Cell division machinery drives cell-specific gene activation during differentiation in Bacillus subtilis

Significance The differential transcription of genes in daughter cells underlies cellular differentiation. During bacterial spore formation, a progenitor cell asymmetrically divides and differentiates into two different cell types. We show that differential transcription in the smaller daughter cell is achieved by a biased localization of a transcriptional activator on one face of the asymmetric division septum and that the correct placement of this factor is dependent on the core cell division machinery. Thus, an intrinsic asymmetry provided by a division septum can provide the basis for the differential transcription of genes during development.

When faced with starvation, the bacterium Bacillus subtilis transforms itself into a dormant cell type called a "spore".Sporulation initiates with an asymmetric division event, which requires the relocation of the core divisome components FtsA and FtsZ, after which the sigma factor σ F is exclusively activated in the smaller daughter cell.Compartment-specific activation of σ F requires the SpoIIE phosphatase, which displays a biased localization on one side of the asymmetric division septum and associates with the structural protein DivIVA, but the mechanism by which this preferential localization is achieved is unclear.Here, we isolated a variant of DivIVA that indiscriminately activates σ F in both daughter cells due to promiscuous localization of SpoIIE, which was corrected by overproduction of FtsA and FtsZ.We propose that the core components of the redeployed cell division machinery drive the asymmetric localization of DivIVA and SpoIIE to trigger the initiation of the sporulation program.

C. difficile | E. coli | RacA | MinCD
During cellular differentiation, a progenitor cell changes to assume the characteristics required for a specific function.For instance, in eukaryotes, stem cells are known for their ability to differentiate into multiple cell types (1)(2)(3).Two hallmarks of this process are differential gene expression and asymmetric division, which result in two unequal daughter cells that behave dissimilarly.A relatively simple model system to study these processes is bacterial endospore formation ("sporulation"), wherein a progenitor cell differentiates into two cell types that each display a different cell fate.Bacillus subtilis is a gram-positive rod-shaped bacterium that normally divides by binary fission when grown in rich medium, resulting in two identical daughter cells that exhibit similar cell fates (4-6).However, when the bacterium senses the onset of starvation, it initiates the sporulation program, which results in the transformation of the cell into an ellipsoidal, dormant cell type called a "spore" that is resistant to myriad environmental assaults (7)(8)(9).Sporulation commences with an asymmetric division event near one pole of the bacterium that produces two dissimilarly sized progeny: a larger mother cell and a smaller forespore, which initially lie side by side, separated by the "polar septum" (Fig. 1A).The asymmetric positioning of the polar septum is achieved by redeployment of FtsZ, a bacterial tubulin homolog that is the core component of the divisome (10,11).Redeployment requires the slight overexpression of the ftsZ gene, driven by a second sporulation-specific promoter upstream of the ftsAZ operon (12,13), which also encodes for FtsA, an actin homolog that tethers FtsZ to the membrane (14).Additionally, the sporulation protein SpoIIE reinforces polar redeployment of FtsZ (15)(16)(17)(18).Ultimately, the forespore will mature into the spore, and the mother cell will lyse after nourishing the forespore into dormancy.This transformation is driven by the sequential, compartment-specific activation of sigma factors in the forespore and mother cell (19).
The first sigma factor that sets off this cascade is σ F , which is activated specifically in the forespore shortly after polar septation (20).To prevent promiscuous or premature activation, σ F is held inactive by an anti-sigma factor (SpoIIAB) (21,22).Activation of σ F is achieved by dephosphorylating an anti-anti-sigma factor (SpoIIAA), which sequesters SpoIIAB and thereby liberates σ F (23,24).The phosphatase responsible for activating SpoIIAA is the FtsZ-redeploying protein SpoIIE, whose phosphatase activity has been studied in detail (25)(26)(27)(28)(29).A puzzling feature of this mechanism is that the three components that regulate σ F are transcribed and translated in the progenitor cell before asymmetric septation, so multiple mechanisms have been proposed to explain the specific activation of σ F specifically in the forespore.One model builds on the observation that SpoIIE displays a biased localization on the forespore face of the polar septum (30)(31)(32), thereby preferentially exerting its phosphatase activity on forespore-localized SpoIIAA, resulting in forespore-specific σ F activation.The preferential localization of SpoIIE in the forespore is dependent on a membrane-associated structural protein called DivIVA that also participates in positioning proteins during cell division and anchoring replicated chromosomes during sporulation (33).DivIVA directly interacts with SpoIIE at the polar septum and likewise also displays a biased localization on the forespore face of the polar septum (32).DivIVA preferentially accumulates on highly negatively curved membranes (34,35) and as such localizes to either side of vegetative septa (36).However, the reason underlying the biased localization of DivIVA on just one side of the polar septum during sporulation, which also harbors negative membrane curvature on either side of the septum, was unclear.A second model invokes the selective degradation of SpoIIE in the mother cell via the preferential localization of the membrane-bound protease FtsH.This model also invokes a requirement for DivIVA, but in a protective role that sequesters SpoIIE from degradation by FtsH (37).
A recent report revealed yet another level of asymmetry at the polar septum (38).Whereas DivIVA and SpoIIE localize to the forespore face of the polar septum, the core components of the cell division machinery itself (FtsA and FtsZ) reportedly display a biased localization on the mother cell face of the polar septum.This asymmetric distribution of the cell division machinery was proposed to cause the unusual thinness of the polar septum compared to vegetative septa (39,40).In this work, we investigated how the asymmetric distribution of four different proteins on either side of a division septum is established and whether the biased localization of DivIVA at the polar septum is principally responsible for the compartment-specific activation of σ F .We first identified a DivIVA variant that did not display forespore-biased localization at the polar septum.Cells harboring this variant promiscuously activated σ F in both the mother cell and forespore, displayed a sporulation defect, and formed a relatively thick polar septum that resembled vegetative septa.The mislocalization of this DivIVA variant was corrected by the overproduction of FtsA and FtsZ, which restored septum thickness, forespore-biased SpoIIE localization, and proper σ F activation.We propose that a unique feature of the polar septum, resulting from the overexpression of ftsA and ftsZ at the onset of sporulation, drives the biased localization of DivIVA to establish an intrinsic asymmetry that initiates the cascade of differential transcription that drives the sporulation program.

