Team:Potsdam Bioware/Project/Summary

From 2011.igem.org

(Difference between revisions)
(Modeling)
(Microviridin)
 
(100 intermediate revisions not shown)
Line 4: Line 4:
===Modification, Selection and Production of Cyclic Peptides for Therapy===
===Modification, Selection and Production of Cyclic Peptides for Therapy===
-
One key task of biopharmaceuticals is the binding and blocking of deregulated proteins. Picking the right lead structure for biopharmaceuticals is very important for success. This year, we developed a novel system for the modification, selection and optimization of peptides showing potential for protease inhibition. These promising inhibitors are found in cyanobacteria and are called Microviridins. They belong to the class peptides characterized by unusual ω- ester and ω-amide bonds between the amino-acid side chains, which are also referred to as depsipeptides. These modifications are introduced post translationally by a set of enzymes and result in an extraordinary tricyclic cage structure.  
+
One key task of biopharmaceuticals is the binding and blocking of deregulated proteins. Picking the right lead structure for biopharmaceuticals is very important for success. This year, we developed a novel system for the modification, selection and optimization of peptides showing potential for protease inhibition. These promising inhibitors are found in cyanobacteria and are called Microviridins. They belong to a class of peptides characterized by unusual ω-ester and ω-amide bonds between the amino-acid side chains, which are also referred to as depsipeptides. These modifications are introduced post translationally by a set of enzymes and result in an extraordinary tricyclic cage structure.  
-
<br><br>Proteases are a large enzyme family and comprise about 647 human gene products. Therefore they are important drug targets. Additional many harmful bacteria, viruses and fungi use proteases in their reproduction cycle and growth. The ability of blocking these proteases is highly relevant therapeutic application. Prominent examples of targets are the angiotensin-converting enzyme (ACE) and HIV proteases.
+
<br>
-
<br>We chose a 6.5 kb Microviridin gene cluster named ''mdnABCDE'' from Mycrocystis aeruginosa NIES843. The mdn gene cluster comprises a gene encoding a precursor peptide named ''MdnA'', which is then modified to form the Microviridin, two genes encoding ATP-grasp-type ligases named ''mdnB'' and ''mdnC'', an ABC- transporter encoding gene named ''mdnE'' as well as one gene encoding an N- acetyltransferase of the GNAT family named ''mdnD'' (Ziemert et al., 2008). Our major aim was to modify the mdnA precursor peptide and optimize its protease inhibiting properties. Towards this goal, we synthesized semi-rational ''mdnA'' gene libraries with partially randomized oligonucleotides. The library was successfully cloned and verified by mass spectrometry.
+
<br>
-
<br><br>To identify the best candidates inhibiting various given proteases, we established two different selection systems: Phage Display and a new developed ''in-vivo'' selection assay. Phage display is a frequently used technique in laboratories, yet to our knowledge, phage display of cellularly cyclized peptides is new. We constructed a fusion between the surface protein geneIII of the phage and the ''mdnA'' gene within the ''mdnA'' gene cluster. This was verified by an ELISA test and a trypsin assay.
+
Proteases are a large enzyme family and comprise about 647 human gene products. Therefore they are important drug targets. In addition, many harmful bacteria, viruses and fungi use proteases in their reproduction cycle and growth. The ability to block these proteases is highly relevant for therapy. Prominent examples of targets are the angiotensin-converting enzyme (ACE) and HIV proteases.
-
<br><br>We also constructed a recombinant ''in-vivo'' selection system linking protease degradation to antibiotic resistance. The designed device is divided into two parts: first a protease activity generator and second a protease detector. For the protease detector we fused various protease cleavage sites between a signal sequence and the antibiotic resistance conferring enzyme β-lactamse. β-Lactamse confers only resistance when transferred to the periplasm of ''E. coli''. Thus, when the protease cleaves the signal sequence from the lactamase antibiotic resistance is abolished and cells die under selective pressure. However, when our Microviridin variants inhibit the protease, cells survive, allowing for easy and efficient inhibitor selection. We achieved different growth rates with increasing Ampicillin concentrations.
+
<br><br>We chose a 6.5 kb Microviridin gene cluster named ''mdnABCDE'' from ''Mycrocystis aeruginosa'' NIES843. The <i>mdn</i> gene cluster comprises (i) a gene encoding a precursor peptide named MdnA, which is then modified to form the Microviridin, (ii) two genes encoding ATP-grasp-type ligases named ''mdnB'' and ''mdnC'', (iii) an ABC- transporter encoding gene named ''mdnE'' as well as (iv) one gene encoding an N- acetyltransferase of the GNAT family named ''mdnD'' (Ziemert et al., 2008). Our major aim was to modify the MdnA precursor peptide and optimize its protease inhibiting properties. Towards this goal, we synthesized semi-rational ''mdnA'' gene libraries with partially randomized oligonucleotides. One library was cloned, verified by sequencing, and successfully used for selection.
