that streamlines the calculation process and conserves
computational power. However, we acknowledge that the algorithm's relative simplicity may not always
reflect the true scenario with accuracy. To ensure precision in our calculations, we incorporated MOE
(Molecular Operating Environment) to conduct semi-flexible docking of the crRNA full length and
LwaCas13a proteins. This approach enabled us to determine the interaction between the crRNA full length
and LwaCas13a with utmost accuracy.
Figure 6 LwaCas13a/crRNA full length complex
Figure 7 Interactions and Hydrogen Bonds between LwaCas13a and crRNA
Our findings from molecular docking experiments suggest that crRNA predominantly interacts with the REC
lobe of the LwaCas13a protein, which is consistent with the current literature.
We utilized MOE software to dock miRNA to Cas13a/crRNA complexes, and subsequently visualized the docking
results via Discovery Studio Visualizer.The
three-dimensional structure of the Cas13a/crRNA/miRNA
complex was obtained, and their interactions and schema were calculated.
Figure 8 The predicted structure of LwaCas13a/crRNA/miRNA ternary complex
| Name |
Number of Binding Sites |
| miR-21-5p |
Helical-1 |
crRNA |
| 4 |
8 |
Table 4 Binding sites between miR-21-5p and LwaCas13a/crRNA complex
Figure 9 Schematic diagram of interaction among LwaCas13a, crRNA and miRNA
Our simulation results demonstrate that miRNAs exhibit strong binding affinity to Cas13a/crRNA complexes,
resulting in robust interactions. It is postulated that in practical experiments, miRNAs exhibit a high
affinity towards the Cas13a/crRNA complex, thereby triggering the nuclease activity of LwaCas13a and
facilitating rapid and efficient cleavage reactions.
Molecular docking provides a static view of how molecules interact
with each
other, much like a still
photograph. On the other hand, molecular dynamics simulation gives a dynamic and continuous perspective
of the same interaction, providing a more detailed and thorough understanding of the underlying
processes. This shift from a static to dynamic approach allows for a deeper exploration of the behavior
and movement of molecules in a system.
Section 2 Molecular Dynamic Simulation
To assess the stability of the CRISPR/Cas13a system, we conducted molecular dynamics simulations
utilizing the GROMACS software. We set various parameters to ensure the simulation conditions closely
resembled the actual scenario.
After completing the simulation, GROMACS tools were used to analyze the trajectory and calculate the root
mean square deviation (RMSD), root mean square fluctuation (RMSF), solvent accessible area (SASA), and
hydrogen bonds (Hbonds).
The Root Mean Square Deviation (RMSD) is an important measure of the positional variance of the atoms in
a molecule with respect to a reference position. A small and stable RMSD value indicates a high degree
of stability in the molecule.
The Root Mean Square Fluctuation (RMSF) is a measure of the average position of atoms in a molecule over
time, which indicates the degree of flexibility of the molecule being analyzed during the simulation.
Generally, a larger RMSF value implies a higher degree of flexibility of a particular region of the
molecule and indicates its relative distance from the active position.
Hydrogen bonds are capable of forming both within individual molecules and between multiple molecules.
The stability of a molecule can be determined by the number of hydrogen bonds present. In the context of
molecular dynamics simulations, a consistent count of hydrogen bonds is a strong indication of the
molecule's stability.
The solvent accessible surface area (SASA) is a parameter that is closely associated with the solvation
free energy of a molecule. Additionally, it can serve as a measure of the stability of macromolecules
when placed in different solvents.
The calculation of four parameters for the crRNA was the first step of our analysis.
Figure 10 Molecular dynamics simulation parameters of crRNA full length. ( A.
RMSD of crRNA full length, B. RMSF of crRNA full length, C. SASA of crRNA full length, D. Hbonds of
crRNA full length )
The findings demonstrate that the parameters of crRNA are characterized by good stability, indicating
that the three-dimensional structure of crRNA is relatively stable. Furthermore, we executed simulations
and calculations to determine the parameters of LwaCas13a protein, yielding the following results.
Figure 11 Molecular dynamics simulation parameters of LwaCas13a. ( A. RMSD of
LwaCas13a, B. RMSF of LwaCas13a, C. SASA of LwaCas13a, D. Hbonds of LwaCas13a )
The outcome of the simulation experiment indicated that the LwaCas13a protein parameters remained
relatively stable. This suggests that the structure of the LwaCas13a 3D model was robust and well-suited
for the subsequent docking process.
