Team:IISER Berhampur/Model

IISER-BPR IGEM

Intro

Our journey throughout our project was not that easy. We got a lot of questions and doubts at every step of our project. Many of them were easy to answer through experience or logic, but some needed to be worked on. Hence, we took help of modelling to answer a few of them. We took an approach of ‘Imagine each step -> Ask Questions: Why? How? -> Modelling to answer the questions -> Integrate the answers into the project -> Proceed ahead.


The flow chart above gives an overview of all such questions and how we tried to answer them. More details on each modelling part is described in the following sections.The whole modelling aspect is divided into 6 sub-sections which are as follows-


  • Epidemiological Studies- To understand the problem of growing dengue cases and why it is an important topic to move ahead with. Further details are below.
  • Mutational Analysis- To understand and justify why targeting Non-structural proteins can be a good target for targeting the inhibition of the protein-protein interactions.
  • Target PPI studies- A brief literature review and studies were done to select a set of protein-protein interactions. This part mainly aims towards the validation of the reporter system.
  • Peptide Inhibitor Designing- After a particular PPI was selected for the validation of the system, the peptide inhibitors were designed and appropriate chemical modifications were done. The designed peptide inhibitors were then to be ranked by our reporter system to check the efficiency of the system.
  • Molecular dynamics analysis- To facilitate the designing of the reporter system.
  • Mathematical Formalisms- To mathematically depict the pipeline for using our reporter system to quantify the efficiency of peptide inhibitors.
 Analysis Graphs Reference 






1. Mutational Analysis


Before going into this section, let’s first have a look at the various types of DENV Proteins.




The RNA genome of DENV encodes 3 structural proteins (C, prM and E) and 7 non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5). The structural proteins are the molecules that the virus particle comes equipped with, such as the proteins comprising their capsids or envelope. The non-structural proteins are encoded by the viral genome and expressed in the infected host cells. These play important roles in several processes such as replication of the virus.


Now let’s come back to the topic of Mutational analysis.


To answer the question ‘Which proteins to select as the target of the peptide inhibitors: Non-Structural or Structural proteins?’, we studied the mutational landscape of the DENV genome, to infer which proteins are relatively stable in terms of their mutational entropy so that the inhibitor is not quickly rendered ineffective by mutations in the genome.


To study the genome data of dengue virus, we used the data available in the Nextstrain (https://nextstrain.org/dengue) open-source project website. This website provides the data for the genetic diversity of the DENV genome over several years. It compares the number of mutation events and Shannon Entropy values(which gives a measure of how frequently a gene mutates) of each codon that codes for an amino acid in the structural and non-structural proteins of the Dengue virus. It takes into account results from numerous studies from all over the world and provides a timeline for the evolution of the DENV genome and geographical distribution of the different genotypes.


To achieve our aim to understand the difference between the rate of mutations in NS Proteins and that in structural proteins, we took into account the number of mutation events that have occurred in each codon of all the genes of DENV. Out of these, we took those positions where the number of mutational events had been 10 or greater than 10. Some number of mutations is a common occurrence in most genes due to background mutational events. But if this number is significantly high in a gene, it shows that there is some kind of selection pressure. So this bar of 10 mutations was set as our threshold to isolate those positions where mutations were greater than normally occurring mutations, which may suggest a greater tendency to mutate. These positions would be more prone to undergo changes, thus rendering our inhibitor ineffective if targeted against these sites.


After this, the selected codon positions in each of the genes were plotted against ‘The Number of mutations’ and compared.


The results obtained were as follow:


  • For each of the four serotypes, the C and E structural proteins, along with the NS1 protein contained at least one codon that had undergone 10 or more than 10 mutations, implying that these proteins have evolved the most over the course of time. All the other NS Proteins did not have mutations greater than 10, in some serotype or the other.
  • Comparatively less number of positions in the structural proteins where mutational events are high, can be accounted for by the fact that Structural proteins are shorter in length and have less number of amino acids, as it is. Moreover, this fact further proves that rate of mutations is high in structural proteins because despite shorter length, C and E proteins have comparably high number of mutations in all 4 serotypes.

Hence, it can be concluded that Non-Structural Proteins would be a better target for our Peptide Inhibitor since it was seen that Structural proteins are more prone to mutations. If chosen as a target for our peptide inhibitor, a more mutationally stable protein, would ensure that the peptide inhibitor is viable for a longer time.


  • Among the Non-structural proteins, it was seen that our target protein NS5 (which was selected later on) did have a significantly high number of mutations, which probably could mean that NS5 too faces high selection pressure. But this could again be partially due to the long length of NS5. Moreover, on taking a closer look, it was found that that the region where STAT-2 interacts with NS5, there is a high number of mutations (In DENV1 and especially DENV2, in which this interaction has been shown). This shows that our chosen interaction is really an important interaction in the DENV pathogenesis since selection pressure is high in that region. Due to this, we stayed with our choice of targeting NS5 and STAT2 interaction (more on this in the ‘Protein-Protein Interaction Studies’ section) in spite of NS5 having a greater number of mutations.