-
<br><br>In addition to the practical work we established a mathematical model of the ''in-vivo'' selection system, which helped us to understand the selection process. We model the reaction kinetics by ordinary differential equations and coded them in matlab.  
+
<br>
-
<br><br>Currently we are selecting our Microviridin library against a panel of proteases by phage display and with our ''in-vivo'' selection system.
+
<br>
 +
To identify the best candidates inhibiting various given proteases, we established two different selection systems: Phage Display and a new developed ''in-vivo'' selection assay. Phage display is a frequently used technique in laboratories, yet to our knowledge, phage display of cellularly cyclized peptides is new. We constructed a fusion between the surface protein geneIII of the phage and the ''mdnA'' gene within the ''mdnA'' gene cluster. This system was verified by western blotting, Phage-ELISA and controlled phage panning experiments.
 +
<br>
 +
<br>
 +
We also devised and constructed a recombinant ''in-vivo'' selection system linking protease degradation to antibiotic resistance. This system is divided into three parts: first a protease activity detector device, second a protease generator device, and third a protease blocking device. For the protease activity detector device we fused various protease cleavage sites between a signal sequence and the antibiotic resistance conferring enzyme β-lactamase. β-lactamase provides resistance when its transferred into the periplasm. Thus, when the protease cleaves the signal sequence from the β-lactamase antibiotic resistance is abolished and cells die under selective pressure. However, when our Microviridin variants inhibit the protease, cells survive, allowing for an easy and efficient selection of protease inhibitors. We demonstrated a wide dynamic range of the system between expressed and non-expressed protease in the presence of increasing ampicillin concentrations. Importantly, after co-transformation with our ''mdnA''-libary we were able to isolate several clones, which are currently being characterized.
 +
<br>
 +
<br>
 +
In addition to the practical work, we established a mathematical model of the ''in-vivo'' selection system. This model, which we were able to fit to experimental data, helped us to understand the selection process. We modeled the reaction kinetics with ordinary differential equations and simulated and fitted data with matlab.
 +
<br>
 +
<br>
 +
Last but not least, we had several human practice projects. We send a survey to all members of the German parliament, visited a member of the parliament, held seminars on ethics and invited children to the lab.
 +
<br>
 +
*Ziemert, N., Ishida, K., Liaimer, A., Hertweck, C. & Dittmann, E. (2008). Ribosomal synthesis of tricyclic depsipeptides in bloom-forming cyanobacteria. Angewandte Chemie (International ed. in English) 47, 7756-9
<br><br>
<br><br>
Line 17: Line 29:
=== Microviridin ===
=== Microviridin ===
-
[[File:UP_Expression_Backbones_Ara.png|left|200px|thumb|'''Figure x:''' Fluorescent ''E. coli'': Induction of the reporter gene YFP-mVenus in our generated Expression Backbone]]  
+
[[File:UP_HPLC_mvd_01.jpg|left|300px|thumb|'''Figure 1:''' <b>HPLC chromatogram of heterologously expressed mdnA</b>]]  
-
The major aim of the microviridin group was to modify the mdnA such that the protease inhibiting activity is enhanced. Therefore we used random mutagenesis as well as focused oligonucleotids for creating a library, which is ready for being screened for mdnA with a therapeutically promising set of mutations. For further experiments we also fused the mdnA to a myc-tag. So in the future we will be able to purify and isolate the mdnA.<br>
+
<br>
-
Due to the applicability of the whole mdn-cluster the creation of several BioBricks was possible. The construction was done using a given template vector containing the mdn-genes and sophisticated design of primers. Characterization of the BioBricks was done via HPLC analysis, mass spectrometry and western blot.<br>
+
The microviridin sub-project aimed for modifications at the genetic level such that protease inhibiting activities of the resulting peptides are enhanced. We used random mutagenesis or focused randomization of oligonucleotides for the creation of gene libraries, which can be screened for <i>mdnA</i>-variants with therapeutically promising mutations. For further experiments, we also fused mdnA to a myc-tag.<br>
-
In a subproject we also tried to build auxiliary expression backbones with inducible promoters for easy cloning via the iGEM restriction enzyme sites. We already have the construct but the process of induction needs to be improved.'''[[:Team:Potsdam_Bioware/Project/Details_Microviridin|[more]]]'''
+
In addition, all coding genes of the <i>mdn</i>-cluster were converted to BioBricks. The microviridin from <i>mdnA</i> was purified and characterized by HPLC and cyclization was verified by mass spectrometry.<br>
 +
For the quick expression of several library clones, we also built auxiliary plasmid backbones with inducible promoters according to a proposed extension of the iGEM cloning standard.
 +
'''[[:Team:Potsdam_Bioware/Project/Details_Microviridin|[more]]]'''
<br>
<br>
<br>
<br>
Line 26: Line 40:
=== Phage Display ===
=== Phage Display ===
-
 
