DNA segregation by the bacterial actin AlfA during Bacillus subtilis growth and development

Abstract

We here identify a protein (AlfA; actin like filament) that defines a new family of actins that are only distantly related to MreB and ParM. AlfA is required for segregation of Bacillus subtilis plasmid pBET131 (a mini pLS32-derivative) during growth and sporulation. A 3-kb DNA fragment encoding alfA and a downstream gene (alfB) is necessary and sufficient for plasmid stability. AlfA-GFP assembles dynamic cytoskeletal filaments that rapidly turn over (t(1/2)< approximately 45 s) in fluorescence recovery after photobleaching experiments. A point mutation (alfA D168A) that completely inhibits AlfA subunit exchange in vivo is strongly defective for plasmid segregation, demonstrating that dynamic polymerization of AlfA is necessary for function. During sporulation, plasmid segregation occurs before septation and independently of the DNA translocase SpoIIIE and the chromosomal Par proteins Soj and Spo0J. The absence of the RacA chromosome anchoring protein reduces the efficiency of plasmid segregation (by about two-fold), suggesting that it might contribute to anchoring the plasmid at the pole during sporulation. Our results suggest that the dynamic polymerization of AlfA mediates plasmid separation during both growth and sporulation.

Figure 1

A model depicting DNA segregation and cell division during sporulation. (A) Enlargement of a cell pole, early during sporulation, depicting how the Soj/Spo0J proteins act to condense DNA surrounding the replication origin and how this region is further compacted and tethered to the pole by RacA (Ben-Yehuda et al, 2003, 2005; Wu and Errington, 2003). (B) A vegetative B. subtilis cell containing two chromosomes and a medial FtsZ ring (1st cell, blue ring). At the onset of sporulation, chromosomes rearrange into an axial filament where each chromosome condenses into two visually distinct domains separated by a gap (second cell, black lines). FtsZ relocalizes as two rings that are positioned near the poles between the gaps in the axial filament (second cell, blue rings). As division occurs at one of the FtsZ complexes, a ring of SpoIIIE localizes to the leading edge of the invaginating septum (third cell, green ring). Septation then delivers the DNA translocase SpoIIIE to the chromosome (fourth cell, green dot), where it completes chromosome partitioning (fifth cell). (C) Three models depicting how low copy number plasmids could be actively partitioned into the forespore (see text for details).

Figure 2

Cell biological assays for plasmid segregation. (A) A GFP reporter for plasmid segregation that can distinguish between sporangia with the pGAP100 plasmid in the mother-cell (MC) (left) or in the forespore (FS) (right). Only plasmids that reside in the forespore will express GFP from the forespore-specific promoter (P_FS). (B–F) Segregation of DNA during sporulation using a GFP-based reporter assay. Cells were sporulated in the presence of the membrane stain FM 4-64 (red), harvested at t₂ (shown) and t₄ (images not shown) and stained with the DNA stain DAPI (blue). The percentage of cells expressing GFP (green) at t₂ and t₄ is reported at the bottom. (B) GFP expressed from the forespore specific reporter (P_FS-GFP) on pBET131 derivative pGAP100 in a wild type-strain (arrow; strain EBS111). Scale bar=1 μm. (C) GFP expressed from P_FS-GFP on pJSA6, a derivative containing only the origin of replication of pBET131 (EBS114). (D) GFP expressed from P_FS-GFP integrated near the oriC region of the chromosome in a wild-type strain (KP646). (E) P_FS-GFP expression from pGAP100 in spoIIIE73–11 (EBS375) or in (F) ΔracA strain (EBS102). In the racA strain, forespores lacking chromosomal DNA can still trap pGAP100 (arrowhead). (Bottom) Percentage of forespores with GFP at t₂ or t₄. Total sporangia scored are indicated in parentheses and the strains corresponding to the data sets remain in the same column as indicated for the images.

Figure 3

Assay for plasmid segregation postseptation. (A) ‘Molecular memory' cassette. Cre recombinase activates the forespore specific GFP reporter by splicing out the loxP flanked kan resistance gene thereby restoring the reading frame of GFP. The circular recombinant product cannot replicate and will be lost from the cell. (B) Cartoon depicting two possible scenarios of DNA segregation during sporulation. Expression of the Cre recombinase in the mother cell will activate the reporter gene only if it resides in the same cell. When the reporter is trapped in the forespore and Cre is expressed in the mother cell (top cell), GFP is not expressed. When the reporter is present in the mother cell where Cre is expressed, excision of the kan cassette and subsequent translocation of the activated reporter to the forespore will then result in GFP expression.

