HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 279, Issue 47, 48821-48829, November 19, 2004

This Article Abstract Full Text (PDF) All Versions of this Article: 279/47/48821    most recent M404944200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Marrington, R. Articles by Addinall, S. G. Search for Related Content PubMed PubMed Citation Articles by Marrington, R. Articles by Addinall, S. G. Social Bookmarking                      What's this?

FtsZ Fiber Bundling Is Triggered by a Conformational Change in Bound GTP^*

Rachel Marrington¶, Elaine Small||, Alison Rodger, Timothy R. Dafforn**, and Stephen G. Addinall||

From the Department of Chemistry, University of Warwick, Coventry CV4 7AL, ^||School of Biological Sciences, Michael Smith Building, Oxford Road, Manchester M13 9PT, and ^**Department of Biosciences, University of Birmingham, Birmingham B15 2TT, United Kingdom

Received for publication, May 4, 2004 , and in revised form, August 23, 2004.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Polymer formation by the essential FtsZ protein plays a crucial role in the cytokinesis of most prokaryotes. Lateral associations between these FtsZ polymers to form bundles or sheets are widely predicted to be extremely important for FtsZ function in vivo. We have carried out a study in vitro of FtsZ polymer formation and bundling using linear dichroism (LD) to assess structural properties of the polymers. We demonstrate proof-of-principle experiments to show that LD can be used as a technique to follow FtsZ polymerization, and we present the LD spectra of FtsZ polymers. Our subsequent examination of FtsZ polymer bundling induced by calcium reveals a substantial increase in the LD signal indicative of increased polymer length and rigidity. We also detect a specific conformational change in the guanine moiety associated with bundling, whereas the conformation and configuration of the FtsZ monomers within the polymer remain largely unchanged. We demonstrate that other divalent cations can induce this conformational change in FtsZ-bound GTP coincident with polymer bundling. Therefore, we present "flipping" of the guanine moiety in FtsZ-bound GTP as a mechanism that explains the link between reduced GTPase activity, increased polymer stability, and polymer bundling.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The FtsZ protein, a structural and functional homologue of eukaryotic tubulins, is a polymer-forming GTPase that plays a critical role in the cell division process in bacteria. Before the onset of cell division in the -proteobacterium Escherichia coli, FtsZ polymers form a dynamic ring (the Z-ring) around the cell center on the inner surface of the cytoplasmic membrane (1–3). The Z-ring recruits at least ten accessory proteins which are essential for cell division to proceed (4–7), and then reduces in diameter until cytokinesis is complete. FtsZ is extensively conserved among prokaryotes with only one phylum of archaebacteria (Crenarchaeota) and three free-living species of bacteria (Ureaplasma urealyticum, Prosthecobacter dejongii, and Pirellula sp.) proven to lack ftsZ genes thus far (8, 9). In addition, FtsZ is required for chloroplast division and may be essential for division of some mitochondria (10–12). Elucidating the mechanism by which the Z-ring drives constriction is, therefore, critical to our understanding of cell division in prokaryotes and cytokinesis in eukaryotic organelles.

Monomers of FtsZ protein associate in vitro in a GTP-dependent fashion into linear, unbranched, polymeric fibers whose dynamics result in the hydrolysis of GTP (13, 14). There is considerable evidence to suggest that the interface between two FtsZ monomers in the polymeric form is crucial for this GTPase activity, with the "top" and "bottom" of consecutive monomers combining to form the GTP-binding pocket (15–19) (Fig. 1). Although GTP binding favors polymer formation, the hydrolysis of GTP to GDP promotes shortening and bending of the FtsZ polymers (20). The simplest polymer form of FtsZ, a one-monomer wide protofilament, has been observed in vitro (21); however, FtsZ fibers are more commonly seen as pairs of parallel protofilaments called thick filaments e.g. Refs. 18, 21, and 22. More complex FtsZ oligomeric structures, which include sheets (or ribbons), tubes, and bundles, have also been observed in vitro (18, 22–29). However, in each case where detailed structure has been obtained, these consist of thick filaments in different arrangements (parallel, anti-parallel, twisted, straight, etc.) (18, 22, 24). Critically, the structure of the Z-ring in vivo remains uncharacterized; however, it seems likely from these in vitro studies that the Z-ring consists of thick filaments arranged in either sheets/ribbons or bundles.

View larger version (50K): [in this window] [in a new window]   FIG. 1. Left, proposed structure of the FtsZ polymer fiber (18). The fiber axis runs vertically up the page. Right, an expanded view showing GDP (in stick representation) bound to FtsZ. The long axis of the guanine moiety extends out of the plane of the page, perpendicular to the fiber axis. The short axis is inclined to the fiber axis. The figure was produced using SwissPDB (59).

The alteration in lateral association caused by these extraneous factors is a process that is little understood at a structural level. It is likely that such associations are caused by alterations of the surface properties of the FtsZ polymer that encourage fiber-fiber interactions. It has been proposed that such changes occur by charge shielding of negatively charged FtsZ polymers (28). However, other mechanisms are possible. A conformational change within the FtsZ monomers making up the polymer or a change in the configuration of the monomer within polymers could each lead to exposure of inter-fiber binding sites. Alternatively, a combination of mechanisms may be responsible for the bundling of FtsZ fibers.

Techniques for examining solution phase conformational changes in proteins are fairly well established, the most commonly used being circular dichroism (CD) and Fourier transform infrared spectroscopy. Both provide spectroscopic information that can be deconvolved to provide measures of secondary structure content and, to some extent, side chain arrangement. Unfortunately, information on the orientation of monomeric protein units within protein polymers is less easy to obtain. In general, determining this sort of information has relied on either indirect biophysical measurement (e.g. using resonance energy transfer to triangulate monomer positions relative to one another (38)) or direct visualization of the polymer using electron microscopy (e.g. Ref. 18).

