A coating of lipoproteins provides a stabilizing environment on the inner membrane of Bacillus subtilis spores

ABSTRACT

A new study by M. J. Flores, K. Duricy, S. Choudhary, M. Laue, and D. L. Popham (J Bacteriol 205:e00142-23, 2023, https://doi.org/10.1128/jb.00142-23) demonstrates a role for the YlaJ/YhcN family of lipoproteins in the immobilization of the spore’s inner membrane. In the absence of these lipoproteins, membrane fluidity increases and membrane-associated proteins like the GerA receptor complexes are more exposed to inimical conditions. The role of these proteins in stabilizing the Bacillus spore inner membrane is now being explored.

COMMENTARY

Even the most modest student of microbiology would instantly recognize an electron micrograph of a bacterial spore. With their striated textured coat peripheral to the electron-transparent cortex and the densely stained protoplast or core, spores are striking and unmistakable. The image in Fig. 5 of Flores et al. (1) is an excellent example. The core itself may appear pockmarked with ribosomes and well-ordered patches of nucleoid; the latter is a consequence of chromosomal encrustment by small acid soluble proteins (SASPs) and an environment that is highly mineralized with dipicolinate-chelated calcium ions (CaDPA). Closer inspection often reveals a membrane layer that surrounds the core, enveloping and buttressing the contents from the extracellular environment. This inner membrane is internal to the germ cell wall and the thick (and crucial) layer of peptidoglycan—or cortex—which serve as the filling between the major inner- and outermost morphological features. The coat represents an exquisite example of macromolecular assembly and attracts much interest from researchers interested in the molecular mechanisms that underpin such processes. Importantly, practitioners who perhaps more prosaically are of the opinion that the only good spore is a dead spore are concerned with its contribution to spore resistance. Evidently, as the spore’s point of contact with the environment, the coat often has to be breached if the desired cidal effect is to be achieved. This is not trivial since the coat presents a durable molecular mesh designed to rebuff the ingress of many potentially deleterious agents, including enzymes and antibodies. However, more seasoned spore observers might argue, with some justification, that the inner membrane is the true business end of the spore.

What do we mean by “inner membrane,” and why is it of such interest to spore biologists? As a reminder, the process of sporulation is triggered in response to nutrient starvation by some members of the Bacillota—generally Bacillales and Clostridiales species—resulting ultimately in the formation of a dormant, refractile, and highly resistant spore (2). In early sporulation, a division septum forms toward one pole of the cell (3). Rather than progressing to fission and the creation of two dissimilarly sized daughter cells, the peptidoglycan in this septum is thinned and the membrane of the larger, mother, cell progressively engulfs, in a ratchet-like phagocytic manner, the smaller compartment (4). The resultant forespore is a double membrane-bound structure, essentially adrift in the mother cell cytoplasm. In a race against time against a background of diminishing nutrients, sporulation proceeds over the next few hours with both mother cell and forespore contributing to the development of the spore. Of the two membranes associated with the forespore, the outer provides the synthetic machinery for the cortex peptidoglycan and serves as the matrix upon which the various proteins that form the coat are steadily deposited. The outer membrane is not visible in electron micrographs of the mature spore and is thought not to present a permeability barrier, although this has not been established definitively. The inner membrane, on the other hand, is of much more significance in the final spore.

For instance, if the raison d être of the spore is to preserve its DNA and metabolic capacity until somehow it finds itself in a growth supportive environment—conceivably many decades in the future—then it needs to exclude potentially damaging exogenous chemicals from entering the core. The molecular sieving and detoxifying properties of the coat extend only so far in this regard, and it is actually the inner membrane that presents the major permeability barrier against reactive chemicals that can damage core DNA, including acids and bases. Indeed, transit to and from the core of even small uncharged molecules, including water, is severely impeded but not entirely excluded. Notably, a list of the principal area(s) of damage associated with regimens that kill spores reveals that, more often than not, the inner membrane is implicated (5).