2. Graphs


1) DENV-1



2) DENV-2



3) DENV-3



4)- DENV-4










3. References


Hadfield et al., Nextstrain: real-time tracking of pathogen evolution, Bioinformatics (2018)Nextstrain: real-time tracking of pathogen evolution



 PPI_Studies/Validations Brief References 






1. Protein-Protein Interaction Studies and Validations


To validate the working of FRaPPe, we thought of selecting a particular DENV associated Protein-Protein Interaction, design peptide inhibitors against them and finally validate the efficiency of the designed peptide inhibitors in inhibiting the target PPI using our reporter system. This will also serve as an exemplar pipeline which can be followed while using our reporter system for building peptide inhibitors against various PPIs.


As we had decided to go ahead with non-structural proteins, hence we searched through all the available literature on DENV interactomes to find out various PPIs involving DENV NSPs (and the interacting partners being either DENV proteins or human proteins) and their specific details (e.g. interacting domains, amino acids involved in the interactions etc.). Based on this, we prepared diagrams representing the interactomes of each NSP.




Based on all the data we had collected, we wanted to select a particular PPI as our target PPI. For this selection, we appointed several criteria such as the amount of data available, whether the structures are available, how significant the role of the PPI is in DENV pathogenesis etc. Finally, by all these, we narrowed down to the interaction between Dengue Virus Non-structural Protein 5 (DENV NS5) and Human Signal Transducer and Activator of Transcription-2 (hSTAT2).


Just after the target PPI was selected, we again jumped into the literature to mine more data on DENV NS5-hSTAT2 interaction.









Brief


Here we will give a brief overview (mostly structural details) of what we obtained through the data-mining (the molecular mechanisms and pathway through which NS5-STAT2 interaction helps in DENV pathogenesis and the detailed biology relating it to our project will be discussed separately):


  • hSTAT2 consists of 5 domains which are as follow (from N to C terminal): Amino-terminal domain, Coiled-coil domain (139-316), DNA-binding domain (317-458), -SH2 domain (568-686, 581-700) and Carboxy terminal domain.
  • In contrast, DENV NS5 consists of 2 domains N-terminal Methyl Transferase domain and a C-terminal RNA dependent RNA polymerase domain.
  • Ihere are four serotypes of DENV (1-4) between which the structure, functions and mechanisms vary, but there is a lot of overlap too. The interaction between NS5-STAT2 has been mostly studied in DENV2 but has been also studied somewhat in DENV1. But this particular PPI is thought to be applicable for all serotypes because of the relatively high percentage of NS5 sequence identity when comparing DENV2 to the other 3 serotypes.
  • It is hypothesized that NS5 and STAT2 first bind with each other through specific regions present on each of them, then NS5 mediates STAT2 degradation and in this also specific regions on NS5 and STAT2 are responsible. But in this project, we are interested only in inhibiting the binding between NS5 and STAT2 (which will eventually inhibit the degradation of STAT2). Though several studies support this, evidence regarding direct interaction between NS5 and STAT2 is still lacking (some speculations suggest that NS5 and STAT2 may be actually binding to a third protein instead of binding to each other).
  • In spite of lack of any concrete evidence regarding NS5 and STAT2 direct interaction, people have mapped the ability of NS5 to bind STAT2 to a region within the hSTAT2 coiled-coil domain. Studies show that amino acids between 181-200 of STAT2 are indispensable for this interaction. Similarly, amino acids between residues 202 and 306 from the N-terminal region of the protein from the dengue 2 serotype have been shown to interact directly with STAT2.

The next job was to create a model of the NS5-STAT2 complex (as no models based on experimentally solved structures were available in the literature). For that, first, we went through the RCSB Protein Data Bank website (http://www.rcsb.org/) and searched for the available structures of DENV NS5 and hSTAT2, based on which we finally selected these two PDB structures for modelling:


  • PDB ID 5ZQK: Dengue Virus Non Structural Protein 5 (Dengue Virus 2) This was used to extract the DENV NS5 (monomeric) structure.
  • PDB ID 6WCZ: CryoEM Structure of full-length ZIKV NS5-hSTAT2 complex. This was used to extract STAT2 structure.
  • Then, using HawkDock Server (http://cadd.zju.edu.cn/hawkdock/), we docked (global docking) the extracted NS5 and STAT2 structure. Then we filtered the output models on the basis of how well they satisfy the experimental results mentioned earlier. The rank 7 model satisfied the experimental results very well with a score of -4037.07. The binding free energy was predicted to be -15.24 kcal/mol by MM/GBSA analysis. We named it as ‘model 1’.

But the problem with the available PDB structures for DENV NS5 and hSTAT2 was that none of them were for the complete proteins. Hence the docked structure was also made of these partial proteins. Hence, to overcome this problem we later thought of using homology modeling for building the complete structure. For that, first we obtained the sequence for full-length DENV NS5 (Accession: 5ZQK_A) and full-length hSTAT2 (Accession: 6WCZ_A) from NCBI Protein (https://www.ncbi.nlm.nih.gov/protein/). Then using SWISS-MODEL (https://swissmodel.expasy.org/) we created structures of complete DENV NS5 (Template: 5zqk.1.A) and hSTAT2 (Template: 6ux2.1.A) proteins by homology-modeling. Now using these two structures, we prepared the model for FL DENV NS5- FL hSTAT2 complex (Model 2) using a procedure very similar to what was adopted for the complex model created earlier (Model 1).