We have developed the technique linear dichroism (LD)¹ (which is technically related to CD) that provides information on the orientation of secondary structural elements and aromatic side chains within protein fibers. LD is the measure of the difference in absorbance by a sample of light polarized in orthogonal directions. To obtain an LD signal the fibers are usually aligned perpendicular to the incident light beam. We have achieved this by the use of a Couette flow cell which induces alignment as a result of shear flow in a liquid. Our own advances in cuvette design have only recently permitted the use of LD to examine relatively low volume (200 µl) biological samples (39, 40). Our initial studies of some common protein fibers, including actin, amyloid, and collagen, have allowed us for the first time to assign the signals observed to known secondary structure types (41). The study presented here represents a further advancement of the technique, with reduction of sample volume to 25 µl (42).

In this work we have used a combination of linear and circular dichroism to examine FtsZ polymerization and bundling. We demonstrate that LD provides a measure of FtsZ polymerization kinetics that is consistent with accepted results from light-scattering measurements. We show that bundling of FtsZ protofilaments induced by the presence of millimolar amounts of calcium is detectable as a substantial increase in LD signal. Importantly, we also detect a significant change in the linear dichroism signal in the guanine region of the spectrum under conditions where polymer bundles are formed. We also show that elevated levels of Mg²⁺ can mimic the bundling properties of Ca²⁺. We have, therefore, detected a change in the conformation of the FtsZ-bound guanine moiety that correlates with (and probably accounts for) the decrease in GTPase activity associated with FtsZ polymer bundling. Hence, we provide the first putative mechanism for the link between increased polymer bundling and reductions in polymer dynamics and GTPase activity.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
FtsZ Purification—FtsZ was purified as described previously (36) using a modification of protocols described by Mukherjee and Lutkenhaus (43) and Lu and Erickson (44).

Spectroscopic Measurements—Light-scattering measurements were performed as described previously (36) with the exception of the direct side-by-side comparisons with LD for Fig. 2. These were performed at room temperature (23 °C) using a PerkinElmer Life Sciences LS50B spectrofluorimeter with excitation and emission wavelengths of 400 nm and a slit width of 2.5 nm with a 0.3-cm path length rectangular cuvette.

View larger version (16K): [in this window] [in a new window]   FIG. 2. Polymerization of FtsZ at 11 µM with 0.2 mM GTP added at 0 s monitored by the change in LD signal at 210 nm (open circles) or light-scattering (filled circles).

LD measurements were performed at room temperature using a Jasco J-715 spectropolarimeter adapted for LD spectroscopy (42). Samples were aligned in the light beam using a custom made Couette cell. The cell was constructed by Crystal Precision Optics, Rugby, UK and consists of a cylindrical cross-section sleeve with a removable quartz capillary (sealed at one end with Araldite Rapid®) held centrally with respect to its circular face by a rubber O-ring. A quartz rod was suspended from the demountable lid into the capillary, creating an annular gap between rod and inner capillary wall of 0.25 mm. The rod and the center of the capillary were aligned so that the capillary was able to rotate freely. The sample (25–60 µl) was placed into the capillary before loading into the cylindrical sleeve. The lid and rod units are then lowered into position. Upon rotation of the capillary (the rod remains stationary) a shear force is induced across the sample. The cylindrical sleeve has two windows for the light to pass through to make contact with the sample. This configuration allows the use of smaller sample volumes (25–60 µl) than has previously been possible (previous cells have required 200 µl (39, 40) or 2000 µl (45) of sample). The voltage applied to the motor that rotates the outer quartz cylinder is controlled electronically to allow the sample solution to be maintained with the highest possible degree of alignment without inducing turbulent flow and Taylor vortices. Data were collected using an interval scan measurement program available within the Jasco software. This enabled single full wavelength scans from 350 to 190 nm to be collected every minute at a scanning speed of 200 nm min^–1, data pitch 0.5 nm, and with a response of 0.5 s, thus, monitoring of the kinetics of polymerization across the whole wavelength spectrum. The spectra are of suitable intensity such that only one scan was necessary. Base lines of data collected after de-polymerization of FtsZ were subtracted from all spectra. The wavelength ranges of the LD spectra were in some cases restricted by the high absorbance of the samples at low wavelength. Data reported in this paper are truncated at a point where the Beer-Lambert law is valid. LD spectra were corrected for light-scattering as per Nordh et al. (46).

FtsZ polymerization assays were performed essentially as described previously (36). Briefly, FtsZ at the concentration specified in individual experiments was incubated in a standard polymerization buffer (50 mM MES, pH 6.5, 10 mM MgCl₂, 50 mM KCl), and GTP (the disodium salt adjusted to neutral pH) was added to a final concentration of 0.05–0.2 mM (see individual experiments) to initiate polymerization. Where specified, MgCl₂ was omitted from the standard buffer, or its concentration was altered. In some experiments CaCl₂ or CuCl₂ were added to the basic polymerization buffer to give the final concentrations described for individual experiments in the text and figure legends.

For all spectroscopic measurements samples were placed in the cuvette/capillary cell immediately upon preparation. Of the three techniques used, the longest dead time (about 40 s) was that taken to load and assemble the capillary LD unit and start the analysis.

