New Roles for HAMP Domains: the Tri-HAMP Region of Pseudomonas aeruginosa Aer2 Controls Receptor Signaling and Cellular Localization

Abstract

The Aer2 chemoreceptor from Pseudomonas aeruginosa is an O₂ sensor involved in stress responses, virulence, and tuning the behavior of the chemotaxis (Che) system. Aer2 is the sole receptor of the Che2 system. It is soluble, but membrane associated, and forms complexes at the cell pole during stationary phase. The domain arrangement of Aer2 is unusual, with a PAS sensing domain sandwiched between five HAMP domains, followed by a C-terminal kinase-control output domain. The first three HAMP domains form a poly-HAMP chain N-terminal to the PAS sensing domain. HAMP domains are often located between signal input and output domains, where they transduce signals. Given that HAMP1 to 3 reside N-terminal to the input-output pathway, we undertook a systematic examination of their function in Aer2. We found that HAMP1 to 3 influence PAS signaling over a considerable distance, as the majority of HAMP1, 2 and 3 mutations, and deletions of helical phase stutters, led to nonresponsive signal-off or off-biased receptors. PAS signal-on lesions that mimic activated Aer2 also failed to override N-terminal HAMP signal-off replacements. This indicates that HAMP1 to 3 are critical coupling partners for PAS signaling and likely function as a cohesive unit and moveable scaffold to correctly orient and poise PAS dimers for O₂-mediated signaling in Aer2. HAMP1 additionally controlled the clustering and polar localization of Aer2 in P. aeruginosa. Localization was not driven by HAMP1 charge, and HAMP1 signal-off mutants still localized. Employing HAMP as a clustering and localization determinant, as well as a facilitator of PAS signaling, are newly recognized roles for HAMP domains. IMPORTANCE P. aeruginosa is an opportunistic pathogen that interprets environmental stimuli via 26 chemoreceptors that signal through 4 distinct chemosensory systems. The second chemosensory system, Che2, contains a receptor named Aer2 that senses O₂ and mediates stress responses and virulence and tunes chemotactic behavior. Aer2 is membrane associated, but soluble, and has three N-terminal HAMP domains (HAMP1 to 3) that reside outside the signal input-output pathway of Aer2. In this study, we determined that HAMP1 to 3 facilitate O₂-dependent signaling from the PAS sensing domain and that HAMP1 controls the formation of Aer2-containing polar foci in P. aeruginosa. Both of these are newly recognized roles for HAMP domains that may be applicable to other non-signal-transducing HAMP domains and poly-HAMP chains.

Keywords: Aer2; HAMP domain; Pseudomonas aeruginosa; cellular localization; chemoreceptor signaling; fluorescence microscopy.

FIG 1

P. aeruginosa Aer2 models and structures. (A) Aer2 associates with Che2 cluster proteins (CheA2 and CheW2) to form a membrane-associated array near the cell pole (10, 11). Aer2 dimers contain an N-terminal tri-HAMP region (HAMP1 to 3), followed by a PAS domain, a di-HAMP unit (HAMP4 to 5), and a kinase control domain. An Aer2 dimer model, as predicted by AlphaFold (in two parts: res. 1 to 285 and 286 to 679) is shown, as are the crystal structures of N-terminal HAMP1 to 3 (res. 1 to 156; PDB 3LNR [17]) and a PAS-heme dimer (res. 173 to 287; PDB 4HI4 [31]). The similarity between the AlphaFold models and crystal structures is striking, with all structural elements correctly identified by AlphaFold (the RMSD of the HAMP1 to 3 dimer is 0.747 Å, whereas the PAS monomer is 1.244 Å). Individual HAMP monomers comprise two parallel α-helices, AS1 and AS2, separated by a semistructured connector. (B) Aer2 HAMP1 to 3 sequence alignment. Buried core residues (based on PBD 3LNR) are highlighted in gray, whereas important connector residues are highlighted in red and brown. See text for details. Note the two-residue overlap between HAMP2 and 3.

FIG 2

HAMP residues targeted for mutagenesis and their impact on Aer2 signaling in E. coli BT3388. (A) HAMP1 residues that bias signaling in both Tsr-HAMP and a Tar-Aer2 HAMP1 chimera that were introduced into full-length Aer2 (24, 25). (B) Charged residues at the HAMP2 to 3 interface that were neutralized by substitution with Ser. (C) HAMP2 residues that are highly conserved in poly-HAMP proteins (see Fig. S3) and were targeted by site-specific random mutagenesis. (D) Hydrophobic connector residues that pack against or between HAMP helices and that were substituted with Ser, which impairs Tsr function at the same positions (19). In all cases, the HAMP structure is rotated –45° on the x axis in comparison with Fig. 1.

