Tài liệu Báo cáo khoa học: ATPase activity of magnesium chelatase subunit I is required to maintain subunit D in vivo ppt

ATPase activity of magnesium chelatase subunit I is required to maintain subunit D in vivo Vanessa Lake 1,2 , Ulf Olsson 2 , Robert D. Willows 1 and Mats Hansson 2 1 Department of Biological Science, Macquarie University, North Ryde, Australia; 2 Department of Biochemistry, Lund University, Sweden During biosynthesis of chlorophyll, Mg 2+ is inserted into protoporphyrin IX by magnesium chelatase. This enzyme consists of three different subunits of  40, 70 and 140 kDa. Seven barley mutants deficient in the 40 kDa magnesium chelatase subunit were analysed and it was found that this subunit is essential for the maintenance of the 70 kDa subunit, but not the 140 kDa subunit. The 40 kDa subunit has been shown to belong to the family of proteins called ÔATPases associated with various cellu- lar activitiesÕ, known to form ring-shaped oligomeric complexes working as molecular chaperones. Three of the seven barley mutants are semidominant mis-sense muta- tions leading to changes of conserved amino acid residues in the 40 kDa protein. Using the Rhodobacter capsulatus 40 and 70 kDa magnesium chelatase subunits we have analysed the effect of these mutations. Although having no ATPase activity, the deficient 40 kDa subunit could still associate with the 70 kDa protein. The binding was dependent on Mg 2+ and ATP or ADP. Our study dem- onstrates that the 40 kDa subunit functions as a chaperon that is essential for the survival of the 70 kDa subunit in vivo. We conclude that the ATPase activity of the 40 kDa subunit is essential for this function and that binding between the two subunits is not sufficient to maintain the 70 kDa subunit in the cell. The ATPase deficient 40 kDa proteins fail to participate in chelation in a step after the association of the 40 and 70 kDa subunits. This step presumably involves a conformational change of the complex in response to ATP hydrolysis. Keywords: AAA; barley; chlorophyll; magnesium chela- tase; Rhodobacter capsulatus. The first unique enzymatic reaction of the (bacterio)chloro- phyll biosynthetic pathway is the insertion of Mg 2+ into protoporphyrin IX. Three different polypeptides participate in the catalytic reaction and these constitute the subunits of magnesium chelatase (Fig. 1). The subunits are designated BchI, BchD and BchH in bacteriochlorophyll-synthesizing organisms such as Rhodobacter and Chlorobium, while in plants, algae and cyanobacteria, the homologous proteins are generally named ChlI, ChlD and ChlH [1]. The average molecular masses of BchI/ChlI, BchD/ChlD and BchH/ ChlH are 40, 70 and 140 kDa, respectively. The largest subunit is red upon purification due to bound protopor- phyrin IX [2–4] and binding studies of deuteroporphyrin IX to the H-subunit show a K d value of 0.53–1.2 l M [5]. The large subunit has therefore been suggested to be the catalytic subunit. The exact role of the other two subunits is not understood. It is known that they form a complex in the presence of Mg 2+ and ATP [2,3,6,7]. The complex forma- tion does not require hydrolysis of ATP, as ADP and nonhydrolysable ATP analogues (but not AMP) allowed complex formation [8]. It is clear, however, that the overall magnesium chelatase reaction requires ATP hydrolysis. The observations are consistent with earlier observations with pea magnesium chelatase where the magnesium chelatase reaction was demonstrated to be a two-step reaction, consisting of an activation step followed by the actual Mg 2+ insertion step [9]. The activation step could proceed with ATP-c-S, whereas ATP was required for the chelation. The three-dimensional structure of the Rhodobacter capsulatus BchI has recently been determined and it was found to belong to the large family of ÔATPases associated with various cellular activitiesÕ, or AAA + proteins [10]. AAA + proteins are important mechanoenzymes that transform chemical energy into biological events and they are usually found in various multimeric states [11,12]. They play essential roles in a broad range of cellular activities, including DNA replication, membrane fusion, cytoskeletal regulation, protein folding and proteolysis [11–13], and now also in porphyrin metallation [10]. The N-terminal half of the BchD subunit is homologous to BchI, while the C-terminal half includes a metal ion coordination motif characteristic for integrin I domains [10]. Integrins are known to participate in cell–matrix and cell–cell interactions [14,15]. They are involved in signalling to and from cells in various physiological processes, including morphogenesis, cell migration, immunity and wound healing [16,17]. The proposed models for the magnesium chelatase reaction all involve an Mg 2+ - and ATP-dependent complex formation of the 40 and 70 kDa subunits. In a subsequent step the complex triggers the Mg 2+ insertion into protoporphyrin by the 140 kDa subunit [2,4,18–20]. Our present model for Correspondence to M. Hansson, Department of Biochemistry, Lund University, Box 124, S-22100 Lund, Sweden. Fax: + 46 46 2224534, Tel.: + 46 46 2220105, E-mail: mats.hansson@biokem.lu.se Abbreviations: AAA + proteins, ATPases associated with various cellular activities. (Received 1 December 2003, revised 20 February 2004, accepted 2 April 2004) Eur. J. Biochem. 271, 2182–2188 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04143.x the magnesium chelatase reaction mechanism also takes into account the structural data of the R. capsulatus BchI and general functional aspects of AAA + proteins [10]. In this model not only the BchI proteins are organized in an AAA + hexamer, but also BchD as the amino acid sequence of the BchD N-terminal half is homologous to BchI. The interac- tions between the BchI and BchD proteins are suggested to occur via three b-hairpin elements, which protrude from the core of the BchI structure and which do not belong to the traditional structure of an AAA + protein. In the double- ringed BchI-BchD structure the ATPase activity of BchI is blocked. A conformational transition upon binding to BchH may bring the integrin I domain of BchD into contact with the integrin-binding motif of BchH, simultaneously triggering porphyrin metallation. This would also lead to a release of the blockade of the ATP-binding site of BchI by the integrin I domain, triggering ATP hydrolysis [10]. It is still an open question which subunit provides the Mg 2+ to be inserted into protoporphyrin, as all three subunits have some relationship with Mg 2+ . Concerning the 40 kDa subunit, it is well-known that Mg 2+ is required to perform the ATP hydrolysis. In addition, kinetic analysis had shown binding of Mg 2+ to this subunit [21]. An integrin I domain, suggested to exist in the C-terminus of the 70 kDa subunit, is a metal binding site (MIDAS motif) that can be expected to bind Mg 2+ [10]. The 140 kDa subunit shows consider- able sequence homology to the CobN subunit of the aerobic cobaltochelatase. CobN binds both the Co 2+ and the hydrogenobyrinic acid a,c-diamide substrate [22]. It can therefore be expected that the 140 kDa magnesium chela- tase subunit, in analogy with CobN, binds the two substrates of the magnesium chelatase reaction. Several barley (Hordeum vulgare L) mutants deficient in magnesium chelatase activity were isolated during the 1950s andreferredtoastheXantha-f, -g and -h loci [23]. It is now known that the 40 kDa subunit is encoded by Xantha-h,the 70 kDa subunit by Xantha-g and the 140 kDa subunit by Xantha-f [24]. The mutations are all lethal. Among the known seven mutant alleles of the Xantha-h gene encoding the smallest subunit of barley magnesium chelatase (corresponding to R. capsulatus BchI), four are recessive (xantha-h 30 , -h 38 , -h 56 and -h 57 ) and three are semidominant (Xantha-h clo 125 , -h clo 157 and -h clo 161 ). The homozygous mutant plants are all yellow and lack chlorophyll. On the other hand, the heterozygous mutants carrying the recessive allele are all green and indistinguishable from the wild-type plants. In contrast the heterozygous plants carrying the semidominant allele are pale green. It has been shown that the recessive mutations prevent transcription of the Xantha-h gene [24], while the semidominant alleles are mis-sense mutations leading to changes of single amino acid residues [25]. The mis-sense mutations have previously been con- structed in the corresponding gene, bchI,ofR. capsulatus [26]. The amino acid exchanges in the three mutants are D207N, R289K and L111F (numbered according to R. capsulatus BchI). These three residues are close to the ATP-binding site located at the interface between two BchI subunits in a presumed oligomeric ring [10]. The mutations D207N and R289K are located at one side of the ATPase active site, while L111F is found at the opposite side at the neighbouring subunit. The deficient BchI proteins also showed a dominant effect in vitro with respect to magnesium chelatase activity. In contrast, they were recessive with respect to the ATPase activity, but could still