GTPase Assay—GTPase activity of newly purified FtsZ protein was determined using -³²P-labeled GTP (Amersham Biosciences) and thinlayer chromatography as previously described (36). For titration of the effects of various ion concentrations on FtsZ GTPase activity, we measured production of NADPH from an enzyme-linked assay that has been used previously with FtsZ (47). This assay gave comparable values for FtsZ GTPase activity to the radioactive method but was more appropriate for multiple repeat experiments.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Measurement of FtsZ Polymerization Kinetics Using LD— For a linear dichroism signal to be measured, the molecule studied has to be aligned to some degree. In the case of the flow alignment Couette system used in these studies, the molecule is aligned by virtue of it having a shape with a high aspect ratio. We predicted that LD would be an ideal technique for studying FtsZ, which polymerizes from monomers (or short oligomers) into long, unbranched, linear polymers. Monomeric units of FtsZ (which roughly approximate to spheres and have an aspect ratio close to 1) will not align and, hence, will not produce an LD signal. Polymeric forms of FtsZ were expected to align and, thus, have a flow LD signal.

To test this prediction FtsZ polymerization reactions were followed by both LD and light-scattering. Samples to be observed by LD were placed in the LD cell, and their LD spectrum was measured in the UV, where the transitions of the chromophores of the protein backbone can be observed. GTP was then added to induce fiber formation, and the change in the LD signal was measured at a wavelength that is sensitive to the presence of aligned -helical structure (FtsZ has a high -helix content that should become apparent in the LD if fibers are formed and align). Control samples, also at room temperature, were studied by light-scattering as described under "Materials and Methods."

The measurement of the LD signal of FtsZ polymerization (Fig. 2) with time showed a very similar-shaped trace to that observed using right-angled light-scattering. It can be seen (Fig. 2) that there is an immediate large LD signal induced by the fast GTP-dependent association of FtsZ monomers to form polymers. There is then a plateau in the trace as the GTP in solution is turned over followed by a sigmoidal decrease as the GTP is exhausted and the polymers dissociate. The consistency between light-scattering and LD validates the use of LD to monitor the polymerization of FtsZ since the observed reaction kinetic is very similar to that observed previously by light-scattering (and other) techniques, e.g. Refs. 26, 33, 34, and 36.

CD Spectra of FtsZ and FtsZ with GTP—To understand the conformational changes that occur within FtsZ when the protein polymerizes, CD spectra were recorded of FtsZ in the presence and absence of GTP (Fig. 3a). These data show no discernable alteration in the signal collected above 200 nm as reported previously (48), with the only alteration between the two preparations being a reduction in signal intensity below 200 nm for the polymerized sample. Results from the deconvolution of the spectra of FtsZ using CDsstr (Table I) showed no significant differences from the secondary structure composition calculated from the x-ray crystal structure of FtsZ (Protein Data Bank code 1FSZ [PDB] ). The decrease in signal intensity of the CD spectra of FtsZ plus GTP compared with FtsZ alone can be attributed to light-scattering from the polymers that have formed upon addition of GTP.

View larger version (14K): [in this window] [in a new window]   FIG. 3. CD spectra of FtsZ at 11 µM in 50 mM MES buffer, pH 6.5, 10 mM MgCl2, and 50 mM KCl at pH 6.5 at 23 °C alone (thick line) and in the presence of 0.2 mM GTP (thin line) (a) and in the presence of 0.2 mM GTP (thick line) and 1 mM (thin line), 5 mM (dashed line), and 10 mM (thick dashed line) Ca2+ (b). Experiments were carried out in a 0.1-mm path length demountable cuvette.

LD Spectra of FtsZ and FtsZ with GTP—Having been unable to demonstrate significant changes in FtsZ conformation associated with either polymerization or polymer bundling by CD, we examined LD spectra of FtsZ under various conditions. We first obtained LD spectra of FtsZ in monomeric and polymeric form. The spectrum of the unpolymerized material is zero, as predicted, because molecular alignment is required to produce a signal (data not shown). Spectra from polymeric FtsZ (Fig. 4a) were obtained during the plateau phase of the reaction and have been truncated at a point where the Beer-Lambert law is followed. The 220-nm n- * transition signal of the -helix in the protein is apparent as a shoulder under the large lower wavelength signal due to the - * transition (41). The sign of this transition is positive so its transition polarization is more parallel than perpendicular to the fiber axis. Because this transition lies perpendicular to the -helix axis, this means that on average the helices within the protein lie more perpendicular than parallel to the fiber axis (39, 40).

View larger version (24K): [in this window] [in a new window]   FIG. 4. a, the LD spectrum of FtsZ at 11 µM in 50 mM MES buffer, pH 6.5, 10 mM MgCl2, and 50 mM KCl at 23 °C 1 min after the addition of 0.2 mM GTP. b, the change in LD at 222 nm of FtsZ as a function of time. Measurement was begun after the addition of 0.2 mM GTP to 11 µM FtsZ at 23 °C in 50 mM MES buffer, pH 6.5, 10 mM MgCl2, and 50 mM KCl in the presence of 0 (open circles), 1 (filled circles), 3(open squares), 5(open diamonds), and 10 mM Ca2+ (filled squares). c, the effect of Ca2+ on the backbone region of LD spectra. d, the effect of Ca2+ on the aromatic region of LD spectra. Both c and d refer to FtsZ at 11 µM in 50 mM MES buffer, pH 6.5, 10 mM MgCl2, and 50 mM KCl at 23 °C alone (thin line) and in the presence of 1 (thin-dashed line), 3 (thick line), 5 (thick-dashed line), and 10 mM Ca2+ (dotted line).

