INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

FtsA is an essential cell division protein of Escherichia coli that is widely conserved in bacteria. Together with ftsZ, which codes for a GTPase analog of the eukaryotic tubulin, ftsA forms one of the most frequently conserved gene pairs among the cell division genes in the eubacteria. Based on sequence homology it has been proposed that FtsA belongs to the sugar kinase/hsp70/actin superfamily (4). This superfamily comprises several proteins with a common two-domain topology and the ability to bind and hydrolyze ATP. FtsA binds to columns of ATP-agarose and can be isolated from cells either as a phosphorylated or a nonphosphorylated form (29), but so far no other biochemical function has been described for this protein.

FtsA is present both in the cytoplasm and in the cytoplasmic membrane (29), where it forms a structural part of the septum (32). It has been proposed that FtsA is a component of a membrane-associated complex (septator or divisome), which would include periplasmic, transmembrane, and cytoplasmic proteins acting coordinately to perform septation (27, 35). Genetic analysis suggests that FtsA may interact, directly or indirectly, with other cell division proteins, such as FtsZ, PBP3, FtsQ, and FtsN (9, 10, 24, 33, 34). The FtsZ/FtsA ratio is important for cell division, and the localization of FtsA to a central position along the cell length depends on the formation of the FtsZ ring (1, 8, 11, 15, 23). Accordingly, the interaction between FtsA and FtsZ from different bacteria has been established using the yeast two-hybrid system (12, 25, 38, 40). In addition the presence of functional FtsA is required for the localization of FtsK, PBP3 (also known as FtsI), FtsQ, and FtsN to the septal ring (2, 7, 37, 41), suggesting that FtsA may indeed be a part of a multiprotein complex (27).

Interaction of FtsA with itself has been suggested based on the ability of ftsA104 (coding for FtsA104: T215A) to complement efficiently two ftsA(Ts) mutations (ftsA2 and ftsA3), but with less efficiency an amber one (ftsA16) (29), and has been recently reported by Yan et al. (40) using the two-hybrid system. Involvement of the protein carboxy terminus in the interaction was suspected from previous results showing that overproduction of a carboxy-terminally truncated form of FtsA causes the formation of long fibers in the cells that also show a curved morphology phenotype (17).

To study the role of the carboxy terminus in the function of FtsA a series of mutants with sequential deletions spanning from 1 to 27 residues was constructed. We have found that this region is essential for the biological activity and for the correct self-interaction of the protein and that the residue W415 is essential for the correct localization of FtsA into septal rings.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial and yeast strains, media, and growth conditions. E. coli strain DH5 [F endA1 hsdR17 supE44 thi1 recA1 gyrA relA1 (lacZYA-argF) U169, (80lacZM15) (19)] was used as host for cloning. Strain OV2 [F ilv leu thyA (deo), his, ara(Am), lac125(Am) galKu42(Am) galE trp(Am) tsx(Am) tyrT] [also known as supFA81(Ts)] was used as wild type for ftsA (31). Strains OV16 [as OV2 ftsA16(Am) (13)] and D2 [as OV2 ftsA2(Ts) leu⁺ (31)] were used for cell division complementation assays. BL21(DE3) pLysS [F ompT hsdSB (r_B m_B) gal dcm (DE3); obtained from Novagen] was used for expression of His-tagged FtsA forms. Luria-Bertani (LB) broth (28) and LB agar, supplemented with antibiotics when required (ampicillin, 100 µg/ml; kanamycin, 50 µg/ml; chloramphenicol, 50 µg/ml; and rifampin, 25 µg/ml) were used for routine cultures of E. coli at 37°C (except when indicated otherwise). Strains carrying ftsA(Ts) alleles used for complementation assays were grown in nutrient broth medium (Oxoid no. 2 nutrient broth) supplemented with thymine (50 µg/ml) (NBT) and antibiotics when required and incubated at the permissive (30°C) or restrictive (42°C) temperatures. Complementation tests were done as described by Sánchez et al (29). VIP386 is BL21(DE3) harboring pMFV12 (expressing an N-terminal fusion of ftsA⁺ to His tag under the control of a T7 promoter contained in a pET28a plasmid purchased from Novagen; construction and strain obtained from M. J. Ferrándiz). VIP516 is BL21(DE3) harboring pSRV2 (expressing an N-terminal fusion of ftsA27 to His tag under the control of a T7 promoter contained in a pET28a plasmid).

