Evaluating targets for chimeric antigen receptor therapy

Introduction: In this exercise, you will use the online bioinformatics tools to evaluate whether the protein mesothelin (MSLN) is a suitable target for CAR cell therapy of hepatocellular carcinoma You will need to manually keep track of everything you find using these tools to give you final recommendation regarding its suitability.

Task 1: Extract the sequence of the canonical protein from UniProt

Task 2: Investigate the expression of MSLN in hepatocellular carcinoma and in all healthy tissues using Xenabrowser
Q: In what tissues is the gene recorded expressed?

Task 3: Predict the subcellular location of the protein using DeepLoc
Q: Where does the protein localize to in the cell?

Task 4: Predict the membrane topology of MSLN using TopCons2
Q: Does the protein (provided that it localizes to the outer cell membrane) appear to have a reasonably sized ectodomain for targeting using CARs/antibodies?

Task 5: Predict whether the protein contains signal peptides using SignalP
Q: Does the protein appear to be secreted?

Task 6: Predict whether the ectodomain(s) of the protein are likely to harbor structural epitopes using BepiPred. When you have done this, isolate the prediction pertaining to the predicted ectodomains
Q: Are there any predicted epitopes in the ectodomain(s)?
Q: Why do you think that we isolate the ectodomain after predicting epitopes and not before?

Task 7 Predict phosphorylation sites of MSLN using NetPhosP
Q: Are there any phosphorylation sites in the ectodomain?

Task 8 Check back with the UniProt entry for MSLN
Q: Are any of your predicted features experimentally validated?

Task 9 Isolate a promising ectodomain target from the sequence of MSLN and use BLAST
Q: Do any other human proteins have similar ectodomains that could lead to potential off-target effects?
Q: If yes, do these domains also localize to the membrane?

Task 10 Check if MSLN is expressed in different isoforms
Q: If yes, do all isoforms express the ectodomain and are they equally suitable on all other features? Q: Would you recommend MSLN as a good target for hepatocellular carcinoma? Why? Why not?