The Effect of Ca²⁺ on the Kinetics of FtsZ Assembly Measured by LD—We examined the LD spectra of FtsZ in the absence and presence of Ca²⁺. First we used LD to compare the polymerization of FtsZ in the presence of increasing concentrations of Ca²⁺. These data show two results of increased Ca²⁺ concentration; first, the length of the plateau phase of the polymerization is increased (Fig. 4b), and second, the amplitude of the peaks in the backbone region of the spectra were significantly increased (Fig. 4, b and c). The increase in plateau duration has been observed previously by light-scattering and corresponds to a decrease in GTPase activity (Refs. 27, 33, and 50 and Table II). The increase in the amplitude of the backbone region of the LD spectra (Fig. 4, b and c) is consistent with polymer bundling improving the degree of alignment of the polymer in the Couette cell because of an increase in length and/or rigidity of the polymers.

Flipping of the Guanine Moiety Can Be Induced by Both Ca²⁺ and Mg²⁺—Because these experiments, along with all previous reports of FtsZ polymer bundling by Ca²⁺, have related to conditions where both Mg²⁺ and Ca²⁺ were present, we sought to clarify the role of the different divalent cations in the guanine-flipping that we had observed upon FtsZ polymer bundling. Our standard polymerization conditions include 10 mM Mg²⁺ and no Ca²⁺ (see under "Materials and Methods"), and it has been demonstrated previously that polymerization of FtsZ does not require the presence of Ca²⁺ (33). Although Mg²⁺ is required for the FtsZ GTPase activity, it is not required for polymerization per se, since short, thin polymers are formed when GTP is added to FtsZ in the absence of Mg²⁺ (Ref. 26 and Fig. 5a). Therefore we examined GTP-dependent FtsZ polymerization in the presence of Ca²⁺ only. To rule out complications arising from the possible presence of residual amounts of Mg²⁺ ions in our purified FtsZ preparation, we added 0.1 mM EDTA to reaction mixtures before the addition of CaCl₂. Experiments were then examined by light-scattering, negative stain electron microscopy, and LD.

View larger version (135K): [in this window] [in a new window]   FIG. 5. Morphology of FtsZ polymers. FtsZ at 8.3 µM in 50 mM MES buffer, pH 6.5, and 50 mM KCl at 30 °C after the addition of 0.1 mM GTP in the presence of no divalent cations (a), 10 mM MgCl2 (b), 10 mM MgCl2 + 10 mM CaCl2 (c), 10 mM CaCl2 (d) 40 mM MgCl2, (e), and 0.1 10 mM MgCl2 + 0.1 mM CuCl2 (f).

View larger version (18K): [in this window] [in a new window]   FIG. 6. Ca2+ alone or elevated levels of Mg2+ can also induce guanine flipping and FtsZ polymer bundling. a, LD spectra of FtsZ at 11 µM in 50 mM MES buffer, pH 6.5, and 50 mM KCl at 23 °C after the addition of GTP to 0.2 mM in the presence of no divalent cations (thin line), 10 mM MgCl2 (thin-dotted line), 10 mM CaCl2 (thick line), 10 mM MgCl2 + 10 mM CaCl2 (thick-dotted line), or 40 mM MgCl2 (thick-dashed line). The inset shows the spectrum in the presence of 0.1 mM CuCl2. b, FtsZ polymerization monitored by light-scattering after the addition of 0.1 mM GTP in the presence of 10 mM MgCl2 (open diamonds), 10 mM MgCl2 + 10 mM CaCl2 (open circles), 30 mM MgCl2 (filled diamonds), 40 mM MgCl2 (filled circles), or 60 mM MgCl2 (open squares).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
We have observed that the LD spectra of FtsZ are influenced by increases in Mg²⁺ and Ca²⁺ concentration, leading to a significant increase in signal at 210 nm (Fig. 4). The LD magnitude increase suggests an increase in either the rigidity or length of the polymer (hence, increased alignment). Alternatively, the -helices of the monomers could orient more perpendicular to the fiber axis or increase in number, but if this was the case (that is, Ca²⁺ induced a large change in the fold of the FtsZ monomer) it would be expected that the CD spectra of the backbone region would be altered significantly, which is not the case (Fig. 3). Thus, we observe a stiffening/lengthening of the fibers in the presence of Ca²⁺. This correlates well with electron microscopy images of FtsZ polymers formed in the presence of 10 mM Mg²⁺ and 10 mM Ca²⁺ (Refs. 27 and 33 and Fig. 5c) or 40 mM Mg²⁺ (Fig. 5e), which, as well as forming frequent bundles, are demonstrably straighter in appearance than those formed in standard reactions (e.g. Refs. 33 and 36 and Fig. 5b).

Our LD data also show a significant change in one part of the near UV region in response to the presence of Ca²⁺ or elevated concentrations of Mg²⁺. The inversion of a peak at 250 nm indicates a large movement of the guanine chromophore within the fiber, altering its orientation from having its long axis approximately perpendicular to approximately parallel to the fiber axis (strictly from >>55° to <<55°). This result elegantly demonstrates one of the major advantages of LD over other optical techniques; it can probe the orientations of key units within the fiber. Free GTP and GDP, which in this case represent >96% of the total nucleotide present at any one time in the reaction mixture, do not give an LD signal because they are not able to be aligned. Hence, LD provides an exquisite method for probing the on-enzyme conformation of nucleotide.