associate in oligomeric complexes with themselves as well as with wild- type BchI [26]. It was concluded that an intact BchI oligomer is required to support magnesium chelation, whereas ATP hydrolysis is achieved by autonomously working BchI subunit interfaces. In the present work we have expanded the study of the BchI subunits with the exchanges D207N, R289K and L111F, and analysed their ability to interact with BchD. The ATPase-deficient BchI proteins provide a tool to dissect the interaction between BchI and BchD and rule out the importance of the ATPase activity in this process. Although the ATPase-deficient 40 kDa BchI subunits can bind the 70 kDa BchD protein in vitro, our in vivo analysis show that the barley 70 kDa subunit is absent in homo- zygous mutants of the Xantha-h gene encoding the 40 kDa protein. Materials and methods Biological material Barley wild-type (cv. Svalo ¨ f’s Bonus) and barley magnesium chelatase mutants [23] were grown in moist vermiculite at 20 °C in 12 h dark/light cycles for 8 days. Lights were turned on at 07:00 h. Yellow homozygous mutant leaves were sorted from green wild-type leaves and put in liquid nitrogen. Total barley protein was isolated from frozen leaves according to [27]. Recombinant R. capsulatus BchI and BchD magnesium chelatase subunits were used in the study. The BchD protein was expressed as a His-tagged fusion protein. The BchI and BchD proteins were produced and purified as described previously [4]. Ni-affinity chromatography The Ni-affinity chromatography system of Novagen was used to immobilize the His-tagged BchD. Ni 2+ was bound to 1 mL HiTrap Ni-affinity columns (Pharmacia). The wash buffer contained 20 m M imidazole instead of the recommended 60 m M . Four separate columns were used for the interaction analysis of His-tagged BchD with the three BchI mutant proteins and the BchI wild-type. Fig. 1. The reaction catalyzed by magnesium chelatase. The insertion of Mg 2+ into protoporphyrin IX is the first unique reaction of the chlorophyll biosynthetic pathway. The reaction requires ATP hydro- lysis and is catalyzed by magnesium chelatase, which consists of three different subunits. Ó FEBS 2004 ATPase activity is required to maintain subunit D (Eur. J. Biochem. 271) 2183 SDS/PAGE and Western blot analysis SDS/PAGE [10% (w/v) acrylamide] was performed accord- ingtoFlingandGregerson[28]withtheTris/Tricinebuffer system of Scha ¨ gger and von Jagow [29]. SDS/PAGE loading buffer consisted of 200 m M Tris/HCl pH 8.8, 20% (v/v) glycerol, 4% (w/v) SDS, 200 m M dithiothreitol and 0.01% (w/v) Bromophenol blue. Proteins on SDS/ PAGE were visualized by staining with colloidal Coomassie Brilliant Blue G-250 [30]. For Western blot analysis, 5 lgof total protein was separated on SDS/PAGE and electro- transferred to Immobilon P filters (Millipore) according to Towbin et al. [31] using a semidry electroblotter. Polyclonal antibodies against the three barley magnesium chelatase subunits were from rabbit. Goat anti-rabbit IgG conjugated to alkaline phosphatase was used as secondary antibody. Antigens on filters were visualized using a chemilumines- cence detection system (Clontech Laboratory Inc.). Magnesium chelatase antisera Antibodies were produced against truncated His-tagged versions of the three barley magnesium chelatase subunits expressed from derivatives of plasmid pET15b. The plas- mids containing Xantha-f, -g and -h were named pAntF1:1, pAntG1 and pAntH, respectively. Plasmid pAntF1:1 con- tains 741 bp of the Xantha-f gene. The produced polypep- tide corresponds to amino acid residues E541 to E781 of the full length XAN-F polypeptide of 1381 amino acid residues (numbered according to [24]). Plasmid pAntG1 has an insert of 717 bp of genomic Xantha-g DNA and produces 54 residues of the C-terminal half of the XAN-G protein. TheXAN-GspecificresiduesAVRVGLNAEKSGDVG RIMIVAITDGRANVSLKKSNDPEAAAASDAPRPST QELK follow after the His-tag. Plasmid pAntH contains 749 bp of Xantha-h,  70% of the gene. The XAN-H- specific amino acid sequence of 239 residues starts with EVMGP after the His-tag and ends with DISTV. The fusion proteins were produced in Escherichia coli BL21(DE3) using the inducible T7 RNA polymerase system [32]. Cells from 1 L cultures were harvested and lysed by sonication. His-tagged magnesium chelatase polypeptides were purified from crude cell extracts accord- ing to Novagen. All buffers used for the purification of the XAN-G polypeptide had to contain 6 M urea to prevent the protein from precipitation. The proteins were desalted into 10 