What factors might account for the prominence of the inner membrane in terms of regulating spore permeability? Atypical morphological features come to the fore again. In contrast to a conventional “fluid” membrane, the lipids that constitute the spore inner membrane are largely immobile, meaning they do not permit significant lateral diffusion of probes or proteins within the membrane. Nor do they support a proton motive force, as is commensurate with a metabolically dormant cell type. As such, the inner membrane has attracted various rheological descriptors, including semi-crystalline and gel-like, with the latter now probably more precise. Regardless, reduced lipid mobility evidently underpins low diffusivity coefficients associated with the inner membrane but at the same time raises questions concerning how this status is both achieved and maintained during dormancy. While one might expect an unusual preponderance of saturated fatty acids to facilitate close molecular packing, it seems this is not the case; the phospholipid and fatty acid content of the inner membrane is not dissimilar to the composition of the growing cell’s plasma membrane. Indeed, attempts at modulating the lipid composition of the membrane by targeted genetic manipulation resulted in only modest differences in the properties of the resultant spores, with cardiolipin—no stranger to unusual environments—depleted spores showing minor reductions in resistance to wet heat and oxidative damage and slower germination (6). Other factors that may have a role in immobilizing inner membrane lipids include charge interactions between phospholipid headgroups on the inner leaflet and calcium ions that are otherwise chelated in the crystalline CaDPA lattice that comprises the major core analyte (7). Similarly, the coat has been implicated in exerting a degree of compression on the membrane since fluidity is partially restored in chemically de-coated spores.

The significance of the inner membrane is not restricted to its lipid content since it is also host to a diverse and functionally crucial proteome. A number of proteins and/or protein complexes important for germination, and therefore spore viability, have been localized to the inner membrane. The germinant receptor complexes, for example, recently identified as ligand-gated ion channels (8), generally form a discrete cluster within the inner membrane of spores and are stimulated to open upon binding of certain amino acids or related germinant molecules. The subsequent release of cations from the core is the first detectable event of germination. In contrast, the SpoVA proteins, which interact to form channels that act as conduits for the movement of CaDPA both in (during sporulation) and out (during germination) of spores, are more abundant and dispersed throughout the inner membrane. Other membrane constituents with roles in germination include the major cortical peptidoglycan lysin, SleB, and receptor-associated GerD, which somehow facilitates receptor clustering. Key proteins associated with metabolism have also been identified from proteomic analyses as constituents of the inner membrane (9). This is in keeping with the ultimate destination of this structure as the plasma membrane in the post-germinative vegetative cell, which has to be primed to meet the energy requirements of early outgrowth in the absence of de novo protein synthesis. The subject of this commentary (1) reveals that a group of related lipoproteins present in large amounts in the inner membrane influence membrane properties and germination behavior. Arguably, the fate of all of these proteins during spore dormancy may be influenced to some extent by the presence of these lipoproteins.

The YlaJ/YhcN family of conserved lipoprotein components are expressed in the forespore and would be lipid-anchored at the outer surface of the inner membrane. Bagyan et al. (10) recognized that the absence of a forespore-expressed lipoprotein, YhcN, reduced the rate of spore outgrowth in Bacillus subtilis. In Bacillus cereus, the ylaJ gene, which encodes a paralog of YhcN, is located in an operon with sleB and ypeB, raising the possibility of some relationship with this cortex lytic enzyme and its associated anchor protein. The genomes of all species of Bacilli and Clostridia that use SleB as a cortex lytic enzyme also encode a YlaJ protein, emphasizing this association. A null mutation in yhcN or in ylaJ reduced the rate of germination in a spore population, in a non-additive manner (11). It seemed that both early and later stages of germination were affected, in some manner that was not at all understood. Two more lipoproteins in B. subtilis, YutC and YrbB (CoxA), also contain a YhcN/YlaJ domain and are also present in the inner membrane fraction of dormant spores (12).

