![[LEFT PANEL] Typically membrane proteins are assayed by expressing recombinant
fragments, often containing only the active site of the enzyme, and are interrogated
in a high-throughput solution-based assay. While cost-effective and expeditious, this
format ignores any organization, structure and topology imparted by membrane.
[RIGHT PANEL] An alternative assay is to examine endogenous or over-expressed
enzymes in a living cellular system. While this system faithfully replicates the
membrane environment and contains the full compliment of every other relevant
animal protein, it is slow, very expensive, and not readily adaptable to high-
throughout testing of a chemical library.
Importantly, efficacy of compounds identified in a solution assay tends to correlate
poorly with the efficacy in cellular assays.
*MIDDLE PANEL+ TDA 2.0™ is an enabling technology which bridges the gap between
these two formats, providing the context afforded by a biological membrane in a
platform fully compatible with HTS and all detection formats.
3](https://image.slidesharecdn.com/blueskytdasupportingdata08nov10-12967467321427-phpapp02/85/SmartScreen-Technology-for-Building-a-Better-Assay-3-320.jpg)
SmartScreen Technology for Building a Better Assay
- 1. 1
- 2. There are several classes of membrane proteins: single- and multi- pass transmembrane proteins, proteins which are associated with the membrane via lipid anchors (such as myrisoylation, palmitoylation or GPI anchors) or electrostatic interactions, and proteins which are normally cytosolic but form complexes with membrane proteins. TDA 2.0™ is not suitable for use with multi-pass transmembrane proteins in general, however, all other membrane proteins which have distinct domains on one side of the membrane would work with TDA 2.0™, regardless of which subcellular membrane, or face of the membrane, the protein is associated with. 2
- 3. [LEFT PANEL] Typically membrane proteins are assayed by expressing recombinant fragments, often containing only the active site of the enzyme, and are interrogated in a high-throughput solution-based assay. While cost-effective and expeditious, this format ignores any organization, structure and topology imparted by membrane. [RIGHT PANEL] An alternative assay is to examine endogenous or over-expressed enzymes in a living cellular system. While this system faithfully replicates the membrane environment and contains the full compliment of every other relevant animal protein, it is slow, very expensive, and not readily adaptable to high- throughout testing of a chemical library. Importantly, efficacy of compounds identified in a solution assay tends to correlate poorly with the efficacy in cellular assays. *MIDDLE PANEL+ TDA 2.0™ is an enabling technology which bridges the gap between these two formats, providing the context afforded by a biological membrane in a platform fully compatible with HTS and all detection formats. 3
- 4. How does template directed assembly work? Engineered recombinant HIS-tagged proteins are produced such that the HIS-tag is on the correct terminus of the protein to reflect the polarity of the enzyme with respect to the membrane (for example, ecto-domains of a receptor would be C-terminally tagged while intracellular domains are N-terminally tagged). TDA 2.0™ is a soluble stable liposome which is made from derivitized lipids which have Ni-NTA covalently attached to the lipid head group. This allows the HIS-tagged proteins to bind to the liposome creating an environment much like a cellular membrane. Unlike a sepharose or agarose bead where HIS-tagged proteins must detach and reattach to translate across the surface, the lipids are fully fluid within the 2-dimensional surface of the liposome, so associated proteins can translate and rotate freely. The spatial and relational organization provided by the membrane surface, combined with this fluidity, promotes the formation of higher- order structures, such as homo- or hetero- dimers or multimers, and allows recruitment of accessory factors. 4
- 5. 5
- 6. Insulin receptor is activated by dimerization which leads to trans-phosphorylation on several tyrosines. The phosphorylated dimer is the active RTK. Typically, in order to see activation in solution, manganese is included in the reaction, which creates a non-physiological environment. Above, the left panel shows a radiometric filter binding assay for autophosphorylation, while the right panel shows autophosphorylation and phosphorylation of an IRS1-derived peptide substrate. As you can see, addition of 10mM MnCl2 increases autophosphorylation as well as substrate phosphorylation. However, addition of TDA 2.0™ in a physiological relevant buffer without MnCl2 shows robust activation. We interpret this data to show enhanced functional dimerization of InR on TDA 2.0™ in a physiologic buffer. 6
- 7. Zap-70 is a T-Cell receptor effector that is normally activated by recruitment to the phosphorylated ITAM domains of CD3-zeta. In vitro, recombinant Zap-70 is difficult to screen as it has been reported to have a very long and variable lag phase. We see this same effect as shown in the top three panels where after two hours we see dramatically different activity of Zap-70 in three different experiments. However, addition of TDA 2.0™ reduces the lag time significantly and increases the predictability of activation, creating a very robust assay for Zap-70. This data was generated using a Caliper EZ reader. 7
- 8. Many enzymes show enhanced activity when assayed in the presence of TDA 2.0™. While secondary in importance to the effects TDA 2.0™ has on improving the biology of an enzyme, this is nonetheless another very beneficial feature of TDA 2.0™. 8
