Our therapeutic element using SEND (Selective Endogenous eNcapsidation for cellular Delivery) can only be conducted under the condition of definitive diagnosis of PKU-patients. That’s why tackling the early diagnosis and detection of phenylketonuria counts significantly while targeting such a disease. Our novel therapy provides the patients with RNA copies of the phenylalanine hydroxylase enzyme (PAH) that’s already deficient in their liver.
We provided practical proof and evidence of the success of our genetic circuits and our theranostic concept based on parts characterization, in addition to experimental and laboratory validation of each step. We identified all the possible aspects in our project that require further lab validation to make sure that all our used parts work properly in harmony in each circuit. That’s exactly why we carried further verification through extensive computational analysis integrated with lab experiments. Choosing a prototype for our diagnostic and therapeutic platforms was essential as a proof for our novel solution for the newborn screening of In-born errors of metabolism (IEM). We found that Phenylketonuria is one of the main challenging metabolic disorders among the panel of IEM.
We transfected hippocampal cell lines with the vector plasmid of our therapeutic circuit (pcDNA3.1 ligated with the therapeutic parts). Transfected cells are incubated at 37o C in a 5% CO2 for 72 hours. One day before conducting the experiment, the Hippocampi cells were seeded in a 6-well culture plate. And dissected into 3 experimental groups: (1) naïve cells (negative control), (2) cells transfected with empty vector (pcDNA3.1+), and (3): cells transfected with pcDNA3.1+ therapeutic circuit) (pcDNA3.1+).
To validate the efficacy of the therapeutic circuit, the phenylalanine hydroxylase gene was measured in the cell using Sybr-green-based Real-time PCR. The following steps are performed:
From the table we found that the fold change of PAH gene (FC) is higher in the third group, which contains cells transfected with PcDNA3.1/circuit than other groups, and we found the lowest FC value in the group containing naive cells.
Ct: cycle threshold, Δ: delta, FC: fold change of PAH gene after normalization to untreated Hippocampi cells, ACTB: β-actin housekeeper gene used for normalization
Following bacterial transformation and colony growth as described in the protocols page , beta-galactosidase expression was determined as follows: For each agar, we added 0.1 mL of bacterial suspension and inverted several times to disperse the bacteria. Next, we added 40 µL of X-GAL to each plate. As a following step, we allowed the soft agar to harden at room temperature, then incubated the plates at 37°C inverted for 12–16 hours. After incubation we added phenylalanine samples and stored the plates for several hours at 4˚C, allowing the blue color to develop. This process was repeated 3 times each time with different phenylalanine concentrations as follows ( 2mg/d, 10 mg/dl, 20 mg/dl ). As well as using tyrosine as a control in a different plate which showed no change of color and to ensure that the sole cause of color production in X-gal is administration of phenylalanine
The following figures illustrate the final result of the transformation, serving as both a proof of concept and a representation of the several iterations:
This plate (Apt-Phe complex +ve/X-gal +ve) shows that the WCB emits a low signal of the blue color at 20 mg/dL of phenylalanine. which validates the specificity of the TyrR for Phe only, not the Apt-Phe complex.
Figure shows the concept of providing Whole-cell based biosensor & Consumption line (Jury line) in the lateral flow assay.
The consumption line (Jury line) is based on a duplex dissociation mechanism through a competitive binding with the analyte. This approach consumes the normal phe-levels in the blood which is between 2-20 milligrams per deciliter (mg/dL). This allows the excess concentration to flow to the test line which reduces the false positives that can results from the normal levels of Phe. This line contains highly sensitive AuNP-labelled aptamers that consume the normal level of phenylalanine from the blood sample with a cut-off value of 20 mg/dL to avoid any false positives by the whole-cell biosensor. The consumption line would hypothetically consume the normal levels of phenylalanine in the provided sample. As a result, the chances of false interpretation of the normal phenylalanine levels by the e.coli-based biosensor as being excess are minimized.
The architecture of the aptamer's binding site prediction model is shown in this figure, together with confusion matrices and the effectiveness of our classification model as evaluated by ROC curves. The model's training produces an AUC of 0.96, while testing on a different dataset yields a 95% success rate.
This figure show the outcomes of our affinity regression model when it was used to analyze a dataset of aptamer-protein complexes, which included both DNA and RNA aptamers. (A) Demonstrates a decline in the Mean Absolute Squared Error, which can be used as a reference to predict the training outcomes' accuracy. Root-mean-square deviation (RMSE) is shown to be decreasing in (B), indicating that the projected values are getting closer to the regression line. (C) Demonstrates a rise in the training and validation scores' correctness over epochs. (D) Demonstrates a rise in the training and validation r-squared values over epochs, demonstrating that our model effectively explains the observed data.
This figure illustrates how well our models can forecast the G values of aptamer-protein complexes. (A) Displays the model's training process's prediction outcomes. (B) demonstrates how well our models performed when used with a different testing set and their capacity to correctly forecast the G values of the generated complexes.
