Computational Binder Design

Our protein modeling team designed binders using machine learning-based tools, as described on our model page. The sequence results for novel albumin binding moieties are attached below:

> 1512K44C|DIFFUSION NOVEL STRUCTURE LENGTH 53
MSRKKERAEELYNTALLSARRGNKKAAERAAEIILEDTGDEEAACKAREALKAI

> 2630E2C|PARTIAL DIFFUSION STRUCTURE LENGTH 53
MSCEEKKIEEHKKKILAELDALGINNKLIKAEIRKSKIPEDMETLFEEIKAERA

> 1712E8C|DIFFUSION NOVEL STRUCTURE LENGTH 40
MDLLKKADCKAKEANELQRKGGKLSDIMKLVKEAEELREAA

Laboratory Results

Description of Image — Figure 1. Purified ABD035_C22, ABD035_C29, ND54_C45, and PD54_C3 protein Gel showing the presence of our modified protein.

Due to the difficulty of procuring plasmids in time for our project, we were unable to test the binding affinity of our modified albumin binding moieties (ABM). If we were to test the binding affinity we would measure the amount of ABM that successfully bound to human serum albumin through equilibrium dialysis and compare the binding percentage between the different modified ABM and our wild-type ABM.

Next Steps

Once we confirm that our modified ABM exhibits improved affinity for human serum albumin, our next step will be to connect this complex with a fourth-generation cephalosporin. We plan to begin by synthesizing a linker capable of facilitating click chemistry. Using Cefepime as a representative fourth-generation cephalosporin, we will then bind the antibiotic to the albumin-ABM complex and measure whether the antibiotic is successfully incorporated into the complex.

Please view this wiki on a computer for the best experience!