NlpD links cell wall remodeling and outer membrane invagination during cytokinesis in Escherichia coli

Abstract

Cytokinesis in gram-negative bacteria requires the constriction of all three cell envelope layers: the inner membrane (IM), the peptidoglycan (PG) cell wall and the outer membrane (OM). In order to avoid potentially lethal breaches in cell integrity, this dramatic reshaping of the cell surface requires tight coordination of the different envelope remodeling activities of the cytokinetic ring. However, the mechanisms responsible for this coordination remain poorly defined. One of the few characterized regulatory points in the envelope remodeling process is the activation of cell wall hydrolytic enzymes called amidases. These enzymes split cell wall material shared by developing daughter cells to facilitate their eventual separation. In Escherichia coli, amidase activity requires stimulation by one of two partially redundant activators: EnvC, which is associated with the IM, and NlpD, a lipoprotein anchored in the OM. Here, we investigate the regulation of amidase activation by NlpD. Structure-function analysis revealed that the OM localization of NlpD is critical for regulating its amidase activation activity. To identify additional factors involved in the NlpD cell separation pathway, we also developed a genetic screen using a flow cytometry-based enrichment procedure. This strategy allowed us to isolate mutants that form long chains of unseparated cells specifically when the redundant EnvC pathway is inactivated. The screen implicated the Tol-Pal system and YraP in NlpD activation. The Tol-Pal system is thought to promote OM invagination at the division site. YraP is a conserved protein of unknown function that we have identified as a new OM-localized component of the cytokinetic ring. Overall, our results support a model in which OM and PG remodeling events at the division site are coordinated in part through the coupling of NlpD activation with OM invagination.

Fig 1. Structure-function analysis of NlpD.

The domain organization of NlpD is illustrated. Indicated are the signal sequence (SS; yellow), lysin motif (LysM; blue), and the degenerate LytM domain (dLytM; green). Also shown are the NlpD truncations that were expressed under the control of the IPTG-inducible lactose promoter either as an untagged protein or as a C-terminal mCherry fusion. Truncations lacking ^SSNlpD are expressed as soluble periplasmic proteins fused to the DsbA signal peptide. Columns indicate (i) the NlpD residues present in each truncation, (ii) whether the fusion to mCherry accumulated at division sites strongly (+++), poorly (+), or appeared evenly distributed along the periphery of the cell (-), and (iii) whether the untagged truncation could (+) or could not (-) compensate for the loss of endogenous NlpD for proper cell separation. ND, not determined.

Fig 2. Localization of NlpD fusions in ^ΔSSnlpD cells.

Overnight cultures of MT47 (^ΔSSnlpD) harboring the integrated expression constructs (A) attHKNP20 (P_lac::nlpD^WT-mCherry), (B) attHKMT101 (P_lac::nlpD^(1–189)-mCherry), (C) attHKMT103 (P_lac::nlpD^(1–115)-mCherry), (D) attHKMT178 (P_lac::^ssdsbA-nlpD^(102–175)-mCherry), (E) attHKMT180 (Pl_ac::^ssdsbA-nlpD^(250–379)-mCherry), (F) attHKMT182 (P_lac::^ssdsbA-nlpD^(189–379)-mCherry), or (G) attHKMT149 (P_lac::nlpD^(1–30)-mCherry) were diluted in minimal M9-maltose medium supplemented with 25μM (E-F), 50μM (A-B), 100μM (C-D), or 150μM (G) IPTG. Cells were grown at 30°C to an OD₆₀₀ of 0.15–0.25 before they were visualized on 2% agarose pads by phase contrast and fluorescence microscopy. Arrows indicate localization of the protein fusion to division sites. Bar = 4μm.

Fig 3. Only full-length NlpD properly promotes cell separation.