Results
A DivIVA Variant Is Specifically Impaired in σ F Activation.After polar septation, σ F is activated specifically in the forespore, which requires a phosphatase, SpoIIE (20,25,41,42).We previously reported that selective activation of σ F in the forespore is likely due to the asymmetric localization of SpoIIE phosphatase on the forespore face of the polar septum.SpoIIE directly interacts with DivIVA, which also asymmetrically localizes on the forespore face of the polar septum (32).Deletion of divIVA resulted in the promiscuous activation of σ F , suggesting a role for DivIVA in positioning SpoIIE and proper activation of σ F .However, this approach could have altered other concurrently occurring DivIVA functions.To genetically separate the contribution of DivIVA in σ F activation from its vegetative and chromosome anchoring roles, we mutagenized selected codons surrounding position 99 of DivIVA, which had previously been reported to specifically affect sporulation, but nonetheless had a chromosome anchoring defect (43), and screened, using fluorescence microscopy, for DivIVA variants that misactivated σ F using fluorescence microscopy.At 1.5 h after induction of sporulation, 82% of otherwise WT cells (n = 272) that harbored gfp under control of a σ Fdependent promoter produced GFP exclusively in the forespore (Fig. 1 B and B').However, substituting Arg at position 94 of DivIVA with Asn (DivIVA R94N ) resulted in production of GFP in both the mother cell and forespore in ~60% of cells (n = 227; Fig. 1 C and C').Immunoblotting cell extracts harvested from B. subtilis cells growing exponentially and during sporulation revealed that DivIVA R94N was present at similar levels as WT DivIVA (SI Appendix, Fig. S1A).This promiscuous activation of σ F resulted in a reduced sporulation efficiency: Whereas cells harboring a divIVA deletion that was complemented with WT divIVA sporulated at 0.94 relative to WT, cells expressing divIVA R94N sporulated at only 0.18 relative to WT (Fig. 1K).We next tested whether DivIVA R94N was impaired in its other known functions.Unlike cells harboring a deletion of divIVA, cells expressing divIVA R94N did not display an obvious exponential growth defect (Fig. 1D) and displayed a cell length that was similar to WT (Fig. 1E), suggesting that this allele was not impaired in the vegetative function of DivIVA.Consistent with this, DivIVA R94N -GFP primarily localized at midcell during vegetative growth, similar to the localization pattern of DivIVA-GFP (36) (Fig. 1 F-G").At the onset of sporulation, DivIVA-GFP localizes to both the hemispherical cell poles and the polar septum; similarly, DivIVA R94N -GFP (at least as measured by diffraction-limited fluorescence microscopy) localized to the cell poles and polar septum (Fig. 1 H-I"), suggesting that there is no gross localization defect of DivIVA R94N during vegetative growth or during sporulation.Degradation of DivIVA at the onset of sporulation prevents polar septation (32).To test whether cells expressing divIVA R94N displayed a defect in polar septation, we examined, using fluorescence microscopy, the morphology of cells stained with a fluorescent membrane dye 1.5 h after the induction of sporulation and quantified the fraction of cells that displayed polar or medial septa, no septa at all, or two polar septa at both poles.The results in Fig. 1J indicate that the population of cell morphologies present at this early stage of sporulation was similar between cells producing DivIVA or DivIVA R94N .Finally, we examined the chromosome anchoring efficiency in cells expressing divIVA R94N by visualizing chromosomes of sporulating cells using the fluorescent dye DAPI and observing the number of forespores that were devoid of any DNA, indicating a chromosome anchoring defect (SI Appendix, Fig. S1 B-F").51% of cells (n = 260) harboring a deletion in racA, which is known to cause a chromosome anchoring defect, produced chromosome-free forespores.In contrast, only 4% of WT cells and 19% of cells expressing divIVA R94N displayed a chromosome anchoring defect (n = 191 and 311, respectively; SI Appendix, Fig. S1  A-C).To test whether the σ F misactivation caused by DivIVA R94N could be due to the slightly elevated chromosome anchoring defect we observed in this strain, we measured the fidelity of σ F activation in cells harboring a racA deletion.Similar to WT, 83% of ΔracA cells correctly activated σ F exclusively in the forespore, indicating that a chromosome anchoring defect is not responsible for promiscuous σ F activation (SI Appendix, Fig. S1B"-F").Taken together, we conclude that the divIVA R94N allele is specifically impaired in σ F activation, which occurs after the previously described roles for DivIVA in vegetative growth, chromosome anchoring, and polar septation.

