Description

We tried to mimic different operations for our systems, Trim and HTRA1 by focusing on computer simulations. We wrote a python program to facilitate the ranking process after the prediction of the 3D structure of any protein to be ready for further processes such as molecular docking, molecular dynamic simulation, and mathematical modeling to predict and test our project. our systems consist of various parts that bind or interact with each other. To validate the trim21 system's ability to call ubiquitin molecules and recruit proteasomal degradation of the targeted protein (tau) intercellularly in the early stage. The interactions were validated using pull-down assay, and NATIVE-PAGE. Then the ability of the system to recruit ubiquitin was assayed using in-vitro ubiquitination and analyzed by western blotting. regarding validation of the HTRA1 system's ability and specificity to degrade either tau or β-amyloid intercellularly and extracellularly in the late stage, the binding affinities of the system parts were validated using pull-down assay, while the switchable ability and specificity of HTRA1 were validated by protease assay and compared with different controls to prove that all systems are valid and effective in treating Alzheimer’s disease whether in early or late stages.

Snitch system


Fig. 1: Graphical illustration showing snitch system composition.



Overview


We were inspired by (team NUDT China 2020) for using the PREDATOR Pro system with some modifications for the degradation of Alzheimer’s pathogenesis which are tau tangles. The Trim system is carried out based on PROTACS. Throughout the past decade, Intense drug discovery research has focused on a type of bifunctional compound known as proteolysis-targeting chimeras or PROTACs. These molecules, a subtype of the larger categories of molecular glues and chemical inducers of proximity, act by non-covalently linking a targeted protein to the substrate adaptor protein of an E3 ligase (Trim21), which naturally functions to polyubiquitinate proteins in vivo, the labeled targeted protein which is tau, transported to the Ubiquitin Proteasome System (UPS), where it is degraded by the 26S proteasome [8].



The snitch system consists of three parts


1. A moiety fused with TBP (TD28rev and www) that binds to the target protein (Coh2).

2. Flexible linker (G4S linker).

3. A second moiety (DocS) fused with TRIM21, once binding occurs, polyubiquitination of targeted protein (tau) will be activated which in turn recruits proteasomal degradation.

Note


In order to target tau proteins, we tried to search for binding domains that could tightly bind tau with the linker, we settled on using tau fibril binding peptides obtained by employing mirror image phage display. The chosen peptides which are TD rev28 (PPYYLRMQLSTT) and WWW (DVWWWNKKRK) consist of D-enantiomeric amino acids. We consider using D-peptides since they can’t be affected by proteases which means that they are preferably stable. In addition, they don’t exhibit toxic immunogenic response like L-peptides [6,29].


Note


PROTAC linkers was a critical choice for us. The physical and biological features of PROTACs are highly dependent on the linker's length and composition. The overall degradation efficiency does not depend on the affinities of Coh2 and DocS for E3 and tau, respectively, but rather on the combination of Coh2 and DocS with a proper linker, allowing productive ternary complex (TC) formation and tau ubiquitination [34]. That’s why linker choice was not easy, we had several discussions about linkers trying to answer two questions

(1) Should we use flexible or rigid linkers?


(2) How we can optimize the length of the linkers that suit our target without any alterations in the function of DocS and Coh2?



We came up with the following points. First, decided to predict the 3D structure of linkers fused with the following parts, Trim-Linker-DocS and Coh2- Linker- TBP with a library of flexible and rigid linkers obtained from literature and iGEM linkers library. We used the previously mentioned servers (step 1) to predict the 3D structure and the ranking code (step 3) to evaluate the quality of the structure. According to our investigations and the available literature, we determined that flexible linkers mainly (G4S linker) would work best for our system. Second, we followed the rational of Team iGEM_NUDT for linker length [17,24].

Proof of concept from dry lab aspect



Fig. 2: Schematic illustration of dry lab workflow



In synthetic biology, theoretical models are frequently employed for gaining insights, making predictions, and enhancing experiments. In our project, we focused on computationally simulating the trim system through molecular docking to test the binding affinity between the listed parts in table1. For molecular docking we had to choose protein-protein docking servers. We settled on using 3 different software, LightDock based on the Glowworm Swarm Optimization (GSO) algorithm [13,23,25]. Cluspro 2.0 based on the Fast Fourier Transform (FFT) algorithm [7,15,16,35] and GalaxyTongDock that performs ab-initio rigid-body docking to predict complex structure of two proteins [22].



Procedures



Step 1


The first step in performing molecular docking is to obtain PDB files for the proteins. The 3D structure for most of trim system’s parts were unknown. We used different servers for 3D structure prediction such as:

1. TrRosetta [9,35] and RoseTTAFold [19] for homology modeling. Both tools use a deep learning-based method to construct protein structures based on direct energy minimizations with a restrained Rosetta

2. MODELLER for homology or comparative modeling of protein 3D structure [10].

3. AlphaFold [14] uses the deep learning algorithm to predict protein structure

Step 2


Evaluation of the predicted 3D structure obtained by the above stated servers through using SWISS-MODEL structure assessment to provide further structural details [3].

Step 3


Ranking code (you can find python script file in programming club page with further explanation of how you can optimize it upon your needs). This code works by ranking the predicted 3D structures and giving them a score out of 6. In addition, we took into consideration the guidelines of the structure assessment (step 2) since the evaluation of the predicted structure depend on the values of Ramachandran Plots, MolProbity, Clash Score, C-Beta Deviations and QMEAN.