m M Na-phosphate pH 7.4, 150 m M NaCl and dispensedinto100lg aliquots of which four portions were given to the rabbit. The desalted XAN-G also contained 1 M urea. mRNA analysis The presence of Xantha-g mRNA was analysed by cDNA synthesis from total RNA. First, 1 lgtotalRNA,1lL dNTP (10 m M )and1lL oligo(dT) 15 (0.5 mgÆmL )1 )were mixed with water to a total volume of 10 lL followed by heating to 65 °C for 5 min. Then 4 lL5· first strand buffer, 4 lLMgCl 2 (25 m M )and2lL dithiothreitol (0.1 M )were added. After 2 min at 42 °C, 0.5 lL Superscript II reverse transcriptase (200 unitsÆlL )1 ; Life Technologies) was added followed by incubation at 42 °C for 50 min and 70 °Cfor 15 min. One microlitre of RNaseH was added and incuba- ted for 20 min at 37 °C. The synthesized first strand cDNA was used as template in a PCR amplification, where Xantha-g-specific primers were utilized. The primers EXgLp67 (5¢-CGTAGATACAAACTTGTTCTCGGT AT-3¢) and EXgUp70 (5¢-GCATTTATTCCCTTCCGTG GAGACT-3¢) are separated by two introns in the chromo- somal Xantha-g DNA. Therefore a DNA fragment ampli- fied from genomic DNA is 566 bp, whereas a fragment amplified from cDNA is 378 bp. The 50 lL reaction contained 2 lL first strand cDNA, 5 lL10· reaction buffer, 3 lLMgCl 2 (25 m M ), 0.5 lLdNTP(20m M ), 2 lL of each primer (10 l M )and0.5lL Taq DNA polymerase (5 unitsÆlL )1 ). Thirty-five cycles were per- formed: 94 °C, 30 s; 58 °C, 30 s; 68 °C, 40 s. After the PCR was completed the DNA fragments were analysed with agarose-gel electrophoresis and DNA sequencing. Results Presence of magnesium chelatase subunits in barley xantha-h mutants In a previous study of the 70 kDa barley XAN-G magnes- ium chelatase subunit, it was found that the XAN-G protein was missing in total cell extracts of the xantha-h 56 and xantha-h 57 mutants [33]. The two mutant plants are suggestedtolackexpressionoftheXAN-Hproteinasno Xantha-h mRNA could be detected in these strains [24]. In contrast, the 70 kDa ChlD protein of Arabidopsis thaliana accumulates to wild-type levels under conditions where no 40 kDa ChlI protein could be detected [34]. Our analysis was extended to all of the barley xantha-h mutants to determine if the lack of XAN-G is a general feature in these mutants. Crude cell extract was isolated from leaves of mutants grown in 12 h dark/light cycles for 7 days and the presence of XAN-G was analysed by Western blotting, using antibodies raised against the C-terminal half of the barley XAN-G protein. High amounts of XAN-G could only be detected in the wild-type. The seven xantha-h mutants probably lack XAN-G totally or contain only trace amounts of the protein (Fig. 2A). The XAN-H protein was found at wild-type level in the semidominant Xantha- h clo 125 ,-h clo 157 and -h clo 161 mutants (Fig. 2B), which have altered amino acid residues in their resulting 40 kDa protein. No XAN-H could be found in the recessive xantha-h 30 , -h 38 ,-h 56 and -h 57 mutants (Fig. 2B). This is in agreement with the lack of Xantha-h mRNA in these mutants [24]. The large 140 kDa XAN-F protein was not affected by the mutations and was detected in all of the seven xantha-h mutants (Fig. 2C). Binding of mutated BchI to BchD The barley Xantha-h clo 125 , -h clo 157 and -h clo 161 mutations have been constructed in the orthologous R. capsulatus 40 kDa magnesium chelatase subunit, BchI, in a previous study [26]. We analysed the ability of these BchI proteins, with the exchanges D207N, R289K and L111F, to bind tothe70kDaBchDproteinwithanN-terminalHis-tag. His-tagged proteins have affinity to immobilized Ni 2+ and usually remain bound to the column when it is washed 2184 V. Lake et al.(Eur. J. Biochem. 271) Ó FEBS 2004 with 60 m M imidazole-containing buffer. The His-tagged BchD, however, eluted at 60 m M imidazole and 20 m M imidazole-containing buffers had to be used in the wash steps (Fig. 3). In the experiment 50 lL of the BchI protein (3 mgÆmL )1 )in50m M Tricine/NaOH pH 8.0 was mixed with 50 lLof50m M Tricine/NaOH pH 8.0, 8 m M ATP, 8m M dithiothreitol, 30 m M MgCl 2 .Themixturewas addedto10lL His-tagged BchD (5 mgÆmL )1 in 50 m M Tricine/NaOH pH 8.0, 4 m M ATP, 4 m M dithiothreitol, 15 m M MgCl 2 ) and left on ice for 90 min. The resulting 110 lL were mixed with 4 mL binding-buffer (20 m M Tris/HCl pH 7.9, 0.5 M NaCl, 4 m M ATP, 15 m M MgCl 2 ) and loaded on a Ni 2+ -containing HiTrap Ni-affinity column equilibrated with binding-buffer. The column was washed with 8 mL wash-buffer (binding-buffer containing 20 m M imidazole) before bound proteins were eluted