- 9. When looking at the ability of an enzyme to phosphorylate a peptide substrate, researchers often use synthetically derived peptides, such as poly-Glu(4)-Tyr. We noticed that when examining activity of ErbB4, a receptor tyrosine kinase (RTK), towards PolyGlu, there was no improvement in activity in the presence of TDA 2.0™. However, when examining activity towards peptides derived from natural substrates of ErbB4, such as Abl and Src, a considerable enhancement in activity is noted in the presence of TDA 2.0™. We interpret this as an indication that the substrate preference of the enzyme is altered when in the membrane context, perhaps selecting more biologically relevant substrates. 9
- 10. To investigate that systematically, we used a different RTK, TrkB, and contracted Molecular Devices to screen a library of peptides in the absence (red) and presence (blue) of TDA 2.0™. The first observation is that the best substrates of TrkB are completely different in the presence and absence of TDA 2.0™. This indicates to us that a key biological property of the enzyme, namely substrate selection, is significantly altered by TDA 2.0™. The second observation from this data set comes from analysis of the sequences of the peptides substrates above. In the absence of TDA 2.0™, comparison of substrate sequence to the non-redundant protein database reveals these substrates either fail to match anything in the database (they are synthetic peptides) or they match viral proteins, which are not likely to be natural substrates of TrkB. In the presence of TDA 2.0™ many synthetic or non-relevant substrates are also identified, however, peptides derived from IRS1 and EGFR, known substrates of TrkB, are identified as substrates. This indicates that presentation of TrkB in the context of TDA 2.0™ biases the substrate selectivity towards relevant substrates of the enzyme. 10
- 11. Initial rates were measured and Km for ATP determined for various enzymes in the presence and absence of TDA 2.0™ using the Caliper EZ reader platform. All enzymes examined to date show significantly lower Km ATP in the presence of TDA 2.0™. 11
- 12. Data presented by Dr. Kathleen Seyb (Marcie Glicksman’s group at Harvard Neuroscience/Brigham and Women’s) at the Society for Biomolecular Sciences annual meeting. They had a grant-driven project to screen Lyn kinase through their library of 75,000 compounds. In order to get high enough signal in their solution-based assay, high concentrations of enzyme (>200 nM) had to be used which made the screen financially impossible. Addition of TDA 2.0™ reduced the amount of enzyme required for a good signal over 25-fold and reduced the cost per well by 50%, leading to successful completion of the screen. 12
- 13. This graph shows a subset of the compounds which were screened both in the presence and absence of TDA 2.0™. It’s not a direct comparison as the enzyme concentration required to get data in the absence of TDA 2.0™ is not the same, and the signal was much lower, but is instructive nevertheless. The graph plots % inhibition in the absence of TDA 2.0™ along the X-axis and in the presence along the Y. Noticeably, TDA 2.0™ alters the pharmacology of Lyn as there are hits unique to each condition. We’re in the process now of examining these compounds in follow- up cellular assays to try to determine if using TDA 2.0™ leads to better quality lead compounds and is more predictive of cell-based assays. Regardless, this data shows that TDA 2.0™ reveals differences in compound SAR (structure-activity relationship), 13
- 14. Insulin signal in the cell initiates upon receptor dimerization and activation by transphosphorylation on specific tyrosines. The phosphotyrosines serve as binding sites for PI3k which is recruited and propagates the signal by converting PIP2 to PIP3, in turn recruiting PDK1 and AKT to the membrane. Activation of Akt by PDK1 and mTOR leads to phosphorylation of many AKT-substrates when ultimately lead to biological effects such as lipolysis, glucose update, growth or proliferation. We are developing an assay which replicates many of these steps in a chemically defined system. 14
- 15. The first step we’re replicating is the phosphorylation of AKT by PDK1. We’ve made a HIS tagged construct of AKT, and well as a HIS-tagged form of PDK1 to deliver these enzymes to the membrane without PIP3. To extend the utility of this assay format, we included GST-tagged mTOR which phosphorylates AKT on S473. When AKT is phosphorylated on both S473 and T308, AKT kinase activity is increased several orders of magnitude. 15
- 16. This figure shows Western blots of reactions containing different combinations of PDK, mTOR and AKT with and without TDA 2.0™. The top panels are anti-GST Westerns showing consistent loading of GST-tagged mTOR. The second set of panels show consistent loading of AKT (lower band) and PDK1 (upper band) using an anti-HIS tag antibody. The third set of panels employs a phospho-specific anti-AKT-pS473 antibody to show phosphorylation of AKT by GST-tagged mTOR. The lower set of panels employs a phospho-specific anti-AKT-pT308 antibody to show phosphorylation of AKT by HIS-tagged PDK1. In the presence of TDA 2.0™, when AKT and PDK are combined, phosphorylation of AKT increases 2-3 fold (compare lanes 5 to 6). This is not solely due to co-localization of AKT and PDK as the same result is obtained using a FLAG-tagged version of PDK1 (data not shown). Further, when AKT and mTOR are combined in the presence of TDA 2.0™, phosphorylation on S473 increases 4-5 fold (compare lanes 7 to 8). Since mTOR is GST-tagged and not co-localized, this confirms our result with FLAG-tagged PDK1 and indicates that AKT is a better substrate for it’s upstream activators when associated with a membrane such as TDA 2.0™. 16
- 17. Finally, when examining the ability of AKT to phosphorylate CROSSTide™, only when AKT, PDK1 and mTOR are combined in the presence of TDA 2.0™ do we see robust kinase activity. 17
- 18. 18