This figure Shows. A) the structure of the Phenylalanine amino acid, generated using PyMOL. B) the structure of the Phenylalanine Aptamer molecule. C) Docking model of the Phenylalanine-Aptamer, which was generated using Poseview. D) the docking model of the Phenylalanine-Aptamer, which was generated using HDOCK.
This figure shows colourimetric change based on the concentration of phenylalanine added at each trial, showing Red color reflecting maximum saturation of the 3ng of aptamers when phenylalanine is added at the given concentration of 20 mg/dL.
We thought about the possibility of getting false negative results due to the inappropriate activity of the e.coli-based biosensor itself and how this could affect our results. Pertaining to avoiding these false negative results, we decided to add another control line that acts as a reference for the e.coli cell viability. This line reflects the vitality of the cells fixed on the paper strip of the lateral flow assay and can indicate certain requirements related to the proper storage of this LFA diagnostic device.
This figure Illustrates the idea of introducing Double-control lines as a verification of the overall test viability.
After establishing the circuit design for our therapeutic approach for phenylketonuria. We considered using an effective method for the PAH-gene delivery to the mammalian cells. That’s why we used a novel gene delivery system, called SEND (Selective Endogenous eNcapsidaton for Gene Delivery) to ensure proper and successful delivery of the PAH-gene to the Hepatocytes.
We prepared our cargo to deliver the PAH-RNA copies using the SEND approach based on introducing recently discovered retroelements (Peg10) and the derivatives of Vesicular stomatitis virus G (VSVg) that are involved in the construction of our mammalian genetic circuit. We then performed sequence improvement across many mutational trials using computational directed evolution. This was done to enhance the functionality of these sequences to be integrated as a packaging system for our therapeutic circuit. Our novel delivery system depended on flanking the PAH coding sequence with the MmPeg10 untranslated regions to enable viral-like particle (VLP) assembly of the cargo mRNA, thus ensure functional intracellular transfer and successful endocytosis of the mRNA copies of the PAH gene. The expression of the PAH after using this principle of SEND gene delivery was measured. After that, we added another element to the VLP-structure of the cargo by introducing the VSVg part that codes for fusogenic protein mediating cellular tropism and directs the VLPs-pseudotyped with VSVg to the liver cells. We also predicted the antigenic property of these proteins including Peg10 and VSVg to avoid any immunogenic probability of our full endogenous gene delivery platform, SEND. This helped us both predict the required dosages and keep the antigenicity of our novel therapeutic delivery system to the minimum immunogenic susceptibility. Our system is ultimately modular and has great opportunity to expand its applications to be used in other nucleic-acid based therapies.
Creating a programmable system is another challenge to be able to control the SEND delivery and packaging of our RNA cargo. For that reason we have set an epitranscriptomic regulatory mechanism for the expression of the PAH RNA. Our design is inspired by the use of CRISPR-cas-induced modulation of the translation of the introduced RNA copies of our gene of interest. This Safety system can control the free release of the messenger RNAs of the phenylalanine hydroxylase enzyme based on the phe-levels in the patient’s blood. As we provided dcas-12g as an extra tuning method. The aim of that transcription-control mechanism, via the use of the catalytically dead type V–G CRISPR–Cas effector, is to enable the dcas 12g to bind to the ssRNA transcripts stopping their translation without the cleavage of the RNA strand itself. We avoided the use of cas9 itself for two main reasons, which are:
I- CRISPR-cas9 system cleaves the DNA and may cause possible mutations due to the cell’s own repair machinery following the cleavage process and this may disrupt the normal gene itself by inducing frameshift mutations in the whole sequence following the cleavage site.
II- The application of the cas12g is that it can potentially bind to the ssRNA or ssDNA strands. Thus ensuring an epitranscriptome-control rather than binding to the DNA level itself.
We furtherly added a Cas programmable RNA binding system by adding L7ae as another regulatory module for the protein expression of this RNA-targeting system (cas12g). L7ae is an RNA-binding protein that acts as a riboswitch controlling the release of the Crispr-cas system, depending on the level of phenylalanine itself. As the level of phe is elevated, exceeding the normal levels, the TyrR regulon will stimulate both ptyrp and paroF promoters, leading to expression of the PAH-RNA transcripts, as well as, the L7Ae protein that will then bind to its kink-turn, thus preventing the translation of the dcas12g. This in turns causes free expression and translation of our therapeutic elements followed by the Virus-like particle (VLP) assembly components.
On the other hand, if the enzyme (PAH) is successfully produced and converts the abnormally elevated phenylalanine molecules into tyrosine. The Tyrosine itself will act on the same circuit regulating the release of the circuit components to prevent over-accumulation of the PAH protein. Where excess Tyrosine will induce the TyrR-regulon to act as a repressor, inhibiting the transcription of the part following the ParoF-promoter, which in our case represents the L7Ae protein. Consequently the L7Ae will not be expressed, so the Kink-turn will be free and the dCas12g will be in an active state.