Cell parameter measurements, photography, and immunofluorescence microscopy. Cultures were grown at the permissive temperature in liquid media in a shaking water bath so that balanced growth was maintained for several doublings (not less than four) before the beginning of the experiment. The optical density at 600 nm (OD₆₀₀) was measured using a Shimadzu UV-1203 spectrophotometer; the optical density was always kept below 0.40 to 0.50 by appropriate dilution with prewarmed medium. Particles were fixed in 0.75% formaldehyde and counted using a ZM Coulter counter, with a 30-µm-diameter orifice connected to a C1000 Channelyzer (both from Coulter Electronics). Cells were spread on thin agar-azide layers and photographed under phase-contrast optics using a Sensys charge-coupled device camera (Photometrics) coupled to a Zeiss Axiolab HBO 50 microscope. The software used for image capture was IPLab Spectrum, and Adobe Photoshop 5.0.2 software was used for processing.

Plasmid constructions and DNA manipulation. Plasmid DNA isolation, cloning techniques, and transformation procedures were done as described by Sambrook et al. (28). Restriction endonucleases and other enzymes were purchased from and used as recommended by Roche Molecular Biochemicals. DNA was sequenced by the dideoxynucleotide method using Sequenase version 2.0 (U.S. Biochemicals). Two-hybrid system cloning vectors (pGAD424 and pGBT9) were obtained from Clontech.

ftsA27

ftsA2336

ftsA2336

Yeast two-hybrid assays. Yeast strains were transformed using the Li acetate method (18) and selected in SD medium supplemented with the required amino acids and glucose at 30°C. Qualitative -galactosidase assays were performed using Whatman no. 5 filters as described in the Matchmaker kit manual (Clontech). For color development, filters were incubated up to 16 h at 30°C (although no changes were usually observed after 3 h). For quantitative analysis of the interactions -galactosidase assays were done in liquid cultures at exponential growth phase using ONPG (o-nitrophenyl--D-galactopyranoside) as the substrate as described by Miller (26). Cell lysis was achieved by two freeze (dry ice-ethanol) and thaw (37°C) cycles. The values presented are the average of at least three independent experiments, carried out in duplicate each time.

Expression and purification of wild-type and carboxy-terminally truncated FtsA proteins used in Fourier transform infrared (FTIR) and circular dichroism (CD) spectroscopy measurements. E. coli strains VIP386 (His-FtsA⁺) and VIP516 (His-FtsA27) were grown at 37°C in LB medium supplemented with antibiotics when required (50 µg of kanamycin and/or chloramphenicol per ml) to an OD₆₀₀ of 0.4. Overexpression of the proteins was induced with 0.5 mM IPTG (isopropyl--D-thiogalactopyranoside). Thirty minutes after addition of the inducer, rifampin (25 µg/ml) was added to the cell culture. Growth was continued for 1 or 2 h depending on the strain. Cells were harvested by centrifugation and resuspended in ice-cold buffer A (5 mM imidazole, 10 mM HEPES [pH 7.9]). The bacteria were lysed by sonic disruption and centrifuged for 15 min at 10,000 rpm at 4°C (10,800 × g) (Sorvall SA-300 rotor). The His-tagged proteins recovered in the supernatant were purified by metal affinity chromatography on a cobalt column (TALON resin; Clontech) equilibrated with buffer A. The proteins were eluted with a step gradient of imidazole ranging from 30 to 200 mM in a 10 mM HEPES, pH 7.9, buffer. Integrity and purity of proteins were checked by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis (SDS-12% PAGE), and fractions containing pure proteins were pooled and dialyzed against 2 mM HEPES, pH 7.9, before analysis.

Overlay assays. Sample proteins were blotted to nitrocellulose membranes (Schleicher and Schuell) using a slot blot filtration manifold (Amersham Pharmacia Biotech AB). The membranes were blocked by incubation for 1 h with 3% skimm milk in blotting buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 5 mM EDTA, 0.05% IGEPAL CA-630). Purified His-FtsA⁺ was biotinylated in vitro with D-biotin-N-hydroxysuccinimide ester (Roche Molecular Biochemicals) following the manufacturer's instructions. The nitrocellulose membranes were washed with blotting buffer and then incubated for 1 h at room temperature with biotinylated His-FtsA⁺ (0.1 µg/ml) in blotting buffer. The membranes were then washed and incubated for 1 h with streptavidin-peroxidase conjugate (1:10,000; Roche Molecular Biochemicals). After extensive washing, biotinylated FtsA bound to the membranes was detected using the BM chemiluminiscence blotting substrate (POD; Roche Molecular Biochemicals).

