Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M610479200 on December 6, 2006

J. Biol. Chem., Vol. 282, Issue 6, 3809-3818, February 9, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
282/6/3809    most recent
M610479200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Whitney, S. M.
Right arrow Articles by Sharwood, R. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Whitney, S. M.
Right arrow Articles by Sharwood, R. E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Linked Rubisco Subunits Can Assemble into Functional Oligomers without Impeding Catalytic Performance*Formula

Spencer M. Whitney1 and Robert E. Sharwood

From the Molecular Plant Physiology, Research School of Biological Sciences, Australian National University, P. O. Box 475, Canberra, Australian Capital Territory 2601, Australia

Received for publication, November 10, 2006 , and in revised form, December 6, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Although transgenic manipulation in higher plants of the catalytic large subunit (L) of the photosynthetic CO2-fixing enzyme ribulose 1,5-bisphospahte carboxylase/oxygenase (Rubisco) is now possible, the manipulation of its cognate small subunit (S) is frustrated by the nuclear location of its multiple gene copies. To examine whether L and S can be engineered simultaneously by fusing them together, the subunits from Synechococcus PCC6301 Rubisco were tethered together by different linker sequences, producing variant fusion peptides. In Escherichia coli the variant PCC6301 LS fusions assembled into catalytically functional octameric ([LS]8) and hexadecameric ([[LS]8]2) quaternary structures that excluded the integration of co-expressed unfused S. Assembly of the LS fusions into Rubisco complexes was impaired 50–90% relative to the assembly of unlinked L and S into L8S8 enzyme. Assembly in E. coli was not emulated using tobacco SL fusions that accumulated entirely as insoluble protein. Catalytic measurements showed the CO2/O2 specificity, carboxylation rate, and Michaelis constants for CO2 and ribulose 1,5-bisphosphate for the cyanobacterial Rubisco complexes comprising fusions where the S was linked to the N terminus of L closely matched those of the wild-type L8S8 enzyme. In contrast, the substrate affinities and carboxylation rate of the Rubisco complexes comprising fusions where L was fused to the N terminus of S or a six-histidine tag was appended to the C terminus of L were compromised. Overall this work provides a framework for implementing an alternative strategy for exploring simultaneous engineering of modified, or foreign, Rubisco L and S subunits in higher plant plastids.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The catalytic efficiency of the photosynthetic CO2-fixing enzyme D-ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco, EC 4.1.1.39 [EC] ) has a central role in determining how efficiently plants use their resources of water, fertilizer nutrient, and light (1–3). As the primary port of entry of inorganic carbon into the biosphere, it is surprising that plant Rubisco is not very efficient, particularly at limiting CO2 concentrations where the rate of catalytic turnover is less than one-thousandth that of many other plant enzymes (4). The tendency of Rubiscos to confuse CO2 with the more abundant atmospheric gas, O2, encumbers photosynthesis in higher plants with both a requirement to invest large amounts of protein in Rubisco and also a requirement for an energy-intensive photorespiratory metabolism to recycle the oxygenated waste product. However, higher plant Rubisco is not the pinnacle of evolution as more efficient and specific forms are found in nature, particularly in a number of red algae (5). The benefits of replacing the Rubisco in C3 crops with these natural variants have been modeled and show substantial improvements that come at no additional energy or resource cost (1, 2, 6). Against this background, it is no surprise efforts are continuing into understanding Rubisco's catalytic mechanism, how it has evolved, and how improvements in its efficiency might be engineered by mutating it or replacing it in plants with more efficient homologs (7–10).