Step 4


Now we have the 3D structure for further investigations/ operations like molecular docking or molecular dynamic simulation.

Step 5


In our project, we used three distinct programs: lightdock, cluspro, and galaxy tongdock for protein-protein docking. We required a scoring algorithm to quantify the binding affinity between protein-protein complexes to obtain the most stable docked conformation. We used PRODIGY to estimate the binding affinity score to rank all complexes generated by various docking programs.

Step 6


Post-MD analysis.


Mathematical modeling for Translation and Transcription

Description


The transcription model is based on the IPTG induction pathway of recombinant protein under the T7 promoter and lac operator. The model assumes that the transcription repressor (lacI dimer) is transcribed and translated from the plasmid [1]. So, in the beginning, the model simulates the transcription and translation of lacI followed by dimerization. During the time needed for lacI dimer formation, the recombinant DNA is transcribed normally, then lacI dimer binds to lacO and stops the elongation of RNA polymerase. However, since the lacI/lacO complex formation is reversible at a certain time, lacI may separate from lacO, leaving RNA polymerase to work and form mRNA; the amount of mRNA formed due to this process is called lacI leakage. lacI leakage is integrated into our model. Therefore, when IPTG initial concentration is set to zero, there is a slight observed mRNA concentration and consequently protein. In the model, we also assume that the IPTG diffuses into the cell through the cell membrane by passive diffusion only. The IPTG induction starts at the same time as lacI transcription. The process of repression and derepression coexist simultaneously; after a certain period, the system reaches equilibrium, and the mRNA and protein concentration tends to stay the same.


Methodology


The model also neglects the effect of protein and mRNA concentration on bacterial metabolism, which may reduce their concentration after a certain period. But it considers the degradation of mRNA and protein. Each step in the process is written as a chemical reaction, then the rate laws of each step are formulated and the change of concentration of each compound with time is calculated using ODE equations which is the summation of all rate laws the compounds are involved in by being formed or consumed. The code was written in python 3.9 and ran on PyChram Community Edition v2021.1.1 . It uses external python packages NumPy, SciPy and Matplotlib. NumPy is used to create time array, SciPy solves the ODEs in the defined function using the function odeint, matplotlib package for plotting the data.


Results



I. His TRIM




Fig.1: this figure shows the behaviour and the concentration of mRna and protein for His trim (1Mm) according to our Mathematical Model.


II. His Trim-Linker-DOCs



Fig.2: this figure shows the behaviour and the concentration of mRna and protein for His Tau-linker-Docs with 2 different IPTG concentration 1mM(on the left) , 2mM (on the right) according to our Mathematical Model.


III. His DOC



Fig.3: this figure shows the behaviour and the concentration of mRna and protein for His Coh (1mM ) according to our Mathematical Model.


IV. GST DOCs



Fig.4: this figure shows the behaviour and the concentration of mRna and protein for GST Docss with 2 different IPTG concentration 1mM (on the left), 2mM (on the right) according to our Mathematical Model.


V. His COH



Fig. 5: this figure shows the behaviour and the concentration of mRna and protein for His Coh (1mM ) according to our Mathematical Model.


VI. GST COH



Fig.6: this figure shows the behaviour and the concentration of mRna and protein for GSt Coh (1mM) according to our Mathematical Model.


VII. GST COH linker TD28rev



Fig.7: this figure shows the behaviour and the concentration of mRna and protein for Coh-linker-TD28rev (1Mm) according to our Mathematical Model.


VIII. GST COH linker WWW



Fig.8: this figure shows the behaviour and the concentration of mRna and protein for Coh-linker-www (1Mm) according to our Mathematical Model.


IX. His Tau



Fig.9: this figure shows the behaviour and the concentration of mRna and protein for His tau (1Mm) according to our Mathematical Model.

Proof of binding from dry and wet lab aspects

Table1. List of docked complexes in trim system

Part

Description

DocS- Coh2

Two complementary modules.

Tau- Tau binding peptide (TBP)

Tau: Major Alzheimer’s pathogenesis.
TBP: Binding peptides for tau fibrils, which are TDrev28 and WWW

His- trim21, linker and DocS

His: tag for protein purifications
Trim: E3 ligase
Linker: G4S flexible linker
Docs S: module that binds tightly to Coh2

GST- coh2, linker and TBP

GST: tag for protein purifications
Coh2: module that binds tightly to DocS
linker: G4S flexible linker
TBP: Binding peptides for tau fibrils.

Whole PROTAC system with tau protein

PROTAC system: illustrated in fig.1

Part 1: The interaction between DocS and
Coh2


Both domains must be tagged for purification procedures. We selected the 6xHis tag, however we discovered in the literature that DocS has low yield and stability, necessitating that it should be linked to a tag other than 6xHis for greater stability. Therefore, we choose to combine DocS with GST tag. Then, we chose to express both domains using GST and 6xHis tags in order to assess their yield and stability and determine which one is preferable.

Procedures


We got the 3D structure for both fusion proteins, the highest ranked 3D structure for His- Docs / His-Coh2 was modeled by (TrRosetta), and GST-coh2 / GST-DocS were both modeled by Alphafold

Results


Molecular docking was used to analyze the interaction between Coh2 and DocS; using Galaxy and ClusPro the scores for GST-Coh2 and His-Doc using were -13.153 and -11.635 kcal/mol, respectively. And the corresponding results for GST-DocS and His-Coh2 were -13.488 and -14.026 kcal/mol. (Fig. 1.) represents the docked pose between the interacted domains.