After determining the stability of crRNA full length and LwaCas13a, we proceeded with molecular dynamics
simulation. The same approach was employed for the complex of crRNA full length and LwaCas13a molecules,
and the parameters were calculated. The results indicate that the crRNA full length and LwaCas13a
complexes are stable.
Figure 12 Molecular dynamics simulation parameters of complex. ( A. Binding Free
Energy of complex, B. RMSD of complex, C. Hbonds of complex, D. RMSF of complex, E SASA of complex. )
Together, low and stable RMSD, stable SASA and Hbonds indicate that the LwaCas13a/crRNA complex is
stable. Furthermore, the complex has low binding energy, which further confirms that the LwaCas13a
protein and crRNA form a stable complex.
Section 3 Reaction kinetics simulation
Reaction kinetics simulation is a mathematical modeling method that helps predict the concentration of
reactants or products under different conditions. It involves constructing a system of ordinary
differential equations.
Our designed CRISPR/Cas13a system cleaves reporter RNA. We simulated this process using the
Michaelis-Menten equation. By altering various parameters, we can predict the trend of reactant or
product concentration.
Once miR-21-5p binds to crRNA, it activates the cleavage activity of the Cas13a/crRNA complex, which
cleaves not only the target RNA but also other RNA the reporter RNA. Therefore, the Cas13a/crRNA/miRNA
ternary complex actually can be treated as an enzyme with RNA cleavage activity. We use a fluorophore
(Carboxyfluorescein,FAM) and a quenching group (Black hole quencher, BHQ1) double attached 6U
oligonucleotides as a reporter RNA to evaluate the cleavage efficiency of the Cas13a/crRNA/miRNA ternary
complex, since cleavage of the reporter RNA results in the release of fluorescence quenched by BHQ1,
which can be measured easily. Such process can be simplified as the reaction below.
The positive reaction rate constant for the binding reaction between the Cas13a/crRNA/miRNA ternary
complex and the reporter RNA is k1, the reverse reaction rate constant is k2, and the reaction rate
constant for fluorescence generated by cleavage is k3. We can use Michaelis-Menten equation to study the
rate of cleavage activities.
\[
v_0 = \dfrac{dF}{dt} = \dfrac{V_m[S]}{K_m+[S]}
\]
\(v_0\) represents reaction rate, which is experimentally expressed as the rate of increase in
fluorescence
intensity, and [S] represents the concentration of the substrate, which is the reporter RNA. In actual
reactions, fluorescence gradually degrades. Therefore, the equation can be rewritten as below.
In this equation, the \(k_{cleavage}\) represents the rate of the enzyme activity, which is the cleavage
activity of Cas13a/crRNA/miRNA complex in our case. \(k_{cleavage} = k3\), \( K_m=\frac{k2+k3}{k1}\). We
set the initial concentration of the ternary complex as 20 nM and the initial concentration of the
reporter RNA as 1000 nM. The parameter settings are as follows.
\[
k1 = 10 nanomole^{-1} {hour}^{-1}
\]
\[
k2 = 0.1 {hour}^{-1}
\]
\[
k3 = 1000 {hour}^{-1}
\]
\[
r = 0.5{minute}^{-1}
\]
By changing the initial concentration of substrate, enzyme, and the values of different parameters, the
following simulation results can be obtained.
Figure 13 Curves of fluorescence production with different association rate
constants
Figure 14 Curves of fluorescence production with different decomposition rates
Figure 15 Curves of fluorescence production with different degradation constants
Figure 16 Curves of fluorescence production with different reporter RNA
concentrations
It can be seen from (1) that the larger the forward reaction rate constants (\(k_1, k_3\)), the faster
the fluorescence reaches its peak. The larger the degradation constant, the faster the fluorescence
reaches its peak. Considering the degradation of fluorescence, the fluorescence intensity ultimately
tends to zero. Since the main difference in actual reactions is the miRNA concentration, and under
physiological conditions miRNA will not be saturated. Therefore we use miRNA concentration to represent
the concentration of the ternary complex in our model. Setting the r value to 0.1 and changing the
initial concentration of the ternary complex, the following curves are obtained.
Figure 17 Curves of fluorescence production with different ternary complex
concentrations
Considering fluorescence degradation, there is an approximate linear relationship between fluorescence
intensity and miRNA concentration within a large range of miRNA concentration (Figure 17).