DivIVA
R94N Does Not Asymmetrically Position SpoIIE on the Forespore Side of the Polar Septum.Since DivIVA anchors SpoIIE to the forespore face of the polar septum, we wondered whether the σ F activation defect could be attributed to incorrect subcellular localization of either DivIVA or SpoIIE.To distinguish between the two faces of the polar septum, which is separated by ~80 nm, we employed dual-color 3D structured illumination microscopy (SIM) (44,45).Additionally, to examine early events immediately after polar septation, we employed a mutant strain (ΔspoIID/ΔspoIIM) that was arrested before engulfment and thus displayed a flat septum.Further, since SpoIIE is released into the forespore and subsequently recaptured at the polar septum, we used a strain that did not produce the recapturing protein SpoIIQ to ensure that we were not detecting released, then recaptured, SpoIIE-GFP.At the onset of polar septation, WT DivIVA-GFP exhibited a biased localization to the forespore side of the invaginating septum in 72% of cells (n = 67) (32) (Fig. 2 A-A"), which was evidenced by plotting the fluorescence intensity of the membrane stain relative to that of the DivIVA-GFP fluorescence to visualize the offset peak of the DivIVA-GFP intensity toward the forespore.This asymmetric positioning of DivIVA-GFP remained in cells that had completed construction of the polar septum (Fig. 2 B-B").Consistent with this localization pattern, SpoIIE-GFP also displayed a forespore-biased localization in 71% (n = 60) of cells during initial septal invagination (Fig. 2 E-E"), and in cells that had completed septation (Fig. 2 F-F").In contrast, DivIVA R94N -GFP did not display a forespore-biased localization, with 54% of cells displaying DivIVA R94N -GFP fluorescence that overlapped with the membrane fluorescence (Fig. 2 C-D").Concomitantly, in 66% of cells harboring divIVA R94N (n = 54), SpoIIE-GFP also did not asymmetrically localize to the forespore face of the polar septum (Fig. 2 G-H").The results suggest that the inability of DivIVA R94N  to display a biased localization on the forespore face of the polar septum leads to improper positioning of SpoIIE, which leads to the promiscuous activation of σ F in both compartments.We wondered whether this mislocalization of SpoIIE in the presence of DivIVA R94N , was due to a reduced interaction between both proteins.Since DivIVA and SpoIIE interact weakly in vitro (32,37), we initially measured the interaction between these proteins using a two-hybrid assay in which SpoIIE and DivIVA were fused to the T18 and T25 subunits of adenylate cyclase, respectively, and heterologously produced in Escherichia coli.SpoIIE-T25 and DivIVA WT -T18 interacted in this assay, as evidenced by increased βgalactosidase activity (Fig. 2I, lanes 1 and 2).By comparison, the interaction of SpoIIE-T25 with DivIVA R94N -T18 similarly showed increased βgalactosidase activity (Fig. 2I, lane 3), suggesting that the R94N substitution did not abrogate the interaction between DivIVA and SpoIIE.Similarly, the R94N substitution also did not appreciably affect DivIVA self-interaction, since we observed increased βgalactosidase activities between DivIVA WT -T25 and DivIVA WT -T18, and DivIVA WT -T25 and DivIVA R94N -T18 compared to the empty vector control (Fig. 2I, lanes 4 to 6).To confirm this interaction biochemically, we constructed B. subtilis strains that produced SpoIIE-GFP and either WT DivIVA with a C-terminally appended FLAG tag or DivIVA R94N -FLAG.After inducing sporulation, we purified the DivIVA variant using an anti-FLAG antibody from detergent-solubilized cell extracts and examined the copurification of SpoIIE-GFP by immunoblotting various fractions collected during purification (Fig. 2J).SpoIIE-GFP copurified with DivIVA-FLAG, as reported previously (32), and with DivIVA R94N -FLAG.In contrast, SigA did not copurify with either construct.Furthermore, in the absence of the FLAG tag on DivIVA, neither DivIVA nor SpoIIE-GFP were retained on the column.The results suggest that the R94N substitution affects an intrinsic ability of DivIVA to asymmetrically localize to the forespore face of the polar septum, in a manner that does not abrogate interaction between DivIVA and SpoIIE.A-A", C-C", E-E", and G-G") nascent or (B-B", D-D", F-F", and H-H (38).Whereas DivIVA and SpoIIE display a forespore-biased localization, FtsA and FtsZ preferentially localize to the mother cell face of the polar septum (Fig. 3A).Moreover, the redeployment of FtsA and FtsZ to polar cell division sites itself is reinforced by SpoIIE, which directly interacts with FtsZ polymers at the onset of sporulation (17,38,46,47).We therefore wondered whether FtsA and FtsZ could reciprocally impact DivIVA R94N and SpoIIE placement at the polar septum.Since the ftsAZ operon is up-regulated at the onset of sporulation (12), we tested whether the slight additional overproduction of FtsA and FtsZ by an extra copy of ftsAZ engineered at an ectopic chromosomal site could correct the sporulation defect caused by DivIVA R94N .Immunoblotting cell extracts of sporulating B. subtilis cells revealed that an additional copy of ftsAZ resulted in a 1.36 ± 0.13 (n = 3)fold increase in steady-state levels of FtsZ protein (SI Appendix, Fig. S1G).Whereas cells producing DivIVA R94N sporulated with an efficiency of 0.18 relative to WT, production of FtsA and FtsZ from the second locus increased sporulation efficiency in this strain to 0.70 relative to WT (Fig. 3B, lanes 4 and 5).The overproduction of FtsA and FtsZ also corrected the σ F activation defect caused by DivIVA R94N : in cells expressing divIVA R94N , only 40% of cells displayed foresporespecific activation of σ F , but slight overproduction of FtsAZ in this strain restored proper σ F activation to 78%, similar to WT levels (Fig. 3C).However, overproduction of FtsA or FtsZ alone did not correct the sporulation defect caused by DivIVA R94N (Fig. 3B, lanes 6 and 7), or restore forespore-specific activation of σ F (Fig. 3C), indicating a combined requirement for and FtsZ overproduction to suppress the DivIVA R94N defects.
We next tested whether the slight overproduction of FtsAZ could correct the mislocalization of DivIVA R94N and SpoIIE at the polar septum using dual-color 3D-SIM.DivIVA R94N -GFP displayed forespore-biased localization in only 46% of cells (n = 126; Fig. 4 B-B"), compared to WT DivIVA, which preferentially localized to the forespore side of the polar septum in 72% of cells (n = 44; Fig. 4 A-A").However, overproduction of FtsA and FtsZ restored DivIVA R94N -GFP to the forespore face of the polar septum to near WT levels (68%; Fig. 4 C-C").We next observed the localization of SpoIIE-GFP.In the presence of WT DivIVA, 71% (n = 60) of cells displayed SpoIIE-GFP on the forespore face of the polar septum (Fig. 4 D-D"); in the presence of DivIVA R94N , only 33% (n = 56) displayed preferential localization of SpoIIE-GFP on the forespore (Fig. 4 E-E").Overproduction of FtsA and FtsZ, though, restored the forespore-biased localization of SpoIIE-GFP in 64% (n = 65) of cells, despite these cells harboring DivIVA R94N (Fig. 4 F-F").The data are consistent with a model in which SpoIIE initially reinforces the redeployment of FtsA and FtsZ to polar positions (12,17,18), but reciprocally, FtsA and FtsZ drive the asymmetric distribution of SpoIIE on one face of the polar septum.