Photoaffinity labeling. Proteins were purified by Ni-nitrilotriacetic acid chromatography. For labeling, the imidazole elution buffer was changed in a G-50 column (Amersham Pharmacia Biotech AB) to labeling buffer (50 mM Tris [pH 8.0], 50 mM NaCl, 5 mM MgCl₂, 2 mM dithiothreitol). Three micrograms of protein was mixed with 2 µCi of 8-azido[-³²P]ATP (20 Ci/mmol; ICN Biochemicals) in 20 µl. After a 5-min incubation in the dark the samples were placed on a paraffin film and irradiated for 10 min with a UVP model GL-25 UV lamp at a distance of 10 cm from the samples. After irradiation, samples were mixed with 50 µl of Ni-nitrilotriacetic acid in binding buffer, washed twice, and mixed with SDS-PAGE loading buffer containing 10 mM EDTA. Labeled proteins were analyzed by SDS-PAGE and autoradiography of the dried gels.

ATR-FTIR spectroscopy. Attenuated total reflection (ATR)-FTIR spectra were recorded at 20°C on a Bruker IFS-55 spectrometer, equipped with a liquid nitrogen-cooled mercury cadmium telluride detector with a nominal resolution of 2 cm¹. The spectrometer was continuously purged with dry air. The proteins were spread on a germanium ATR plate with an aperture angle of 45° (50 by 20 by 2 mm; Harrick EJ2121) by slowly evaporating the sample under nitrogen. A total of 1,024 scans were averaged to improve the signal/noise ratio and corrected for the spectrum of the clean ATR plate.

CD spectroscopy. Spectra of FtsA and FtsA27 were recorded in 2 mM HEPES, pH 7.9, at room temperature, using a Jasco J-710 spectropolarimeter in quartz cells with a 0.02-cm path length. The spectra were monitored from 185 to 260 nm, using a scan speed of 50 nm/min, a response time of 2 s, a band width of 1.0 nm, and a resolution of 2 data points/nm. Eight spectra were averaged and subsequently corrected for the spectrum of the buffer.

Analysis of secondary structure. The determination of protein secondary structure was based on a multivariate statistical analysis method of band shape recognition (Oberg et al., unpublished data). This analysis used a commercially available partial least-squares package to determine fractional secondary structure composition (PLSPlus 2.1 for Spectra Calc; Galactic Industries, Salem, N.H.). A reference set (RaSP50) was generated by collecting ATR-FTIR and CD spectra of 50 proteins with known X-ray structures selected to represent a wide range of helix and sheet compositions as well as 60 different protein domain folds. To analyze combined CD and IR data, hybrid spectra were built by placing side-by-side CD (185 to 260 nm) and IR (1,720 to 1,500 cm¹) spectra in a single array. Before analysis, all spectra were normalized so that the PLS algorithm was forced to use band shapes, rather than absolute intensities. Cross validation procedures using the RaSP50 protein spectra have shown root mean square structure determination errors of ±4.5% for -helix and ±6.3% for -sheet conformations.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Deletion mutants of FtsA lacking five or more residues from the carboxy terminus are not biologically active in E. coli. To study the role of the carboxy terminus of E. coli FtsA on its function, a series of sequential ftsA deletion mutants comprising codons coding for 1, 5, 6, 13, and 27 amino acid residues from the carboxy terminus were constructed (Fig. 1). A spontaneous mutant in which the last 23 residues have been replaced by a 36-residue unrelated amino acid sequence, FtsA2336, was also studied. The functionality of these mutants was analyzed by complementation assays of strains containing either a thermosensitive ftsA mutation (D2) or an amber mutation in ftsA in a temperature-sensitive suppressor background (OV16). Strains D2 and OV16 were transformed with plasmids expressing the deletion or wild-type ftsA alleles at levels similar to the expression of ftsA from the chromosome (pLYV57, -58, -59, -60, -61, and pMSV20), and their ability to grow on NBT plates at 30 or 42°C was tested. Western blot analysis using an affinity-purified antibody raised against FtsA indicated that, when produced under the same gene expression conditions, the cellular levels of all the truncated forms were similar to those of the wild-type protein at both growth temperatures (see Materials and Methods); consequently, the possibility that expression from an artificial system, or in vivo thermal instability of the mutant forms can cause unspecific side effects on viability can be excluded. The results of the complementation analysis show that only ftsA⁺ and ftsA1 are able to rescue the two strains (Table 1). The ftsA5 allele in its turn is able to partially complement ftsA2 but not ftsA16. No complementation of any of the two mutations was observed when the ftsA6, fts13, fts27, and fts2336 were tested. We conclude that deletion of five or more FtsA carboxy-terminal residues yields proteins that are not functional in vivo, suggesting that this part of the molecule has an important role on the biological activity of FtsA.