Fig. 1. illustrates the interaction between Coh2 and DocS when fused with both tags GST and 6xHis, using ClusPro and Galaxy. Structures were visualized by PyMol.



Part 2: Tau and tau binding peptide


Results


Top models of tau binding peptides were WWW peptide modeled by i-TASSER and TD28rev modeled by AppTest.


Fig. 2. represents the interaction between tau whole filaments and TD28rev peptide using Galaxy software. The estimated binding score was -175.066 kcal/mol.


Fig. 3. represents the interaction between tau whole filaments and WWW peptide using Galaxy software. The estimated binding score was -171.237 kcal/mol.

Part 3 and 4: His-Trim21-linker-DocS Vs Gst-
Coh2-linker-TBP


Procedures


We get the 3D structure for both fusion proteins, the highest ranked 3D structure for His, trim2, linker, Docs was modeled by (TrRosetta, AlphaFold, Robetta, Modeller), For the second moiety, Gst, coh2, linker, TBP (WWW and TD28rev) was modeled by trRosetta.

Results


Table 2. Demonstrates the result of docking poses with the corresponding affinity score.

Moiety

His, trim21, linker, docS

GST, Coh, linker, TBP

TBP

WWW

TD28rev

Affinity score

(-14.139) and (-15.235) kcal/mol, using lightdock and cluspro respectively

(-10.43) and (-18.729) kcal/mol, using lightdock and cluspro respectively




Fig. 4. illustrates the docking mode between the fusion protein. These structures were visualized using PyMol.



Fig. 5. Graphical representation showing the change in concentration of Trim21-linker-DocS, GST-Coh2- linker-tau binding peptide and their complex with time.


Part 5: Whole trim system with tau protein


In the final docking stage of the Trim system, we wanted to make sure that the entire system (Trim-linker-DocS , GST-Coh2-linker-TBP) is bound to tau protein. Initial docking was performed blindly, but the findings showed that tau models did not connect to the binding peptide region. Therefore, docking is repeated by establishing a restrain file and evaluating the affinity to determine whether tau will bind or not.

Fig. 6. represents the docking result between the trim system and tau.


Proof of concept from WET lab perspective


Our system consists of various parts that bind or interact with each other. To validate the system's ability to call ubiquitin molecules and recruit proteasomal degradation of the targeted protein (tau), we have to determine the binding and interaction of each part alone and then the interaction of the whole system to prove that all system is valid and effective in treating Alzheimer disease.

I. DOCs-COH2


Description:


In order to test the binding affinity between our PROTAC domains, we choose to express DOC and COH two times fused with two different tags (His-tag and Gst-tag) in Bl21 therefore we can con detect the binding affinity using pulldown assay using (Invitrogen Ni-NTA Agarose, cat.no:R901-15) and overcoming the low expression yield of these domains per se as well as determine the optimum expression conditions to get maximum yield and binding affinity. [2,5].

Methodology


The pull-down technique is a method for performing ligand binding assays with recombinant proteins capable of binding to an affinity matrix in the presence or absence of a denaturing agent (that can be replaced by a physiological buffer). The method entails coupling a known amount of protein to a known amount of affinity matrix and then using the suspension of this protein-coupled matrix as the source of the recombinant protein to analyze direct protein-protein interactions. That will include a purified and tagged protein as "bait" to bind any interacting proteins(prey). The procedure begins with immobilizing the tagged protein (DOCs) on an affinity ligand unique to the tag, followed by the construction of affinity support to capture and purify additional proteins (COH2) that interact with the (DOCs). Following incubation of the prey (COH2) proteins with an immobilized bait (DOCS) protein, interaction complexes are eluted with an affinity ligand-specific eluting solution. SDS-PAGE will be used to detect binding. [18,30]

Results



Fig. 7.: BCA assay shows binding affinities between our PROTAC domains DOCs and COH2.

According to binding validation between our PROTAC domains, BCA results show that interaction between (His-DOCs) and (GST-COH2) stabilizes the binding between the two domains rather than the interaction between (HIS-COH2) and (GST-DOCs).

II. GST-COH-TBP against aggregated His-Tau


Description


Tau and tau binding peptides will be expressed in Escherichia coli strain BL21 (DE3) and induced with isopropyl-β-d-thiogalactopyranoside (IPTG) then, the interaction between them will be tested by Pull-down assay and visualized on SDS-PAGE. 1st why D-Peptides? Several strategies have already been used to overcome peptides' high protease sensitivity and improve blood-brain barrier permeability. The use of D-enantiomeric amino acids, which are assumed to be more protease resistant and often less immunogenic than the corresponding L-peptides, becomes a strategy for improving peptide stability. [27,28]

• Tau vs Td28rev

Using Thioflavin fluorescence assays, the peptideTD28rev was tested for its ability to inhibit the fibrillization and aggregation of PHF6, tau3RD (three repeat domain construct of tau protein, also known as K19), and full-length tau, and it showed an inhibitory effect, which was confirmed by DLS. [12]

• Tau vs WWW

WWW, a phase 2 inhibitor derived from one of VQIINK inhibitors (MINK) and containing Trp at positions 3, 4, and 5 but it failed to inhibit, possibly due to a destabilizing effect of multiple tryptophans in series on capping onto the amyloid chain. [29] In order to detect the binding between TBP and TAU, we choose to express TBP(TDrev28) fussed with [GST-COH] and express tau tagged with His-tag therefore we can confirm the binding affinity using pulldown assay using (Invitrogen Ni-NTA Agarose, cat.no: R901-15). [2,5,32].