Based on this complex structure (Model 2), after analysing the position of N and C terminals, it seemed that it is better to have heavy protein tags only at N terminals of both the proteins so that those tags won’t constraint or affect the protein structure or their interaction between NS5-STAT2 to a very great extent. This insight was taken into consideration while designing the constructs so as to avoid false results as much as possible and to make our reporter more effective.









3. References:


1. Ashour, Joseph, et al. "Mouse STAT2 restricts early dengue virus replication." Cell host & microbe 8.5 (2010): 410-421.


2. Ashour, Joseph, et al. "NS5 of dengue virus mediates STAT2 binding and degradation." Journal of virology 83.11 (2009): 5408-5418.


3. Aslam, B., et al. "Structural modeling and analysis of dengue-mediated inhibition of interferon signaling pathway." Genetics and molecular research: GMR 14.2 (2015): 4215-37.


4. Boxiao, W., et al. (2020) CryoEM structure of full-length ZIKV NS5-hSTAT2 complex doi: http://doi.org/10.2210/pdb6WCZ/pdb


5. Chen F, Liu H, Sun HY, Pan PC, Li YY, Li D, Hou TJ. Assessing the performance of the MM/PBSA and MM/GBSA methods. 6. Capability to predict protein-protein binding free energies and re-rank binding poses generated by protein-protein docking. Physical Chemistry Chemical Physics, 2016, 18(18):22129-22139.


6. Chew, Miaw-Fang, Keat-Seong Poh, and Chit-Laa Poh. "Peptides as therapeutic agents for dengue virus." International journal of medical sciences 14.13 (2017): 1342.


7. Choubey, Sanjay Kumar, et al. "Structural and functional insights of STAT2-NS5 interaction for the identification of NS5 antagonist–An approach for restoring interferon signaling." Computational Biology and Chemistry 88 (2020): 10733


8. El Sahili, Abbas, and Julien Lescar. "Dengue virus non-structural protein 5." Viruses 9.4 (2017): 91.


9. El Sahili, Abbas, et al. "NS5 from dengue virus serotype 2 can adopt a conformation analogous to that of its Zika virus and Japanese encephalitis virus homologues." Journal of virology 94.1 (2019).


10. Feng T, Chen F, Kang Y, Sun HY, Liu H, Li D, Zhu F, Hou TJ. HawkRank: a new scoring function for protein-protein docking based on weighted energy terms. Journal of Cheminformatics, 2017, 9(1):66.


11. Guex, N., Peitsch, M.C., Schwede, T. Automated comparative protein structure modeling with SWISS-MODEL and Swiss-PdbViewer: A historical perspective. Electrophoresis 30, S162-S173 (2009).


12. H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig, I.N. Shindyalov, P.E. Bourne. (2000) The Protein Data Bank Nucleic Acids Research, 28: 235-242


13. H.M. Berman, K. Henrick, H. Nakamura (2003) Announcing the worldwide Protein Data Bank Nature Structural Biology 10 (12): 980


14.Hou TJ, Wang JM, Li YY, Wang W. Assessing the performance of the MM/PBSA and MM/GBSA methods: I. The accuracy of binding free energy calculations based on molecular dynamics simulations. Journal of Chemical Information & Modeling, 2011, 51(1):69-82.


15. Hou TJ, Qiao XB, Zhang W, Xu XJ. Empirical Aqueous Solvation Models Based on Accessible Surface Areas with Implicit Electrostatics. Journal of Physical Chemistry B, 2002, 106(43):11295-11304.


17. Mazzon, Michela, et al. "Dengue virus NS5 inhibits interferon-α signaling by blocking signal transducer and activator of transcription 2 phosphorylation." The Journal of infectious diseases 200.8 (2009): 1261-1270.


18. Morrison, Juliet, and Adolfo García-Sastre. "STAT2 signaling and dengue virus infection." Jak-stat 3.1 (2014): e27715.


19. Protein [Internet]. Bethesda (MD): National Library of Medicine (US), National Center for Biotechnology Information; [1988] - [cited 2020 Oct 02]. Available from:https://www.ncbi.nlm.nih.gov/protein/


20. ASun HY, Li YY, Tian S, Xu L, Hou, TJ. Assessing the performance of MM/PBSA and MM/GBSA methods. 4. Accuracies of MM/PBSA and MM/GBSA methodologies evaluated by various simulation protocols using PDBbind data set. Physical Chemistry Chemical Physics, 2014, 16(31):16719-16729.


21. Waterhouse, A., Bertoni, M., Bienert, S., Studer, G., Tauriello, G., Gumienny, R., Heer, F.T., de Beer, T.A.P., Rempfer, C., Bordoli, L., Lepore, R., Schwede, T. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 46, W296-W303 (2018).