+
[[File:UP_Phage_Display_02.jpg|left|290px|thumb|'''Figure 2: Phage Display of microviridin.''' Enrichment of MdnA carrying phages upon panning towards a protease was demonstrated.]]
-
Phage display is a powerful tool for  selecting peptides or proteins that bind and regulate the function of target proteins. It is defined as a system in which the protein and its encoding gene are physically linked. Because of the therapeutic interest of microviridin as protease inhibitors a selection system for screening recombinant mdnA-libraries is of great importence. In our project the fundamental suitability of phage display for this purpose was shown. Therefore a appropriate phagemid was constructed. The production of phage particles carrying mdnA was determined by ELISA and phage display. '''[[:Team:Potsdam_Bioware/Project/Details_Phage|[more]]]'''
+
<br>
 +
Phage display is a powerful tool for  selecting peptides or proteins that bind and regulate the function of target proteins. Phages robustly link the genotype to the phenotype of the presented peptide allowing for selection of protease binders under controlled ''in vitro'' conditions. We established genetics and panning of therapeutically interesting ''mdnA''-libraries towards a panel of several purified proteases. First, the general suitability of phage display for this purpose was shown. An appropriate phagemid vector harboring an <i>mdnA-myc-geneIII</i> fusion gene, was constructed. The expression of the MdnA-myc-geneIII protein in ''E. coli'' was shown by western blot. Production of phage particles carrying MdnA were analyzed by Phage-ELISA. Finally, phage display procedures were optimized and an enrichment of MdnA modified phages over a control was shown. '''[[:Team:Potsdam_Bioware/Project/Details_Phage|[more]]]'''
 +
<br>
 +
<br>
 +
<br>
 +
<br>
 +
<br>
=== In Vivo Selection ===
=== In Vivo Selection ===
 +
[[Image:UP_SurvTEV_onepla.jpg|left|290px|thumb|'''Figure 3: Ampicillin resistance test.'''
 +
Survival screen, without induced (blue) and with induced (magenta) protease. From left to right increasing ampicillin concentrations are shown.]]
-
In addition to the Phage Display we developed a novel selection system. The design aimed for a cheap and time-saving alternative in contrast to an <i>in vitro</i> screen of protease inhibition kinetics. The assay allows us to select effective inhibitors for a any protease, among the billions of randomly generated mutants of the Microviridin. For this purpose we designed a plasmid containing two devices, first a protease activity detector and second a protease generator. We used the BioBrick <partinfo>I757010</partinfo> (β-lactamase) and <partinfo>K208005</partinfo> (ssTorA) and fused them together via a linker peptide to create our first device. The protease generator is an Arabinose inducible protease in iGEM standard.<br>
+
A proper designed ''in vivo'' selection is very powerful, quick and entails an inherent selection against general toxicity. We designed and employed a multi-component system enabling inexpensive and time-saving selection of microviridin libraries blocking various proteases. Our genetic circuit links with high sensitivity and wide dynamic range a protease generator, a protease detector and a protease inhibitor device. This is achieved by coupling the export of the enzyme ß-lactamase to a signal sequence via an interspersed modularly cloned protease cleavage site. Inducible lactamase and protease generators are united on a single plasmid while the <i>mdnA</i> library is provided in trans. Functionality of the system was characterized under a multitude of conditions for different proteases. Ultimately, a one step selection of a genetic <i>mdnA</i> library against the protease TEV with increasing concentrations of the antibiotic ampicillin yielded inhibiting clones. '''[[:Team:Potsdam_Bioware/Project/Details_Selection|[more]]]'''
-
[[File:UP_images_TorA-construct-design2_20.9.2011.png|center|500px|thumb|'''Figure x:''' Schema of our constructed selection vector containing protease generator and protease activity detector]]<br>
+
<br><br>
-
The system works in a combined manner of the two devices. In order to confer β-lactam antibiotic resistance to cells the β-lactamase has to be exported in the periplasm. This transport is mediated by the TorA export sequence via the Twin-Arginine Translocation (TAT) system (DeLisa, 2008). The linker peptide between the TorA export sequence and the β-lactamase displays the corresponding protease cleavage site. In addition the linker peptide is chosen as short as possible to imitate folding because the TAT pathway only allows transport of correctly folded substrates. The linker peptide as well as the protease are designed as exchangeable parts.<br>
+
<br><br>
-
 
