This work provides insights into the molecular mechanism of phosphorelay systems in fungi. Various hybrid histidine kinases (hHK) are able to phosphoryl transfer to the same His-phosphotransfer protein (HPt) in order to develop a response. However, demonstration of that function has not been completely elucidated. So far, the best characterized phosphorelay system at molecular and structural level in fungi is formed by the only hybrid histidine kinase from Saccharomyces cerevisiae named Sln1 (from group VI) and HPt Ypd1^1,2. Thus, to understand the phosphotransfer from many to one we have used the phosphorylatable receiver domains (REC-1) from five hybrid histidine kinases belonging to different groups (III, IV, V, VI, and XI), present in the thermophilic fungus Chaetomium thermophilum (Fig. 1a).

Fig.1. Phosphorelay systems in C. thermophilum. a) Schematic representation of domain organization of 5 hHKs that should phosphotransfer to HPt. b) Phosphotransfer experiments from 3 REC-1 domains phosphorylated with acetyl phosphate (AcP) or phosphoramidate (PAM) to HPt.

Using native PAGE electrophoresis we have observed phosphorylation of HPt upon incubation with the REC-1 domains from group III (REC_hHK3), VI (REC_hHK6) and V (REC_hHK5), previously phosphorylated with phosphodonors (Fig. 1b).  Whereas the REC-1 domains from hybrid histidine kinases from group IV (REC_hHK4) and XI (REC_hHK11) were not, probably due to lack of structural stability. Interestingly, we did not observe complex formation in the native PAGE between REC-1 domains and HPt, and  using microscale thermophoresis the Kd ranged from 10 to 100 µM and indication of lower affinity for recognition compared to other complexes between bacterial cognate two-component systems that show phosphotransfer³.

 We obtained a complex between the REC_hHK6 and HPt in the presence of the phosphomimetic beryllium trifluoride that was bound to the phosphorylatable D1221 in REC_hHK6 at a phosphotransfer distance of H105 of HPt (Fig. 2a). The structure of this complex is similar to the previous structure of the REC-1 domain of Sln1 from S. cerevisiae and Ypd1 also bound to the phosphomimetic. We also obtained the structure of REC_hHK3 alone and the REC-1 from Sln1 from Candida albicans (REC_{Cal_Sln1}), however, none of them contained phosphomimetic bound even if it was added during the crystallization conditions.  The superposition of REC_hHK3 over REC_hHK6 in the complex did not caused major clashes and 75% of the residues involved in interactions with HPt were conserved in nature and sequence explaining the phosphotransfer.

Interestingly, we observed that the phosphomimetic bound in REC_hHK6 was not stabilized by contacts with the conserved Thr in β4 as in the complex REC_{Sc_Sln1} -Ypd1 (Fig. 2b)². Also, Thr in β4 and Phe in β5 were not oriented towards the active site as to follow the Y-T mechanism observed in REC_{Sc_Sln1} and other bacterial REC domains⁴ (Fig. 2b). Moreover, the residue before the conserved Thr, a conserved Leu, was not adopting the appropriate rotamer to follow the Thr movement. We hypothesize that the absence of this Leu-Thr switch prevents the stabilization of phosphoryl groups at the active center which may be a useful strategy to phosphotransfer when the affinity for complex formation is low. The structures of REC_hHK3 and REC_{Cal_Sln1} do not show the L-T switch and the Y-T mechanism, thus, supporting our findings.

Fig. 2. Complex structure of REC_hHK6 bound to Ct_HPt. a) Complex containing phosphomimetic and stabilized by interacting residues. b) Zoom in on the active center of the complex for REC-1 from C. thermophilum hHK6 and S. cerevisiae Sln1.

 We also have observed that the structure of HPt in the complex differs from Ypd1 from S. cerevisiae as it contains a long N-terminal and a short loop between helix αD and αE (Fig. 3a). In Sc_Ypd1, the N-terminal is short and the loop is very long compacted to the helix bundle. Surprisingly, HPt of other human pathogenic fungi show long N-terminals and short loops while Candida species show short N-terminal and variable length of loops, being longer in C. albicans and shorter in C. auris. We solved the envelope structure of HPt Ypd1 from C. albicans by Small Angle X-ray Scattering coupled to Size Exclusion Chromatography that allowed to model a structure were the long loop had an extended conformation in contrast to the Sc_Ypd1 (Fig. 3b). Finally, phosphotransfer experiments from REC_{Cal_Sln1} to WT and to a mutant lacking the loop (∆107-147) demonstrated that the loop is not involved in phosphotransfer (Fig. 3c). Fig. 3. Structure of HPt. a) Structural comparison between HPt from C. thermophilum and Ypd1 from S. cerevisiae. b) SEC-SAXS structure of Ypd1 from C. albicans fitting its model structure.

Finally, the structure of HPt isolated was compared with its structure in the complex which barely showed changes. However, we observed the formation of salt bridges in the crystal stabilizing dimers through an ionic interface that seemed forced by crystal packing but which is conserved in other HPt (Fig. 3d).

References:

1. Xu, Q., Porter, S. W. & West, A. H. The yeast YPD1/SLN1 complex: insights into molecular recognition in Two-component signaling systems. Structure 11, 1569–1581 (2003).

2. Zhao, X., Copeland, D. M., Soares, A. S. & West, A. H. Crystal structure of a complex between the phosphorelay protein YPD1 and the response regulator domain of SLN1 bound to a phosphoryl analog. J Mol Biol 375, 1141–1151 (2008).

3. Willett, J. W. et al. Specificity residues determine binding affinity for Two-component signal transduction systems. mBio 4, e00420-13 (2013).

4. Gao, R., Bouillet, S. & Stock, A. M. Structural basis of response regulator function. Annu Rev Microbiol 73, 175–197 (2019).