22. Weng GQ, Wang EC, Wang Z, Liu H, Li D, Zhu F, Hou TJ. HawkDock: a web server to predict and analyze the structures of protein-protein complexes based on computational docking and MM/GBSA. Nucleic Acids Research, 2019, 47(W1): W322-W330.


23. Zacharias M. Protein–protein docking with a reduced protein model accounting for side-chain flexibility. Protein Science, 2003, 12(6):1271-1282.


24. The PyMOL Molecular Graphics System, Version 2.3.2 Schrödinger, LLC.



 iPep_Design Potential_Peptide Modifications GRAVY,Instability_Index,pI References 






Peptide Inhibitor (IPep) Designing


So once we were done with the target PPI selection and building its complex model, the next job was to design peptide inhibitors against it.


The first question which came to us was ‘Our IPeps should bind to which protein among NS5 and STAT2 so as to interrupt their interaction?’ Since the pathogenic protein NS5 used to bind and inhibit the function of the helpful STAT2 protein, we decided to target NS5 using IPeps i.e. our designed IPeps will be binding the residues on NS5 that are involved in its interaction with STAT2. In this way interaction of NS5 with STAT2 will be inhibited, hence allowing STAT2 to carry out its normal function, and in addition, since some residues of NS5 are now masked, it may inhibit other pathogenic functions of NS5 apart from its effect on STAT2. Hence, NS5 will be considered as a receptor and STAT2 as it’s ligand and our target is to block the receptor.









Potential Peptide


Then, with the Model 1 of NS5-STAT2 complex, we derived potential peptide sequences for IPeps using Rosetta Petiderive Protocol (https://rosie.rosettacommons.org/peptiderive/). We obtained 3 sequences as output as follows (Lower the score, better is the peptide in binding to the receptor):


1: KEQKILQE (Score: -10.131 REU)

2: QTKEQKILQE (Score: -12.029 REU)

3: QTKEQKILQETL (Score: -12.115 REU)

These 3 sequences then entered into an in-silico IPep designing Pipeline we adopted for this project. While peptide based drugs are slowly growing in use, they have 2 major disadvantages associated with them.


  • In-vivo instability - Degradation by various natural proteolytic enzymes present inside the body decreases the half-life of the peptide drugs as well as decreases its biochemical concentration.
  • Membrane impermeability- Membrane impermeability is one of the major problems associated with peptide drugs and even though peptides can be natural ligands for GPCR but very few of the peptides are naturally taken up by cells.







Modifications


After deducing the peptide sequences which can act as inhibitors, a number of chemical modifications were done in-silico to increase the half-life of the peptides and to stabilise it more.


Some of the common modifications which were done and their purposes are mentioned below. The list is prepared according to various literature studies which focused on the same aspect. The modifications which were necessary and useful for the peptide inhibitors were chosen.


  • Termini protection- Methionine was added at the end terminus as without them the peptides are usually more prone to degradation. So addition of a chain of Methionine ensured that the termini are well protected from any degradation.
  • Amino acid substitution- Certain Amino Acid residues such as arginine were replaced with lysine to modify the rigidity and the conformation of the peptide and make it more suitable for our own purposes.
  • Identification of hydrophobic Amino acids which were replaced by polar or charged residues for the modulation of the biological pI while maintaining the bioactivity.
  • The chemical changes to polar or charged residues were also helpful to stop the formation of Hydrophobic patches, which usually makes them poorly soluble.
  • For the delivery of the peptide drugs, Cell Penetrating Peptides can be designed which in conjugation with the peptide can be used for drug delivery.

The 3 initial peptide sequences and the necessary modifications carried out on them to generate the final peptide inhibitor library are mentioned below-



1.1: KEQKILQE (Without Modifications)


1.2 KEQKILQEMMMMMMMMMMM

1.3 KEQKISQEMMMMMMMMMMM

1.4 MKEQKILQEMMMMMMMMMMMM

1.5 MKEQKISQEMMMMMMMMMMMM

1.6 MKEQKISQEMMMMMMMMMMM

2.1: QTKEQKILQE (Without Modifications)

3.1: QTKEQKILQETL (Without Modifications)

3.2 QTKEQKILQETLMMMMMMMMM

3.3 MMMMMQTKEQKILQETLMMM

3.4 QTKEQKILQETLMMMMMMMMMM

3.5 MMMMMMQTKEQKILQETLMMMM







GRAVY, Instability Index , pi and Half life


The various parameters such as Instability Index, GRAVY ( The GRAVY value for a protein or peptide is calculated by adding the hydropathy values of each amino acid residue and dividing it by the total number of residues in the sequence or length of the sequence, Increasing positive scores indicates increasing hydrophobicity), Theoretical pI and Half Life of the designed peptide inhibitors were checked with the ProtParam tool of ExPasy (https://web.expasy.org/protparam/). The values obtained for each peptide are mentioned in table 1.