+
<br><br><br> <br>
-
If the protease device is functional, it will cleave the linker peptide between TorA and β-lactamase construct which leaves the cell without any antibiotic resistance. The construct was tested with increasing Ampicillin concentrations. The number of surviving colonies was depended on the export rate of the β-lactamase into the periplasm. A high cleavage rate of the linker peptide leads to a reduced Ampicillin resistance. With expressed protease a dramatically drop of the number of surviving colonies could be observed.<br>
+
-
The survival assay was carried out with and without expressed protease. By increasing of the Ampicillin concentration we could detect a cutoff Ampicillin concentration. The colony forming units (CFU) were counted and compared.<br>
+
-
 
+
-
 
+
-
Bilder über survival screen<br>
+
-
We could show that our system works. The next transformation of the constructed microviridin library and our vector into competent <i>E. coli</i>.<br>
+
-
 
+
-
'''[[:Team:Potsdam_Bioware/Project/Details_Selection|[more]]]'''
+
=== Modeling ===
=== Modeling ===
-
 
+
[[File: UP_3Dplot_Lact_06.png|left|300px|thumb|'''Figure 4:''' <b>Plot predicting lactamase concentrations in the periplasm</b> (and thus the cell fittness) in dependence of the enzyme inhibition reaction coefficient K<sub>D</sub>.]]
-
There is no synthetic biology without modeling, of course. We focused on systems modeling in which the reaction kinetics of the whole system is analyzed and outcomes are predicted. Thus a synthetic biology approach can be chosen because a better understanding of the system is achieved and further changes can be planed - just like in engineering.<br> The following schema shows the major reactions taking place in our cell system. The Roman numbers indicate the place of reactions that were written down as equations and then numerically propagated through time. We were able to see that our system works very well in theory. We learned about correct time-scales for our triggering and we were able to identify expected cell-division rates as a reference for the lab work. '''[[:Team:Potsdam_Bioware/Project/Details_Modeling|[more]]]'''<br>  
+
<br>
-
[[File:UP_schema_modeling2v.png|center|600px|thumb|Simplified schema of our cell system including labels and markers for the chemical reactions in it. Marked reactions are considered in our system modeling.]]<br>
+
There is no synthetic biology without modeling, of course. We focused on systems modeling of our ''in vivo'' selection system in which reaction kinetics are analyzed and outcomes are predicted or fitted to experimental data. On one hand, predictions based on known parameters enable the rational choice of optimized conditions (e.g. induction levels of the protease generator). On the other hand, fitting to biological readouts (e.g. survival) enables deciphering of intrinsic parameters (e.g. interactions constants).<br><br>
 +
Reactions were coded as ordinary differential equations under consideration of sequential induction time points. Substance concentrations were numerically propagated through time. A useful dynamic range and also a certain robustness of our system were confirmed. We learned about correct time-scales for triggering and we predicted cell-division rates as a reference for the lab work. In a final step, we fitted our model to wet-lab measurements. '''[[:Team:Potsdam_Bioware/Project/Details_Modeling|[more]]]'''<br><br><br><br><br>
=== Ethics ===
=== Ethics ===