(A-C) MT50 (^ΔSSnlpD ΔenvC) or TB140 (ΔenvC) cells were grown in minimal M9-maltose medium at 37°C to an OD₆₀₀ of 0.2–0.3 before visualization on 2% agarose pads with DIC optics (A-B) or analyzed by flow cytometry (C). Bar = 10μm. (D-H) Overnight cultures of MT50 harboring the integrated expression constructs (D) attHKMT20 (P_lac::nlpD^WT), (E) attHKMT102 (P_lac::nlpD^(1–189)), (F) attHKMT104 (P_lac::nlpD^(1–115)), (G) attHKMT179 (Pl_ac::^ssdsbA-nlpD^(250–379)), or (H) attHKMT181 (P_lac::^ssdsbA-nlpD^(189–379)) were diluted in minimal M9-maltose medium and grown at 37°C. Mid-log cultures were then backdiluted into M9-maltose medium with or without the inducer IPTG. Cells were further grown at 37°C to an OD₆₀₀ of 0.2–0.3 before flow cytometry analysis. Histograms of cultures that were either uninduced (grey) or induced with 150μM (D, G-H, blue) or 1mM (E-F, red) IPTG were overlayed. (I) Cells of MT122 (^ΔSSnlpD) and MT123 (^ΔSSnlpD ΔamiC) alone or harboring the integrated constructs attHKMT20 (P_lac::nlpD^WT), attHKMT179 (Pl_ac::^ssdsbA-nlpD^(250–379)), or attHKMT181 (P_lac::^ssdsbA-nlpD^(189–379)) were grown in LB at 30°C. Following normalization for cell density (OD₆₀₀ = 0.5), 5 μl of the resulting cultures was spotted on LB agar containing 150μM IPTG and 20 μg/ml CPRG. The plates were incubated at 30°C and photographed after 14 hours.

Fig 4. OM localization of NlpD is required for proper cell separation.

(A) The domain structure of NlpD is illustrated as in Fig 1. Details of the signal sequence are presented with the lipobox in red and the arrow indicating the cleavage site just before the acylated cysteine. The IM-retained variant (NlpD^(S27D)) contains a mutated signal sequence (indicated by the asterisk) with an aspartate at the +2 position after the acylated cysteine (underlined). The soluble periplasmic variant (NlpD^(27–379)) is fused to the DsbA signal peptide (purple) that is cleaved upon export to the periplasm via the Sec system. (B-D) Cytological assay to determine the subcellular localization of the NlpD variants. Overnight cultures of MT47 (^ΔSSnlpD) expressing different NlpD-mCherry fusions from the integrated constructs (B) attHKNP20 (P_lac::nlpD^WT-mCherry), (C) attHKMT21 (P_lac::nlpD ^(S27D)-mCherry), or (D) attHKMT147 (P_lac::^ssdsbA-nlpD ^(27–379)-mCherry) were diluted in minimal M9-maltose medium supplemented with 25 μM (D) or 150 μM (B, C) IPTG. Cells were grown at 30°C to an OD₆₀₀ of 0.4, washed and then osmotically shocked by resuspension in plasmolysis buffer and the plasmolyzed cells were visualized by phase contrast and fluorescence microscopy. Arrows indicate signals that display a smooth OM peripheral signal in (B), track with the inner membrane in (C), or fill the increased periplasmic spaces of plasmolysis bays in (D). Bar = 4μm. (E-G) Overnight cultures of MT50 (^ΔSSnlpD ΔenvC) harboring the integrated expression constructs (E) attHKMT20 (P_lac::nlpD^WT), (F) attHKMT12 (P_lac::nlpD ^(S27D)), or (G) attHKMT121 (P_lac::^ssdsbA-nlpD ^(27–379)) were diluted in minimal M9-maltose medium and grown at 37°C until mid-log. Cultures were then backdiluted into M9-maltose medium without (grey histogram) or with 150 μM (blue histogram) or 1 mM (red histogram) IPTG and grown at 37°C to an OD₆₀₀ of 0.2–0.3 before flow cytometry analysis. (H) Cells of MT122 (^ΔSSnlpD) and MT123 (^ΔSSnlpD ΔamiC) alone or harboring the integrated constructs described above were grown and spotted on CPRG agar as described in Fig 3I. The subcellular localization of each NlpD variant is indicated in the square brackets: OM, outer membrane; IM, inner membrane; peri, periplasm.

Fig 5. Flow cytometry-based enrichment and screen to identify potential NlpD regulators.