Methodology


The pull-down technique is a method for performing ligand binding assays with recombinant proteins capable of binding to an affinity matrix in the presence or absence of a denaturing agent (that can be replaced by a physiological buffer). The method entails coupling a known amount of protein to a known amount of affinity matrix and then using the suspension of this protein-coupled matrix as the source of the recombinant protein to analyze direct protein-protein interactions. That will technique includes a purified and tagged protein as "bait" to bind any interacting proteins(prey). The procedure begins with immobilizing the tagged protein (his-tau) on an affinity ligand unique to the tag, followed by the construction of affinity support to capture and purify additional proteins (GST-COH2-TBP) that interact with the (his-tau). Following incubation of the prey (GST-COH2-TBP) proteins with an immobilized bait (his-tau) protein, interaction complexes are eluted with an affinity ligand-specific eluting solution. SDS-PAGE will be used to detect binding [18,30]

Results



Fig. 8.: BCA assay shows that tau aggregated has more bunding affinity to bind to the WWW binding peptide than TD28REV

According to validation of tau aggregated -TBPs binding affinity, we can conclude from BCA results that tau aggregated bind successfully to TBPs with more binding affinity to (www)rather than (TD28REV).

III. [TRIM-DOC] against [COH-TBP]


Description


In order to confirm the binding affinity [TRIM-DOC] and [COH-TBP], we choose to express trim21 fussed with DOC. Domain and tagged with His-tag, unlike TBP(TDrev28) express it fused with COH. Domain and tagged with Gst tag, therefore, we can confirm the binding affinity using pulldown assay using (Invitrogen Ni-NTA Agarose, cat.no: R901-15) . [4,21,32].

Methodology


The pull-down technique is a method for performing ligand binding assays with recombinant proteins capable of binding to an affinity matrix in the presence or absence of a denaturing agent (that can be replaced by a physiological buffer). The method entails coupling a known amount of protein to a known amount of affinity matrix and then using the suspension of this protein-coupled matrix as the source of the recombinant protein to analyze direct protein-protein interactions. That will technique includes a purified and tagged protein as "bait" to bind any interacting proteins(prey). The procedure begins with immobilizing the tagged protein (trim-docs) on an affinity ligand unique to the tag, followed by the construction of affinity support to capture and purify additional proteins(coh2-TBP) that interact with the (trim-doc). Following incubation of the prey (coh2-TBP) proteins with an immobilized bait(trim-Docs) protein, interaction complexes are eluted with an affinity ligand-specific eluting solution. SDS-PAGE will be used to detect binding. [18,30]

Results



Fig. 9.: BCA assay shows that (TRIM-L-DOCs) has more bunding affinity to bind to the WWW binding peptide than TD28REV

The results show that the Trim-L-Doc can bind the Coh2 fused to WWW more efficiently than that fused with TD28rev peptide, proving that in both cases TBP didn’t hinder the binding affinity of our two interacting modules.

IV. Trim-21 whole system vs Tau aggregated


Description


Protein degraders that target proteolysis (PROTAC). These are heterobifunctional small molecules made up of two ligands linked together by a linker: one recruits and binds a protein of interest (POI), while the other recruits and binds an E3 ubiquitin ligase. The simultaneous binding of the POI and ligase by the PROTAC causes ubiquitylation and subsequent degradation of the POI by the ubiquitin-proteasome system (UPS), after which the PROTAC is recycled to target another copy of the POI [1].Ubiquitin can be added to a substrate protein as a protein tag through the coordinated actions of ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzyme (E2), and ubiquitin-protein ligase (E3). The formation of thioesters can determine E2 activity in the presence of E1 and ubiquitin. In vitro ubiquitination assays can be used to investigate a putative protein's E3 activity as well as its E2/E3 or E3/substrate specificities. A western blot with a specific antibody can detect the result. [37] The Trim system is inspired by (PROTAC) and it consists of Trim fused with Docs, Coh2, and tau binding peptide.

Methodology


Trim-21, Docs, Coh2, and tau binding peptide will be expressed fused together in Escherichia Coli BL21 (DE3) to form two composite parts [trim-21-Docs] and[coh2-TBP] & induced with isopropyl-β-d-thiogalactopyranoside (IPTG) and tau will be expressed in Escherichia Coli BL21 (DE3) & induced with isopropyl-β-d-thiogalactopyranoside (IPTG)as well. aggregates will be formed by using 3 methods (SAPK4 & PKa – SAPK4 & PKa + heparin – heparin only – self-aggregation) then, the invitro ubiquitination assay will be done to confirm that our system can ubiquitinate misfolded and aggregated protein.


Result



Fig. 10.: BCA assay shows that the whole system (TRIM-L-DOCs and GST-coh2-www) could bind to tau aggregated.

Whereas the elutions of the pulldown assay proved that the total system is working perfectly, and each binding partner can interact with the other and fusion using the flexible linker allowed the system appropriate configuration without interfering with binding affinity


Therefore, we can say that our snitch system is valid and effective to recognize misfolded tau aggregations according to binding affinities and validation of the interaction between each part alone, and also this proved that the whole system is working, but unfortunately, we did not receive our Promega grant before the wiki-freeze day so, this did not enable us to do an In-vitro Ubiquitination assay to confirm that our system is able to call ubiquitin which in turn will activate the proteasomal degradation cascade as we hypothesized. You can find more about challenges we faced in the obstacles page.


