Evaluation of Competing J domain:Hsp70 Complex Models in Light of Existing Mutational and NMR Data

📝 Abstract

Ahmad et al. recently presented an NMR-based model for a bacterial DnaJ J domain:DnaK(Hsp70):ADP complex(1) that differs significantly from the crystal structure of a disulfide linked mammalian auxilin J domain:Hsc70 complex that we previously published(2). They claimed that their model could better account for existing mutational data, was in better agreement with previous NMR studies, and that the presence of a cross-link in our structure made it irrelevant to understanding J:Hsp70 interactions. Here we detail extensive NMR and mutational data relevant to understanding J:Hsp70 function and show that, in fact, our structure is much better able to account for the mutational data and is in much better agreement with a previous NMR study of a mammalian polyoma virus T-ag J domain:Hsc70 complex than is the Ahmad et al. complex, and that our structure is predictive and provides insight into J:Hsp70 interactions and mechanism of ATPase activation.

💡 Analysis

Ahmad et al. recently presented an NMR-based model for a bacterial DnaJ J domain:DnaK(Hsp70):ADP complex(1) that differs significantly from the crystal structure of a disulfide linked mammalian auxilin J domain:Hsc70 complex that we previously published(2). They claimed that their model could better account for existing mutational data, was in better agreement with previous NMR studies, and that the presence of a cross-link in our structure made it irrelevant to understanding J:Hsp70 interactions. Here we detail extensive NMR and mutational data relevant to understanding J:Hsp70 function and show that, in fact, our structure is much better able to account for the mutational data and is in much better agreement with a previous NMR study of a mammalian polyoma virus T-ag J domain:Hsc70 complex than is the Ahmad et al. complex, and that our structure is predictive and provides insight into J:Hsp70 interactions and mechanism of ATPase activation.

📄 Content

1 Evaluation of Competing J domain:Hsp70 Complex Models
in Light of Existing Mutational and NMR Data

Rui Sousa*, Jianwen Jiang, Eileen M. Lafer, Andrew P. Hinck,
Liping Wang, Alexander B. Taylor, and E. Guy Maes

Dept. of Biochemistry, U. of TX Health Sci. Ctr., 7703 Floyd Curl Drive, San Antonio TX 78229-3900 *Correspondence: sousa@biochem.uthscsa.edu, 210-567-2506, 210-567-6595 (fax)

Ahmad et al. recently presented an NMR-based model for a bacterial DnaJ J domain:DnaK(Hsp70):ADP complex(1) that differs significantly from the crystal structure of a disulfide linked mammalian auxilin J domain:Hsc70 complex that we previously published(2). They claimed that their model could better account for existing mutational data, was in better agreement with previous NMR studies, and that the presence of a cross-link in our structure made it irrelevant to understanding J:Hsp70 interactions. Here we detail extensive NMR and mutational data relevant to understanding J:Hsp70 function and show that, in fact, our structure is much better able to account for the mutational data and is in much better agreement with a previous NMR study of a mammalian polyoma virus T-ag J domain:Hsc70 complex than is the Ahmad et al. complex, and that our structure is predictive and provides insight into J:Hsp70 interactions and mechanism of ATPase activation.

I. Comparison with previous NMR studies of J-Hsp70 interactions: The regions of the J domains of 2 J proteins that interact with Hsp70 have been previously mapped by NMR chemical shift analysis. Greene et al.(3) mapped them in a complex of bacterial DnaJ and DnaK and Garimella et al.(4) mapped them in a complex of polyoma virus T-ag J domain and bovine Hsc70. Ahmad et al. state that the data from their complex agrees with the data of Greene et al. and that our structure does not. The Ahmad et al. complex does show better agreement with the Greene et al. study, though there are discrepancies: Greene et al. observe shifts in the functionally critical H and D residues of the invariant HPD motif, while Ahmad et al. conclude that their own data do not support a role for the HPD loop in binding DnaK:ADP, and say that they observed “no chemical shift changes for residues Asp35 (and His33) of the HPD loop…It appears that the HPD motif is exclusively involved in the interaction with the ATP state, although it could also have a purely structural role.”
Ahmad et al. suggest that the Greene et al. results for D35 could have been due to asp-specific pH effects.
However Greene et al. observed significant peak broadening for H33 as well as D35 upon titration with DnaK:ADP so asp-specific pH effects or differences in nucleotide state cannot explain this discrepancy.

More importantly, our structure is of a mammalian Hsc70:J complex, and specifically we use bovine Hsc70, as was used by Garimella et al. When our structure is compared to the chemical shift/peak broadening/protection data from Garimella et al. we find excellent agreement: 8 of 13 residues in PyJ identified as shifted/broadended by Garimella et al. correspond to auxilin residues that are close to Hsc70 in our structure. In contrast, the Ahmad et al. complex is entirely inconsistent with the Garimella data: none of the residues identified as broadened/shifted in Garimella correspond to J residues identified in the Ahmad et al. complex as close to labeled DnaK residues. This is expected. The Garimella and Greene studies indicated that prokaryotic Hsp70 binds DnaJ differently than how mammalian Hsc70 binds J protein. Bacterial DnaJ uses primarily J domain helix II to bind bacterial DnaK, while the mammalian (viral) protein uses primarily helix III to bind mammalian Hsc70. The different binding mode of Hsc70:PyJ vs. DnaK:DnaJ was the central conclusion of the Garimella study, being incorporated in the title, addressed in the abstract (“our novel evidence implicating helix III differs from evidence that Escherichia coli DnaK primarily affects helix II and the HPD loop of DnaJ.”), and in the introduction, results, and discussion of that study. This conclusion is also supported by extensive mutational data (see below). It is primarily auxilin J domain helix III which contacts Hsc70 in our structure. Since our complex is of a mammalian Hsc70 and a mammalian J, it should be compared, and is expected to be more similar to, the mammalian complex studied by Garimella, rather than the bacterial complex studied by Greene.
The excellent agreement between the Garimella et al. results and our structure is strong evidence for the physiological relevance of the latter. These NMR data and relevant references are summarized in table 1.

2

II. Mutational data:

1. At least 38 single, multiple point and deletion mutants in the bacterial DnaJ J domain have been characterized for complementation in vivo, with a subset also characterized for binding to and ability to stimulate DnaK in

This content is AI-processed based on ArXiv data.