with 4 mL elute-buffer (20 m M Tris/HCl pH 7.9, 0.5 M NaCl, 1 M imidazole, 6 M urea). Two 2 mL fractions were collected from the run-through, four 2 mL fractions were collected from the wash and four 1 mL fractions were collected from the elute. The collected proteins were precipitated by addition of 100% (w/v) trichloroacetic acid to a final concentration of 20% (w/v). Imidazole at 1 M concentration in the elute-buffer inhibited trichloro- acetic acid precipitation, but the problem was overcome by including urea in the buffer. After a wash with acetone the precipitated proteins were dried and resolved in 200 lL SDS/PAGE loading buffer. Ten microlitres were analysed by SDS/PAGE followed by staining with colloidal Coomassie Brilliant Blue. Pure His-tagged BchD andpurewild-typeBchIwereloadedoneachgelto identify the proteins in the run-through, wash and elute fractions. The analysis showed that the His-tagged BchD bound to the column, as His-tagged BchD was only found in the elute fractions and not, or to very little extent, in the run-through and wash fractions. The three BchI proteins with the exchanges D207N, R289K and L111F, as well as the wild-type BchI, were found in the run-through fractions and the first wash fractions, but also in the elute fraction (Fig. 4A). The experiment was also performed without His-tagged BchD. The various BchI proteins probably show some affinity to the HiTrap Ni-affinity column (Fig. 4C). However, as the amount of BchI in the elute fractions were much higher when His- tagged BchD was present in the experiment we conclude that the four different BchI proteins can all bind to His- tagged BchD. Further experiments showed that the binding of wild-type BchI to His-tagged BchD was dependent on Mg 2+ and that ADP, but not AMP, could be used instead of ATP. The three modified BchI proteins could also bind to His-tagged BchD when ADP was used instead of ATP (Fig. 4B). Presence of Xantha-g mRNA in xantha-h mutants A possible explanation for the absence of 70 kDa XAN-G protein in the xantha-h mutants could be that the XAN-H protein affects the level of Xantha-g mRNA. Therefore, the presence of Xantha-g mRNA was analysed in one semidominant and one recessive xantha-h mutant (Xantha- h clo 157 and xantha-h 57 , respectively). First strand cDNA was synthesized from total RNA of the mutants. Total RNA of a wild-type strain grown in parallel with the mutants was used as a positive control. The first strand Fig. 3. Coomassie-stained SDS/polyacrylamide gels. The strength of His-tagged BchD binding to Ni 2+ immobilized on a HiTrap Ni-affinity column was analysed. (A) The column was washed with buffer containing 60 m M imidazole before being eluted with 1 M imidazole. (B) The wash buffer contained only 20 m M imidazole. The wash buffer containing 20 m M imidazole was used in the following experimentsas60m M imidazole elutes the His-tagged BchD from the column. W1, wash fraction 1; W2, wash fraction 2; W3, wash fraction 3; W4, wash fraction 4; E1, elute fraction 1; E2, elute fraction 2; E3, elute fraction 3. The arrows indicate the His-tagged BchD. Fig. 2. Western blot analysis. Analysis of magnesium chelatase sub- units XAN-G (70 kDa; A), XAN-H (40 kDa; B) and XAN-F (140 kDa; C) in barley wild-type (Wt) and mutants xantha-h 30 ,-h 38 , -h 56 ,-h 57 ,-h clo 125 (DN), -h clo 157 (RK) and -h clo 161 (LF). The arrows indicate the XAN-G, XAN-H and XAN-F antigens. Ó FEBS 2004 ATPase activity is required to maintain subunit D (Eur. J. Biochem. 271) 2185 cDNAwasthenusedinanordinaryPCRamplification and the resulting DNA fragments were isolated after agarose gel electrophoresis and analysed by DNA sequence analysis. The oligonucleotides used as primers are separated by two introns in the genomic DNA fragment. The expected size of a DNA fragment amplified from the cDNA was 378 bp, whereas the size of a DNA fragment amplified from genomic Xantha-g DNA was 566bp.DNAfragmentsof378bpcouldbeisolatedfrom the wild-type as well as the two mutants, demonstrating that the absence of XAN-G protein in the mutants cannot be explained by abnormalities at the transcriptional level (Fig. 5). Discussion The structural analysis of BchI clearly revealed it as an AAA + protein [10]. It is therefore logical to search for the function of BchI among the various functions of AAA + proteins. The AAA + proteins represent one type of molecular chaperones and their function is to control the fate of proteins or DNA. This is done by facilitating protein folding and unfolding, assembly and disassembly of