HTRA1 overview

Regarding validation of the HTRA1 system's ability and specificity to degrade either tau or β-amyloid intracellularly and extracellularly in the late stage, the binding affinities of each interacting partner were validated both on the dry lab level by a specified pipeline starting with 3D modeling till reaching to predict the protease activity of HTRA and on experimental level following cloning workflow then performing quantitative and qualitative estimation of the binding using BCA assy and pull-down assay, respectively. while the switchable ability and specificity of HTRA1 were validated by protease assay and compared with different controls to prove that all systems are valid and effective in treating Alzheimer’s disease, whether in the early or late stages.



Fig. 1.: A graphical illustration of HTRA1 system



HtrA1 switchable system

Trimerization of HTRA1


HTRA1 is a trimer in nature. Each monomer is composed of four different domains: a C-terminal PDZ domain, a trypsin-like catalytic domain, a Kazal-like domain, and an IGFBP-like domain. We could not find the complete trimer structure in the protein data bank. We had to rely on modeling tools to obtain the complete form of HTRA for further processes such as molecular docking and molecular dynamic simulation to validate our system. We used the GalaxyHeteromer webserver for the trimerization [32]


Fig. 2. The 3D structure of HTRA1 as a trimer



Inhibitory domain


To regulate HTRA1 activity, extensive research for HTRA1 inhibitors was performed. We compiled a sizable inhibitors library and followed the same workflow described previously (3D structure prediction and quality assessment ranking code). The top-rated structures were WAP four-disulfide core domain protein retrieved from UniProt (WFDC2_HUMAN) and human serine protease inhibitor known as (SPINK8) in Uniprot . We wanted to make sure that the inhibitors bound to the catalytic domain of HTRA1 to stop its activity. To accomplish this, we used molecular docking simulation.



Fig. 3. illustrates the binding between HTRA1 and WAP inhibitors. The top docked model was obtained by Lightdock with a restrained file (more about this can be found on the docking page) and the estimated binding affinity was -71.937 kcal/mol.




Fig. 4. illustrates the binding between HTRA1 and SPINK8.The top docked model obtained by ClusPro also with a restrained file, and the estimated binding affinity was -25.0 kcal/mol.


To validate our dry lab results, we chose to build two HTRA1 inhibitors, one with a high binding affinity and the other with a low binding affinity, to discover which one would be superior when employed in the wet lab. This concept will be tested using the protease assay technique

Affinity clamps


We designed the HTRA1 system to be a dual system for degrading both beta-amyloid and tau fibrils. To accomplish this, we started looking for peptides to be appended to the protein of interest (POI) [31] revealed that rational design and assembly of peptides enabled their use as modules to control the activity of proteases such as HTRA1 and modulate downstream signals. A modular design technique was used to create a transducer protease coupled to an auto-inhibitory (AI) domain [31].

Here, fig. 5. represents the affinity clamp for beta-amyloid. Both peptides are colored in blue and appended to a linker (L2) from its C/N terminus. The linker length covered a distance of 4 to 3 amino acids. After many modeling iterations with different linker lengths found in the literature. We found that the optimum length was 25 and 32 Å to connect the seed clamp and TD28REV with WWW, respectively.


Fig. 5. Clamp of beta amyloid




Fig. 6. clamp of tau protein



Binding peptides for HTRA1


A well-studied mechanism for activating the HTRA1 family reveals that the attachment of hydrophobic peptides to the PDZ domain induces conformational changes that activate the proteolytic activity of HTRA1. Ultimately this process facilitates the recognition and destruction of misfolded proteins [26]. Based on the literature, we obtained a wide variety of binding peptides. (To get their 3D structure, we used benchmarked modeling tools, followed by running a quality assessment check on Swiss Structure assessment then all models were ranked by our ranking code)



Fig. 7. represents the docked pose of HTRA1 with HTRA1 binding peptide. The top model obtained by galaxy webserver and the estimated binding affinity was -33.622 Kcal/mol

Engineering switchable system.


To activate HTRA1, a conformational change in the linker is required, which eliminates the attached inhibitor from the active site. The conformational rearrangement can be mediated through the affinity clamp binding and stabilizing the inhibitor away from the active site of these two domains (inhibitor and affinity clamp connected with L1). Additionally, (H1A) binding peptide bound to the PDZ domain and is connected to the affinity clamp on the other side with L3.

To proceed with the design of HTRA1 system, we must answer two questions:

1. How can we calculate the length of Linker 1 and Linker 3?
2. How can we be sure about the on/off switch of the system?
3. Designing Linker 1 and Linker 3.


Fig. 8. represents the 3D structural modeling of the switches we used in HTRA1 system



We choose to align the model of HTRA1 bound to the inhibitor with the model of H1A peptide bound to the PDZ domain of HTRA. Subsequently, the distance between the C/N terminus of both the inhibitor and H1A peptide was calculated and subtracted from the length of L2 (which links the affinity clamp). The distance of L1 and L2 were 12 and 21, respectively, which equal to 4 to 7 Amino acids approximately. The following figure represents our chosen switches which are comprised of (Inhibitor-Linker1-clamp harbors linker2- linker3-HTRA1 binding peptide)

Switching on/off


To prove that the HTRA1 system would be a switchable device triggered when the affinity clamp harbors the targeted proteins, tau, or beta-amyloid. We settled on evaluating the binding energy between HTRA1 and all the domains associated with it as the evidence that will prove the validity of our hypothesis. Affinity scores obtained by prodigy have proven our hypothesis illustrated in (Fig. 9.)