After the in silico chemical modifications, the docking of the peptide inhibitors to NS5 was performed by Dr Malay Kumar Rana using HADDOCK. The active site residues (constraints for docking) were given based on literature and Model 1. 10 residues ranging from 262 to 272 of NS5 were used to define the active site and hence the docking of the peptide inhibitors was done to that part. The docking scores obtained from HADDOCK for each peptide are mentioned in table


1. Lower HADDOCK score indicates better binding of the peptide inhibitor to NS5


Table 1: Results from various in-silico characterizations performed on each peptide inhibitor


1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 50000000 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7
1 2 3 4 5 6 7







References


Bruzzoni-Giovanelli, Heriberto, et al. "Interfering peptides targeting protein–protein interactions: the next generation of drugs?." Drug Discovery Today 23.2 (2018): 272-285.


Lee, Andy Chi-Lung, et al. "A comprehensive review on current advances in peptide drug development and design." International journal of molecular sciences 20.10 (2019): 2383.


London N, Raveh B, Movshovitz-Attias D, Schueler-Furman O. (2010). Can self-inhibitory peptides be derived from the interfaces of globular protein-protein interactions? Proteins, 78:3140-49.


Lyskov S, Chou FC, Conchúir SÓ, Der BS, Drew K, Kuroda D, Xu J, Weitzner BD, Renfrew PD, Sripakdeevong P, Borgo B, Havranek JJ, Kuhlman B, Kortemme T, Bonneau R, Gray JJ, Das R., "Serverification of Molecular Modeling Applications: The Rosetta Online Server That Includes Everyone (ROSIE)". PLoS One. 2013 May 22;8(5):e63906. doi: 10.1371/journal.pone.0063906. Print 2013.


Yuval Sedan, Orly Marcu,Sergey Lyskov, Ora Schueler-Furman Peptiderive server: derive peptide inhibitors from protein–protein interactions Nucleic Acids Research 2016; doi: 10.1093/nar/gkw385


Gasteiger E., Gattiker A., Hoogland C., Ivanyi I., Appel R.D., Bairoch A. ,ExPASy: the proteomics server for in-depth protein knowledge and analysis Nucleic Acids Res. 31:3784-3788(2003).



 Literature Dengue_in_India Equations Results References 






Literature


Before we started to work on our project we needed to gather existing information on the spread of the dengue . Therefore we began a search to find relevant data on the disease


.

Some key factors which we looked into were:


  • Seroprevalence, which is a good indicator of endemicity, was found to be approximately 57% in India based on pooled estimates of seven studies.
  • In the 69 studies which adopted WHO 1997 dengue case classification, the pooled Case Fatality Ratio was 1.1%. The pooled CFR for 8 studies that used the WHO 2009 case definition was 1.6%.
  • Proportion of Severe Dengue Cases: The overall percentage of severe Dengue among laboratory-confirmed studies was 28.9%.
  • Distribution of Serotypes: The published studies from India indicated the circulation of all the four-dengue serotypes, with DEN-2 and DEN-3 being the more commonly reported serotypes.
  • We then analyzed the distribution of Dengue in India with respect to several correlated factors: Age of infected person, Geographical region, and seasonality of Dengue.







DENGUE - A CLOSER LOOK AT HOME


We looked further into government records to see how the number of dengue cases changed over time . As we can see in the plot below the number of cases shows an increase over the years [2009-2017].




But the yearly data obtained does not reveal the complete picture. The following plot time vs Reported Dengue cases take over a period of 13 months gives us a better idea of the yearly trend . As we can see the number of cases starts to increase rapidly around August until it peaks around October. This increase in cases can be associated with the monsoon season in India which leads to more breeding grounds for the mosquitoes.


How does this help us ?


Knowing that the distribution of dengue cases is not uniform over the span of 1 year can help the government design policies which are more efficient . Robust sanitising and cleaning operations around August can also significantly decrease the disease burden.




India is a vast country. To see whether dengue cases were spread uniformly over the country we dived into government records again . We expected that higher population density would result in a higher number of cases but we found some exceptions . Kerala even though having a high population density had comparatively low cases . This is because of the strict measures taken up by the Kerala government to prevent the spread of the disease which includes indoor space spraying and maintaining environmental cleanliness.




Well a much more deeper understanding of the dynamics of the disease was required and hence we tried to model the disease burden of dengue.


Before we move on to the inferences we made let us look at the model in detail. The basic SIR model we all know and love ( the pandemic has made us all look at it one time or the other ) does not work for a disease like dengue . This is because dengue cannot spread via direct contact , there is a third party involved (i.e. the mosquitoes). Therefore the basic model has to be modified to include the mosquitoes, which makes things a bit more difficult as then we have to consider mosquito birth rate , death rate the data for which is not readily available. But even with these problems we can still make interesting inferences.









Equations


N=Sh+Ih+Rh n=Sv+Iv dSh_dt= pih-(alpha*Sh*Iv)/N - ah*Sh dIh_dt= ((alpha*Sh*Iv)/N - gamma*Ih -ah*Ih -phih*Ih) dRh_dt= gamma*Ih - ah*Rh dSv_dt= piv-(beta*Sv*Ih)/n - av*Sv +av*n dIv_dt= (beta*Sv*Ih)/n - av*Iv



THE INITIAL VALUES

The equations require some variables to be considered. In this section we introduce the values that we considered for running the model.


How does this help us ?