Latest revision as of 03:02, 29 October 2011

Summary

Modification, Selection and Production of Cyclic Peptides for Therapy

One key task of biopharmaceuticals is the binding and blocking of deregulated proteins. Picking the right lead structure for biopharmaceuticals is very important for success. This year, we developed a novel system for the modification, selection and optimization of peptides showing potential for protease inhibition. These promising inhibitors are found in cyanobacteria and are called Microviridins. They belong to a class of peptides characterized by unusual ω-ester and ω-amide bonds between the amino-acid side chains, which are also referred to as depsipeptides. These modifications are introduced post translationally by a set of enzymes and result in an extraordinary tricyclic cage structure.

Proteases are a large enzyme family and comprise about 647 human gene products. Therefore they are important drug targets. In addition, many harmful bacteria, viruses and fungi use proteases in their reproduction cycle and growth. The ability to block these proteases is highly relevant for therapy. Prominent examples of targets are the angiotensin-converting enzyme (ACE) and HIV proteases.

We chose a 6.5 kb Microviridin gene cluster named mdnABCDE from Mycrocystis aeruginosa NIES843. The mdn gene cluster comprises (i) a gene encoding a precursor peptide named MdnA, which is then modified to form the Microviridin, (ii) two genes encoding ATP-grasp-type ligases named mdnB and mdnC, (iii) an ABC- transporter encoding gene named mdnE as well as (iv) one gene encoding an N- acetyltransferase of the GNAT family named mdnD (Ziemert et al., 2008). Our major aim was to modify the MdnA precursor peptide and optimize its protease inhibiting properties. Towards this goal, we synthesized semi-rational mdnA gene libraries with partially randomized oligonucleotides. One library was cloned, verified by sequencing, and successfully used for selection.

To identify the best candidates inhibiting various given proteases, we established two different selection systems: Phage Display and a new developed in-vivo selection assay. Phage display is a frequently used technique in laboratories, yet to our knowledge, phage display of cellularly cyclized peptides is new. We constructed a fusion between the surface protein geneIII of the phage and the mdnA gene within the mdnA gene cluster. This system was verified by western blotting, Phage-ELISA and controlled phage panning experiments.

We also devised and constructed a recombinant in-vivo selection system linking protease degradation to antibiotic resistance. This system is divided into three parts: first a protease activity detector device, second a protease generator device, and third a protease blocking device. For the protease activity detector device we fused various protease cleavage sites between a signal sequence and the antibiotic resistance conferring enzyme β-lactamase. β-lactamase provides resistance when its transferred into the periplasm. Thus, when the protease cleaves the signal sequence from the β-lactamase antibiotic resistance is abolished and cells die under selective pressure. However, when our Microviridin variants inhibit the protease, cells survive, allowing for an easy and efficient selection of protease inhibitors. We demonstrated a wide dynamic range of the system between expressed and non-expressed protease in the presence of increasing ampicillin concentrations. Importantly, after co-transformation with our mdnA-libary we were able to isolate several clones, which are currently being characterized.

In addition to the practical work, we established a mathematical model of the in-vivo selection system. This model, which we were able to fit to experimental data, helped us to understand the selection process. We modeled the reaction kinetics with ordinary differential equations and simulated and fitted data with matlab.