(A-C) Dot plots of (A) a control non-chaining strain, TB140 (ΔenvC), (B) a control chaining strain, MT50 (^ΔSSnlpD ΔenvC), and (C) the transposon (EZ-Tn5 <Kan-2>) mutagenized TB140 library, with side scatter pulse height (SSC-H; y-axis) versus side scatter pulse width (SSC-W; x-axis). The gate used for sorting is shown on the plot, with the number of events present in that gate indicated as a percentage of the total population. (D-J) Overnight cultures of (D) TB28 (wt), TB140 (ΔenvC), (E) MT47 (^ΔSSnlpD), MT50 (ΔenvC ^ΔSSnlpD), (F) MT140 (ΔyraP), MT135 (ΔenvC ΔyraP), (G) MT51 (ΔtolQ-pal), MT55 (ΔenvC ΔtolQ-pal), or (H) MT141 (^ΔSSnlpD ΔyraP), MT53 (^ΔSSnlpD ΔtolQ-pal) were diluted in LB medium. Cells were grown at 30°C to an OD₆₀₀ of 0.3–0.5 before they were visualized on 2% agarose pads with DIC optics (D-H) or analyzed by flow cytometry (I-J). Bar = 10μm. (K-L) Overnight cultures of MT50 (ΔenvC ^ΔSSnlpD) and (K) MT135 (ΔenvC ΔyraP) and MT179 (ΔenvC ΔyraP ^ΔSSnlpD) or (L) MT55 (ΔenvC ΔtolQ-pal) and MT96 (ΔenvC ΔtolQ-pal ^ΔSSnlpD) were diluted in LB medium. Cells were grown at 30°C to an OD₆₀₀ of 0.15–0.25 before analysis by flow cytometry.

Fig 6. NlpD and AmiC localize to division sites independently of YraP and/or the Tol-Pal system.

(A-D) Overnight cultures of (A) MT47 (^ΔSSnlpD), (B) MT141 (^ΔSSnlpD ΔyraP), (C) MT53 (^ΔSSnlpD ΔtolQ-pal), or (D) MT147 (^ΔSSnlpD ΔtolQ-pal ΔyraP) harboring the integrated expression construct attHKNP20 (P_lac::nlpD-mCherry) were diluted in minimal M9-maltose medium supplemented with 100μM IPTG. (E-H) Overnight cultures of (E) TB143 (ΔamiC), (F) MT150 (ΔamiC ΔyraP), (G) MT149 (ΔamiC ΔtolQ-pal), or (H) MT178 (ΔamiC ΔtolQ-pal ΔyraP) harboring the integrated expression construct attHKNP16 (P_lac-m3::amiC-gfp) were diluted in minimal M9-maltose medium supplemented with 25μM IPTG. For all strains, cells were grown at 30°C to an OD₆₀₀ of 0.2–0.3 before they were visualized on 2% agarose pads by phase contrast and fluorescence microscopy. In all images, arrows indicate the localization of the protein fusion to division sites. Bar = 4μm.

Fig 7. OM localization of YraP is required for cell separation in the absence of EnvC.

(A) The domain structure of YraP is illustrated. Indicated are the signal sequence (SS; red) and the two bacterial OsmY and nodulation domains (BON1/2; orange). Details of the signal sequence are presented with the lipobox in red and the arrow indicating the cleavage site just before the acylated cysteine. The IM-retained variant (YraP^IM) contains a mutated signal sequence (indicated by the asterisk) with an aspartate and glutamate at the +2 and +3 positions after the acylated cysteine (underlined). The soluble periplasmic variant (YraP^peri) is fused to the DsbA signal peptide (purple) that is cleaved upon export to the periplasm via the Sec system. (B-I) Overnight cultures of (B) TB140 (ΔenvC) or MT135 (ΔenvC ΔyraP) either (C) alone or harboring the integrated construct (D) attλMT196 (P_lac::yraP^WT), (E) attλMT198 (P_lac::yraP^IM), (F) attλMT209 (P_lac::^ssdsbA-yraP^peri), (G) attλMT197 (P_lac::yraP^WT-mCherry), (H) attλMT199 (P_lac::yraP^IM-mCherry), or (I) attλMT210 (P_lac::^ssdsbA-yraP^peri-mCherry) were diluted in minimal M9-maltose medium only (B-C) or supplemented with 10μM (D, F-G, I) or 20μM (E, H) IPTG. Cells were grown at 30°C to an OD₆₀₀ of 0.2–0.25 before they were visualized on 2% agarose pads with DIC optics. Bar = 10μm.