Consequences of Elevated Mg²⁺ on GTPase Activity—It is established that GTPases require the presence of a divalent cation (usually Mg²⁺) coordinating the phosphates of the GTP for efficient catalysis to proceed. The metal ion usually bridges between the oxygen atoms of the terminal 2 phosphates of the nucleotide and oxygens from bulk solvent or protein side chains. GTPases are also exquisitely sensitive to the type of divalent cation that performs this coordination. The GTPase activity of FtsZ, for example, is completely abolished in the presence of Ca²⁺ instead of Mg²⁺ (Table II). It is, however, often erroneously assumed that metal-chelated GTP in solution only occurs in the extended conformation appropriate for catalysis. In fact, studies by Sigel et al. (54) demonstrate that GTP can exist in two very different metal bound conformations, extended and a folded macrochelate. The macrochelate occurs as a result of an interaction between the metal ion and the N7 nitrogen on the guanine, leading to a folded conformation of the nucleotide. This conformation is restricted to purines and is more prevalent in guanine compared with adenine. The interaction between a metal ion (M²⁺) and GTP can, thus, be written,

(REACTION I)

where M²⁺ is a divalent metal ion.

At saturating concentrations of metal ion (dissociation constants for Ca²⁺, Mg²⁺, and Cu²⁺ between 0.05 and 2 mM, 0.1 and 2 mM, and 40 nM and 0.25 mM, respectively, depending on the ionization of the GTP), the relative proportions of the nucleotide in each conformation was determined by the equilibrium constant K_eq. Sigel et al. show (54) that this constant is determined by the type of metal ion present, with the maximum conversion to the macrochelate state occurring in the presence of Cu²⁺. Solutions of GTP with Ca²⁺or Mg²⁺, thus, contain 24 and 21% of the macrochelate form, respectively, at full saturation. The question is now posed as to what implications this has for the on-enzyme population of GTP. It has been shown that the FtsZ GTPase activity increases as Mg²⁺ concentration increases up to an optimal level (26). This is consistent with the formation of a catalytically competent GTP-Mg²⁺ complex on the enzyme (including chelation of the metal by protein side chains). At these elevated concentrations of metal ion, formation of the GTP-Mg²⁺ complex becomes favored in solution until almost all GTP is in complex with Mg²⁺. For Mg²⁺, 20% of this GTP-Mg²⁺ complex is in the form of the macrochelate. When this situation is reached in the presence of FtsZ we have the set of potential steps shown in Reaction II. The upper row of reactions represents the binding of GTP-Mg²⁺ in the extended conformation to FtsZ, which is the productive pathway leading to GTP hydrolysis. The binding of this conformation is analogous to the binding of the catalytically competent conformation observed in crystal structures of GTPases with bound GTP and Mg²⁺. The lower row of reactions represents an equivalent set of events for the macrochelate form of GTP-Mg²⁺. Although macrochelate formation has been detected and measured in solution (54), there are no data on its interactions with GTPases. Our data from near UV LD at elevated concentrations of Mg²⁺, however, indicates the presence of GTP in the active site in a radically altered conformation compared with that seen at lower Mg²⁺ concentrations. We propose that these data obtained with FtsZ provide the first evidence of GTP-Mg²⁺ in the macrochelate conformation (or in a macrochelate-like conformation) bound to a GTPase. Data from structural studies of the GTPase active site suggest that hydrolysis of macrochelate GTP is unlikely to occur because the precise positioning of the phosphates and metal ion in the active site is crucial for catalysis. Our data support this because in each case we detected the signature LD signal for the alternative GTP conformation. At elevated Mg²⁺ concentrations we also detected a decrease in GTPase activity that

(REACTION II)

matches the onset of bundling. This is a consequence of the changes in the amounts of free Mg²⁺ and free GTP as well as the relative amounts of the two possible GTP-Mg²⁺ complexes. From this it is a logical extension that these in turn are related to the values for the association constants for formation of the GTP-Mg²⁺ complex (both on and off enzyme) as well as the equilibrium constant for the interconversion between the extended and macrochelate form of the Mg²⁺-GTP complex. Although direct determination of all these constants is beyond the realm of this study, it seems likely that the increase in the absolute concentration of the macrochelate form at high magnesium concentrations (tending to 20% of all GTP) is responsible for the decrease in GTPase. This is particularly likely as the FtsZ-GTP (macrochelate) conformation is likely to be catalytically inactive and can, thus, be treated as a competitive inhibitor of FtsZ.

Consequences of elevated Mg²⁺ and Ca²⁺ on FtsZ Bundling—It has previously been observed that the presence of high concentrations of the catalytically inert metal ion Ca²⁺ can induce the bundling of FtsZ polymers (27, 33), and we observe similar bundling at elevated Mg²⁺ concentrations (Fig. 5e). In both cases we also observe flipping of the guanine group indicative of GTP bound to FtsZ adopting an alternative conformation. As outlined in the previous section, elevated concentrations of divalent metal cations in solution with GTP lead to the formation of increasing amounts of macrochelate containing metal and nucleotide. We, therefore, suggest that polymer bundling and macrochelate formation may be linked. However, we recognize a number of possible mechanisms by which calcium ions could induce FtsZ polymer bundling; 1) Ca²⁺ acts to buffer negative charges on the surface of FtsZ and, hence, reduces repulsive forces and induces association between polymers (28), 2) Ca²⁺ allows the formation of stable FtsZ polymers by virtue of the ion acting as an analogue of Mg²⁺ but without allowing hydrolysis (the increased persistence of these polymers allows more time for lateral bundling to occur), 3) Ca²⁺ binds to the FtsZ polymer, resulting in a conformational change that aids bundling (the conformational change occurs in either FtsZ or GTP or both).