protein complexes, degradation of protein, replication and tran- scription of DNA, etc. [11–13]. Here we found that mRNA encoding the 70 kDa XAN-G subunit is present in both a recessive and a semidominant barley xantha-h mutant. This is in accordance with an earlier study, where wild-type levels of Xantha-g mRNA were detected by Northern blot analysis in the four recessive xantha-h mutants[33].Itistherefore likely that the lack of XAN-G protein in xantha-h mutants is due to failure of the mutated 40 kDa XAN-H proteins to interact with XAN-G in a normal way. This protective interaction seems to be specific for the XAN-H protein because wild-type levels of XAN-G are found in eight available barley Xantha-f mutants deficient in the 140 kDa XAN-F magnesium chelatase subunit [35]. In the xantha- h 30 , -h 38 , -h 56 and -h 57 mutants the failure to maintain XAN-G is easily explained by the absence of XAN-H protein. In the Xantha-h clo 125 , -h clo 157 and -h clo 161 mutants, however, the lack of XAN-G has to be explained by an inhibited activity of the deficient XAN-H proteins. Inter- estingly, the recombinant R. capsulatus BchI proteins with exchanged amino acid residues orthologous to the Xantha- h clo 125 , -h clo 157 and -h clo 161 mutations could still bind to the His-tagged BchD protein. In addition, the binding of BchD Fig. 4. Coomassie-stained SDS/polyacrylamide gels. The ability of wild-type BchI and BchI with modifications D207N, L111F and R289K to bind His-tagged BchD was analysed in a so-called pull-down experiment. A new column was used for experiments with individual mutant samples. The columns were stripped and recharged with Ni 2+ between each experiment. (A) Buffers contained ATP. (B) Buffers contained ADP. (C) Control experiment without His-tagged BchD. Buffers contained ATP. W1, wash fraction 1; W2, wash fraction 2; W3, wash fraction 3; W4, wash fraction 4; E1, elute fraction 1; E2, elute fraction 2; R, run-through fraction 2. Fig. 5. Presence of Xantha-g mRNA in barley wild-type, the recessive mutant xantha-h 57 and the semidominant mutant Xantha-h clo 157 . First strand cDNA synthesis was performed with total RNA, followed by ordinary PCR. DNA fragments were separated with agarose gel electrophoresis and stained with ethidium bromide. The amplified 378 bp fragment indicates the presence of Xantha-g mRNA in all three strains tested. DNA fragments of 566 bp originate from the amplifi- cation of contaminating genomic DNA. 2186 V. Lake et al.(Eur. J. Biochem. 271) Ó FEBS 2004 showed the same dependence on ATP or ADP as the wild- type BchI. Previously, the mutant BchI proteins were found to associate in oligomeric complexes with themselves as well as with wild-type BchI [26]. Thus, the deficient BchI proteins interact in a similar way to wild-type BchI although they cannot contribute to magnesium chelation. The lack of ATPase activity is their major divergence. Our study demonstrates that the 40 kDa subunit is a chaperone that is essential for the survival of the 70 kDa subunit in vivo.We conclude that the ATPase activity of the 40 kDa subunit is essential for the function of this subunit as a chaperone and that binding of I to D is not enough to maintain the D subunit in the cell. Our study suggests that ATP hydrolysis is important for a mechanistic step after the formation of an ID complex. This is supported by studies performed with N-ethylmaleimide-treated 40 kDa ChlI subunit of Synechocystis [19]. Similarly to the effects of the barley Xantha-h clo 125 , -h clo 157 and -h clo 161 mutations studied here, N-ethylmaleimide treatment abolished ATP hydrolysis and magnesium chelatase activity, but still allowed complex formation between the 40 and the 70 kDa subunits. It should be noted that there are examples of AAA + proteins that might function without ATP hydrolysis [12]. On the other hand it has been shown for several AAA + proteins that a significant change of conformation occurs during the ATP hydrolysis cycle and it has been suggested that this may be a general feature of these proteins [12,36–41]. Further studies have to be performed to understand the reason and a possible function of the instability of the 70 kDa subunit. Obviously, the D subunit is a substrate of the I subunit, which led to questions concerning the assembly of the I and D building blocks into the suggested hexameric double-ring structure [10]. Several pathways are possible, among which are that (a) the I and D subunits first assemble into pure I and pure D hexamers, which then form an ID complex; (b) an