Engineering Rubisco in higher plants is complicated by the separate locations of the genes coding for the large (L) and small (S)2 subunits and the complex assembly mechanism that necessitates the coordinated expression, post-translational modifications, and assembly of both subunits into a hexadecamer (L8S8) within the chloroplast stroma (Fig. 1) (11–13). The L subunit contains the catalytic site and the S subunits, whose precise role in the structure and function of Rubisco remains poorly understood and which are essential for catalytic viability (7, 8, 14). Indeed, when stripped of S or assembled with foreign S, the catalytic properties of Rubiscos are substantially impaired. Genetically engineering Rubisco in plastids, therefore, needs to attend to both L and S. For the Rubisco L subunit gene (rbcL) located in the plastid genome (plastome), genetic manipulation by plastome transformation in tobacco is a routine but protracted process requiring typically 4–12 months to obtain mature homoplasmic transformants (15–19). In contrast, an appropriate means for engineering the native (or foreign) S subunit genes (RbcS) in higher plants has remained an elusive challenge due to the multiple RbcS copies in higher plant nuclei that essentially precludes them from targeted mutagenic or replacement strategies. Moreover, attempts to incorporate recombinant S into higher plants by transplastomic methods have highlighted how circumventing the assembly of cytosolic-synthesized S with plastid synthesized L into L8S8 complexes is almost immutable unless the endogenous levels of S have been substantially reduced by antisense (17, 20, 21). Resilient expression of the native S is also problematic when transplanting in foreign L subunits, as highlighted by the transplastomic replacement of the tobacco rbcL with sunflower rbcL that produced tobacco-sunflower transformants whose hybrid sunflower L8-tobacco S8 enzyme was kinetically impaired and unable to support autotrophic growth in air (18).

Here a novel strategy for simultaneously engineering Rubisco S and L subunits was examined using the Synechococcus sp. PCC6301 enzyme that, unlike higher plant Rubiscos, is coded by a single rbcL-rbcS operon and can be functionally expressed in Escherichia coli (12, 22–25). Using different peptide linkers, the PCC6301 L and S were tethered together to produce an array of SL and LS fusion peptides. Presented here are results that show the subunit fusions can correctly fold and assemble into functional Rubisco oligomers in E. coli and that unlinked S is excluded from assembly. Moreover the kinetic properties of the Rubisco oligomers were evaluated and found that the catalytic prowess of most fusion-Rubiscos closely mimicked the wild-type PCC6301 L8S8 enzyme.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Materials—Unlabeled and carboxyl-14C-labeled carboxypentitol-P2 was synthesized as described (26). Ribulose-P2 was synthesized and purified as described (27). All cloned DNA sequences were fully sequenced using BigDye terminator sequencing (Applied Biosystems) on an ABI 3730 sequencer (Biomolecular Resource Facility, The John Curtin School of Medical Research, Australian National University) following the manufacturer's protocol (Applied Biosystems).

Amplification of Linker Genes—A gene coding for the linker-60 peptide was assembled by splice overlap extension using successive PCR reactions with Herculase Enhanced DNA polymerase (Stratagene). The initial PCR reaction contained 2 µM concentrations each of primers Linker1, Linker2, Linker3, and Linker4 that overlapped with adjacent primers by 20–23 nucleotides (supplemental Fig. 1). The PCR products were diluted 100-fold in a subsequent PCR reaction that amplified the 193-bp full-length linker-60 gene using primers 5'EcoRVlinker and 3'Nde60mer (Fig. 2A). A 208-bp gene coding for a tetrahistidine tagged linker-60 peptide (60H4) was assembled using primer Linker2His-4 instead of Linker2 in the initial PCR reaction (supplemental Fig. 1). Both linker-60 genes were cloned into pGEM-T Easy (Promega) to give plasmids plinker60TVE and plinker60H4TVE. Genes coding for shorter linker peptides of 20 and 40 amino acids were amplified from plinker60TVE using the primer pairs 5'EcoRVlinker/3'Nde20mer and 5'EcoRVfusion/3'Nde40mer and cloned into pGEM-T Easy to give plasmids plinker20TVE and plinker40TVE.

Cloning Wild-type PCC6301 Rubisco Genes—The Synechococcus PCC6301 Rubisco rbcL-S operon was amplified from pSH1 (28) using primers 5'NdeCrbcL (5'-CATATGCCCAAGACGCAATCTGCCGCAG-3', the NdeI site is underlined, and the rbcL initiator codon is in bold) and 3'SacISynS (5'-TGAGCTCTTAGTATCGGCCGGGACGATGAACGAT-3', the SacI site is underlined, and the complement of rbcS terminator codon is in bold) and the 1851-bp NdeI-SacI product was cloned into pET30Xa/LIC (pET30, Novagen) to give plasmid pETCLS (Fig. 2B).