View larger version (56K): [in this window] [in a new window]   FIG. 1.   Sequence of the carboxy terminus of FtsA. The top part of the figure shows the alignment of the carboxy end of the published FtsA sequences from 25 bacterial species, including E. coli, obtained using the program ClustalW. The residues have been highlighted accordingly to their average BLOMSUM62 scores as follows: black for maximum (3.0)-score residues, medium gray for mid-range (1.5)-score residues, and light grey for low (0.5)-score residues. Similar residues according to BLOMSUM62 are shaded as the most conserved one. The bottom part of the figure shows the sequences of the E. coli FtsA carboxy end derived from the wild type and from the deleted ftsA alleles used in this work. The alien residues in FtsA2336 appear in italics. Shading codes are as described above; data were provided by L. Sánchez and A. Valencia.

ftsA1

View larger version (53K): [in this window] [in a new window]   FIG. 2.   Localization of FtsA in E. coli cells expressing different carboxy-terminally truncated forms of FtsA. OV16 strains containing plasmids pLYV29 (ftsA27), pLYV35 (ftsA5) or pLYV41 (ftsA1) were grown at 30°C until an OD600 of 0.15 was reached. At time zero each culture was divided into two portions; one portion was transferred to 42°C, and the other was left at 30°C. Samples were withdrawn after 2 h, processed for immunostaining with purified MVM1 FtsA antiserum, and observed as described in the text. All the frames were corrected to minimize background differences using the Photoshop Levels adjustment. The reference bar marks 5 µm. (A) OV16/pLYV30 (FtsA+); (B) OV16/pLYV41 (FtsA1); (C) OV16/pLYV35 (FtsA5); (D) OV16/pLYV29 (FtsA27). Note the different scale in the D frames.

Effect of the overexpression of FtsA carboxy terminus deletion mutants on cell morphology. Increasing amounts of FtsA cause cell filamentation, a phenotype that has been interpreted as the result of an imbalance in the FtsA/FtsZ ratio (8, 11). However, the overproduction of FtsA carboxy-terminal truncations causes a peculiar phenotype, in which the cells coil forming C-shaped (short) and coiled (longer) cells (CC phenotype). This phenotype has already been observed by Gayda et al. (17) in wild-type cells containing a form of FtsA lacking the last 28 residues encoded by a plasmid.

ftsA5

ftsA6

ftsA13

ftsA27

ftsA27

ftsA6

ftsA13

ftsA27

ftsA2336

ftsA5

View larger version (93K): [in this window] [in a new window]   FIG. 3.   Morphology of E. coli cells expressing two different carboxy-terminal deletion mutant forms of FtsA. OV2 and OV16 strains containing plasmid pLYV35 (ftsA5) or pLYV29 (ftsA27) were grown at 30°C until an OD600 of 0.15 was reached. At time zero they were split in four portions; two of them were transferred to 42°C, while the remaining two were left at 30°C. After 15 min IPTG (50 µM) was added to one portion at each temperature. Samples were withdrawn 2 h later, fixed, and observed as described in the text. Only the induced samples are shown. The reference bar marks 5 µm. (A) OV16/pLYV35; (B) OV16/pLYV29; (C) OV2/pLYV29.