  • Ih0=500
  • Rh0=200
  • Iv0=300000
  • N=960000000 (based on 1996 values [4] )
  • n=10000000000 [ Considering on average ]
  • pih=10
  • piv=25
  • phih=0.01
  • ah=0.000046
  • av=0.25






Results


The calculated plot for the number of infected people perfectly fits with the available data for a gamma value of 0.75 ..


The gamma value as we described earlier is the recovery constant which is the fraction of the infected people that recover from the disease . The following plot shows the maximum number of infected people reached by our defined population [check the initial variables section] versus the recovery constant . From here we can see that an increase in the gamma value brought on by new drugs and better treatment can bring down the maximum number of infected people and hence deaths significantly .


This SIR plot depicts several features:









References


Bruzzoni-Giovanelli, Heriberto, et al. "Interfering peptides targeting protein–protein interactions: the next generation of drugs?." Drug Discovery Today 23.2 (2018): 272-285.


Lee, Andy Chi-Lung, et al. "A comprehensive review on current advances in peptide drug development and design." International journal of molecular sciences 20.10 (2019): 2383.


London N, Raveh B, Movshovitz-Attias D, Schueler-Furman O. (2010). Can self-inhibitory peptides be derived from the interfaces of globular protein-protein interactions? Proteins, 78:3140-49.


Lyskov S, Chou FC, Conchúir SÓ, Der BS, Drew K, Kuroda D, Xu J, Weitzner BD, Renfrew PD, Sripakdeevong P, Borgo B, Havranek JJ, Kuhlman B, Kortemme T, Bonneau R, Gray JJ, Das R., "Serverification of Molecular Modeling Applications: The Rosetta Online Server That Includes Everyone (ROSIE)". PLoS One. 2013 May 22;8(5):e63906. doi: 10.1371/journal.pone.0063906. Print 2013.


Yuval Sedan, Orly Marcu,Sergey Lyskov, Ora Schueler-Furman Peptiderive server: derive peptide inhibitors from protein–protein interactions Nucleic Acids Research 2016; doi: 10.1093/nar/gkw385


Gasteiger E., Gattiker A., Hoogland C., Ivanyi I., Appel R.D., Bairoch A. ,ExPASy: the proteomics server for in-depth protein knowledge and analysis Nucleic Acids Res. 31:3784-3788(2003).



 iPep_Design Potential_Peptide Modifications GRAVY,Instability_Index,pI References 






Peptide Inhibitor (IPep) Designing


So once we were done with the target PPI selection and building its complex model, the next job was to design peptide inhibitors against it.


The first question which came to us was ‘Our IPeps should bind to which protein among NS5 and STAT2 so as to interrupt their interaction?’ Since the pathogenic protein NS5 used to bind and inhibit the function of the helpful STAT2 protein, we decided to target NS5 using IPeps i.e. our designed IPeps will be binding the residues on NS5 that are involved in its interaction with STAT2. In this way interaction of NS5 with STAT2 will be inhibited, hence allowing STAT2 to carry out its normal function, and in addition, since some residues of NS5 are now masked, it may inhibit other pathogenic functions of NS5 apart from its effect on STAT2. Hence, NS5 will be considered as a receptor and STAT2 as it’s ligand and our target is to block the receptor.









Potential Peptide


Then, with the Model 1 of NS5-STAT2 complex, we derived potential peptide sequences for IPeps using Rosetta Petiderive Protocol (https://rosie.rosettacommons.org/peptiderive/). We obtained 3 sequences as output as follows (Lower the score, better is the peptide in binding to the receptor):


1: KEQKILQE (Score: -10.131 REU)

2: QTKEQKILQE (Score: -12.029 REU)

3: QTKEQKILQETL (Score: -12.115 REU)

These 3 sequences then entered into an in-silico IPep designing Pipeline we adopted for this project. While peptide based drugs are slowly growing in use, they have 2 major disadvantages associated with them.


  • In-vivo instability - Degradation by various natural proteolytic enzymes present inside the body decreases the half-life of the peptide drugs as well as decreases its biochemical concentration.
  • Membrane impermeability- Membrane impermeability is one of the major problems associated with peptide drugs and even though peptides can be natural ligands for GPCR but very few of the peptides are naturally taken up by cells.







Modifications


After deducing the peptide sequences which can act as inhibitors, a number of chemical modifications were done in-silico to increase the half-life of the peptides and to stabilise it more.


Some of the common modifications which were done and their purposes are mentioned below. The list is prepared according to various literature studies which focused on the same aspect. The modifications which were necessary and useful for the peptide inhibitors were chosen.


  • Termini protection- Methionine was added at the end terminus as without them the peptides are usually more prone to degradation. So addition of a chain of Methionine ensured that the termini are well protected from any degradation.
  • Amino acid substitution- Certain Amino Acid residues such as arginine were replaced with lysine to modify the rigidity and the conformation of the peptide and make it more suitable for our own purposes.
  • Identification of hydrophobic Amino acids which were replaced by polar or charged residues for the modulation of the biological pI while maintaining the bioactivity.
  • The chemical changes to polar or charged residues were also helpful to stop the formation of Hydrophobic patches, which usually makes them poorly soluble.
  • For the delivery of the peptide drugs, Cell Penetrating Peptides can be designed which in conjugation with the peptide can be used for drug delivery.