I-hexamer is first formed, which then helps in the stepwise formation of a D-hexamer; (c) the hexameric double-ring is assembled in total random order, i.e. any combinations of pure I, pure D or mixed ID complexes are likely to exist. Acknowledgements Dr Salam Al-Karadaghi is acknowledged for fruitful discussions. This work was made possible thanks to generous support from the Swedish Research Council, the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning, and the Magn. Bergvall Foundation to M. H., and the Australian Research Council (grant no. A09905713) and an IREX award (grant no. X00001636) to R. D. W. References 1. Willows, R.D. & Hansson, M. (2003) Mechanism, Structure, and Regulation of Magnesium Chelatase in the Tetrapyrrole Handbook II, pp. 1–48. Academic Press, San Diego, CA. 2. Jensen, P.E., Gibson, L.C.D. & Hunter, C.N. (1998) Determinants of catalytic activity with the use of purified I, D and H subunits of the magnesium protoporphyrin IX chelatase from Synechocystis PCC6803. Biochem. J. 334, 335–344. 3. Willows, R.D., Gibson, L.C.D., Kanangara, C.G., Hunter, C.N. & von Wettstein, D. (1996) Three separate proteins constitute the magnesium chelatase of Rhodobacter sphaeroides. Eur. J. Biochem. 235, 438–443. 4. Willows, R.D. & Beale, S.I. (1998) Heterologous expression of the Rhodobacter capsulatus bchI-D,and-H genes that encode mag- nesium chelatase subunits and characterization of the recon- stituted enzyme. J. Biol. Chem. 273, 34206–34213. 5. Karger, G.A., Reid, J.D. & Hunter, C.N. (2001) Characterization of the binding of deuteroporphyrin IX to the magnesium chelatase H subunit and spectroscopic properties of the complex. Bio- chemistry 40, 9291–9299. 6. Papenbrock, J., Gra ¨ fe, S., Kruse, E., Ha ¨ nel, F. & Grimm, B. (1997) Mg-chelatase of tobacco: identification of a Chl D cDNA sequence encoding a third subunit, analysis of the interaction of the three subunits with the yeast two-hybrid system, and recon- stitution of the enzyme activity by co-expression of recombinant CHL D, CHL H and CHL I. Plant J. 12, 981–990. 7. Gibson, L.C.D., Jensen, P.E. & Hunter, C.N. (1999) Magnesium chelatase from Rhodobacter sphaeroides: initial characterization of the enzyme using purified subunits and evidence for a BchI-BchD complex. Biochem. J. 337, 243–251. 8. Jensen, P.E., Gibson, L.C.D. & Hunter, C.N. (1999) ATPase acti- vity associated with the magnesium-protoporphyrin IX chelatase enzyme of Synechocystis PCC6803: evidence for ATP hydrolysis during Mg 2+ insertion, and the MgATP-dependent interaction of the ChlI and ChlD subunits. Biochem. J. 339, 127–134. 9. Walker, C.J. & Weinstein, J.D. (1994) The magnesium-insertion step of chlorophyll biosynthesis is a two-stage reaction. Biochem. J. 299, 277–284. 10. Fodje, M.N., Hansson, A., Hansson, M., Olsen, J.G., Gough, S., Willows, R.D. & Al-Karadaghi, S. (2001) Interplay between an AAA module and an integrin I domain may regulate the function of magnesium chelatase. J. Mol. Biol. 311, 111–122. 11. Vale, R.D. (2000) AAA proteins: Lords of the ring. J. Cell Biol. 150, F13–F19. 12. Ogura, T. & Wilkinson, A.J. (2001) AAA + superfamily ATPases: common structure-diverse function. Genes Cells 6, 575–597. 13. Neuwald,A.F.,Aravind,L.,Spouge,J.L.&Koonin,E.V.(1999) AAA + : a class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res. 9, 27–43. 14. Hynes, R.O. (2002) Integrins: bidirectional, allosteric signaling machines. Cell 110, 673–687. 15. Emsley, J., Knight, C.G., Farndale, R.W., Barnes, M.J. & Liddington, R.C. (2000) Structural basis of collagen recognition by integrin a2b1. Cell 101, 47–56. 16. Bella, J. & Berman, H.M. (2000) Integrin-collagen complex: a metal-glutamate handshake. Structure Fold. Des. 8, R121–R126. 17. van der Flier, A. & Sonnenberg, A. (2001) Function and inter- actions of integrins. Cell Tissue Res. 305, 285–298. 18. Walker, C.J. & Willows, R.D. (1997) Mechanism and regulation of Mg-chelatase. Biochem. J. 327, 321–333. 19. Jensen,P.E.,Reid,J.D.&Hunter,C.N.(2000)Modificationof cysteine residues in the ChlI and ChlH subunits of magnesium chelatase results in enzyme inactivation. Biochem. J. 352, 435–441. 20. Gra ¨ fe,S.,Saluz,H.P.,Grimm,B.&Ha ¨ nel, F. (1999) Mg- chelatase of tobacco: the role of the subunit CHL D in the che- lation step of protoporphyrin IX. Proc. Natl Acad. Sci. USA 96, 1941–1946. 21. Reid, J.D., Siebert, C.A., Bullough, P.A. & Hunter, C.N. (2003) The ATPase activity of the ChlI subunit of magnesium chelatase and formation of a heptameric AAA + ring. Biochemistry 42, 6912–6920. 