Flores et al. (1) have now made a careful and detailed study of mutants lacking any or all of these four homologs in B. subtilis and have identified phenotypes associated with individual or multiple mutants. They demonstrate that these lipoproteins contribute to the restriction of lipid mobility in the spore’s inner membrane and potentially may stabilize or otherwise protect other proteins embedded in the inner membrane, such as those controlling germination. Firstly, mutations in any one, or all, of the four lipoprotein genes reduced the proportion of spores that germinate in valine, via the GerA receptor, to the same extent, implying that they shared the same function. In the single mutants, adding a second copy of any one at another locus restored the germination behavior to the wild type.

In addition, the GerA germination apparatus appeared slightly more heat sensitive, as reducing the temperature of pre-activation of the spores from 75°C to 55°C increased the proportion of germinating spores by an additional 10%, though not restoring it to the wild type. This led to a more rigorous test of the heat sensitivity of dormant spores, using conditions (85°C for 90 min) that kills 85% of wild-type spores and measuring the proportion that could germinate, outgrow, and form colonies on a rich medium. Losing one or even all four of the lipoproteins had little effect. Introducing a defect in the CwlJ cortex lytic enzyme uncovered a significant phenotype, however. Losing one had little effect on the percent survival/recovery. Losing two had a little more, and losing three or four reduced the efficiency of colony formation to less than 1%. It might be assumed that this was demonstrating an effect on SleB activity, as that alternative cortex lytic enzyme is essential for cortex hydrolysis but providing an exogenous enzyme to lyse the cortex did not restore viability, so the spores had been inactivated rather than merely trapped in an indigestible cortex as a result of loss of SleB function; there must have been other damage.

The YhcN protein was known to be one of the most abundant in spore membranes (9, 12) and appears to localize uniformly around the forespore (13). The levels of all four paralogs, FLAG-tagged, were compared in the Flores et al. study by immunoblotting (1). Protein levels were generally similar, with the lowest, CoxA, being approximately one-third lower in abundance than YhcN. Importantly, the loss of any one did not lead to an increased expression of any other. There was no visible defect in the spores of any mutants as judged by electron microscopy of thin sections. Finally, one important change in spore properties was noted. Dormant spores were prepared in the presence of Laurdan, a fluorescent dye that is incorporated into the membranes of sporulating cells, and then treated to remove the coat and outer membrane. An emission shift (generalized polarization) reports on the fluidity of the membrane. The membrane fluidity was in all cases much less than that in vegetative cells, but it was progressively increased on the deletion of one, two, or all four lipoprotein genes.

To further emphasize the importance of accessory proteins in spore inner membrane stability and function, another family of proteins have been described recently that affect spore heat resistance (14). Exceptionally heat resistant B. subtilis spores are produced by strains that have acquired the transposon Tn1546, which confers increased heat resistance and slower germination on the spore (15, 16). This encodes, among others, a SpoVA^mob operon, a YhcN homolog, and another protein called 2Duf, as it contains two domains of unknown function. Introduction of the transposon Tn1546 into B. subtilis increases dramatically the level of heat and peroxide resistance, whereas an otherwise isogenic transposon lacking 2Duf confers less, but still significant, protection. 2Duf decreases inner membrane permeability and lowers lipid probe mobility (17). Though 2Duf is absent from most B. subtilis strains, partial homologs carrying one of the domains are encoded by yetF and ydfS (14). Both proteins are present in the inner membrane proteome (12) but are less abundant than the YlaJ/YhcN family. Each would possess an integral membrane three helix domain and a potentially surface-exposed domain and probably exists as a tetramer; loss of YetF, the more highly expressed, considerably reduces spore heat resistance and the efficiency of germination via both GerA and GerB/K receptors (14).

It is clearly time for a renewed focus on the composition and properties of the spore inner membrane, especially as two families of protein paralogs that help to stabilize the membrane and its embedded or associated proteins have been recently identified. How such proteins can modify membrane fluidity, how this might stabilize embedded proteins, and whether such protein stabilization is selective, all remain as questions for further investigation.