Team:Moscow/Engineering
Engineering success
Overview
Last year, we conceived the concept of split-system improvement. However, while implementing the idea, our team met several complications with its purification. It turned out that a few other teams faced the same problem, so we made an attempt to discover why. We hypothesized that the split-system might be too large and, thus, complicated our work. During our further research, we found a much smaller and relatively new protein - CasX, which we used to design a new test system. To explore our device properties, we developed its kinetic model available in detail on our Modelling page. Here, a comprehensive report on our molecular design, along with hardware, is represented.
Molecular design №1: split-system
Imagine and Design
Firstly, we decided to improve our previous year’s (Moscow 2019) design for HCV genotypes detection. This system is supposed to work as follows: two dCas9 proteins fused with N- and C-ends of β-lactamase bind to HCV’s cDNA target sequence. Binding results in the reunion of β-lactamase providing visible enzyme activity - oxidation reaction of nitrocefin. To determine the distance between two dCas9 proteins that will provide the most effective reunion of β-lactamase, we made molecular modeling.
Build and Test: Molecular modeling
We had the results of molecular modeling from the previous year, but we decided to perform a more complex and full one. Additionally, we already knew that PAM-out (PAM - protospacer adjacent motif, PAM-out - PAM regions are located on different DNA strands and oriented outward of the complex) orientation is more suitable for the split system [1]. This year we simulated the structures of two dCas9 proteins in the nucleoprotein complex and calculated the optimal interval in nucleotides between them. For this purpose, we wrote Python scripts for the Chimera to build models of two dCas9 proteins. To design these systems, we used the SpCas9 protein structure isolated from Streptococcus pyogenes (PDB id 5Y36)[2]. The structure of an ideal B-DNA was used for modeling the DNA separating dCas9 proteins. For PAM-out orientation, we generated models with different numbers of nucleotides between PAM regions. For choosing the best interval between Cas complexes, the graphs of the dependence of the distance between the geometric centers of the C-end, N-end, and the angles between them on the interval in nucleotides between the PAM regions were built (see Fig.1B). A model with the smallest angle and distance was chosen (the angle value was a priority when choosing a model), which also did not contain atomic overlaps between SpCas9 complexes, namely, the model with an interval between PAM regions of 63 nucleotides (see Fig.1B).
Figure 1. A - the structure of two SpCas9 proteins in a nucleo-protein complex in a PAM-out orientation with an interval between PAM regions of 63 nucleotides; the green arrow marks the N-end (N) of the first SpCas9 protein, the red arrow marks the C-end (C) of the second. B - a graph of the dependence of the distance between the geometric centers of the N-end (N) of the first SpCas9 and the C-end of the second for the PAM-out orientation, as well as the angles between them on the interval between the PAM regions; red zone - models in which there are atomic overlaps of dSpCas9 complexes; the interval between PAM regions at which the minimum distance and angle between amino acid residues of interest is observed is plotted and marked with a dot.
Build and Test: complications
However, after the definition of molecular distance for our split-system, we met experimental difficulties in β-lactamase C-end protein purification. We tried to find a solution to our problem and figured out that the team of UiOslo_Norway 2018 had a similar problem with dCas9-beta-lactamase construction. Hence, this part required more attention and seems not so executable as we thought.
Hardware №1: photometer
Moreover, we improved a portable photometer for our system. In last year's project (Moscow 2019), we developed several prototypes of photometers aimed to detect the reaction of nitrocefin with β-lactamase which N- and C-ends were fused with two dCas proteins (Fig.2). Our task was to develop a simple device capable of measuring optical density at a specific wavelength. During the oxidation reaction of nitrocefin with β-lactamase, its optical properties change. Namely, the absorption spectrum of light moves to a longer wavelength range. And this year we thought to utilize the same principle with decreased measurement error.
Figure 2. We improved the photometer which we first made in 2019.
Molecular design №2: CasX collateral activity
Learn, Improve and Research
We also took into account that SpCas9 has NGG PAM, which is located on the 3' end of the targeted sequence. So our system is limited by the 2 PAM sequences that are located on defined distance (to reach an angle between N-ends of Cas proteins less than 20 degree number of nucleotides between them may be 22, 43, 53, 63, 73, 84, 94).
We thought about other CRISPR-Cas constructions, for example with one Cas protein. We choose CasX with more variable PAM sequence TTCN that is located on the 5' end of target. In other words we overcame 2 limitations - PAMs on defined distance and searching for more rare PAMs.
Design of guide RNAs for detection and distinguishing a variety of HCV genotypes (8 main types and more than 80 subtypes) is non-trivial and requires more bioinformatics work (particularly, analyse literature and alignments of great amounts isolated from the HCV database).
Also there are known Resistance-Associated Substitutions (RASs) in proteins of HCV, which increased a possible number of detected SNPs.
So, we searched targets in our first approximation that CasX will have a similar number of target sequences near detected SNPs as a system with 2 SpCas9. And it was approved.
Number of possible targets for SpCas9_SpCas9_PAM-out is 299 targets, in NS5A - 38 targets.
For a system with 1 CasX - 223 targets in the whole genome and 32 targets in NS5A. For searching targets was used python scripts (iGEM Team Moscow 2019) and NC_004102.1 genome.
Imagine and Design
We designed the following principle: viral RNA from a patient's blood is converted into DNA and amplified with help of the LAMP method. In the presence of genetic material from a certain genotype of pathogen, CasX cuts the reporter molecule labeled with biotin and the FAM fluorophore from opposite sides. Resulting solution is placed on a test strip with two binding zones. The uncleaved molecules will bind to the control site on the test strip, while the cleaved ones will bind to the second site. Thus, the presence of two bands on the test strip indicates the presence of a specific viral genotype in the sample, one indicates its absence.