22. Debussche, L., Couder, M., Thibaut, D., Cameron, B., Crouzet, J. & Blanche, F. (1992) Assay, purification, and characterization of cobaltochelatase, a unique complex enzyme catalyzing cobalt insertion in hydrogenobyrinic acid a,c-diamide during coenzyme Ó FEBS 2004 ATPase activity is required to maintain subunit D (Eur. J. Biochem. 271) 2187 B 12 biosynthesis in Pseudomonas denitrificans. J. Bacteriol. 174, 7445–7451. 23. Henningsen, K.W., Boynton, J.E. & von Wettstein, D. (1993) Mutants at xantha and albina loci in relation to chloroplast bio- genesis in barley (Hordeum vulgare L.). The Royal Danish Acad. Sci. Lett., Copenhagen 42, 1–349. 24. Jensen, P.E., Willows, R.D., Petersen, B.L., Vothknecht, U.C., Stummann,B.M.,Kannangara,C.G.,vonWettstein,D.&Hen- ningsen, K.W. (1996) Structural genes for Mg-chelatase subunits in barley: Xantha-f -g and -h. Mol. Gen. Genet. 250, 383–394. 25. Hansson, A., Kannangara, C.G., von Wettstein, D. & Hansson, M. (1999) Molecular basis for semidominance of missense muta- tions in the XANTHA-H (42-kDa) subunit of magnesium chela- tase. Proc. Natl Acad. Sci. USA 96, 1744–1749. 26. Hansson, A., Willows, R.D., Roberts, T.H. & Hansson, M. (2002) Three semidominant barley mutants with single amino acid sub- stitutions in the smallest magnesium chelatase subunit form defective AAA + hexamers. Proc. Natl Acad. Sci. USA 99, 13944– 13949. 27. Hansson, M., Gough, S.P., Kannangara, C.G. & von Wettstein, D. (1997) Analysis of RNA and enzymes of potential importance for regulation of 5-aminolevulinic acid synthesis in the proto- chlorophyllide accumulating barley mutant tigrina-d 12 . Plant Physiol. Biochem. 35, 827–836. 28. Fling, S.P. & Gregerson, D.S. (1986) Peptide and protein molecular weight determination by electrophoresis using a high-molarity Tris buffer system without urea. Anal. Biochem. 155, 83–88. 29. Scha ¨ gger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379. 30. Neuhoff, V., Arold, N., Taube, D. & Ehrhardt, W. (1988) Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis 9, 255–262. 31. Towbin, H., Staehelin, T. & Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl Acad. Sci. USA 76, 4350–4354. 32. Studier, F.W. & Moffatt, B.A. (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189, 113–130. 33. Petersen, B.L., Møller, M.G., Jensen, P.E. & Henningsen, K.W. (1999) Identification of the Xan-g gene and expression of the Mg- chelatase encoding genes Xan-f -g and -h in mutant and wild-type barley (Hordeum vulgare L.). Hereditas 131, 165–170. 34. Rissler, H.M., Collakova, E., DellaPenna, D., Whelan, J. & Pogson, B.J. (2002) Chlorophyll biosynthesis. Expression of a second Chl I gene of magnesium chelatase in Arabidopsis supports only limited chlorophyll synthesis. Plant Physiol. 128, 770–779. 35. Olsson, U., Sirijovski, N. & Hansson, M. (2004) Characterization of eight barley xantha-f mutants, deficient in magnesium chelatase. Plant Physiol. Biochem.,inpress. 36. Hanson,P.I.,Roth,R.,Morisaki,H.,Jahn,R.&Heuser,J.E. (1997) Structure and conformational changes in NSF and its membrane receptor complexes visualized by quick-freeze/deep- etch electron microscopy. Cell 90, 523–535. 37. Rouiller, I., Butel, V.M., Latterich, M., Milligan, R.A. & Wilson- Kubalek, E.M. (2000) A major conformational change in p97 AAA ATPase upon ATP binding. Mol. Cell 6, 1485–1490. 38. Rouiller, I., DeLaBarre, B., May, A.P., Weis, W.I., Brunger, A.T., Milligan, R.A. & Wilson-Kubalek, E.M. (2002) Conformational changes of the multifunction p97 AAA ATPase during its ATPase cycle. Nat. Struct. Biol. 9, 950–957. 39. Zhang,X.,Shaw,A.,Bates,P.A.,Newman,R.H.,Gowen,B., Orlova, E., Gorman, M.A., Kondo, H., Dokurno, P., Lally, J., Leonard, G., Meyer, H., van Heel, M. & Freemont, P.S. (2000) Structure of the AAA ATPase p97. Mol. Cell 6, 1473–1484. 40.Song,H.K.,Hartmann,C.,Ramachandran,R.,Bochtler,M., Behrendt, R., Moroder, L. & Huber, R. (2000) Mutational studies on HslU and its docking mode with HslV. Proc.NatlAcad.Sci. USA 97, 14103–14108. 41. Lenzen, C.U., Steinmann, D., Whiteheart, S.W. & Weis, W.I. (1998) Crystal structure of the hexamerization domain of N-ethylmaleimide-sensitive fusion protein. Cell 94, 525–536. 2188 V. Lake et al.(Eur. J. Biochem. 271) Ó FEBS 2004 . Mg 2+ is required to perform the ATP hydrolysis. In addition, kinetic analysis had shown binding of Mg 2+ to this subunit [21]. An integrin I domain, suggested. could still bind to the His-tagged BchD protein. In addition, the binding of BchD Fig. 4. Coomassie-stained SDS/polyacrylamide gels. The ability of wild-type