Our data allow us to distinguish between these possible mechanisms. We observe very little change in the backbone conformation of FtsZ between bundled and non-bundled polymers. This indicates that if a conformation change in the protein architecture is important (mechanism 3) it involves only small changes in backbone torsions and perhaps side chain movements. We observe a signature LD signal for an alternative GTP conformation at concentrations of Ca²⁺ that induce bundling (Fig. 4d). We propose that this represents a nucleotide conformation that is quite different to the extended conformations observed in x-ray crystal structures. The question is then whether Ca²⁺ induces the conformational change directly or by binding to FtsZ and in some way alters the conformation of the nucleotide binding site with a resulting "knock on" effect on the cofactor conformation. Because an alternative conformation of GTP is populated in solution at high metal ion concentrations (54), we favor a direct interaction between Ca²⁺ and GTP leading to the observed LD signal. In each case where we have detected the signature LD signal for the alternative GTP conformation, we have also detected polymer bundling. We therefore conclude that the presence of GTP in the alternative (macrochelate-like) conformation is linked to fiber bundling. A second link, between bundling and a decrease in GTPase, still holds and it seems likely that the cause of the reduction of the GTPase is the adoption of the alternative conformation of GTP bound to FtsZ.

This model has allowed us to make and test a prediction. If the alternative GTP conformation is similar to the macrochelate structure and if the formation of this structure is important in bundling, then an agent that induces the macrochelate conformation more efficiently than Ca²⁺ or Mg²⁺ would be a more potent bundler of FtsZ polymers. Cu²⁺ is the most potent macrochelate-inducing metal ion so far studied (>90% of GTP-Cu²⁺ is in the macrochelate conformation in solution) (54). We, therefore, tested the effect of adding CuCl₂ to standard FtsZ polymerization reactions. Our results showed that Cu²⁺ induces the formation of FtsZ polymer bundles at much lower concentrations (down to 0.1 mM) than Ca²⁺ (Fig. 5f). Incubation of FtsZ with CuCl₂ at higher concentrations (10 mM) resulted in extensive protein aggregation as reported previously for copper and the other divalent cations Mn²⁺, Fe²⁺, Co²⁺, and Ni²⁺ (27). It is significant that Cu²⁺, Mn²⁺, Fe²⁺, Co²⁺, and Ni²⁺ all have more than three times the preponderance for forming the macrochelate than do either Ca²⁺ or Mg²⁺ (54). Furthermore, we find that FtsZ polymers formed in the presence of Cu²⁺ show the alternative GTP conformation signature LD signal at a lower concentration (0.1 mM) than either Mg²⁺ or Ca²⁺ (Fig. 6a, inset). This provides strong support for the proposal that formation of a macrochelate conformation of GTP bound to FtsZ leads to fiber bundling. It may also be considered that the effect of Cu²⁺ may be due to an interaction between the Cu²⁺and cysteine residues in the protein (a well characterized interaction in other systems). However, examination of the primary sequence of E. coli FtsZ shows the lack of any cysteines.

It should be noted that if bundling occurs due to the effect of cationic buffering, it would be expected that low concentrations of Cu²⁺ would induce little or no bundling. Because we find the converse to be true, we can rule out mechanism 1 as the sole driving force behind FtsZ polymer bundling.

Taken together these results indicate that FtsZ bundling is inherently linked to a conformational change in the GTP. The conditions under which this conformation is formed match closely to those that favor formation of the macrochelate structure of GTP-Me²⁺ in solution. The result of this conformational change in polymers formed using Mg²⁺ is to produce polymers with reduced GTP turnover that are, hence, less dynamic. This could be said to be consistent with mechanism 2 proposed above. However, since a number of situations where reduced GTPase activity does not lead to polymer bundling have been reported (26, 33, 55, 56), a reduction in dynamics must not in itself be responsible for bundle formation (conversely, it is interesting to note that bundling may always result in reduced dynamics, since all studies of FtsZ bundling where GTPase activity was tested report that activity decreases (23, 27, 33, 57, 58).) If FtsZ bundling is thought to be solely the result of a reduced GTPase activity, then it would also be expected that at saturation the equally catalytically inert cations Cu²⁺ and Ca²⁺ should have very similar effects on bundling. As we have seen, this is very much not the case. Thus, the increased bundling potency of the Cu²⁺ over the equally catalytically inert Ca²⁺ can be more easily explained by a secondary, catalytically unrelated effect of the alternative GTP conformation. Examination of the x-ray crystal structure of GDP-FtsZ shows the guanine moiety to be close to the surface of the protein. Therefore, we propose a model whereby the alternative on-enzyme conformation of GTP-M²⁺ creates a site for lateral fiber association near the GTP binding pocket (Fig. 7).

View larger version (16K): [in this window] [in a new window]   FIG. 7. Model for Ca2+-induced FtsZ bundling. Ca2+ binds to FtsZ, altering the bound conformation of GTP, leading to displacement of the phosphates from the active site and, hence, a reduction in GTPase activity. Flipping of the guanine moiety into the macrochelate conformation moves it to a more surface-exposed position, where it forms a new interface for bundling.

	FOOTNOTES

Both authors contributed equally to this work.

^¶ Funded by Syngenta.

A Wellcome Career Development Fellow.

A Medical Research Council Career Development Fellow. To whom correspondence should be addressed. Tel.: 44-121-414-5881; E-mail: T.R.Dafforn{at}bham.ac.uk.

¹ The abbreviations used are: LD, linear dichroism; MES, 4-morpholineethanesulfonic acid.

