Structural basis for kinase inhibition in the tripartite E. coli HipBST toxin-antitoxin system

Reviewer #1 (Public Review):

This work describes a structural analysis of the tripartite HipBST toxin-antitoxin (TA) system, which is related to the canonical two-component HipBA system composed of the HipA serine-threonine kinase toxin and the HipB antitoxin. The crystal structure of the kinase-inactive HipBST complex of the Enteropathogenic E. coli O127:H6 was solved and revealed that HipBST forms a hetero-hexameric complex composed of a dimer of HipBST heterotrimers that interact via the HipB subunit. The HipS antitoxin shows a structural resemblance to HipA N-terminal region and the HipT toxin represents to the core kinase domain of HipA, indicating that in HipBST the hipA toxin gene was likely split in two genes, namely hipS and hipT.

The structure also reveals a conserved and essential Trp residue within the HipS antitoxin, which likely prevents the conserved "Gly-rich loop" of HipT from adopting an inward conformation needed for ATP binding. This work also shows that the regulating Gly-rich loop of the HipT toxin contains conserved phosphoserine residues essential for HipT toxicity that are key players within the HipT active site interacting network and which likely control antitoxin binding and/or activity.

Strengths:

The manuscript is well written and the experimental work well executed. It shows that major features of the classical two-component HipAB TA system have somehow been rerouted in the case of the tripartite HipBST. This includes the N-terminal domain of the HipA toxin, which now functions as bona fide antitoxin, and the partly relegated HipB antitoxin, which could only function as a transcription regulator. In addition, this work shows a new mode of inhibition of a kinase toxin and highlights the impact of the phosphorylation state of key toxin residues in controlling the activity of the antitoxin.

Weaknesses:

A major weakness of this work is the lack of data concerning the role of HipB, which likely does not act as an antitoxin. Does it act as a transcriptional regulator of the hipBST operon and to what extent both HipS and HipT contribute to such regulation? These are still open questions.

In addition, there is no in-depth structural comparison between the structure of the HipBST solved in the work and the two recent structures of HipBST from Legionella. This is also a major weakness of this work.

Reviewer #2 (Public Review):

The work by Bærentsen et al., entitled "Structural basis for regulation of a tripartite toxin-antitoxin system by dual phosphorylation" deals with the structural aspects of the control of the hipBST TA operon, the role of auto-phosphorylation in the activation and neutralisation of the enzyme and the direct effects of HipS and HipB in neutralisation. This is a follow-up to the Vang Nielsen et al., and Gerdes et al., papers from the same authors on this very unique TA module, that brings forth a thorough and well written dissection of an unusually complex regulatory system.

This is a much improved manuscript, the paper is more focused and the message is now clear.

Author Response:

Reviewer #1 (Public Review):

[...] Strengths:

The manuscript is well written and the experimental work well executed. It shows that major features of the classical two-component HipAB TA system have somehow been rerouted in the case of the tripartite HipBST. This includes the N-terminal domain of the HipA toxin, which now functions as bona fide antitoxin, and the partly relegated HipB antitoxin, which could only function as a transcription regulator. In addition, this work shows a new mode of inhibition of a kinase toxin and highlights the impact of the phosphorylation state of key toxin residues in controlling the activity of the antitoxin.

Weaknesses:

A major weakness of this work is the lack of data concerning the role of HipB, which likely does not act as an antitoxin. Does it act as a transcriptional regulator of the hipBST operon and to what extent both HipS and HipT contribute to such regulation? These are still open questions.

We thank the reviewer for their feedback and will include a supplementary figure (Figure 1 supplement 2) and accompanying text that shows the transcriptional role of HipB, and how HipS and HipT influence this regulatory effect.

In addition, there is no in-depth structural comparison between the structure of the HipBST solved in the work and the two recent structures of HipBST from Legionella. This is also a major weakness of this work.

A structural comparison to the recent structures from Legionella will be included in the discussion, including Figure 6 supplement 1.