Besides, this Crispr-based regulation should be furtherly controlled using an extra-safety switch though exogenous controlling factors. In this regard, we designed another regulatory system for the Crispr itself using anti-crispr proteins that can effectively bind to the CRISPR-Cas system at many different sites, each inhibiting a certain function of the CRISPR system. We introduced the Tet-On 3G inducible system as a modified technique for the gene expression control of both the Cre-recombinase and the anti-crispr protein. This platform is based on two central elements: I- (tTA) the tetracycline (tet)-controlled transactivator that is modified to Tet-On 3G. II- Specific responsive promoter (Ptet) which is enhanced to TRE3G promoter controlling the expression of our transgene (Cre recombinase ). The Tet-On 3G doesn't induce TRE3G promoter until the external element, which is (tetracycline (tet) and tet-derivatives) such as doxycycline hydrochloride (Dox), is present. In this case these external factors improve the Tet-On 3G DNA-binding capacity, thereby performing transcription activation of TRE3G promoter. Accordingly, our transgene (Cre recombinase) will be freely expressed and will target the loxP sites, flanking the (STOP) sequence. This stop sequence consists of nLacZ with nuclear localization signal. This system hinders the gene expression of the downstream gene (anti CRISPR). After the deletion of the (STOP) sequence by Cre recombinase activity, the transcription of nLacZ will be replaced by antiCRISPR expression
Thus we ensure a versatile safe design that is modular, adaptive, safe and self-regulated system for crispr cas12g- regulation based on the integration of exogenous switch (tet-ON-3G) that triggers the expression of anti-crispr proteins. This mechanism depends on the action of Cre-recombinase cleaving ability to remove the STOP sequence by cutting the LoxP-sites.
This plate (Phe/X-gal) +ve shows that the WCB emitted a bright blue color at a phenylalanine concentration of 20mg/dL. This is the best outcome after the improvements and making good use of our iterations.
By the end of our directed evolution-phase, we proceeded with dry lab techniques to make sure that we picked the most suitable and optimized anti-crisprs for our desired cas12g. After that, we used lab techniques and went ahead by working on a poly-transformation method to introduce our genetic circuit to the e.coli cells. We prepared 3 agars as shown in the illustration below:
After proving the success of our design through multi-stage assessment of each part in our genetic circuits, we carried out accuracy verification by testing the sensitivity and specificity of the lateral flow assay as a whole.
Our target was to identify all the possible flaws in the results of the diagnostic test. For this reason, We created and prepared standard phenylalanine samples resembling testing and control samples for our experiments. The Phenylalanine concentrations in these samples were determined and analyzed based on mass spectrometry. These standard phenylalanine samples were matched to values from real patients.
We then tested these samples using our diagnostic assay to assess and validate the test results. We found that out of 10 PKU samples 8 of them were detected as positive with only 2 false negative results and we have found that 10 out of 10 Normal Controls were detected as negative results
All of the previously mentioned data proved that our test was accurate, sensitive and specific to phenylketonuria (PKU). The aforementioned results and validation steps represent a strong proof for the accurate and optimum performance of our test that does not depend on the operator (operator-independent).
This in return will support the scaling up of mass testing and mass production for our lateral flow test to aid the process of tracking and tracing the newborns suffering from PKU, thus providing an additional benefit for our health system and national screening programs by delivering accessible and affordable point-of-care diagnostic test that help alleviate the financial burden of these newborn screening programs in our country.
Our next step in improving our diagnostic test is to furtherly proceed with multiple advanced validation steps after getting approvals and consents required to be able to verify and prove the accuracy of our test on real sample from PKU-patients
We have performed computational modeling and analysis of the cas12g to predict a set of putative sgRNA off-targets in Cas12g applications. The main features of the suggested gRNAs are summarized in Table 1:
Table 1. Shows sgRNA sequences are ranked to pick candidate CRISPR-cas12g-sgRNA sequences of maximum on-target activity for the target provided.
Docking score calculation for the anti-crispr proteins binding & affinity to cas12g:
After obtaining the crRNA candidates, we analyzed their sequences to settle on the best gRNA with the highest score and accurate binding ability to the target sequence. We put this gRNA in a thermocycler to perform PCR amplification. We then analyzed the PCR product to verify proper amplification by electrophoresis and ran the sample on a 0.8% agarose gel at 5 V/cm for 40 minutes. We then compared the DNA fragment with the DNA ladder ranging from 100-10000bp.
We have conducted multiple rounds of directed evolution to the anti-crispr protein that showed a high level of docking stability. Consequently we created a multi-stage enhancement and improvement of the coding sequence of the AcrIIA5 v2 anti-crispr protein. This is considered as a computational validation and a strong proof of the AcrIIA5V2-mediated inhibition of the nuclease dead CRISPR-Cas-12g system. We can notice that the mutant variant (R32E) showed the highest score of optimum performance among the other mutated sequences of the anti-crispr protein.
This illustration demonstrates the mutational landscape of the AcrIIA5 v2 as a disabling motif for the dcas 12g. As presented in the part :BBa_K4140021
© 2022 - Content on this site is licensed under a Creative Commons Attribution 4.0 International license.
The repository used to create this website is available at gitlab.igem.org/2022/afcm-egypt.