	ACKNOWLEDGMENTS

 
We thank Dr. Löwe (Cambridge) for the use of the FtsZ protofiber model.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Margolin, W. (2000) FEMS Microbiol. Rev. 24, 531–548[CrossRef][Medline] [Order article via Infotrieve]
2. Lutkenhaus, J., and Addinall, S. G. (1997) Annu. Rev. Biochem. 66, 93–116[CrossRef][Medline] [Order article via Infotrieve]
3. Addinall, S. G., and Holland, B. (2002) J. Mol. Biol. 318, 219–236[CrossRef][Medline] [Order article via Infotrieve]
4. Chen, J. C., and Beckwith, J. (2001) Mol. Microbiol. 42, 395–413[CrossRef][Medline] [Order article via Infotrieve]
5. Hale, C. A., and de Boer, P. A. J. (2002) J. Bacteriol. 184, 2552–2556[Abstract/Free Full Text]
6. Ma, X., Ehrhardt, D. W., and Margolin, W. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 12998–13003[Abstract/Free Full Text]
7. Addinall, S. G., and Lutkenhaus, J. (1996) J. Bacteriol. 178, 7167–7172[Abstract/Free Full Text]
8. Vaughan, S., Wickstead, B., Gull, K., and Addinall, S. G. (2004) J. Mol. Evol. 58, 19–29[CrossRef][Medline] [Order article via Infotrieve]
9. Glockner, F. O., Kube, M., Bauer, M., Teeling, H., Lombardot, T., Ludwig, W., Gade, D., Beck, A., Borzym, K., Heitmann, K., Rabus, R., Schlesner, H., Amann, R., and Reinhardt, R. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 8298–8303[Abstract/Free Full Text]
10. Miyagishima, S.-Y., Nishida, K., and Kuroiwa, T. (2003) Trends Plant Sci. 8, 432–438[CrossRef][Medline] [Order article via Infotrieve]
11. Osteryoung, K. W., and Nunnari, J. (2003) Science 302, 1698–1704[Abstract/Free Full Text]
12. Gilson, P. R., Yu, X.-C., Hereld, D., Barth, C., Savage, A., Kiefel, B. R., Lay, S., Fisher, P. R., Margolin, W., and Beech, P. L. (2003) Eukaryot. Cell 2, 1315–1326[Abstract/Free Full Text]
13. Romberg, L., and Mitchison, T. J. (2004) Biochemistry 43, 282–288[Medline] [Order article via Infotrieve]
14. Romberg, L., and Levin, P. A. (2003) Annu. Rev. Microbiol. 57, 125–154[CrossRef][Medline] [Order article via Infotrieve]
15. Erickson, H. P. (1998) Trends Cell Biol. 8, 133–137[CrossRef][Medline] [Order article via Infotrieve]
16. Scheffers, D. J., de Wit, J. G., den Blaauwen, T., and Driessen, A. J. (2002) Biochemistry 41, 521–529[CrossRef][Medline] [Order article via Infotrieve]
17. Lowe, J., and Amos, L. A. (1998) Nature 391, 203–206[CrossRef][Medline] [Order article via Infotrieve]
18. Lowe, J., and Amos, L. A. (1999) EMBO J. 18, 2364–2371[CrossRef][Medline] [Order article via Infotrieve]
19. Diaz, J. F., Kralicek, A., Mingorance, J., Palacios, J. M., Vicente, M., and Andreu, J. M. (2001) J. Biol. Chem. 276, 17307–17315[Abstract/Free Full Text]
20. Lu, C., Reedy, M., and Erickson, H. P. (2000) J. Bacteriol. 182, 164–170[Abstract/Free Full Text]
21. Romberg, L., Simon, M., and Erickson, H. P. (2001) J. Biol. Chem. 276, 11743–11753[Abstract/Free Full Text]
22. Lowe, J., and Amos, L. A. (2000) Biol. Chem. 381, 993–999[CrossRef][Medline] [Order article via Infotrieve]
23. Gueiros-Filho, F. J., and Losick, R. (2002) Genes Dev. 16, 2544–2556[Abstract/Free Full Text]
24. Oliva, M. A., Huecas, S., Palacios, J. M., Martin-Benito, J., Valpuesta, J. M., and Andreu, J. M. (2003) J. Biol. Chem. 278, 33562–33570[Abstract/Free Full Text]
25. Erickson, H. P., Taylor, D. W., Taylor, K. A., and Bramhill, D. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 519–523[Abstract/Free Full Text]
26. Mukherjee, A., and Lutkenhaus, J. (1998) EMBO J. 17, 462–469[CrossRef][Medline] [Order article via Infotrieve]
27. Yu, X. C., and Margolin, W. (1997) EMBO J. 16, 5455–5463[CrossRef][Medline] [Order article via Infotrieve]
28. Hale, C. A., Rhee, A. C., and de Boer, P. A. (2000) J. Bacteriol. 182, 5153–5166[Abstract/Free Full Text]
29. Gonzalez, J. M., Jimenez, M., Velez, M., Mingorance, J., Andreu, J. M., Vicente, M., and Rivas, G. (2003) J. Biol. Chem. 278, 37664–37671[Abstract/Free Full Text]
30. Addinall, S. G., Cao, C., and Lutkenhaus, J. (1997) J. Bacteriol. 179, 4277–4284[Abstract/Free Full Text]
31. Stricker, J., Maddox, P., Salmon, E. D., and Erickson, H. P. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 3171–3175[Abstract/Free Full Text]
32. Trusca, D., Scott, S., Thompson, C., and Bramhill, D. (1998) J. Bacteriol. 180, 3946–3953[Abstract/Free Full Text]
33. Mukherjee, A., and Lutkenhaus, J. (1999) J. Bacteriol. 181, 823–832[Abstract/Free Full Text]
34. Scheffers, D. J., den Blaauwen, T., and Driessen, A. J. (2000) Mol. Microbiol. 35, 1211–1219[CrossRef][Medline] [Order article via Infotrieve]