Experimental validation
Section 1 Collateral cleavage activity corresponding to different crRNAs
To evaluate the efficiency of the crRNA we designed, we tested four different crRNA sequences for
cleavage of the FAM-BHQ RNA probe. This allowed us to compare the collateral cleavage activity of
various crRNA, as activated by single-stranded RNA such as miRNA.
Figure 18 Real-time fluorescence intensity curve of different crRNA
Our experiment involved the observation of the activity of different crRNA. The findings indicate that
crRNA 1 and crRNA 2 had poor activity when compared to crRNA 3 and crRNA full ength, which displayed
relatively better activity. Computer simulations were also carried out to determine the parameters of
these crRNAs, and it was observed that crRNA full length had better parameters than crRNA 3(Figure 4 ＆
Figure 5 ). Based on our observations and simulations, we strongly recommend using crRNA full length as
the most suitable option for subsequent experiments.
Section 2 Collateral cleavage activity of different miRNA concentrations
We tested the relationship between different miRNA
concentrations (0nM，10nM，25nM，50nM，75nM，100nM，125nM，150nM，175nM，200nM) and the cleavage activity of the
CRISPR/Cas system, using the measured fluorescence intensity to represent the cleavage activity of the
CRISPR/Cas system, and we plotted the following curve.
Figure 19 Fitting curves for different miRNA concentrations and fluorescence
intensities
Based on the experimental results, the system generates the strongest fluorescence intensity when the
miRNA concentration is 100nM. This indicates that the CRISPR/Cas system has the strongest cleavage
activity at this point. As the miRNA concentration increases, the fluorescence intensity decreases,
eventually becoming constant. Our hypothesis is that due to the CRISPR/Cas system's unique collateral
cleavage activity, miRNA may compete with the Reporter RNA substrate. Consequently, a high concentration
of miRNA could lead to a decrease in fluorescence generated by the system's cleavage.
Section 3 Experimental validation of enzymatic reaction kinetics
To ascertain the cleavage activity of our CRISPR-Cas system, we initially utilized computer to simulate
the kinetics of the reaction.
In the next step, we conducted experiments to validate the results of the computer simulation and
calculated the relevant parameters. We tested the changes in fluorescence intensity induced by different
concentration gradients of Reporter RNA(0nM, 100nM, 125nM, 250nM, 500nM) over time.
Figure 20 Real-time fluorescence intensity curve of different Reporter RNA
concentration
The CRISPR/Cas13a system, when activated, can be viewed as an enzyme system. The reporter RNA in the
environment is considered the substrate. The relationship between the two can be described by the
Michaelis-Menten equation.
\[
\dfrac{1}{v_0} = \dfrac{K_m}{V_m}\dfrac{1}{[reporter]}+\dfrac{1}{V_m}
\]
We used the experimentally obtained data for fitting the above equation, and obtained the relationship
between the concentration inverse and the fluorescence intensity inverse.
Figure 21 The fitting curve of Lineweaver-Burk
After using double inverse graphing, the experimental results were found to be consistent with
Michaelis-Menten equation, which is in line with our expectations.
Learn
We have made great efforts in the ENGINEERING part and finally optimized the simulation and validation
pipeline of CRISPR/Cas13a system. Thanks to the development of computer science, we simplified many
experimental processes, and save experimental materials and time.
By evaluating crRNAs, we identified the most efficient crRNA; through molecular docking and molecular
dynamics, we revealed the stability of crRNAs, Cas13a proteins, and crRNA/Cas13a complexes, and
clarified the feasibility of fluorescence experiments; and through reaction kinetics simulations, we
explored the effects of individual reaction conditions. In conclusion, physical simulation played a
crucial role in our project, and we believe that this pipeline can be applied to other iGEM teams, which
also uses CRISPR/Cas13a system, to simplify the experimental process and save experimental materials and
time.
Figure 22 Pipeline of our crRNA engineering
Part II Cloning and Expression of crRNA
Brief Introduction
crRNA is the critical part that determines the target oligonucleotides of CRISPR/Cas system. Since crRNA
is a short single-stranded RNA, which has poor stability, it is difficult to store crRNA for long time.