Cloning PCC6301 S-linker-L (CSLL) Rubisco Fusion Genes— The Synechococcus PCC6301 rbcL and rbcS genes were amplified separately from pSH1 using the primer pairs 5'NdeCrbcL/3'SynrbcL (5'-TTAGAGCTTGTCCATCGTTTCAAATTCGAA-3', the complement of rbcL terminator codon is in bold) and 5'NcoSynS (5'-CCATGGGCATGAAAACTCTGCCCAAAGAG-3', the NcoI site is underlined, and rbcS initiator codon is in bold)/3'EcoRVSynS (5'-TGATATCGGCCGGGACGATGAACGATG-3', the EcoRV site is underlined), respectively. The 1422- and 337-bp DNA fragments were cloned into pGEM-T Easy to give plasmids pCLTVE(NdeI) and pCSTVE(NcoI), respectively. The rbcL gene was excised from pCLTVE(NdeI) with NdeI and NotI, and the 1438-bp fragment was cloned into pET30 to give pETCL. The rbcS gene from pCSTVE(NcoI) was cloned in-frame with different length linker peptides by cloning the 332-bp NcoI-EcoRV fragment into plasmids plinker20TVE, plinker40TVE, plinker60TVE, and plinker60H4TVE to give plasmids pCS20TVE, pCS40TVE, pCS60TVE, and pCS60H4TVE, respectively. The 513-bp NcoI-NdeI rbcS60 gene from pCS60TVE was cloned into pET28a+ (Novagen) to give pETCS60, and then the 552-bp XbaI-NdeI fragment was cloned into pETCL, giving plasmid pETCS60L (Fig. 2B). The 393-bp and 453-bp NcoI-NdeI fragments from pCS20TVE and pCS40TVE were cloned into pETCS60L to give plasmids pETCS20L and pETCS40L, respectively (Fig. 2B). Plasmid pETCS60LS that contains a fused and a non-fused copy of rbcS was assembled by replacing the 1078-bp SphI fragment in pETCLS with the 1589-bp fragment from pETCS60L. The 513-bp NcoI-NdeI rbcS60 fragment was then replaced with the 528-bp NcoI-NdeI rbcS60H4 fragment from pCS60H4TVE to give plasmid pETCS60H4LS (Fig. 2B). Sequence coding for a C-terminal His6 tag was cloned 3' to rbcL by amplifying the gene with primers 5'NdeCrbcL and 3'XhoSynL (5'-CTCGAGCTTGTCCATCGTTGCAAATTCGAAC-3', the XhoI site is underlined) and cloning the 1237-bp KpnI-XhoI fragment into pETCS40L to give pETCS40LH6 (Fig. 2B).

Cloning the PCC6301 CL40S Rubisco Gene—The rbcL gene from pSH1 was amplified using primers Nae5'SynL (5'-GCCGGCGTGAAGGACTACAAAC-3', the NaeI site is underlined) and 3'EcoRVSynL (5'-GATATCCCTTGTCCATCGTTTCGAATTCGAACTTG-3', the EcoRV site is underlined). The 1386-bp DNA product was cloned into pGEM-T Easy to give plasmid pCLEVTVE into which the 121-bp EcoRV-NdeI linker-40 gene from plinker40TVE was cloned to give plasmid pCL40TVE. The rbcS gene was amplified from pSH1 using primers 5'NdeSynS (5'-TCATATGAGCATGAAAACTCTGCCCAAAGA-3', the NdeI site is underlined, and the rbcS initiator codon is in bold) and Syn3SacI, and the 347-bp product was cloned into pGEM-T Easy to give plasmid pCSTVE(NS). The 344-bp NdeI-SacI rbcS fragment was then cloned in-frame 3' to the fused rbcL40 gene in pCL40TVE to give plasmid pCL40S from which the 1757-bp KpnI-SacI fragment was used to replace the corresponding 1665-bp fragment in pETCLS to give pETCL40S (Fig. 2B).

