Crystal Structure of a Chimeric Antigen Receptor (CAR) scFv Domain Rearrangement Forming a VL-VL Dimer

Round 1

Reviewer 1 Report

The manuscript submitted by Philippe Youkharibache and colleagues is a welcome and somewhat old-fashioned description of an interesting crystal structure. Although it is a simple report on crystal structure, I believe it deserves publication.

The Chimeric Antigen Receptor (CAR) T-cell therapy requires the design of transmembrane proteins, which bind specific antigens and cause immune response. The analysis presented in the submitted manuscript indicates two important issues:

a) Although it is usually assumed that CAR-bound Ig domains are in monomeric form to bind a single target, in this case they are dimeric. Although this could be an experimental artifact, it is important to consider this possibility.

b) The inter-molecular interface is not canonical, it is thermodynamically more stable than the canonical one, and it is not predicted by AlphaFold2.

In my opinion, this manuscript can be published like it is.

Author Response

We thank Referee #1 for their review and positive comments on and their support of publication for our manuscript.  We take the comment on our structure description as "a welcome and somewhat old-fashioned" as a compliment.

Reviewer 2 Report

The manuscript by Cheung et al presents a structural study of the anti-CD19 CAR 47G4-CD828Z. Using X-ray crystallography, they determined the high-resolution structure of the IgVL homodimer and discovered that the homodimer was formed via an unexpected inverted VL | VL interface. Furthermore, they conducted MD simulations to demonstrate that this newly discovered inverted interface was more stable than the canonical interface found in the well-known Bence Jones protein. Overall, the study is intriguing and a strong candidate for publication in Crystals. However, I have a few minor comments.

 

1. To demonstrate that the protein was proteolyzed during crystallization, it would be better to perform SDS-PAGE analysis or even mass spectrometry analysis of the crystals. There could be other reasons why adding protease inhibitors did not yield any crystals. Moreover, are there any biological implications for this proteolysis?

 

2. Figure 3A-D is hard to read. Aligning the orientations of A&C and B&D could be beneficial. Changing the color schemes of the structures to a rainbow spectrum (blue for the more N-terminal and red for the more C-terminal) might also improve the readability.

 

3. The authors should discuss the observed inverted dimer or canonical dimer in the full-length CAR proteins, preferably by drawing a cartoon diagram. Would either of the dimer interfaces face steric hindrance?

Author Response

Please see attachment

Author Response File: Author Response.pdf