Figure 4

Identification of an actin related protein encoded by plasmid pLS32. (A) Phylogenetic analysis showing the relationship between MreB, ParM, FtsA, actin and AlfA. The tree was calculated using the neighbor-joining method. Sequences were aligned using TCoffee and ClustalW and the tree constructed using Phylip version 3.6 (Felsenstein, 1989). Sequence abbreviations and accession numbers are as follows. Actin family members (purple): Drosophila melanogaster (AAM50595), Saccharomyces cerevisiae (NP_116614), Caenorhabditis elegans (CAA34718). MreB family members (green): B. subtilis (MreB AAA22397 and Mbl AAA67878), Clostridium tetani (NP_781022), Vibrio cholerae (AAF93588), Escherichia coli (AAA58054). FtsA family members (cyan): B. subtilis (AAA22456), Pseudomonas aeruginosa (AAG07796), Streptococcus pneumoniae (AAL00315), Escherichia coli (NP_414636). ParM family members (red): plasmid R478 Serratia marcescens (NP_941097), plasmid R27 Salmonella typhi StbA (AAD54035), plasmid R1 E. coli (P11904), plasmid R64 Salmonella typhimurium (BAB91612), plasmid pAPEC-02-R E. coli (AAT37581), plasmid pCP301 Shigella flexneri (AAL72301). AlfA family members (purple): plasmid pLS32 B. subtilis (BAA24871), CAC1161 Clostridium acetobutylicum (AAK79133). (B) Alignment of the phosphate 2 motif for selected protein sequences belonging to actin, FtsA, ParM, MreB and AlfA. The aspartic acid is conserved in actins across kingdoms. Sequences and accession numbers are the same as in (A).

Figure 5

AlfA-GFP assembles dynamic filamentous structures. Strains were grown on an agarose pad containing 10% LB at 30°C. Three examples of wild type (A–C) NHB10 (pNCH106 AlfA-GFP) and mutant (F–H) NHB55 (pNCH135 AlfA(D168A)-GFP) are shown. Cell membranes were stained red with FM 4-64. All images are to the same scale. White bar equals 1 micron. (D, E, I, J) FRAP demonstrates that wild-type AlfA-GFP filaments (D, E) are dynamic but filaments assembled by AlfA(D168A)-GFP (I, J) are static. After collecting a prebleach image (−2 s), a small region (red circle) of the filament was exposed to light from an argon laser (488 nm) for 0.5 s and images collected at the indicate times (sec) (D, I). Three-dimensional plots (E, J) show pixel intensity data corresponding to each image. The bleached zone is indicated by a red bracket. (K–M) Quantitative FRAP analysis for wild type (K–L) or AlfA(D68A)-GFP (M). Relative fluorescence intensity of the bleached region (black circles) or an unbleached region (open blue squares) of the filament is plotted versus time. The dashed line indicates the extent of final recovery observed by the end of the experiment. The observed decrease in fluorescence for the unbleached region is due to both photobleaching and to incorporation of bleached subunits into the unbleached region of the polymer.

Figure 6

The alf region contains two genes required for plasmid segregation. (A) A schematic showing the 3 kb alfAB region sufficient for stabilizing a mini-pLS32 replicon. Mutations in either alfA or alfB disrupt segregation. The role of upstream (orf4) or downstream regions (orf7) is unknown. (B) Plasmids containing alfB∷kan mutations are unstable and only poorly transform into B. subtilis. PY79 was transformed with equal amounts (500 ng) of plasmid DNA (pBET131 alfAB+, pJSA3 (ori_{1.5 kb}), pFG6001 alfB∷kan) and plated on LB cam plates. The number of transformants obtained is the average for two experiments and the differences between pBET131 and pFG6001 were reproduced at least five times. (C) Strains containing pFG6001 grow poorly on LB cam plates. PY79 containing pBET131 (wild type), pJSA3 (ori_{1.5 kb}), or pFG6001 (alfB∷kan) were streaked onto LB cam plates and incubated overnight at 30°C.