Figure 3. Detection system based on CasX collateral activity
Imagine and Design
For the system developing we build a kinetic model to solve some problems:
To confirm our expectations of system functioning, such as the potential to detect nucleic acids in a reasonable time;
To determine optimal conditions (sgRNA initial concentration, CasX initial concentration, etc.);
To determine detector sensitivity - the minimum concentration of nucleic acid in a sample that can be detected using our test system;
To determine the time required to carry out one test.
Kinetic model is described in detail on the Modeling page.
Learn
According to our Modeling, further improvements are necessary. Firstly, we will need to determine experimentally the time of fluorescence signal development because in modeling it is hard to determine potential brightness. Unfortunately, we did not have an opportunity to run all experiments in the lab.
On the other hand, modeling demonstrated the necessity of HUDSON as a step aimed to reduce the time required for detection and increase the signal intensity by protecting the RNA in the system from degradation. So we included HUDSON in our final system.
Hardware №2.1: thermostat-fluorimeter
Introduction and research
But we not only redesigned the modeling part but also changed the hardware. In this year's project, the detection is based on FAM fluorescence with additional heating steps, so the photometer we have developed previously is not suitable for the detection. Using the experience and knowledge gained during the development of the photometer, we have developed a new device (Fig.4).
Figure 4. Primary design of the thermostat with the function of fluorimeter.
Imagine and Design
We decided to create a thermostat that contains a separate compartment where fluorescence is measured. We use the Peltier element for heating and maintaining the temperature, as the most accessible and does not require special skills. The device is controlled using an Arduino-based microcontroller. The part of the device that is responsible for detecting fluorescence consists of led emitting light with a wavelength of 470 nm, a photodiode and a light filter transmitting light with a wavelength longer than 500 nm.
Test and Learn
To check the device for maintaining the temperature and its change, a thermocouple is installed in the device. When conducting tests using data obtained from a thermocouple, we would estimate to what temperatures our device is heated and adjust the modes to maintain the temperature.
To test the detection of fluorescence, we would conduct a series of experiments with a control in the form of a clean strip. Using a 505 nm led emitting light, a 510 nm filter, and a photodiode. It is necessary to make sure that the light filter does not pass light with a wavelength of 505 nm or does not pass significantly at all. Then calibrate the photodiode using the Ultrospec 1100 pro laboratory spectrophotometer. We would adjust the sensitivity to light using different concentrations of the coloured solution.
We would have a calibrated part responsible for RT-LAMP reaction and HUDSON, and a part responsible for fluorescence detection.
Discussions with experts from BIOCAD and improvement
We asked the experts of the company BIOCAD to give an expert assessment of our project, including the developed test system and our device. During the discussion, changes were made to the concept of the device, taking into account the comment.
- The heating part may affect the detection of fluorescence. Perhaps we should separate these actions in space.
- We were asked to come up with a mechanism so that we didn't have to add reagents to the test tube while it was in the thermostat.
- Strips placed in the instrument for fluorescence assessment may leave a fluorophore in the instrument itself.
Hardware №2.2: fluorimeter
Research and design
In addition to the suggestions of experts from BIOCAD, we analyzed our device ourselves and made changes.
- During the entire testing process, you need to quickly change the temperature to three different values (95, 65, 32 degrees Celsius). We realized that developing a Peltier-based thermostat would be problematic since conventional Peltier elements do not reach temperatures of 95 degrees. Also, we won't be able to quickly change the temperature using the Peltier element.
- It was proposed to change the light path from the led to the photodiode in the device. To make the led stand at 90 degrees relative to the photodiode. This will reduce the number of photons that will pass through the filter to the photodiode.
During the discussions, we decided to abandon the design of the thermostat together with a fluorimeter in a single device. We decided to make two separate devices: thermostat and fluorimeter. But also we did not set the task to reinvent the wheel: there are a lot of cheap, reliable, and compact thermostats available to buy. So we decided to focus on designing the fluorometer.
Taking into account the comments of BIOCAD, we changed the light scheme of the device (Fig. 5). We also decided to make a removable block where the test strips are placed. After each measurement, the unit is removed and washed with an alcohol solution to avoid a false-positive result.
Figure 5. General view of the fluorimeter
Application instruction
When you turn on the device, use the button to move through the menu. Before starting the measurement, the instrument must be calibrated. To do this, select “On” from the menu, then select “calibrate”. After that, the Device will not record the background. To start the measurement, place the test strip with the reaction mixture applied. Insert it into the cell in the fluorometer. Then select “start " in the menu by clicking on the button. After that, the device begins to Shine an exciting light on the test strip. The measurement results will be displayed on the screen. After measuring one strip, remove the strip and the compartment itself, rinse it thoroughly and load it back into the device dry. After you receive the necessary data, you can turn off the device from the power outlet.
Replacement (Fig. 5) unit for test strips. Several units are included with the device. We recommend washing it thoroughly with an alcohol solution after each measurement.
Conclusion
To conclude, we went through a long way of creating the final version of our test system from a split system based on dCas9 with fused ends of β-lactamase to CasX and its collateral activity. To be able to effectively detect the signal we also developed and improved our fluorimeter.
References
- Yihao Zhang et al., “Paired Design of DCas9 as a Systematic Platform for the Detection of Featured Nucleic Acid Sequences in Pathogenic Strains,” ACS Synthetic Biology 6, no. 2 (February 17, 2017): 211–16, https://doi.org/10.1021/acssynbio.6b00215.
- Cong Huai et al., “Structural Insights into DNA Cleavage Activation of CRISPR-Cas9 System,” Nature Communications 8, no. 1 (November 9, 2017): 1375, https://doi.org/10.1038/s41467-017-01496-2.