A distance-weighted interaction map reveals a previously uncharacterized layer of the Bacillus subtilis spore coat

Abstract

Bacillus subtilis spores are encased in a protein assembly called the spore coat that is made up of at least 70 different proteins. Conventional electron microscopy shows the coat to be organized into two distinct layers. Because the coat is about as wide as the theoretical limit of light microscopy, quantitatively measuring the localization of individual coat proteins within the coat is challenging. We used fusions of coat proteins to green fluorescent protein to map genetic dependencies for coat assembly and to define three independent subnetworks of coat proteins. To complement the genetic data, we measured coat protein localization at subpixel resolution and integrated these two data sets to produce a distance-weighted genetic interaction map. Using these data, we predict that the coat comprises at least four spatially distinct layers, including a previously uncharacterized glycoprotein outermost layer that we name the spore crust. We found that crust assembly depends on proteins we predicted to localize to the crust. The crust may be conserved in all Bacillus spores and may play critical functions in the environment.

Figure 1. The Spore Coat Genetic Interaction Network

Examples of fluorescence microscopy images used in construction of the network. Cells were sporulated at 37°C by suspension in Sterlini-Mandelstam medium, stained with membrane stain (FM4-64, Invitrogen, red) and imaged 3 hours after resuspension. GFP-fusion fluorescence is shown in green. (A) Cells contain yaaH fused to gfp (PE793). (B) Cells contain yaaH fused to gfp, safA is deleted (PE861); YaaH-GFP shows complete mislocalization in safA mutant cells. (C) Cells contain oxdD fused to gfp (PE634). (D) Cells contain oxdD fused to gfp, safA is deleted (PE816); OxdD-GFP shows incomplete localization in safA mutant cells, forming a single dot. (E) The spore coat interaction network was drawn in Cytoscape (www.cytoscape.org) with directional edges (arrows) denoting genetic dependencies among genes and fusion proteins. The safA-dependency data was collected in this study. The remainder of the data is shown in Figure S1. Other genetic interactions were curated from the literature (Table S1).

Figure 2. A Distance-Weighted Genetic Interaction Network

(A) Cells were sporulated by suspension in Sterlini-Mandelstam medium in the presence of Mitotracker Red (Invitrogen). Samples were collected at hour 6, pelleted and resuspended in PBS containing 1 μg mL⁻¹ Mitotracker Red and imaged by phase contrast and fluorescence microscopy. Five section Z-series spaced 0.3μm apart were collected. Images were taken using brightfield, phase contrast, imaging, and epifluorescence with TexasRed and FITC fluorescence filters at each z-step. Stacks were deconvolved for 30 iterations using Autoquant. The middle plane of each stack was analyzed with PSICIC to determine cell contours and to measure the location of the membranes and the fluorescent protein fusions within each cell. All strains are C-terminal GFP-fusions, except SpoIVA (GFP-SpoIVA) which is fused at the N-terminus. CotA (CotA-YFP) is a C-terminal YFP-fusion. All fusions are integrated at the endogenous locus and are expressed from the native promoter, except gfp-spoIVA and cotG-gfp, which are integrated at the amyE locus and expressed from cloned spoIVA and cotG promoters respectively. All fusions are expressed in otherwise wild type cells. Each box is delimited by the first and third quartiles and the red line crossing the box is the median. The whiskers correspond to ± 2.7 standard deviations. Crosses indicate outliers. Data and statistics are included in Table S2 (B) Network of genetic interactions with distance-weighted edges. Edges in the network diagram represent genetic interactions found in fluorescence microscopy-based molecular epistasis experiments. Text boxes contain the names of coat proteins and are centered on the mean of the measurements from the MP to the coat protein-GFP fusion at hour 6 of sporulation (Table S2) on the x-axis. The y-axis at zero nanometers represents the MP. Position along the y-axis is arbitrary. The colored bars above the diagram represent the total spread of data plus and minus one standard deviation from the means of all the members of the corresponding coat layer. The bars represent, in cyan: the safA-independent cotE-independent innermost coat layer (IVA, VM, VID, YhaX), in gold: the safA-dependent inner coat (CotD, CotT, YaaH, YuzC), in blue: the cotE-dependent outer coat (CotA, CotS, CotO, CotM, YtxO), and in brown: the cotE-dependent crust (CotG, CotW, CotZ). The diagram was drawn using Inkscape (www.inkscape.org) with an original scale of 1mm = 1nm.

Figure 3. Characterization of the Outermost Coat Layer

(A) Cells were sporulated at 37°C by suspension in Sterlini-Mandelstam medium, stained with membrane stain (FM4-64, red) and analyzed by fluorescence microscopy at hour 5. One representative field of cells is shown for each condition. The image combines signal from the TexasRed and FITC channels. All strains express cotW-gfp from the endogenous locus. From left to right: CotW-GFP in an otherwise wild type strain (PE599), CotW-GFP in a cotG::erm strain (PE2220) and CotW-GFP in a cotX cotYZ::neo strain (PE2224). (B and D) Wild type and (C and E) cotXYZ mutant spores were fixed and stained with Ruthenium red and analyzed by thin-section transmission electron microscopy. The core, cortex, inner coat, outer coat and crust are labeled (Co, Cx, IC, OC and Cr, respectively). D and E show crust layer material not associated with a spore. The size bars are 500 nm (in B and C) or 100 nm (in D and E).