A carboxy-terminally truncated FtsA protein, FtsA27, retains its secondary structure and its ability to bind ATP. To test if the carboxy-terminal deletions had any effect on the structure of FtsA we analyzed the secondary structure of His-FtsA⁺ and His-FtsA27 (the protein missing the longest carboxy terminus stretch among the ones used in this study). ATR-FTIR and CD spectra were recorded in 2 mM HEPES, pH 7.9. No significant difference was detected between the spectra of both proteins (Fig. 4). Their secondary structure was determined by a new method based on the recognition of the spectral bandshapes (K. Oberg, J.-M. Ruysschaert, and E. Goormaghtigh, unpublished data). The input spectra are made of a combination of ATR-FTIR and CD spectra, and the secondary structure fractions are determined by similarity of the hybrid patterns to the reference set made of 50 selected proteins. The data reported in Table 2 show no obvious change in the secondary structure of the protein resulting from the truncation of the carboxy terminus. We must stress here that the variations observed are in the range of the structure evaluation method error and therefore are not related to any structural modification of the His-FtsA27 polypeptidic chain with respect to the wild type.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   ATR-FTIR (top) and CD (bottom) spectra of His-FtsA+ (grey) and the carboxy-terminal deletion mutant form His-FtsA27 (black) in 2 mM HEPES, pH 7.9.

View this table: [in this window] [in a new window]   TABLE 2.   Secondary structure of wild-type His-FtsA+ and carboxy-terminal deletion form His-FtsA27 determined by a combination of ATR-FTIR and CD spectroscopy

View larger version (111K): [in this window] [in a new window]   FIG. 5.   Photoafinity labeling of His-FtsA+ and His-FtsA27 with the photoreactive ATP analog 8-azido[-32P]ATP. The SDS-PAGE separation of the material obtained after the photoactivation of the cross-linking reagent was performed and developed as described in the text. The reaction mixture loaded in the lane marked as A+ contained His-FtsA+, while His-FtsA27 was contained in the one loaded in lane 27. Lane D210A is a negative control (see text for details). The position of molecular mass markers (in kilodaltons) is indicated at the left.

ftsA27

Role of the carboxy-terminus in interaction of FtsA molecules. The different phenotypes produced by the mutants in OV2 and OV16 at 42°C suggested that the FtsA molecules might establish interactions with one another. To test this possibility the yeast two-hybrid system (14) was used. The wild-type ftsA gene was fused to the GAL4-binding and activation domain sequences, and the resulting plasmids were transformed into the yeast reporter strain SFY526. Control experiments showed that FtsA⁺ alone does not activate transcription from the GAL4 promoter. But when both fusion genes were assayed together, blue colonies were obtained in filter assays (Table 3), and 10.9 ± 1.8 Miller units of -galactosidase activity were measured in liquid culture assays, demonstrating that FtsA⁺ molecules interact in the yeast two-hybrid.

View larger version (45K): [in this window] [in a new window]   FIG. 6.   Interaction of His-FtsA+ and His-FtsA27 measured by an overlay assay. Different quantities (as indicated) of purified His-FtsA+ and His-FtsA27 were blotted to nitrocellulose membranes using a slot blot filtration manifold. Bovine serum albumin (BSA) was included as a control. The filters were then blocked to prevent further binding and incubated with in vitro-biotinylated His-FtsA+. Binding of biotinylated protein to the filter-bound proteins was detected by incubating with streptavidin-peroxidase conjugate and developing with a chemiluminiscence detection reagent.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Several mutants producing carboxy terminus truncations of the FtsA cell division protein that can be grouped in three categories have been constructed. The first one, FtsA1, behaves in all respects like the wild-type protein, showing that the last Phe residue, even though it is conserved in many sequences, is dispensable. Next, FtsA5 is able to localize into mid-cell rings but does not complement ftsA mutations and, when overproduced, generates coiled cells independently of the presence of the full-length protein. Finally, FtsA6, FtsA13, FtsA27, and FtsA2336, which are not functional, do not localize into septal rings, and produce coiled cells only in the presence of a full-length FtsA, but not in its absence (Fig. 3). All together, these results indicate that residues 415 to 419 are necessary for the biological function of FtsA, which is in agreement with the results of Ma et al. (23), who, by using green fluorescent protein fusions, mapped the intracellular localization domains of FtsA to both termini of the protein, and moreover, W415 is essential for the localization of the protein into septal rings.

When overproducing FtsA6, FtsA13, FtsA27, or FtsA2336 in strains OV2 or OV16 at 30°C (with FtsA⁺ levels sufficient to support both normal growth and division) cell morphology is strikingly altered to a CC phenotype, while no such phenotype is observed after overproduction of the same proteins in OV16 at 42°C (containing no more than 1% of the FtsA⁺ levels). This result suggests that an interaction between the truncated proteins and a full-length protein might be necessary for the production of the CC phenotype. FtsA belongs to a protein family (4, 29) that includes members that are able to establish self-interactions. In the case of actin this interaction is essential for its biological function, i.e., the formation of the actin filament that is a major component of the cytoskeleton of all eukaryotic cells (30).