Next, we examined whether FtsA and FtsZ displayed a mother cell-biased localization on the polar septum in the presence of DivIVA R94N using 3D-SIM.In the presence of WT DivIVA, both FtsA-mNeonGreen (54%, n = 23 cells) and FtsZ-mNeonGreen (58%, n = 20 cells) localized on the mother cell face of the polar septum (Fig. 5 A-A (Fig. 5 B-B" and E-E").Introduction of a second chromosomal copy of ftsAZ resulted in restoration of mother cell-biased localization of FtsA (53%, n = 28 cells) and FtsZ (50%, n = 8 cells) (Fig. 5 C-C" and F-F").Khanna et al. reported that the asymmetric positioning of FtsA and FtsZ is responsible for the unusual thinness of the polar septum as compared to vegetative septa (38).Since FtsA and FtsZ influence DivIVA and SpoIIE positioning, we wondered whether DivIVA R94N impacts the thickness of the polar septum.We therefore examined the effect of DivIVA R94N on polar septum thickness using transmission electron microscopy.Similar to what was reported previously, we observed that the thickness of the polar septum in cells producing WT DivIVA was 20 nm (IQR = 45.0 nm) (Fig. 4 G and H).In contrast, cells producing DivIVA R94N displayed thicker polar septa of 36 nm (IQR = 40.5 nm) which, although is thinner than ~80 nm vegetative septa, was nonetheless thicker than a WT polar septum.The overproduction of FtsA and FtsZ, though, largely corrected the thick polar septum defect caused by DivIVA R94N and decreased polar septum thickness to 10 nm (IQR = 58.5 nm).In sum, the results indicate that FtsA and FtsZ influence the correct localization of the DivIVA/SpoIIE complex at the polar septum, likely by changing the architecture of the polar septum (which is manifested at one level by thinness of this septum).Additionally, the observation that DivIVA R94N influences polar septum thickness suggests that the DivIVA/ SpoIIE complex reciprocally influences the function of FtsA and FtsZ at the polar septum.

The Min System Does Not Impact DivIVA
R94N Phenotypes.During vegetative growth, the Min system, composed of MinC and MinD in B. subtilis, is recruited to both sides of the division septum at midcell by DivIVA to inhibit FtsZ polymerization.As a result, aberrant FtsZ ring formation immediately adjacent to the nascent division septum is prevented (36,48).The interaction between DivIVA and MinD is mediated by a third protein, MinJ (49-51).The Min system has been implicated in mediating chromosome translocation events that occur after polar septation (52).However, despite the observation that MinD localizes to the polar septum even earlier (53) its role in the placement of the asymmetric division septum and any other sporulation functions in which the Min system participates has been less clear (54).We therefore wondered whether the Min system is required for the asymmetric distribution of FtsA and FtsZ on the mother cell face of the polar septum.First, we measured sporulation efficiency in the absence of various components of the Min system.Deletion of minJ did not result in a significant decrease in sporulation efficiency (Fig. 6A; compare lanes 3 and 5).However, deletion of minCD resulted in a mild sporulation defect (Fig. 6A, lane 6) that was exacerbated in the presence of divIVA R94N and phenocopied the full divIVA deletion (Fig. 6A, compare lanes 2 and 7).The additive nature of these mutations suggested that the contribution of MinC and MinD in sporulation may be independent of the divIVA R94N phenotype.Next, we investigated whether MinC and MinD affect σ F activation.In the absence of MinC and MinD, 50% of cells (n = 231) displayed misactivation of σ F ; the additional presence of divIVA R94N in ΔminCD cells further increased misactivation of σ F to 90% (n = 103) (Fig. 6 B-C'), further suggesting that MinC and MinD may operate in a separate pathway from DivIVA R94 in the activation of σ F .We next examined whether MinD displays a biased localization at the polar septum using 3D-SIM.In the presence of WT DivIVA, GFP-MinD displayed forespore-biased localization in 64% of cells (n = 25) (Fig. 6D), similar to the biased localization pattern of DivIVA and SpoIIE.In the presence of DivIVA R94N , GFP-MinD continued to display a forespore-biased localization pattern at the polar septum in 66% of cells (n = 27), despite the observation that DivIVA R94N itself increasingly displayed unbiased localization, along with SpoIIE (Fig. 2).Finally, we measured polar septum thickness in the absence of MinC and MinD.Deletion of minCD alone did not affect the thickness of the polar septum compared to WT cells (Fig. 6 F and G).The presence of DivIVA R94N in ΔminCD cells increased septum thickness, indicating that the DivIVA R94N phenotype is independent of the Min system.Furthermore, the introduction of an extra copy of ftsAZ in the chromosome in cells expressing divIVA R94N restored polar septum thickness to near WT levels (Fig. 6 F and G; compare lanes 1, 3, and 5 in Fig. 6G), indicating that the pathway mediated by R94 in DivIVA (involving FtsA and FtsZ and polar septum thickness) is independent of the Min system.As a final test to check whether any other sporulation protein could be involved in the asymmetric positioning of FtsA and FtsZ on the mother cell face of the polar septum, and DivIVA and SpoIIE on the forespore face, we exploited the ability to artificially induce polar septation during vegetative growth by overexpressing ftsAZ and additionally inducing spoIIE expression (37).Using 3D-SIM, we observed that both DivIVA-GFP and SpoIIE-GFP typically localized on the "minicell" face of the induced polar septum (SI Appendix, Fig. S1 H-I").In contrast, FtsZ-GFP typically localized closer to the larger compartment (SI Appendix, Fig. S1 J-J").We therefore conclude that no additional sporulation proteins are required for the asymmetric localization of DivIVA, SpoIIE, FtsA, and FtsZ on either side of the polar septum.

