Alternative Architecture of the E. coli Chemosensory Array

Abstract

Chemotactic responses in motile bacteria are the result of sophisticated signal transduction by large, highly organized arrays of sensory proteins. Despite tremendous progress in the understanding of chemosensory array structure and function, a structural basis for the heightened sensitivity of networked chemoreceptors is not yet complete. Here, we present cryo-electron tomography visualisations of native-state chemosensory arrays in E. coli minicells. Strikingly, these arrays appear to exhibit a p2-symmetric array architecture that differs markedly from the p6-symmetric architecture previously described in E. coli. Based on this data, we propose molecular models of this alternative architecture and the canonical p6-symmetric assembly. We evaluate our observations and each model in the context of previously published data, assessing the functional implications of an alternative architecture and effects for future studies.

Structure of the Native Chemotaxis Core Signalling Unit from E-gene lysedE. colicells

“…In addition to interactions between receptors within the same TOD, we also observed interactions between the two TODs (Figure 4C). Given the symmetry of the extended array architecture, especially in the periplasmic space (31), such interactions presumably also occur between receptors from neighboring CSUs. Notably, however, the observed interactions were transient (<10 ns) and did not form a single, long-range pattern across the CSU on the timescale of our simulation.…”

Studying bacterial chemosensory array with CryoEM

“…V. cholera (PM7, pPM045, PM15, PM16, PM049, PM22) EMD-3398 [49] H. pylori (7.13) EMD-8460 [23] In vitro construction of Tar in nanodisc [15] S. enterica minicell (TH16943) [50] Lysed V. cholerae (C6706) [22] A. brasilense [24] Lysed E. coli (RP437) EMD-4991, EMD-4992, EMD-4993 [33] In vitro reconstitution of chemotaxis 2D array EMD-10050, PDB-6S1K [18] E. coli minicells (WM4196) EMD-10160 [28] T. denticola (Td) ATCC 35405 EMD-11381, EMD-11384, EMD-11385, EMD-11386 [25] V. cholerae (C6706, PM6, PM7), P. aeruginosa (PAO1), S. oneidensis (MR-1), and M. alcaliphilum (20Z R ) [51] E. coli minicell (WM4196) [37] which has a kinase-ON conformation; and EEEE, which has a kinase-OFF conformation [17]. Using cryoET and subtomogram averaging (STA), where repeating subvolumes, such as the CSUs within an array, are aligned and averaged, a 3D density map of the TarCF-QEQE CSU was reconstructed at 11.3 Å resolution [17], a significant improvement from previous attempts [19][20][21].…”

“…Native E. coli cells are usually too thick to extract detailed structural information. One [15,18,33,28,37] with permissions.…”

“…Both the original and the alternative lattices are P2 lattices, with a unit cell containing three CSUs (original, left) or a single CSU (alternative, right). Panels are reproduced from [ 15 , 18 , 33 , 28 , 37 ] with permissions.…”

“…This assembling was also reported in in vitro construction of 2D monolayer arrays and lysed E. coli cells [ 17 ]. Recently, an alternative P2 packing with a unit cell containing a single CSU has been reported from native Treponema denticola [ 25 ] and an E. coli minicell [ 37 ] (see Figure 3E , right). Although the average map of CSU from the alternative packing was not reported, its basic structure appears to be conserved.…”