Last but not least, we had several human practice projects. We send a survey to all members of the German parliament, visited a member of the parliament, held seminars on ethics and invited children to the lab.

  • Ziemert, N., Ishida, K., Liaimer, A., Hertweck, C. & Dittmann, E. (2008). Ribosomal synthesis of tricyclic depsipeptides in bloom-forming cyanobacteria. Angewandte Chemie (International ed. in English) 47, 7756-9



Highlights

Microviridin

Figure 1: HPLC chromatogram of heterologously expressed mdnA


The microviridin sub-project aimed for modifications at the genetic level such that protease inhibiting activities of the resulting peptides are enhanced. We used random mutagenesis or focused randomization of oligonucleotides for the creation of gene libraries, which can be screened for mdnA-variants with therapeutically promising mutations. For further experiments, we also fused mdnA to a myc-tag.
In addition, all coding genes of the mdn-cluster were converted to BioBricks. The microviridin from mdnA was purified and characterized by HPLC and cyclization was verified by mass spectrometry.
For the quick expression of several library clones, we also built auxiliary plasmid backbones with inducible promoters according to a proposed extension of the iGEM cloning standard. [more]


Phage Display

Figure 2: Phage Display of microviridin. Enrichment of MdnA carrying phages upon panning towards a protease was demonstrated.


Phage display is a powerful tool for selecting peptides or proteins that bind and regulate the function of target proteins. Phages robustly link the genotype to the phenotype of the presented peptide allowing for selection of protease binders under controlled in vitro conditions. We established genetics and panning of therapeutically interesting mdnA-libraries towards a panel of several purified proteases. First, the general suitability of phage display for this purpose was shown. An appropriate phagemid vector harboring an mdnA-myc-geneIII fusion gene, was constructed. The expression of the MdnA-myc-geneIII protein in E. coli was shown by western blot. Production of phage particles carrying MdnA were analyzed by Phage-ELISA. Finally, phage display procedures were optimized and an enrichment of MdnA modified phages over a control was shown. [more]




In Vivo Selection

Figure 3: Ampicillin resistance test. Survival screen, without induced (blue) and with induced (magenta) protease. From left to right increasing ampicillin concentrations are shown.

A proper designed in vivo selection is very powerful, quick and entails an inherent selection against general toxicity. We designed and employed a multi-component system enabling inexpensive and time-saving selection of microviridin libraries blocking various proteases. Our genetic circuit links with high sensitivity and wide dynamic range a protease generator, a protease detector and a protease inhibitor device. This is achieved by coupling the export of the enzyme ß-lactamase to a signal sequence via an interspersed modularly cloned protease cleavage site. Inducible lactamase and protease generators are united on a single plasmid while the mdnA library is provided in trans. Functionality of the system was characterized under a multitude of conditions for different proteases. Ultimately, a one step selection of a genetic mdnA library against the protease TEV with increasing concentrations of the antibiotic ampicillin yielded inhibiting clones. [more]







Modeling

Figure 4: Plot predicting lactamase concentrations in the periplasm (and thus the cell fittness) in dependence of the enzyme inhibition reaction coefficient KD.


There is no synthetic biology without modeling, of course. We focused on systems modeling of our in vivo selection system in which reaction kinetics are analyzed and outcomes are predicted or fitted to experimental data. On one hand, predictions based on known parameters enable the rational choice of optimized conditions (e.g. induction levels of the protease generator). On the other hand, fitting to biological readouts (e.g. survival) enables deciphering of intrinsic parameters (e.g. interactions constants).

Reactions were coded as ordinary differential equations under consideration of sequential induction time points. Substance concentrations were numerically propagated through time. A useful dynamic range and also a certain robustness of our system were confirmed. We learned about correct time-scales for triggering and we predicted cell-division rates as a reference for the lab work. In a final step, we fitted our model to wet-lab measurements. [more]




Ethics

For information about an ethics seminar and a survey among politicians see [here].

Software

Have a look at the features and screenshots of our BioLog app on this [page].