Discussion
DivIVA is a scaffolding protein that performs multiple functions in B. subtilis at different points of the cell cycle (33).In this study, we examined an understudied role of DivIVA near the onset of sporulation: the asymmetric tethering of the phosphatase SpoIIE on the forespore face of the polar septum to achieve compartmentspecific activation of σ F in the forespore.Previous studies reported the asymmetric localization of DivIVA and SpoIIE at the polar septum immediately after membrane invagination commenced (32), followed by the preferential release of SpoIIE into the forespore (33), where it is protected from proteolysis by the mother cell localized protease FtsH (37).However, it was not clear whether the asymmetric distribution of DivIVA and SpoIIE was simply coincidental with compartment-specific gene expression in the forespore and whether degradation by FtsH produced the asymmetric distribution of DivIVA and SpoIIE at the polar septum.Based on our genetic and cytological data, we propose that components of the divisome that are overproduced to establish the polar septum, perhaps combined with some unique feature of the sporulation polar septum that is absent in vegetative, medial septa, drive the asymmetric localization of the DivIVA/SpoIIE complex to the forespore face of the polar septum and that this initial asymmetric localization is the primary driver of compartmentalized gene expression in the forespore.This model is consistent with two principal observations.First, a DivIVA variant, DivIVA R94N , promiscuously localized to both sides of the polar septum, and along with it tethered SpoIIE to either side of the polar septum, which resulted in the activation of σ F in both mother cell and forespore compartments.Thus, despite the presence of FtsH in the mother cell compartment, the initial mislocalization of SpoIIE to the mother cell face of the polar septum was sufficient to activate σ F in the mother cell and cause a reduction in sporulation efficiency.Second, correction of this promiscuous mislocalization of the DivIVA R94N /SpoIIE complex was achieved by overproducing two core components of the cell division machinery: FtsZ and FtsA.This suggested that the process of asymmetric cell division itself was linked to the biased localization of DivIVA and SpoIIE.Moreover, the previously described role of SpoIIE in reinforcing the polar localization of FtsZ to achieve asymmetric cell division (15) indicates a reciprocal dependence of DivIVA/SpoIIE and FtsA/FtsZ for proper localization of each complex.A related version of this model is that the overproduction of FtsA and FtsZ results in an exclusion of DivIVA and SpoIIE from one face of the polar septum, and vice versa.
DivIVA preferentially binds to highly negatively curved membranes (55).Consistent with this notion, during vegetative growth, DivIVA localizes to both sides of the division septum and forms two static rings that abut the septum on either side (36).During sporulation, DivIVA initially localizes similarly to both sides of the nascent polar septum (32), but strangely this double-ring pattern is not stably maintained in mature polar septa, despite the presence of negative curvature on either side of this septum (32).What, then, distinguishes the polar septum from a vegetative septum?Other than subcellular placement, a major difference is that the polar septum is much thinner than the vegetative septum, suggesting a fundamentally different architecture (40).Recently, Khanna et al. reported that this reduction in septum thickness is driven by the asymmetric localization of FtsA and FtsZ on the mother cell face of the polar septum, which in turn is reinforced by SpoIIE (38).One possible source for the initial asymmetrical localization of FtsA and FtsZ is the source of both proteins: the medial position of the progenitor cell in the cytosol of what will become the mother cell (Fig. 7).Thus, the redeployment of FtsA and FtsZ from midcell may naturally result in biased localization of both proteins on the mother cell face of the septum once they arrive at the polar location.Similarly, the accumulation of DivIVA near the pole which will form part of the forespore may also influence the biased localization of DivIVA on the forespore face of the polar septum.We therefore propose an integrated model (Fig. 7) in which SpoIIE initially reinforces the redeployment of overproduced FtsA/FtsZ to polar positions in the cell (17,18).The combined action of SpoIIE and FtsA/FtsZ overproduction results in the placement of the divisome on the mother cell face of the invaginating septum, which then ultimately results in a thin septum (Fig. 7, Top Right cell) (38).At the onset of polar cell division, membrane invagination recruits DivIVA to the forming polar septum, where DivIVA binds to SpoIIE.However, the presence of the divisome on the mother cell face excludes DivIVA from the mother cell face of the polar septum, perhaps due to a unique architecture of a thin septum that is produced at the polar position and/or due to crowding by FtsA and FtsZ that prevents DivIVA from localizing on the mother cell side.This exclusion of DivIVA is somehow overcome by the DivIVA R94N variant (which eventually results in a thick polar septum, due to the mislocalization of the divisome on either side of the polar septum).Presumably, the overexpression of ftsAZ corrects the DivIVA R94N defect by reestablishing a thin polar septum and/or excluding DivIVA from this face, thereby restoring the asymmetric localization of FtsA/FtsZ and DivIVA/SpoIIE on either side of the polar septum.Finally, we propose that any remaining SpoIIE on the mother cell face of the polar septum is degraded by FtsH (37) to ensure the exclusive localization of SpoIIE in the forespore to activate σ F in that compartment.Thus, the construction of an unusual division septum, driven by an interplay between divisome components and a cell fate determinant, establishes intrinsic asymmetry to trigger cellular differentiation.

Materials and Methods
Strain Construction and General Methods.All B. subtilis strains are derivatives of B. subtilis PY79 (56); genotypes are listed in SI Appendix, Table S1.Genes of interest were amplified using PCR to include their native promoter and cloned into integration vectors pDG1662 (for insertion into the amyE locus) or pDG1731 (for insertion into the thrC locus) (57,58).The divIVA R94N   generated using the QuikChange site-directed mutagenesis kit (Agilent).For B. subtilis growth curves, one colony on each strain was first inoculated in 2 mL Difco Sporulation Medium (DSM; KD Medical).Then, 150 µL of this suspension was placed in individual wells of a 96-well plate.Cells were grown at 37 °C for 6 h, with shaking, using a microtiter plate reader (Tecan), and the optical density at 600 nm was measured every 15 min.All E. coli strains and plasmids used in this publication are listed in SI Appendix, Table S1.
Sporulation Efficiency Assay.Sporulation efficiency was determined by resistance to wet heat as described previously (59).Strains were grown in DSM for 24 h at 37 °C at 250 rpm and were then subjected to 80 °C for 20 min in a water bath to kill nonsporulating cells and defective spores.Serial dilutions in phosphatebuffered saline (PBS) were plated on Luria-Bertani (LB) agar and incubated at 37 °C overnight.Sporulation efficiency was determined by enumerating colonyforming units (cfu) per ml and reported relative to cfu obtained for a culture of WT (strain PY79) grown in parallel.
Epifluorescence Microscopy.Cells were induced to sporulate using the resuspension method in Sterlini-Mandelstam (SM) medium (60).At different time points, 150 µL of cell cultures was harvested and resuspended in 10 µL PBS containing 5 µg mL −1 FM4-64 membrane dye and/or 2 µg mL −1 DAPI to visualize DNA, as needed.Finally, 3 µL of cell suspension was placed on a glass bottom culture dish (MatTek) and covered with 1% agar pad made with SM medium.Cells were viewed using a DeltaVision core microscope system equipped with an environmental chamber at 22 °C.Cell images were captured with a Photometrics cool snap HQ2 camera.Eight planes were acquired every 0.2 µm and the data were submitted to deconvolution using SoftWorx software (61).