The yeast two-hybrid system (14) has already been used to detect interactions among some of the components of the bacterial cell division machinery, including those of the MinCDE system (20), and those of FtsZ with SulA (20), ZipA (25), SpoIIE (22), or FtsA (12, 24, 25, 38, 40). More recently, Yan et al. (40) have reported the self-interaction of E. coli FtsA using the two-hybrid system, a result that has been confirmed in the present work. Using a protein overlay assay (5) we have also detected in vitro the interaction of FtsA molecules. In these assays biotinylated FtsA was used at a concentration of 2.5 nM in the liquid phase. As this concentration is much lower than the intracellular concentration of FtsA, estimated to be around 100 nM when assuming that 150 molecules are contained in each E. coli cell (36), the observed interaction falls well within the expected physiological range.

As discussed by Yan et al. (40) this interaction might be unique to E. coli FtsA, because in the yeast two-hybrid system, the Staphylococcus aureus FtsA molecules do not interact. But given that the strength of the interaction measured for E. coli FtsA falls in the lower range of the two-hybrid assay (11 Miller units for the wild-type proteins), it is possible that if the S. aureus FtsA interaction is slightly weaker than the E. coli one, it might fall below the significant range of the assay.

When tested in the two-hybrid system, deletions of five or more residues impaired the interaction with the wild-type protein, showing that the carboxy terminus of the protein is important for the self-interaction. Although they did not interact in the yeast two-hybrid assay we could nevertheless detect binding of biotinylated FtsA⁺ to FtsA27 when using the protein overlay assay. A plausible interpretation of these results is that the deletion produces a decrease in the binding affinity between the wild type and the mutant protein, weakening the interaction below the detection threshold of the two-hybrid assay. This might be the reason also for the CC phenotype being expressed only when the truncated proteins are produced at high levels. These alterations of the interaction properties observed in carboxy-terminally truncated FtsA proteins (lacking five residues or more) are accompanied by the simultaneous loss of their function in E. coli division, as evidenced by their inability to complement ftsA mutants D2 and OV16.

Each truncated protein species was able to interact with itself and with other truncated FtsA proteins with the exception of FtsA1. This shows that truncation of the carboxy terminus does not result in an extreme change in the protein structure, as the mutated proteins can still self-interact. Furthermore, a comparative FTIR-CD spectroscopy analysis of the soluble His-FtsA⁺ and His-FtsA27, the more extensive of the truncated FtsA forms analyzed, showed that their secondary structures are similar (Table 2 and Fig. 4). This result indicates that the deletion of the 27 last carboxy-terminal residues, although important for the biological activity, does not affect significantly the structure of the protein. Accordingly, FtsA27 retained the ability to bind ATP (Fig. 5), indicating that the structure of the ATP-binding site remains intact.

The FtsA6, FtsA13, and FtsA27 proteins cannot localize into septal rings, and they express a coiled phenotype only in the presence of FtsA⁺. FtsA5, on the other hand, is able to localize into rings and produces coiled cells independently of the presence of FtsA⁺. These data suggest that both a correct localization of the protein in the septal ring and a correct interaction between FtsA molecules are important for cell division. Alterations in the interaction of FtsA might be responsible for the distorted cell shape observed in these mutants.

We can postulate that the FtsA molecule contains domains able to establish interactions (evidence to include the carboxy terminus among them is shown) and can speculate that removal of a carboxy-terminus segment exposes one or more domains, normally masked in the wild-type protein, that may be able to establish a different and stronger interaction. Alternatively the deletion of the carboxy terminus may affect the properties of the resultant protein to modify its transport into the yeast nucleus, resulting in an alteration of the molecules available for the establishment of the interaction. This trivial explanation would be more difficult to reconcile with the apparent conservation of the protein structure in the deleted forms. In any case the loss of interaction with the wild type is accompanied by a loss of function, but the exact nature of the FtsA interactions and the molecular mechanisms responsible for the impairment of the biological function observed when the carboxy terminus of FtsA is deleted remain, nevertheless, to be discovered.