Cloning the Tobacco Rubisco TSLL Fusions—Sequence coding 83 nucleotides of the tobacco ribosomal RNA (Prrn) promoter, 63 nucleotides of the T7g10 5'-untranslated region, and a codon modified rbcS gene (cmrbcS) was assembled by splice overlap extension using overlapping primers by successive PCR reactions as described above (supplemental Fig. 1). The translated product from cmrbcS matched that coded by a native Nicotiana tabacum (tobacco) Rubisco RbcS (11). The initial PCR reaction used primers ModSun1, ModSun2, ModTob3, ModTob4r, ModTob5, ModTob6r, ModTob7, and ModTob8rH6 that overlapped by 15 nucleotides with adjoining primers (supplemental Fig. 1). The second PCR reaction amplified a 523-bp Prrn-T7g10 5'-untranslated region-cmrbcS sequence using primers 5'HindPrrn and 3'EcoRVmNtS and was cloned into pGEM-T Easy to give plasmid pPcmSTVE. To clone the linker-40 and -60 genes in-frame to the 3' end of cmrbcS, the gene was amplified from pPcmSTVE using primers 5'NcoImNtS and 3'EcoRVmNtS, and the 368-bp NcoI-EcoRV cmrbcS gene cloned into plinker40TVE and plinker60TVE to give pTS40 and pTS60, respectively. The 489-bp cmrbcS40 and 549-bp cmrbcS60 NcoI-NdeI fragments were then cloned into pET28a(+) to give plasmids pETTS40 and pETTS60. A tobacco rbcL was cloned in-frame to the 3' end of the cmrbcS-linker genes by amplifying a copy of rbcL from pLEV3 (15) using primers 5'NdeNtL (5'-TCATATGTCACCACAAACAGAGACTAA-3', the NdeI site is underlined, and the rbcL initiator codon is in bold) and Trb-cLHind (5'-CAAGCTTTTACTTATCCAAAACGTCCAC-3', the HindIII site is underlined, and the complement of rbcL terminator codon is in bold) and cloning the 1436-bp NdeI-Hin-dIII rbcL fragment into pETTS40 and pETTS60 to give plasmids pETTS40L and pETTS60L, respectively (Fig. 2B).

Growth and Expression of Rubisco in E. coli—The Rubisco genes were cloned downstream of the T7 promoter in pET28a(+) and pET30 and transformed into BL21(DE2) cells (Promega). Rubisco expression was auto-induced by growing the cells at 22 °C in Luria-Bertani medium containing 30 µg·ml-1 kanamycin supplemented with additional carbon sources (0.5% (v/v) glycerol, 2.8 mM glucose, 5.6 mM {alpha}-lactose), nutrients (1 mM MgSO4, 25 mM (NH4)2SO4, 40 mM KH2PO4, 50 mM Na2HPO4) and trace metals (50 µM FeCl3, 20 µM CaCl2, 2 µM Na2SeO3, 2 µM H3BO3, 10 µM MnCl2, 10 µM ZnSO4, 2 µM CoCl2, 2 µM CuCl2, 2 µM NiCl2, 2 µM Na2NoO4) (29). The cultures (100 ml) were shaken at 200 oscillations per minute for 120 h with final cell densities measured by absorbance at 600 nm between 6.6 and 9.2. The cells were harvested by centrifugation (6000 x g, 10 min, 4 °C), frozen in liquid nitrogen, and stored at -70 °C.

Protein Extraction—E. coli cells pellets were suspended in ice-cold extraction buffer (100 mM EPPS-NaOH. pH 8, 1 mM EDTA, 20 mM MgCl2, 2 mM dithiothreitol, 0.043% (w/v) protease inhibitor mixture for bacterial cells (Sigma)) and lysed by passage through a French pressure cell (140 megapascals). An aliquot of lysate sample was mixed with an equal volume of SDS buffer (125 mM Tris-Cl, pH 6.8, 4% (w/v) SDS, 20% (v/v) glycerol, 150 mM 2-mercaptoethanol, 0.01% (w/v) bromphenol blue), and the remainder was centrifuged at 38,000 x g for 15 min at 4 °C. Aliquots of the supernatant were either assayed for protein content using the dye binding Pierce Coomassie Plus kit, treated with a equal volume of SDS buffer (soluble protein sample for SDS-PAGE analysis), and diluted 2-fold with Native PAGE buffer (30% (v/v) glycerol, 0.001% (w/v) bromphenol blue) or incubated with 25 mM NaHCO3 at 25 °C for 30 min before measuring Rubisco content ([2-14C]carboxyarabinitol-P2 binding, see below) and carboxylase activity under substrate ribulose-P2-limiting and -saturating conditions (see below).