Fig. 9. represents the energy gradient between HTRA1 and all its associated domains



Proof of concept from WET-lab perspective



I. HTRA1 Activity:


In order to prove that our recombinant HTRA1 is valid and functionally active as a serine protease, therefore the protease activity was assayed by incubating HtrA1 with substrate proteins (β-casein, BSA, or other proteins)

II. HTRA1 inhibition by our two inhibitors (Q8 AND POC.)


In order to prove that our Two inhibitors can inhibit HTRA1 activity from determining the specificity of our system, the protease activity was assayed by incubating HtrA1 and inhibitors with substrate proteins (β-casein, BSA, or other proteins).

III. HTRA1 against tau and β-amyloid


To prove that our recombinant HTRA1 is valid and functionally active as serine protease against tau and β-amyloid, therefore the protease activity was assayed by incubating HtrA1 with substrate proteins (β-amyloid and tau).

IV. HTRA1 whole system against tau and β-amyloid


In order to prove that our assembled system that includes recombinant HTRA1 is valid and functionally active as serine protease against tau and β-amyloid, the protease activity was assayed by incubating HtrA1 whole system with substrate proteins (β-amyloid and tau).

V. HTRA1 vs Whole assembled system


Description


HTRA1 itself and the whole assembled system will be tested against each other by protease assay to detect protease activity and to prove that our system can be switched on and off.

Methodology



The protease activity was assayed by incubating HtrA1 or its derivatives with substrate proteins (5–3 µg of β-casein, BSA, or our target proteins [aggregated tau and amyloid-beta) in a 20 µl mixture containing 50 mM Tris/HCl (pH 7.6) either with or without 1.5 mM DTT for 0.5–8 h at 37 ◦C. Then degradation was detected by SDS-PAGE (MURWANTOKO et al., 2004)

VI. HTRA1 VS H1A


Description


H1A binding peptide is supposed to interact with the PDZ domain in trimeric HTRA1 and it will be proved by pull-down assay. They will be expressed in Escherichia coli strain BL21 (DE3) and induced with isopropyl-β-d-thiogalactopyranoside (IPTG) then, the interaction between them will be tested by Pull-down assay and visualized on SDS-PAGE.

Methodology

The pull-down technique is a method for performing ligand binding assays with recombinant proteins capable of binding to an affinity matrix in the presence or absence of a denaturing agent (that can be replaced by a physiological buffer). The method entails coupling a known amount of protein to a known amount of affinity matrix and then using the suspension of this protein-coupled matrix as the source of the recombinant protein to analyze direct protein-protein interactions. That will include a purified and tagged protein as "bait" to bind any interacting proteins(prey). The procedure begins with immobilizing the tagged protein (HTRA1) on an affinity ligand unique to the tag, followed by the construction of affinity support to capture purify additional proteins(H1A) that interact with the (HTRA1). Following incubation of the prey (H1A) proteins with an immobilized bait (HTRA1) protein, interaction complexes are eluted with affinity ligand-specific eluting solution. SDS-PAGE will be used to detect binding. [18,30]

VII. Amyloid beta vs amyloid binding peptide (Aβ37-42)


Description


It’s hypothesized that a peptide derived from the Aβ42 C terminus would be incorporated into Aβ42 oligomers, disrupt their structure, and thus inhibit their toxicity.

The peptide was discovered to protect cells from A42-induced toxicity. and the selected peptide demonstrated an inhibitory activity and rescued the cells from Aβ42 -induced toxicity. [11]

Methodology



The pull-down technique is a method for performing ligand binding assays with recombinant proteins capable of binding to an affinity matrix in the presence or absence of a denaturing agent (that can be replaced by a physiological buffer). The method entails coupling a known amount of protein to a known amount of affinity matrix and then using the suspension of this protein-coupled matrix as the source of the recombinant protein to analyze direct protein-protein interactions. That will include a purified and tagged protein as "bait" to bind any interacting proteins(prey). The procedure begins with immobilizing the tagged protein (amyloid-beta) on an affinity ligand unique to the tag, followed by the construction of affinity support to capture and purify additional proteins (amyloid-beta peptide) that interact with the (amyloid-beta). Following incubation of the prey (amyloid-beta peptide) proteins with an immobilized bait (amyloid-beta) protein, interaction complexes are eluted with an affinity ligand-specific eluting solution. SDS-PAGE will be used to detect binding. [18,30]

VIII. Amyloid beta vs tau binding peptide AND Tau vs amyloid binding peptide


Description


We are trying to produce a dual-function system to catch tau aggregates and amyloid beta plaques, so the binding affinity between Tau and amyloid-beta binding peptide & amyloid beta and tau binding peptides will be tested by pull-down assay.