More importantly, the general laboratory protocol for crRNA production is also somewhat deficient. To
produce a short ssRNA, the DNA template of this ssRNA is obtained first, and the template is used to
transcribe RNA. However, primers for short-stranded DNA are difficult to design, so PCR of
short-stranded DNA is difficult to achieve. Also, it is difficult to observe whether PCR is successful
or not for short-stranded DNA by gel electrophoresis. Therefore, we set to design an improved laboratory
protocol for crRNA production.
Design
Based on SEVA3.1 standard, we designed this laboratory protocol for crRNA production. As a trackless
assembly, Golden Gate Assembly is better than other cloning methods in accuracy, efficiency and
convenience. Therefore, we chose Golden Gate Assembly as our general assembly method. First, we design
two sets of primer--spacer and loop that are corresponded to the crRNA spacer sequence and loop sequence
respectively. The two sets of primers are designed to bind with pveg, a plasmid recommended by iGEM. The
primers of spacer F are designed with guide sequence and restriction site of SapI enzyme. After PCR,
spacer primers will amplify the products carrying guide sequence of crRNA for Golden Gate Assembly. Loop
R primers are designed with loop sequence of crRNA and sequence of T7 promoter. It will amplify the
products carrying loop sequence for Golden Gate assembly after PCR amplification. Both groups of primers
have BsaI enzyme recognition sites at the ends. After enzymatic digestion by BsaI, they both produce
sticky ends. The two PCR products interact with each other depending on the specific sticky ends, and
then they are ligated with T4 ligase. As a result, a plasmid with full crRNA sequence is obtained.
Figure 23 (A)Primer design schematic (B)Primer sequence description
diagram(C)Plasmid construction flowchart
To get large amount of plasmid with crRNA sequence, the plasmid is transformed into the bacterial
chassis. Then, we will extract the plasmid and digested it with Sapl enzyme to make it linearized. After
purification, T7 RNA polymerase is used for transcription of crRNA in vitro.
Build and Test
We used primers loop and primers spacer, and used the plasmid pveg as the template to conduct PCR
amplification to get the product for Golden Gate Assembly. We used this procedure for PCR amplification
(98℃ 1min/(98℃ 10s/55℃ 5s/72℃ 1min/kb)x35/72℃ 2min ). After PCR amplification, we followed the
instruction of BsaI enzyme for enzyme digestion, and used T4 ligase to ligate the digested PCR product.
By performing gel electrophoresis, we would like to confirm that the PCR experiment was successful, but
did not obtain the expected results. We conducted several rounds of experiments following the control
principle and adjusted the annealing temperature, but the results were still unsatisfactory despite
repeated efforts. Therefore, we thought that the primer Loop R might be self-complementary, resulting in
its inability to bind to the template plasmid. In order to verify this thought, we analyzed the
secondary structure of the primers by computational modeling, and found that the primer indeed have a
great potential too form multiple stem-loop structures.
Figure 24 Secondary structure of primers predicted by UNAfold software
Based on the fact that the primer itself may form a loop, we needed to design new loop R primers to
reduce the possibility of self-complementary pairing. After our analysis, we thought that there might be
two factors that lead to inefficient of old loop R primer. On one hand, the primer Loop R was too long,
which led to the formation of complementary pairs on its own. On the other hand, the banding site was
designed to bind to Biobrick prefix, which has the sequence of GCGGCCGC that was easy to form
self-complementary stem-loop region. So we designed two sets of new loop R primers with different
improvements. One shortened the overhanging sequence, and the other changed the banding site. Finally,
we successfully amplified the desired DNA sequence using these new primers, so we choose one of it
continue our experiment. Result is shown in the following figure:
Figure 25 Gel electrophoresis after PCR of two set primers. (Lane1: DNA marker.
Lane2: guide sequence of crRNA3. Lane3: guide sequence of crRNA2. Lane4: guide sequence of crRNA1.
Lane5: guide sequence of crRNA full length. Lane6: loop region.)
Through gel electrophoresis, we proved that the PCR experiments successfully amplified the DNA sequences
for Golden Gate assembly. After Golden Gate assembly, the plasmid was transferred into the bacterial
chassis for replication, and colony PCR was carried out to select colonies with successful assembly.
Figure 26 Colony gel electrophoresis after PCR. (Lane1: DNA Marker, Lane2-4:
crRNA full length, Lane5-7: crRNA1, Lane8-10: crRNA2, Lane11-13: crRNA3.)
After that, we used SapI enzyme to linearize the plasmid. Finally, T7 RNA polymerase was used to
transcribe crRNA.