PAGE and Immunoblot Analyses—Proteins were separated by SDS-PAGE and non-denaturing PAGE using 4–12% NuPAGE Bis-Tris and 4–12% Tris-glycine gels (Invitrogen), respectively. The Bis-Tris gels were buffered with MES and electrophoresed at 200 V according to the supplier's instructions. Tris-glycine gels were run at 4 °C in 60 mM Tris, 191 mM glycine buffer at 60 volts for 18 h. Protein bands were visualized using Gelcode Blue reagent (Pierce) or blotted onto nitrocellulose (Hybond C, APBiotech) using an Xcell transfer cell (Novex) according to the manufacturer's specifications. Immunoblot analyses were performed as described previously (5) using polyclonal antiserum raised in rabbits to pure spinach Rubisco, tobacco Rubisco, or Synechococcus PCC6301 Rubisco. Immunoreactive proteins were visualized using AttoPhos reagent (Promega) with a Vistra FluorImager.

Rubisco Purification—Rubiscos without histidine tags were purified from E. coli cells by size exclusion chromatography or by ultracentrifugation through sucrose density gradients. Cells that had been stored at -70 °C were suspended in ice-cold extraction buffer, lysed by a French press, and centrifuged as described above. The supernatant was chromatographed through a Superdex 200HR 10/30 column equilibrated with column buffer (50 mM EPPS-NaOH, 100 mM NaCl, pH 8) using an ÁKTA explorer system (APBiotech). Alternatively, a saturated ammonium sulfate solution (pH 7) was slowly added to the supernatant to a final concentration of 20% (w/v) on ice, and the extract was centrifuged at 30,000 x g for 15 min at 2 °C. The supernatant was collected, and Rubisco was precipitated by adding ammonium sulfate to 50% (w/v) and pelleted by centrifugation. The precipitate was dissolved in a 0.7-ml gradient buffer (25 mM EPPS-NaOH, pH 8, 1 mM EDTA) and centrifuged at 31,000 rpm for 26 h at 4 °C through an exponential density gradient (mixing volume, 12.15 ml; gradient volume, 15 ml) of 7–28.8% (w/v) sucrose in gradient buffer using an SW31Ti rotor (Beckman). Fractions collected from the Superdex 200 chromatography (0.3 ml) and the sucrose gradients (1 ml) were assayed for substrate-saturated RuBP carboxylase activity (see below) and an aliquot mixed with an equal volume of SDS or Native PAGE buffer for PAGE analysis. The three fractions with the highest Rubisco activities for each sucrose gradient were pooled and dialyzed for 16 h at 4 °C against 2 liters of gradient buffer and then for 90 min against 0.5 liters of gradient buffer containing 20% (v/v) glycerol. The dialyzed samples were frozen in liquid nitrogen and stored at -70 °C.

Rubiscos with histidine tags were purified using Ni2+-nitrilotriacetic acid-agarose (Qiagen). E. coli cells were suspended in ice-cold affinity extraction buffer (50 mM Tris-Cl, pH 8, 0.3 M NaCl, 10 mM imidazole, 0.043% (w/v) bacterial protease inhibitor mixture), lysed by a French press, and centrifuged (as above). The soluble protein was chromatographed through Ni2+-nitrilotriacetic acid-agarose and washed with 20 bed volumes of affinity extraction buffer. Bound protein was eluted in 1.5–2 ml of elution buffer (50 mM Tris-Cl, pH 8, 0.3 M NaCl, 200 mM imidazole) and immediately dialyzed with successive changes of gradient buffer and stored as described above.

Kinetic Measurements—The Michaelis constant for ribulose-P2 (KmRuBP) and catalytic turnover rate (Vmaxc) were measured in cell free soluble E. coli protein extracts. After lysis and centrifugation (see above), the E. coli-soluble protein extract was preincubated with 25 mM NaHCO3 for 20 min at 25 °C to activate the Rubisco. Carboxylase activities were measured using NaH14CO3 assays (28, 30) containing different amounts of substrate ribulose-P2 (0–2 mM). Assays were buffered with 100 mM EPPS-NaOH, pH 8, containing 20 mM MgCl2 and were performed in duplicate with unbuffered ribulose-P2 added to initiate catalysis. Catalytic turnover rate was calculated by dividing the substrate-saturated carboxylase activity by the concentration of Rubisco active sites measured by the stoichiometric binding of the tight binding inhibitor [2-14C]carboxyarabinitol-P2 as described previously (17, 31). After preincubation with 25 mM NaHCO3, duplicate aliquots of extract were incubated for up to 30 min with 13–39 µM [2-14C]carboxypentitol-P2 (an isomeric mixture of [2-14C]carboxyarabinitol-P2 and [2-14C]carboxyribitol-P2) at 25 °C, and the amount of Rubisco-bound [14C]carboxyarabinitol-P2 was recovered by gel filtration (17).