Immunoblotting and Coimmunoprecipitation.B. subtilis cells were induced to sporulation by resuspension.For immunoblotting cell extracts, cells were harvested 2 h after sporulation induction, cell density (determined by OD 600 ) was adjusted to 0.8, and cells from 1 mL of culture were harvested by centrifugation 7,500 × g for 10 min.Cell pellets were resuspended in protoplast buffer (0.5 M sucrose, 20 mM MgCl 2 , and 10 mM potassium phosphate at pH 6.8) containing 1 mg mL −1 lysozyme and incubated for 30 min at 37 °C (62).Samples were then diluted with two volumes of protoplast buffer containing 1× loading buffer to lyse protoplasts.For immunoblot detection of DivIVA, FtsZ, and SigA, samples were heated at 95 °C for 10 min; for detection of SpoIIE-GFP, samples were incubated at room temperature for 10 min.Samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride membrane, and immunoblotted using rabbit antisera raised against purified DivIVA, FtsZ, SigA, and GFP.Coimmunoprecipitation was performed using the µMACS kit (Miltenyi Biotec).Fifteen-milliliter cultures of cells were harvested by centrifugation 1.5 h after induction of sporulation, and cell pellets were stored at −80 °C.Pellets were resuspended in 1 mL protoplast buffer containing 1 mg mL −1 lysozyme and incubated at 37 °C for 30 min to generate protoplasts.Protoplasts were harvested by centrifugation and resuspended in 1 mL binding buffer (50 mM Tris-HCl at pH 7.5, 150 mM NaCl, 7.5% glycerol, and 0.1% Triton-X-100); sample viscosity was reduced using a syringe needle to shear chromosomal DNA.Cell debris was removed by centrifugation at ~15,000 × g, and the supernatant was combined with an additional 150 µL of binding buffer and added to 50 µL magnetic agarose beads harboring anti-DYKDDDK antibodies (Miltenyi Biotec) equilibrated with binding buffer and incubated at 4 °C overnight.Samples incubated with beads were placed in columns, washed 4 times with 200 µL of binding buffer, and eluted by competition with 50 µL of elution buffer containing FLAG peptides.Loading buffer was added, and proteins were separated by SDS-PAGE and visualized by immunoblotting as described above.
Dual-Color 3D-SIM.Super-resolution imaging was conducted on a custom-built 4-beam SIM system equipped with two lasers (488 nm and 561 nm, Coherent, Sapphire 488 LP-300 mW and Sapphire 561 LP-200 mW), a phase-only nematic spatial light modulator (Meadowlark Optics, MSP1920-400-800-HSP8), a water objective lens (Nikon, CFI SR Plan Apo ×60/1.27NA), and a piezo z-stage (Applied Scientific Instrumentation, PZ-2150, 150µm axial travel) (45).In this work, only the 3D-SIM acquisition mode was used.High-precision #1.5 coverslips (Thorlabs, CG15XH) were cleaned by immersion in 75% ethanol overnight and air-dried before use.To ensure effective adherence of bacteria to the coverslips, a droplet of 10 µL poly-L-lysine solution (Sigma-Aldrich, P8920) was applied to the center of each coverslip in a biosafety cabinet.The solution was air dried at room temperature, after which the coverslips were rinsed with pure ethanol and left to air dry until ready for use.Orange FluoSpheres (Invitrogen, F8800, 0.1µm diameter) were used as the fiducials for correcting local chromatic aberrations in dual-color imaging.Orange beads were dissolved in pure methanol at 1:50,000 dilution, and 2 µL of the solution was applied to the center of the poly-L-lysine-coated coverslips immediately before use.Beaded coverslips were mounted in a magnetic imaging chamber (Warner Instruments, QR-40LP) for imaging.After staining membranes with FM4-64, cells were washed three times with 1× PBS, each wash centrifuging the solution at 3,000 rpm for 3 min.Cells were concentrated in 100 µL of 1× PBS stock solution, and a 2 -µL droplet of the stock was applied to the center of a beaded coverslip.Cells were allowed to settle for 1 min, and the sample was rinsed once by 1× PBS before imaging.To minimize photobleaching, the FM4-64 dye was excited at 561 nm, followed by excitation of the green fluorescent protein at 488 nm.To estimate and correct chromatic aberrations in the system, the apparent position of the orange beads in both color channels was recorded and used to register the images.Image registration was conducted in ImageJ so that the orange beads colocalized in both lateral and axial views for each segmented region of interest.
Images were collected over 2 µm (ensuring we imaged the entire thickness of each bacterium) with an axial step size of 0.125 µm.We used an exposure time of 50 ms per phase, resulting in 27-s imaging time for dual-color volume acquisition.The approximate intensity at the sample plane was 80 W/cm 2 and 60 W/cm 2 for 488 nm and 561 nm, respectively.Further details on 3D-SIM reconstruction algorithms and associated software have been described (45).Following reconstruction, all SIM images had a lateral pixel size of 40.9 nm.Microscopy images represent single plane, and the asymmetric positioning of DivIVA and SpoIIE was highlighted on a graph using line scan tool using Fiji software.Arbitrary axes were chosen in the area where both the GFP and the membrane signals were the strongest.