Fig 8. Recruitment of YraP to division sites is dependent on its OM localization but independent of NlpD, the Tol-Pal complex, and AmiC.

(A-C) Overnight cultures of MT140 (ΔyraP) harboring the integrated construct (A) attλMT197 (P_lac::yraP^WT-mCherry), (B) attλMT199 (P_lac::yraP^IM-mCherry), or (C) attλMT210 (P_lac::^ssdsbA-yraP^peri-mCherry) were diluted in minimal M9-maltose medium supplemented with 10μM (A-C) or 20μM (B) IPTG. (D-G) Overnight cultures of (D) MT140 (ΔyraP), (E) MT141 (ΔyraP ^ΔSSnlpD), (F) MT147 (ΔyraP ^ΔSSnlpD ΔtolQ-pal), and (G) AAY22 (ΔyraP ΔamiC) harboring the integrated construct attλMT197 (P_lac::yraP^WT-mCherry) were diluted in minimal M9-maltose medium supplemented with 10μM IPTG. All cultures were grown at 30°C to an OD₆₀₀ of 0.15–0.2 before cells were visualized on 2% agarose pads by phase contrast and fluorescence microscopy. Arrows indicate localization of the protein fusion to division sites. Bar = 4μm.

Fig 9. Localization of YraP or NlpD in cephalexin-treated cells.

Overnight cultures of (A-B) MT140 (ΔyraP) harboring the integrated construct attλMT197 (P_lac::yraP-mCherry) or (C-D) MT47 (^ΔSSnlpD) harboring the integrated expression construct attHKNP20 (P_lac::nlpD-mCherry) were diluted in minimal M9-maltose medium supplemented with either 25μM (A-B) or 100μM (C-D) IPTG and grown at 30°C until mid-log. Cultures were then backdiluted into M9-maltose medium with the indicated IPTG concentration with or without 10μg/ml cephalexin as indicated. Cells were grown at 30°C to an OD₆₀₀ of 0.2 before they were visualized on 2% agarose pads by phase contrast and fluorescence microscopy. Arrows indicate localization of the protein fusion to division sites. Bar = 10μm.

Fig 10. NlpD inactivation results in OM integrity defects.

Cells of TB28 (WT), TB143 (ΔamiC), MT47 (^ΔSSnlpD), MT176 (^ΔSSnlpD ΔamiC), MT140 (ΔyraP), and MT51 (ΔtolQ-pal) were grown overnight in LB at 30°C. Following normalization for cell density (OD₆₀₀ = 1), the resulting cultures were serially diluted (10⁻¹ to 10⁻⁶), and 5 μl of each dilution was spotted on LB agar only or supplemented with the indicated concentrations of SDS and EDTA. The plates were incubated at 37°C and photographed after ~24 hours.

Fig 11. Model for the regulation of NlpD during cell division.

Shown is a schematic depicting two cycles of a potential coupling mechanism coordinating septal PG splitting by NlpD/AmiC with OM invagination. Prior to the initiation of constriction, we envision that the dLytM domain of OM-anchored NlpD is prevented from accessing the PG layer and/or AmiC, potentially by physical distance or protein conformation constraints. When the divisome is activated, OM constriction promoted by the Tol-Pal system may bring the dLytM domain of NlpD into proximity of the PG layer and AmiC where it can stimulate septal PG splitting. AmiC activation also appears to require YraP, which may activate NlpD directly or indirectly through its yet to be determined role in maintaining OM integrity. PG processing by AmiC is expected to increase the distance between the OM and PG layer, thus returning NlpD to its inactive configuration and requiring another round of OM constriction to trigger further PG processing and so on until cell division is complete and the daughter cells are separated. Shown are hypothetical protein conformations and interactions within the Tol-Pal system based on current literature [–71].