Purified Rubisco preparations were used to measure the Michaelis constants for CO2 (Km(CO)2) and the CO2/O2 specificity (Sc/o at pH 8.3 (32)). Km(CO)2 was measured by 14CO2 fixation at 25 °C, pH 8, according to Andrews (28) in nitrogen-sparged septum-capped scintillation vials. The assays were initiated by adding purified enzyme (that had been preincubated for 20 min in buffer containing 20 mM MgCl2 and 25 mM NaHCO3) into N2-equilibrated assay buffer (100 mM EPPS-NaOH, 20 mM MgCl2, 0.6 mM ribulose-P2, 0.1 mg·ml-1 carbonic anhydrase) containing varying concentrations of NaH14CO3. The assays were stopped after 2 min with 0.5 volumes of 25% (v/v) formic acid and dried at 80 °C, and then the residue was dissolved in 0.5 ml of water before adding 2 volumes of scintillant (UltimaGold, Packard Bioscience) for scintillation counting.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Engineering Linker Peptides of Different Lengths—Genes coding for linker peptides of different lengths for fusing the large (L) and small (S) subunits of Rubisco from Synechococcus PCC6301 (cyanobacteria) and N. tabacum (tobacco) were synthesized by splice overlap extension using PCR (supplemental Fig. 1). The linker sequences were designed based on flexible Ser-Gly-Gly motif repeats with a basic amino acid every 10th residue that had been previously used to assemble E. coli chaperonin complexes from GroES-GroEL fusions (33). Examination of available x-ray structures for different L8S8 Rubisco hexadecamers indicated considerable variability in the spatial separation between the N and C termini of the adjoining L and S subunits (Fig. 1). Because there were numerous permutations for linking the termini of a subunit to its cognate partners, four different sized linkers were tested that comprised 20, 40, 60, and 65 (60H4) amino acids corresponding to flexible peptides of ~ 65, 130, 185, and 200 Å. Restriction sites at the 3' (EcoRV) and 5' (NdeI) ends were incorporated into the coding sequence of each linker gene to facilitate the in-frame cloning of L and S genes (Fig. 2A). An internal His4 tag comprising the sequence His-Ser-His-His-His-His was engineered into the 60H4 fusion peptide to facilitate purification by affinity chromatography.


Figure 1
View larger version (80K):
[in this window]
[in a new window]

  		FIGURE 1.
Quaternary structure of L8S8 Rubisco. Spinach Rubisco hexadecamer comprising four pairs of large (L) subunits (light and dark gray) and two layers of four small (S) subunits (light and dark yellow) viewed down the 2-fold (A) and 4-fold axes (39) (Protein Data Bank code 1RCX) (B). Using DS viewer Pro software, the shortest distance between the C-terminal Val-475 of anL(blue) to the Met-1 of the three closest S (orange) is 48, 57, and 59 Å (dotted white lines). The direct distance from the C-terminal Tyr-123 of an S (red) to the Ala-9 N-terminal residue (no x-ray coordinates are available for N-terminal residues from Pro-3 to Lys-8) of the three closest L (light blue) is 39, 46, and 98 Å (solid white lines).