The 3 initial peptide sequences and the necessary modifications carried out on them to generate the final peptide inhibitor library are mentioned below-



1.1: KEQKILQE (Without Modifications)


1.2 KEQKILQEMMMMMMMMMMM

1.3 KEQKISQEMMMMMMMMMMM

1.4 MKEQKILQEMMMMMMMMMMMM

1.5 MKEQKISQEMMMMMMMMMMMM

1.6 MKEQKISQEMMMMMMMMMMM

2.1: QTKEQKILQE (Without Modifications)

3.1: QTKEQKILQETL (Without Modifications)

3.2 QTKEQKILQETLMMMMMMMMM

3.3 MMMMMQTKEQKILQETLMMM

3.4 QTKEQKILQETLMMMMMMMMMM

3.5 MMMMMMQTKEQKILQETLMMMM







GRAVY, Instability Index , pi and Half life


The various parameters such as Instability Index, GRAVY ( The GRAVY value for a protein or peptide is calculated by adding the hydropathy values of each amino acid residue and dividing it by the total number of residues in the sequence or length of the sequence, Increasing positive scores indicates increasing hydrophobicity), Theoretical pI and Half Life of the designed peptide inhibitors were checked with the ProtParam tool of ExPasy (https://web.expasy.org/protparam/). The values obtained for each peptide are mentioned in table 1.


After the in silico chemical modifications, the docking of the peptide inhibitors to NS5 was performed by Dr Malay Kumar Rana using HADDOCK. The active site residues (constraints for docking) were given based on literature and Model 1. 10 residues ranging from 262 to 272 of NS5 were used to define the active site and hence the docking of the peptide inhibitors was done to that part. The docking scores obtained from HADDOCK for each peptide are mentioned in table


1. Lower HADDOCK score indicates better binding of the peptide inhibitor to NS5


Table 1: Results from various in-silico characterizations performed on each peptide inhibitor

Neehar idhar table add karne hai









References


Bruzzoni-Giovanelli, Heriberto, et al. "Interfering peptides targeting protein–protein interactions: the next generation of drugs?." Drug Discovery Today 23.2 (2018): 272-285.


Lee, Andy Chi-Lung, et al. "A comprehensive review on current advances in peptide drug development and design." International journal of molecular sciences 20.10 (2019): 2383.


London N, Raveh B, Movshovitz-Attias D, Schueler-Furman O. (2010). Can self-inhibitory peptides be derived from the interfaces of globular protein-protein interactions? Proteins, 78:3140-49.


Lyskov S, Chou FC, Conchúir SÓ, Der BS, Drew K, Kuroda D, Xu J, Weitzner BD, Renfrew PD, Sripakdeevong P, Borgo B, Havranek JJ, Kuhlman B, Kortemme T, Bonneau R, Gray JJ, Das R., "Serverification of Molecular Modeling Applications: The Rosetta Online Server That Includes Everyone (ROSIE)". PLoS One. 2013 May 22;8(5):e63906. doi: 10.1371/journal.pone.0063906. Print 2013.


Yuval Sedan, Orly Marcu,Sergey Lyskov, Ora Schueler-Furman Peptiderive server: derive peptide inhibitors from protein–protein interactions Nucleic Acids Research 2016; doi: 10.1093/nar/gkw385


Gasteiger E., Gattiker A., Hoogland C., Ivanyi I., Appel R.D., Bairoch A. ,ExPASy: the proteomics server for in-depth protein knowledge and analysis Nucleic Acids Res. 31:3784-3788(2003).



 iPep_Design Potential_Peptide Modifications GRAVY,Instability_Index,pI References 






Peptide Inhibitor (IPep) Designing


So once we were done with the target PPI selection and building its complex model, the next job was to design peptide inhibitors against it.


The first question which came to us was ‘Our IPeps should bind to which protein among NS5 and STAT2 so as to interrupt their interaction?’ Since the pathogenic protein NS5 used to bind and inhibit the function of the helpful STAT2 protein, we decided to target NS5 using IPeps i.e. our designed IPeps will be binding the residues on NS5 that are involved in its interaction with STAT2. In this way interaction of NS5 with STAT2 will be inhibited, hence allowing STAT2 to carry out its normal function, and in addition, since some residues of NS5 are now masked, it may inhibit other pathogenic functions of NS5 apart from its effect on STAT2. Hence, NS5 will be considered as a receptor and STAT2 as it’s ligand and our target is to block the receptor.









Potential Peptide


Then, with the Model 1 of NS5-STAT2 complex, we derived potential peptide sequences for IPeps using Rosetta Petiderive Protocol (https://rosie.rosettacommons.org/peptiderive/). We obtained 3 sequences as output as follows (Lower the score, better is the peptide in binding to the receptor):


1: KEQKILQE (Score: -10.131 REU)

2: QTKEQKILQE (Score: -12.029 REU)

3: QTKEQKILQETL (Score: -12.115 REU)

These 3 sequences then entered into an in-silico IPep designing Pipeline we adopted for this project. While peptide based drugs are slowly growing in use, they have 2 major disadvantages associated with them.