Bacterial Two-Hybrid Assay.Plasmids harboring spoIIE or divIVA are derived from pKNT25 and pUT18 and were cotransformed in E. coli BTH101 strain (63).Transformants were plated on LB agar containing 100 μg mL −1 ampicillin, 50 μg mL −1 kanamycin, and 1% glucose and incubated at 30 °C for 96 h.For each cotransformation, 10 colonies were pooled together and grown overnight at 30 °C with shaking in LB medium containing ampicillin, kanamycin, and 1 mM Isopropyl ß-D-1-thiogalactopyranoside.Cells were then diluted 1:10 into fresh LB medium in a 96-well plate, and OD 600 was measured using a plate reader (Tecan).Then, 100 µL of the diluted culture was placed in a new 96-well plate and lysed by addition of 10 μL of lysis buffer [1 mg mL −1 lysozyme in 1× BugBuster buffer (Sigma)] at room temperature for 15 min.One hundred microliters of Buffer Z (62 mM Na 2 HPO 4 , 45 mM NaH 2 PO 4 , 10 mM KCl, 1 mM MgSO 4 , and 50 mM β-mercaptoethanol) was added, and the beta-galactosidase reaction was started by adding 2 mM (final concentration) ortho-Nitrophenylβ-galactoside (ONPG).Hydrolysis of ONPG was monitored by measuring A 420 every 5 s for 30 min using a microplate reader (Tecan).Β-galactosidase activity was calculated by measuring the V max of the A 420 appearance divided by the OD 600 .Values were then multiplied by 100,000, a coefficient that was chosen empirically to approximate Miller units.
Transmission Electron Microscopy.Strains were induced to sporulate using the resuspension method.At t = 1.5 h cells were harvested by centrifugation, fixed in fixation buffer (4% formaldehyde, 2% glutaraldehyde in 0.1 M cacodylate buffer), post-fixed using a 1% osmium tetroxide solution, then dehydrated sequentially in 35%, 50%, 75%, 95% and 100% ethanol, followed by 100% propylene oxide.Cells were infiltrated in an equal volume of 100% propylene oxide and epoxy resin overnight and embedded in pure resin the following day.The epoxy resin was cured at 55 °C for 48 h.The cured block was thin-sectioned and stained in uranyl acetate and lead citrate.The sample was imaged with a Hitachi H7600 TEM equipped with a charge-coupled device camera (64).Septum thickness measurements were taken using Fiji software.The line tool was used to evaluate the distance over the septum, and each value was plotted in a violin graph.

Fig. 1 .
Fig. 1.Cells expressing divIVA R94N are specifically defective for σ F activation. (A) Schematic of sporulation initiation in B. subtilis.Sporulation begins with an asymmetric division, producing a smaller "forespore" and a larger "mother cell".The sigma factor σ F is specifically activated in the forespore compartment.Membranes are depicted in yellow; the cell wall is depicted in gray.(B-C') Fluorescence micrographs monitoring σ F activation using promoter fusions (P spoIIQ -gfp, a σ Fcontrolled promoter) in (B and B') otherwise WT cells or (C and C') cells expressing divIVA R94N as the only copy of divIVA at t = 1.5 h after induction of sporulation.(B and C) Fluorescence from GFP production; (B' and C') overlay, GFP and membranes visualized using FM4-64.Strains used: SJC124 and SJC93.(Scale bar: 1 µm.)The percentage of cells (n > 200) displaying forespore-exclusive (left column) or mother cell and forespore (right column) production of GFP is shown to the right of the micrographs.(D) Growth curves of WT (black), ΔdivIVA (red), or ΔdivIVA strain complemented divIVA (blue) or divIVA R94N (orange), as measured by optical density (O.D.) at 600 nm.Data points represent the mean; errors: SD (n = 3 independent cultures).(E) Quantification of cell lengths WT (gray), ΔdivIVA (red), or ΔdivIVA strain complemented divIVA (blue) or divIVA R94N (orange), measured using fluorescence microscopy with membranes visualized with FM4-64.Strains used: PY79, KR546, BRAF22, and APB8.Bars represent the mean; whiskers: I.Q.R. (n = 86 to 265 individual cells).(F-I") Fluorescence micrographs of otherwise WT cells producing (F-F" and H and H") DivIVA-GFP, or (G-G" and I and I") DivIVA R94N -GFP during (F-G") vegetative growth, or (H-I") 1.5 h after induction of sporulation (strains used: SC634 and SC635).(F-I) Fluorescence from GFP; (F'-I') overlay, membranes and GFP; (F"-I") differential interference contrast (DIC).(Scale bar: 1 µm.)Arrows indicate pole-localized DivIVA-GFP; arrowheads indicate DivIVA-GFP at the polar septum.(J) Cell morphologies of cells stained with membrane dye FM4-64 producing WT or R94N variants of DivIVA examined 1.5 h after induction of sporulation.Examined morphologies (disporic, no septa, medial septum, or polar septum) are indicated to the Right, and the fraction of cells exhibiting that morphology is indicated in different shades of gray.(K) Sporulation efficiencies, determined as heat resistance, of WT, ΔdivIVA, or cells expressing divIVA or divIVA R94N as the only copy of divIVA (strains PY79, KR546, BRAF22, and APB8) and reported as relative to WT. thrC is an ectopic chromosomal locus from which indicated alleles of divIVA were expressed.Symbols are individual values from independent cultures; bars represent the mean.Statistical analysis: one-way ANOVA, **** indicates P-value < 0.0001.

FtsA
and FtsZ Influence DivIVA and SpoIIE Placement.The localization pattern for DivIVA and SpoIIE is the opposite of what was recently reported for the central divisome components FtsA and FtsZ at the polar septum

Fig. 4 .
Fig. 4. A second copy of ftsAZ restores forespore-biased localization of DivIVA R94N and SpoIIE as well as septum thinness.Subcellular localization of (A-C") indicated DivIVA-GFP variant or (D-F") SpoIIE-GFP monitored using dual-color 3D-SIM in mutant cells that are blocked before the engulfment stage of sporulation (ΔspoIIM ΔspoIID), in the presence of (A-B" and D-E") one copy of ftsAZ at the native locus, or (C-C" and F-F") in cells harboring a second copy of ftsAZ at an ectopic locus, imaged 2 h after induction of sporulation.Cells in (D-F") harbor an additional deletion of spoIIQ, which prevents the recapture of SpoIIE at the polar septum.(A-F) Membranes visualized using FM4-64; (A'-F') fluorescence from GFP; (A"-F") overlay, membrane and GFP.Strains: SC634, SC635, SC688, SC656, SC657, and SC671.(Scale bar: 1 µm.)Line scan analyses of normalized fluorescence intensity from GFP (green trace) or membrane stain (pink trace) along the axis of the dashed line (indicated in A"-F") in both channels at the selected polar septa are shown at the Right, and the percentage of cells that exhibit forespore-biased GFP localization (left column) or septum colocalized GFP localization (right column) is indicated.(G) Septum thickness of strains harboring DivIVA WT , DivIVA R94N , and DivIVA R94N harboring an extra copy of ftsAZ (strains BRAF22, APB8, and SJC112) observed by transmission electron microscopy.(Scale bar: 500 nm.) (H) Measurements of septal thickness of at least 3 cells per strain.The violin plot represents 10 measurements per cell taken along the length of the septum.