Methodology



The pull-down technique is a method for performing ligand binding assays with recombinant proteins capable of binding to an affinity matrix in the presence or absence of a denaturing agent (that can be replaced by a physiological buffer). The method entails coupling a known amount of protein to a known amount of affinity matrix and then using the suspension of this protein-coupled matrix as the source of the recombinant protein to analyze direct protein-protein interactions. That will include a purified and tagged protein as "bait" to bind any interacting proteins(prey). The procedure begins with immobilizing the tagged protein (amyloid-beta or tau) on an affinity ligand unique to the tag, followed by the construction of affinity support to capture and purify additional proteins (amyloid-beta or tau peptides) that interact with the (amyloid-beta or tau). Following incubation of the prey (amyloid-beta or tau peptide) proteins with an immobilized bait (amyloid-beta or tau) protein, interaction complexes are eluted with an affinity ligand-specific eluting solution. SDS-PAGE will be used to detect binding. [18,30]

Unfortunately, the shipment of Plug-Sink parts has been delayed massively. We did not even get to work on the parts of this system experimentally as we did not receive them before the wiki-freeze day. Also, we did not receive our Promega grant as well. So, this played a major part in our project’s delay.You can find more about challenges we faced in the obstacles page


References

1- Békés, M., Langley, D.R. & Crews, C.M. PROTAC targeted protein degraders: the past is prologue. Nat Rev Drug Discov 21, 181–200 (2022). https://doi.org/10.1038/s41573-021-00371-6

2-Bayer, E., Morag, E., & Lamed, R. (1994). The cellulosome — A treasure-trove for biotechnology. Trends In Biotechnology, 12(9), 379-386. https://doi.org/10.1016/0167-7799(94)90039-6

3-Benkert, P., Biasini, M., & Schwede, T. (2010). Toward the estimation of the absolute quality of individual protein structure models. Bioinformatics, 27(3), 343-350. https://doi.org/10.1093/bioinformatics/btq662

4-Benn, J., Mukadam, A., & McEwan, W. (2022). Targeted protein degradation using intracellular antibodies and its application to neurodegenerative disease. Seminars In Cell &Amp; Developmental Biology, 126, 138-149. https://doi.org/10.1016/j.semcdb.2021.09.012

5-Craig, S., Foong, F., & Nordon, R. (2006). Engineered proteins containing the cohesin and dockerin domains from Clostridium thermocellum provides a reversible, high affinity interaction for biotechnology applications. Journal Of Biotechnology, 121(2), 165-173. https://doi.org/10.1016/j.jbiotec.2005.07.005

6-Dammers, C., Yolcu, D., Kukuk, L., Willbold, D., Pickhardt, M., & Mandelkow, E. et al. (2016). Selection and Characterization of Tau Binding ᴅ-Enantiomeric Peptides with Potential for Therapy of Alzheimer Disease. PLOS ONE, 11(12), e0167432. https://doi.org/10.1371/journal.pone.0167432

7-Desta, I., Porter, K., Xia, B., Kozakov, D., & Vajda, S. (2020). Performance and Its Limits in Rigid Body Protein-Protein Docking. Structure, 28(9), 1071-1081.e3. https://doi.org/10.1016/j.str.2020.06.006

8-Drummond, M., & Williams, C. (2019). In Silico Modeling of PROTAC-Mediated Ternary Complexes: Validation and Application. Journal Of Chemical Information And Modeling, 59(4), 1634-1644. https://doi.org/10.1021/acs.jcim.8b00872

9-Du, Z., Su, H., Wang, W. et al. The trRosetta server for fast and accurate protein structure prediction. Nat Protoc 16, 5634–5651 (2021). https://doi.org/10.1038/s41596-021-00628-9

10-Eswar, N., Webb, B., Marti‐Renom, M., Madhusudhan, M., Eramian, D., & Shen, M. et al. (2006). Comparative Protein Structure Modeling Using Modeller. Current Protocols In Bioinformatics, 15(1). https://doi.org/10.1002/0471250953.bi0506s15

11-Fradinger, E., Monien, B., Urbanc, B., Lomakin, A., Tan, M., & Li, H. et al. (2008). C-terminal peptides coassemble into Aβ42 oligomers and protect neurons against Aβ42-induced neurotoxicity. Proceedings Of The National Academy Of Sciences, 105(37), 14175-14180. https://doi.org/10.1073/pnas.0807163105

12-Funke, S., Dammers, C., Yolcu, D., Kukuk, L., Rudolph, S., & Willbold, D. (2013). O2–08–03: Selection and characterization of tau‐binding D‐enantiomeric peptides for therapeutic applications in neurodegenerative diseases. Alzheimer's &Amp; Dementia, 9(4S_Part_8). https://doi.org/10.1016/j.jalz.2013.04.179

13-Jiménez-García, B., Roel-Touris, J., Romero-Durana, M., Vidal, M., Jiménez-González, D., & Fernández-Recio, J. (2017). LightDock: a new multi-scale approach to protein–protein docking. Bioinformatics, 34(1), 49-55. https://doi.org/10.1093/bioinformatics/btx555

14-Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., & Ronneberger, O. et al. (2021). Highly accurate protein structure prediction with AlphaFold. Nature, 596(7873), 583-589. https://doi.org/10.1038/s41586-021-03819-2

15-Kozakov, D., Hall, D., Xia, B., Porter, K., Padhorny, D., & Yueh, C. et al. (2017). The ClusPro web server for protein–protein docking. Nature Protocols, 12(2), 255-278. https://doi.org/10.1038/nprot.2016.169

16-Lavi, A., Ngan, C., Movshovitz-Attias, D., Bohnuud, T., Yueh, C., & Beglov, D. et al. (2013). Detection of peptide-binding sites on protein surfaces: The first step toward the modeling and targeting of peptide-mediated interactions. Proteins: Structure, Function, And Bioinformatics, 81(12), 2096-2105. https://doi.org/10.1002/prot.24422

17-Li, G., Huang, Z., Zhang, C., Dong, B., Guo, R., & Yue, H. et al. (2015). Construction of a linker library with widely controllable flexibility for fusion protein design. Applied Microbiology And Biotechnology, 100(1), 215-225. https://doi.org/10.1007/s00253-015-6985-3