Learning
This is the first time for us to do molecular cloning. When we designed primers at begining, we only
cared about whether the design of the sticky end was reasonable, and neglected the possibility of
forming self-complementary pairing of the primers. After discovering this problem, we learned the primer
design method and designed new primers. When designing new primers, we paid special attention to the
secondary structure of the primers and the possibility of forming primer dimers between forward and
reverse primers, and we finally succeeded in designing good primers.
References
[1] Abudayyeh, O. O., Gootenberg, J. S., Essletzbichler, P., Han, S., Joung, J., Belanto, J. J.,
Verdine, V., Cox, D. B. T., Kellner, M. J., Regev, A., Lander, E. S., Voytas, D. F., Ting, A. Y., &
Zhang, F. (2017). RNA targeting with CRISPR-Cas13. Nature, 550(7675), 280–284.
https://doi.org/10.1038/nature24049
[2] Damalas SG, Batianis C, Martin-Pascual M, de Lorenzo V, Martins Dos Santos VAP. SEVA 3.1: enabling
interoperability of DNA assembly among the SEVA, BioBricks and Type IIS restriction enzyme standards.
Microb Biotechnol. 2020 Nov;13(6):1793-1806. doi: 10.1111/1751-7915.13609. Epub 2020 Jul 25. PMID:
32710525; PMCID: PMC7533339.
[3] https://2022.igem.wiki/thessaloniki-meta/experiments
[4] Lorenz, R., Bernhart, S. H., Höner Zu Siederdissen, C., Tafer, H., Flamm, C., Stadler, P. F., &
Hofacker, I. L. (2011). ViennaRNA Package 2.0. Algorithms for molecular biology : AMB, 6, 26.
https://doi.org/10.1186/1748-7188-6-26
[5] Liu L, Li X, Wang J, Wang M, Chen P, Yin M, Li J, Sheng G, Wang Y. Two Distant Catalytic Sites Are
Responsible for C2c2 RNase Activities. Cell. 2017 Jan 12;168(1-2):121-134.e12. doi:
10.1016/j.cell.2016.12.031.
[6] UniProt Consortium. UniProt: the Universal Protein Knowledgebase in 2023. Nucleic Acids Res. 2023
Jan 6;51(D1):D523-D531. doi: 10.1093/nar/gkac1052.
[7] Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, Heer FT, de Beer TAP, Rempfer
C, Bordoli L, Lepore R, Schwede T. SWISS-MODEL: homology modelling of protein structures and complexes.
Nucleic Acids Res. 2018 Jul 2;46(W1):W296-W303. doi: 10.1093/nar/gky427.
[8] Schrödinger, L., & DeLano, W. (2020). PyMOL. Retrieved from http://www.pymol.org/pymol.
[9] Adasme MF, Linnemann KL, Bolz SN, Kaiser F, Salentin S, Haupt VJ, Schroeder M. PLIP 2021: expanding
the scope of the protein-ligand interaction profiler to DNA and RNA. Nucleic Acids Res. 2021 Jul
2;49(W1):W530-W534. doi: 10.1093/nar/gkab294.
[10] Sinha S, Molina Vargas AM, Arantes PR, Patel A, O'Connell MR, Palermo G. RNA-mediated Allosteric
Activation in CRISPR-Cas13a. bioRxiv [Preprint]. 2023 Jul 27:2023.07.27.550797. doi:
10.1101/2023.07.27.550797.
[11] Tambe A, East-Seletsky A, Knott GJ, Doudna JA, O'Connell MR. RNA Binding and HEPN-Nuclease
Activation Are Decoupled in CRISPR-Cas13a. Cell Rep. 2018 Jul 24;24(4):1025-1036. doi:
10.1016/j.celrep.2018.06.105.
[12] Accelrys. Discovery Studio Visualizer, v20.1.0.19295, BIOVIA, Dassault Systèmes, San Diego, 2020.
[13] RNA structural dynamics as captured by molecular simulations: A comprehensive overview. doi:
10.1021/acs.chemrev.7b00427.a
[14] MATLAB and Simbiology Toolbox Release 2020b, The MathWorks, Inc., Natick, Massachusetts, United
States.
[15] Ramachandran A, Santiago JG. CRISPR Enzyme Kinetics for Molecular Diagnostics. Anal Chem. 2021 May
25;93(20):7456-7464. doi: 10.1021/acs.analchem.1c00525.