FIG 3

Helical HAMP stutters and deletions to modulate the helical phase. (A) Native helical stutters (an insertion of four, or deletion of three, residues) occur at the N and C termini of HAMP2. Stutters result in helices occupying one of two conformations, conformation A or B, as shown on the helical wheels. The dotted line in conformation B represents the AS2 phase stutter. In conformation A, the AS1 helix is underwound and more dynamic, whereas in conformation B, AS2 is underwound and more dynamic. Four overlapping residue deletions were created (shown in purple and blue in both panels A and B) to bring the helical alignment back into phase. (B) The effects of the stutter deletions on Aer2 behavior in E. coli BT3388. The HAMP structure is rotated –45° on the x axis compared with Fig. 1. H, HAMP.

FIG 4

Assessing alternating HAMP conformations through the poly-HAMP chain. (A) HAMP substitutions in the same locations of contiguous HAMP domains that were tested independently for their effects on Aer2 signaling in E. coli BT3388. Residues are shown on the sequence alignment and on the HAMP1 to 3 structure (rotated –45° on the x axis compared with Fig. 1). Residues in the same series are shown in the same color. (B). The HAMP substitutions in each series and their impact on Aer2 function.

FIG 5

The impact of HAMP signal-off lesions on PAS signaling. Behavioral assays for Aer2-HAMP signal-off mutants, Aer2-PAS signal-on mutants, and Aer2-HAMP/PAS double mutants, showing the average percentage of E. coli BT3388 cells tumbling over a 1-s period 30 s after switching to air or N₂. **, P < 0.01. The hash symbol indicates that smooth swimming was delayed 20 s in N₂ compared to WT Aer2. Behavioral assay results for Aer2-R260A and Aer2-L264A are reproduced with permission from reference . Copyright 2021 American Chemical Society. The PAS-Hβ locations of R260 and L264 are shown on the PAS monomer from Fig. 1, rotated 15° on the y axis.

FIG 6

Localization of full-length Aer2-eGFP and N-terminal truncation mutants in P. aeruginosa. Aer2-eGFP models for full-length and truncated Aer2 proteins created from the AlphaFold model in Fig. 1 and the structure of eGFP (PDB 2Y0G [50]), manually attached to the C terminus of Aer2. Representative fluorescence images are shown for cells with an OD₆₆₀ of 1.9 to 2.1 (2.1 for Aer2_1–679-eGFP, 1.9 for Aer2_38–679-eGFP, and 2.0 for Aer2_57–679-eGFP and Aer2_173–679-eGFP), along with the relative fluorescence intensity of a subset of 70 to 90 cells from 3 to 4 independent experiments.

FIG 7

Electrostatic potential map for HAMP1 to 3 and effects of HAMP1 charge-neutralization on Aer2 behavior. (A) Electrostatic potential map of HAMP1 to 3 (res. 1 to 157) showing positive charge at the N terminus (blue), in contrast to negative charge throughout the remainder of the structure (red). (B) Location of HAMP1 N-terminal charged residues (yellow) in comparison with L21 and L44 (cyan) from Fig. 2A. The effects of charge-neutralization on Aer2 behavior in E. coli BT3388 are also shown.

FIG 8

Model for Aer2 PAS-HAMP signaling. N-terminal HAMP1-3 are coupling partners for PAS signaling that orient and poise PAS dimers for signaling. During signaling, HAMP1 to 3 move as a coordinated unit. When O₂ binds to the PAS domain, it causes the rotation of PAS monomers around the dimer interface with concomitant conformational changes in N-terminal HAMP1 to 3 and C-terminal HAMP4 to 5 (center panel). Distal PAS signaling motions force C-terminal HAMP4-5 to loosen its inhibitory control over the kinase control domain. In the presence of HAMP1 to 3 signal-off lesions (left panel, marked with an X), PAS signaling motions are blocked so that HAMP4 to 5 retain inhibitory control over the kinase control domain. In contrast, distal PAS signal-on lesions (right panel, marked with an X) impose the on-conformation of the PAS domain, which shifts HAMP1 to 3 into the on-conformation and HAMP4 to 5 into a deinhibitory state that can no longer obstruct the activity of the kinase control domain. Solid arrows signify conformational changes. H, HAMP.