Fig. 5 .
Fig. 5.A second copy of ftsAZ in cells producing DivIVAR94N restores mother cell-biased localization of FtsA and FtsZ.Subcellular localization of (A-C") FtsA-mNeonGreen or (D-F") FtsZ-mNeonGreen monitored using dual-color 3D-SIM in mutant cells that are blocked before the engulfment stage of sporulation (ΔspoIIM ΔspoIID), in the presence of (A-B" and D-E") one copy of ftsAZ at the native locus, or (C-C" and F-F") in cells harboring a second copy of ftsAZ at an ectopic locus, imaged 2 h after induction of sporulation.(A-F) Membranes visualized using FM4-64; (A'-F') fluorescence from GFP; (A"-F") overlay, membrane and GFP.Strains: SC800, SC801, SC807, SC802, SC803, and CS809.(Scale bar: 1 µm.)Line scan analyses of normalized fluorescence intensity from GFP (green trace) or membrane stain (pink trace) along the axis of the dashed line (indicated in A"-F") in both channels at the selected polar septa are shown at the Right, and the percentage of cells that exhibit mother cell-biased GFP localization (left column), septum colocalized GFP localization (middle column), or forespore-biased GFP localization (right column) is indicated.

Fig. 6 .
Fig. 6.MinD displays forespore-biased localization independent of DivIVA R94N and is not involved in determining septal thickness.(A) Sporulation efficiencies, determined as heat resistance and reported as relative to WT, of (lanes 1 and 2) WT, ΔdivIVA, or cells expressing divIVA or divIVA R94N from an ectopic chromosomal locus as the only copy of divIVA in (lanes 3 and 4) otherwise WT cells, (lane 5) in the absence minJ, or (lanes 6 and 7) in the absence of minCD.Strains: PY79, KR546, BRAF22, APB8, SC816, SC780, and SC781.Statistical analysis: one-way ANOVA, **** indicates P-value < 0.0001.(B-C') Fluorescence micrographs monitoring σ F activation using promoter fusions (P spoIIQ -gfp, a σ Fcontrolled promoter) in ΔminCD cells in cells expressing (B and B') WT divIVA or (C and C') divIVA R94N as the only copy of divIVA at t = 1.25 h after induction of sporulation.(B and C) Fluorescence from GFP production; (B'-C') overlay, GFP and membranes visualized using FM4-64.Strains used: SC782 and SC783.The fraction of cells (n > 200) displaying forespore-exclusive (left column) or mother cell and forespore (right column) production of GFP is shown to the right of the micrographs.Scale bar: 2 µm (D-E") Subcellular localization of GFP-MinD monitored using 3D-SIM in mutant cells that are blocked before the engulfment stage of sporulation (ΔspoIIM ΔspoIID), in the presence of (D-D") WT DivIVA or (E-E") DivIVA R94N , imaged 2 h after induction of sporulation.(D and E) Membranes visualized using FM4-64; (D'-E') fluorescence from GFP; (D"-E") overlay, membrane and GFP.Strains: SC810 and SC811.(Scale bar: 2 µm.)Line scan analyses of normalized fluorescence intensity from GFP (green trace) or membrane stain (pink trace) along the axis of the dashed line (indicated in D"-E") in both channels at the selected polar septa are shown at the Right, and the percentage of cells that exhibit forespore-biased GFP localization (left column), septum colocalized GFP localization (right column) is indicated.(F) Septum thickness of ΔminCD strains harboring DivIVA WT or DivIVA R94N in the presence or absence of an extra copy of ftsAZ expressed from an ectopic chromosomal locus and observed by transmission electron microscopy.(Scale bar: 500 nm.)Strains: SC831, SC832, SC833, and SC834.(G) Measurements of septal thickness of at least 3 cells per strain.The violin plot represents 10 measurements per cell taken along the length of the septum.

Fig. 7 .
Fig. 7. Model for compartment-specific activation of σ F .Schematic of B. subtilis divisome placement during (Left) vegetative growth and (Right) sporulation.FtsA(purple) and FtsZ (orange) rings initially assemble at midcell.During vegetative growth, the divisome constricts, creating highly negatively curved membranes on either side of the division septum, which recruits DivIVA (blue).DivIVA then forms double rings that flank the septum (depicted as dots).Once constriction completes, DivIVA remains abutted to either side of the mature septum(36).During sporulation, DivIVA localizes to the extreme poles and SpoIIE (green) associates with FtsZ at midcell and reinforces the redeployment of FtsA and FtsZ to the poles of the bacterium, whereupon only one FtsZ ring actively constricts to form the polar septum and the other (gray) remains inactive.In a WT cell, FtsA and FtsZ mediate septation, but localize to the mother cell face of the septum, resulting in a unique septal architecture that is manifested, in part, as a thin septum(38).DivIVA and SpoIIE display a forespore-biased localization at this septum, resulting in preferential activation of σ F in the forespore compartment (depicted in pink).Cells producing DivIVA R94N disrupt the biased localization of FtsA and FtsZ, resulting in the elaboration of a more vegetative-like septum (that is in part manifested as a thicker septum), and in turn results in the unbiased localization of DivIVA R94N and SpoIIE at the polar septum, resulting in the promiscuous activation of σ F in both compartments.
Data, Materials, and Software Availability.This work does not report original code.All other data are included in the manuscript and/or supporting information.F. Soheilian of the Electron Microscopy Laboratory of CCR for TEM sample preparation and imaging.This work was funded by the Intramural Research Program of the NIH, Advanced Imaging and Microscopy Resource (J.C. and H.S.); the National Institute of Biomedical Imaging and Bioengineering (H.S. and J.C.); and the National Cancer Institute, Center for Cancer Research (K.S.R.).This work was supported by the Howard Hughes Medical Institute.
allele was