35. Rivas, G., Fernandez, J. A., and Minton, A. P. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 3150–3155[Abstract/Free Full Text]
36. Small, E., and Addinall, S. G. (2003) Microbiology 149, 2235–2242[Abstract/Free Full Text]
37. Mukherjee, A., Cao, C., and Lutkenhaus, J. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 2885–2890[Abstract/Free Full Text]
38. Sivasothy, P., Dafforn, T. R., Gettins, P. G. W., and Lomas, D. A. (2000) J. Biol. Chem. 275, 33663–33668[Abstract/Free Full Text]
39. Rodger, A., Rajendra, J., Marrington, R., Ardhammar, M., Nordén, B., Hirst, J. B., Gilbert, A. T. B., Dafforn, T. R., Halsall, D. J., Woolhead, C. A., Robinson, C., Pinheiro, T. J., J., K., Seymour, M., Perez, N., and Hannon, M. J. (2002) Phys. Chem. Chem. Phys. 4, 4051–4057[CrossRef]
40. Rodger, A., Rajendra, J., Marrington, R., Mortimer, R., Andrews, T., Hirst, J. B., Gilbert, A. T. B., Halsall, D. J., Dafforn, T. R., Ardhammar, M., Nordén, B., Woolhead, C. A., Robinson, C., Pinheiro, T., J., K., Seymour, M., Perez, N., and Hannon, M. J. (2002) in Biophysical Chemistry: Membranes and Proteins (Templer, R. H., ed) pp. 3–19, Royal Society of Chemistry, Cambridge, UK
41. Dafforn, T. R., Rajendra, J., Halsall, D. J., Serpell, L. C., and Rodger, A. (2004) Biophys. J. 86, 404–410[Medline] [Order article via Infotrieve]
42. Marrington, R., Dafforn, T. R., Halsall, D. J., Andrews, T., and Rodger, A. (2004) Biophys. J. 87, 2002–2012[CrossRef][Medline] [Order article via Infotrieve]
43. Mukherjee, A., and Lutkenhaus, J. (1998) Methods Enzymol. 298, 296–305[Medline] [Order article via Infotrieve]
44. Lu, C., and Erickson, H. P. (1998) Methods Enzymol. 298, 305–313[Medline] [Order article via Infotrieve]
45. Pandya, M. J., Spooner, G. M., Sunde, M., Thorpe, J. R., Rodger, A., and Woolfson, D. N. (2000) Biochemistry 39, 8728–8734[CrossRef][Medline] [Order article via Infotrieve]
46. Nordh, J., Deinum, J., and Norden, B. (1986) Eur. Biophys. J. 14, 113–122[Medline] [Order article via Infotrieve]
47. Lockhart, A., and Kendrick-Jones, J. (1998) FEBS Lett. 430, 278–282[CrossRef][Medline] [Order article via Infotrieve]
48. Sossong, T. M., Jr., Brigham-Burke, M. R., Hensley, P., and Pearce, K. H., Jr. (1999) Biochemistry 38, 14843–14850[CrossRef][Medline] [Order article via Infotrieve]
49. Miki, M., and Mihashi, K. (1976) Biophys. Chem. 6, 101–106[Medline] [Order article via Infotrieve]
50. Lu, C., Stricker, J., and Erickson, H. P. (1998) Cell Motil. Cytoskeleton 40, 71–86[CrossRef][Medline] [Order article via Infotrieve]
51. Clark, L. B. (1977) J. Am. Chem. Soc. 99, 3934–3938[Medline] [Order article via Infotrieve]
52. Zaloudek, F., Novros, J. S., and Clark, L. B. (1984) J. Am. Chem. Soc. 107, 7344–7351
53. Rodger, A., and Nordén, B. (1987) Circular and Linear Dichroism, Oxford University Press, Oxford
54. Sigel, H., Bianchi, E. M., Corfu, N. A., Kinjo, Y., Tribolet, R., and Martin, R. B. (2001) Chemistry 7, 3729–3737[CrossRef][Medline] [Order article via Infotrieve]
55. Lu, C., Stricker, J., and Erickson, H. P. (2001) BMC Microbiol. 1, 7[CrossRef][Medline] [Order article via Infotrieve]
56. White, E. L., Ross, L. J., Reynolds, R. C., Seitz, L. E., Moore, G. D., and Borhani, D. W. (2000) J. Bacteriol. 182, 4028–4034[Abstract/Free Full Text]
57. Beuria, T. K., Krishnakumar, S. S., Sahar, S., Singh, N., Gupta, K., Meshram, M., and Panda, D. (2003) J. Biol. Chem. 278, 3735–3741[Abstract/Free Full Text]
58. Santra, M. K., Beuria, T. K., Banerjee, A., and Panda, D. (2004) J. Biol. Chem. 279, 25959–25965[Abstract/Free Full Text]
59. Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, 2714–2723[CrossRef][Medline] [Order article via Infotrieve]
60. Johnson, W. C., Guex, N., and Peitsch, M. C. (1999) Proteins 35, 307–312[CrossRef][Medline] [Order article via Infotrieve]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				S. G. Addinall, E. Small, D. Whitaker, S. Sturrock, W. D. Donachie, and M. M. Khattar New Temperature-Sensitive Alleles of ftsZ in Escherichia coli J. Bacteriol., January 1, 2005; 187(1): 358 - 365. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 279/47/48821    most recent M404944200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Marrington, R. Articles by Addinall, S. G. Search for Related Content PubMed PubMed Citation Articles by Marrington, R. Articles by Addinall, S. G. Social Bookmarking                      What's this?