18-Louche, A., Salcedo, S., & Bigot, S. (2017). Protein–Protein Interactions: Pull-Down Assays. Methods In Molecular Biology, 247-255. https://doi.org/10.1007/978-1-4939-7033-9_20

19- M, B., F, D., I, A., J, D., S, O., & GR, L. et al. (2022). Accurate prediction of protein structures and interactions using a three-track neural network. Yearbook Of Paediatric Endocrinology. https://doi.org/10.1530/ey.19.15.15

20-MURWANTOKO, YANO, M., UETA, Y., MURASAKI, A., KANDA, H., OKA, C., & KAWAICHI, M. (2004). Binding of proteins to the PDZ domain regulates proteolytic activity of HTRA1 serine protease. Biochemical Journal, 381(3), 895–904. https://doi.org/10.1042/bj20040435

21-Neklesa, T., Winkler, J., & Crews, C. (2017). Targeted protein degradation by PROTACs. Pharmacology &Amp; Therapeutics, 174, 138-144. https://doi.org/10.1016/j.pharmthera.2017.02.027

22-Park, T., Baek, M., Lee, H., & Seok, C. (2019). GalaxyTongDock: Symmetric and asymmetric ab initio protein–protein docking web server with improved energy parameters. Journal Of Computational Chemistry, 40(27), 2413-2417. https://doi.org/10.1002/jcc.25874

23-Roel-Touris, J., Jiménez-García, B., & Bonvin, A. (2020). Integrative modeling of membrane-associated protein assemblies. Nature Communications, 11(1). https://doi.org/10.1038/s41467-020-20076-5

24-Reddy Chichili, V., Kumar, V., & Sivaraman, J. (2013). Linkers in the structural biology of protein-protein interactions. Protein Science, 22(2), 153-167. https://doi.org/10.1002/pro.2206

25-Roel-Touris, J., Bonvin, A., & Jiménez-García, B. (2019). LightDock goes information-driven. Bioinformatics, 36(3), 950-952. https://doi.org/10.1093/bioinformatics/btz642

26-Runyon, S., Zhang, Y., Appleton, B., Sazinsky, S., Wu, P., & Pan, B. et al. (2007). Structural and functional analysis of the PDZ domains of human HtrA1 and HtrA3. Protein Science, 16(11), 2454-2471. https://doi.org/10.1110/ps.073049407

27-Sadowski, M., Pankiewicz, J., Scholtzova, H., Ripellino, J., Li, Y., & Schmidt, S. et al. (2004). A Synthetic Peptide Blocking the Apolipoprotein E/β-Amyloid Binding Mitigates β-Amyloid Toxicity and Fibril Formation in Vitro and reduces β-Amyloid Plaques in Transgenic Mice. The American Journal of Pathology, 165(3), 937-948. https://doi.org/10.1016/s0002-9440(10)63355-x

28-Schumacher, T., Mayr, L., Minor, D., Milhollen, M., Burgess, M., & Kim, P. (1996). Identification of d -Peptide Ligands Through Mirror-Image Phage Display. Science, 271(5257), 1854-1857. https://doi.org/10.1126/science.271.5257.1854

29-Seidler, P., Boyer, D., Rodriguez, J., Sawaya, M., Cascio, D., & Murray, K. et al. (2017). Structure-based inhibitors of tau aggregation. Nature Chemistry, 10(2), 170-176. https://doi.org/10.1038/nchem.2889

30-Sinha D, Bakhshi M, Vora R. Ligand binding assays with recombinant proteins refolded on an affinity matrix. Biotechniques. 1994 Sep;17(3):509-12, 514. PMID: 7818905.

31-Stein, V., & Alexandrov, K. (2014). Protease-based synthetic sensing and signal amplification. Proceedings Of The National Academy Of Sciences, 111(45), 15934-15939. https://doi.org/10.1073/pnas.1405220111

32-Stelzl, U., Worm, U., Lalowski, M., Haenig, C., Brembeck, F., & Goehler, H. et al. (2005). A Human Protein-Protein Interaction Network: A Resource for Annotating the Proteome. Cell, 122(6), 957-968. https://doi.org/10.1016/j.cell.2005.08.029

33-Su, H., Wang, W., Du, Z., Peng, Z., Gao, S., Cheng, M., & Yang, J. (2021). Improved Protein Structure Prediction Using a New Multi‐Scale Network and Homologous Templates. Advanced Science, 8(24), 2102592. https://doi.org/10.1002/advs.202102592

34-Troup, R., Fallan, C., & Baud, M. (2020). Current strategies for the design of PROTAC linkers: a critical review. Exploration Of Targeted Anti-Tumor Therapy, 1(5). https://doi.org/10.37349/etat.2020.00018

35-Vajda, S., Yueh, C., Beglov, D., Bohnuud, T., Mottarella, S., & Xia, B. et al. (2017). New additions to the C lus P ro server motivated by CAPRI. Proteins: Structure, Function, And Bioinformatics, 85(3), 435-444. https://doi.org/10.1002/prot.25219

36-Yang, J., Anishchenko, I., Park, H., Peng, Z., Ovchinnikov, S., & Baker, D. (2020). Improved protein structure prediction using predicted interresidue orientations. Proceedings Of The National Academy Of Sciences, 117(3), 1496-1503. https://doi.org/10.1073/pnas.1914677117

37- Zhao, Q., & Xie, Q. (2022). In vitro Protein Ubiquitination Assays. Bio-protocol.org. Retrieved 8 October 2022, from https://bio-protocol.org/e928.