  • In-vivo instability - Degradation by various natural proteolytic enzymes present inside the body decreases the half-life of the peptide drugs as well as decreases its biochemical concentration.
  • Membrane impermeability- Membrane impermeability is one of the major problems associated with peptide drugs and even though peptides can be natural ligands for GPCR but very few of the peptides are naturally taken up by cells.







Modifications


After deducing the peptide sequences which can act as inhibitors, a number of chemical modifications were done in-silico to increase the half-life of the peptides and to stabilise it more.


Some of the common modifications which were done and their purposes are mentioned below. The list is prepared according to various literature studies which focused on the same aspect. The modifications which were necessary and useful for the peptide inhibitors were chosen.


  • Termini protection- Methionine was added at the end terminus as without them the peptides are usually more prone to degradation. So addition of a chain of Methionine ensured that the termini are well protected from any degradation.
  • Amino acid substitution- Certain Amino Acid residues such as arginine were replaced with lysine to modify the rigidity and the conformation of the peptide and make it more suitable for our own purposes.
  • Identification of hydrophobic Amino acids which were replaced by polar or charged residues for the modulation of the biological pI while maintaining the bioactivity.
  • The chemical changes to polar or charged residues were also helpful to stop the formation of Hydrophobic patches, which usually makes them poorly soluble.
  • For the delivery of the peptide drugs, Cell Penetrating Peptides can be designed which in conjugation with the peptide can be used for drug delivery.

The 3 initial peptide sequences and the necessary modifications carried out on them to generate the final peptide inhibitor library are mentioned below-



1.1: KEQKILQE (Without Modifications)


1.2 KEQKILQEMMMMMMMMMMM

1.3 KEQKISQEMMMMMMMMMMM

1.4 MKEQKILQEMMMMMMMMMMMM

1.5 MKEQKISQEMMMMMMMMMMMM

1.6 MKEQKISQEMMMMMMMMMMM

2.1: QTKEQKILQE (Without Modifications)

3.1: QTKEQKILQETL (Without Modifications)

3.2 QTKEQKILQETLMMMMMMMMM

3.3 MMMMMQTKEQKILQETLMMM

3.4 QTKEQKILQETLMMMMMMMMMM

3.5 MMMMMMQTKEQKILQETLMMMM







GRAVY, Instability Index , pi and Half life


The various parameters such as Instability Index, GRAVY ( The GRAVY value for a protein or peptide is calculated by adding the hydropathy values of each amino acid residue and dividing it by the total number of residues in the sequence or length of the sequence, Increasing positive scores indicates increasing hydrophobicity), Theoretical pI and Half Life of the designed peptide inhibitors were checked with the ProtParam tool of ExPasy (https://web.expasy.org/protparam/). The values obtained for each peptide are mentioned in table 1.


After the in silico chemical modifications, the docking of the peptide inhibitors to NS5 was performed by Dr Malay Kumar Rana using HADDOCK. The active site residues (constraints for docking) were given based on literature and Model 1. 10 residues ranging from 262 to 272 of NS5 were used to define the active site and hence the docking of the peptide inhibitors was done to that part. The docking scores obtained from HADDOCK for each peptide are mentioned in table


1. Lower HADDOCK score indicates better binding of the peptide inhibitor to NS5


Table 1: Results from various in-silico characterizations performed on each peptide inhibitor

Neehar idhar table add karne hai









References


Bruzzoni-Giovanelli, Heriberto, et al. "Interfering peptides targeting protein–protein interactions: the next generation of drugs?." Drug Discovery Today 23.2 (2018): 272-285.


Lee, Andy Chi-Lung, et al. "A comprehensive review on current advances in peptide drug development and design." International journal of molecular sciences 20.10 (2019): 2383.


London N, Raveh B, Movshovitz-Attias D, Schueler-Furman O. (2010). Can self-inhibitory peptides be derived from the interfaces of globular protein-protein interactions? Proteins, 78:3140-49.


Lyskov S, Chou FC, Conchúir SÓ, Der BS, Drew K, Kuroda D, Xu J, Weitzner BD, Renfrew PD, Sripakdeevong P, Borgo B, Havranek JJ, Kuhlman B, Kortemme T, Bonneau R, Gray JJ, Das R., "Serverification of Molecular Modeling Applications: The Rosetta Online Server That Includes Everyone (ROSIE)". PLoS One. 2013 May 22;8(5):e63906. doi: 10.1371/journal.pone.0063906. Print 2013.


Yuval Sedan, Orly Marcu,Sergey Lyskov, Ora Schueler-Furman Peptiderive server: derive peptide inhibitors from protein–protein interactions Nucleic Acids Research 2016; doi: 10.1093/nar/gkw385


Gasteiger E., Gattiker A., Hoogland C., Ivanyi I., Appel R.D., Bairoch A. ,ExPASy: the proteomics server for in-depth protein knowledge and analysis Nucleic Acids Res. 31:3784-3788(2003).



©iGEM IISER Berhampur