 
Subunit Organization of Rubisco Fusions—Because the N and C termini of each L and S are positioned on the surface of L8S8 Rubiscos (Fig. 1), the viability of linking the subunits in either an L-S or S-L arrangement was examined. The genes coding for Synechococcus PCC6301 Rubisco L and S are coded by a rbcL-rbcS operon and can assemble into a functional L8S8 enzyme when expressed in E. coli (28). Fusion peptides containing alternately arranged L and S were assembled by cloning different linker peptides between the codons for the C-terminal Tyr-111 of S to the N-terminal Met-1 of L (CSLL fusions) and the C-terminal Leu-462 of L to the N-terminal Met-1 of S (CL40S fusion) (Fig. 2B). The juxtaposed linker peptides contained either 20, 40, 60, or 65 amino acids, resulting in the production of the CSLL Rubisco fusion peptides CS20L (67.2kDa), CS40L (68.7 kDa), CS60L (70.2 kDa), and CS60H4L (70.8 kDa), respectively. To assess whether an unfused S would assemble with CSLL peptides, a sequence spanning the intergenic region and rbcS from the native cyanobacterial rbcLS operon was retained down-stream of the genes coding for the CS60L and the His4-tagged CS60H4L fusion peptides (Fig. 2B). As well, a gene coding for a 70-kDa CS40LH6 Rubisco fusion containing a His6 tag fused directly to the C-terminal Lys-461 of L was cloned. Compliant with the arrangement of the genes for L and S in the native Synechococcus Rubisco operon (CLS), a gene coding for a 68.9-kDa CL40S fusion peptide was made that contained the 40-amino acid linker sequence.


Figure 2
View larger version (33K):
[in this window]
[in a new window]

  		FIGURE 2.
Linker sequences and arrangement of L and S in the Rubisco fusion genes. A, nucleotide and translated amino acid sequence of the 195-bp linker peptide 60H4. The C-terminal Tyr (Y) and N-terminal Met (M) residues of the adjoining Rubisco subunits coded within the EcoRV and NdeI restriction cloning sites are shown in bold. The sequence spanning the 20 (60 bp), 40 (120 bp), and 60 (180 bp) linker peptides are indicated by arrows. The Ser-(His)4 sequence unique to the 60H4 linker peptide is in italics and shown in parentheses. B, arrangement of the native and fusion Rubisco genes from Synechococcus PCC6301 (cyanobacterium) and N. tabacum (tobacco) cloned into E. coli expression vectors pET28 and pET30 (see "Experimental Procedures").

 
Expression of Rubisco Fusion Peptides in E. coli—Large amounts of mostly unassembled Synechococcus PCC6301 Rubisco accumulated in E. coli cells when the genes were transcriptionally put under the control of the T7 promoter in pET plasmids, and protein expression was stimulated by lactose autoinduction of the T7 polymerase from the DE3 {lambda} lysogen in E. coli BL21(DE3) cells (29). The gradual stimulation of cyanobacterial Rubisco expression through autoinduction improved the amount of functional hexadecameric enzyme recovered from the cellular soluble protein by 3–4 fold compared with the same cells induced with 0.4 mM isopropyl-beta-D-thiogalactopyrano-side for 16 h at 22 °C (data not shown). However, even by auto-induction, SDS-PAGE and immunoblot analyses of the total and soluble E. coli protein showed that only a small proportion (<2%) of the large pool of fused and unfused cyanobacterial Rubisco L and S peptides produced in the cells were correctly folded and soluble (Fig. 3). Immunodetection using an antibody to the cyanobacterial Rubisco also recognized an E. coli protein of ~55 kDa and 2 additional smaller peptides in the extracts of cells producing the CSLL fusions (Fig. 3, C and D). The sizes of these smaller proteins (labeled CSL{Delta}L) varied between 19.5 and 24.5 kDa and appeared to comprise the sequence spanning S, the linker peptide, and the N-terminal residues of L, as the difference of ~1.5 kDa in the measured size of the smallest (~19.5, 21, and 23kDa) and largest (~21, 22.5, and 24.5kDa) peptides, respectively, for each fusion peptide were consistent with the differences in size of the intervening linker sequence. Extrapolating from the translated peptide sequence, the sizes of these smaller peptides correspond to proteolysis in L around residues Lys-11 and Lys-29. If these CSL{Delta}L peptides were the result of proteolysis, an additional 50-kDa cyanobacterial L peptide would be produced; however, immunoblot analysis did not recognize a peptide of this size in either the total or soluble cellular protein fractions (Fig. 3, B and D). This suggests the remaining 50-kDa N-terminal-cleaved L is either rapidly degraded, or alternatively, the CSL{Delta}L species are the products of incorrect translational processing of the mRNA whereupon inappropriate folding of the message has resulted in premature translational termination at similar positions within rbcL. Interestingly, for the CL40S fusion there are no smaller peptides recognized by the Rubisco antibody, indicating in this arrangement that little or no proteolysis (or premature translational termination) is observed.

The CL40S and CSLL Fusions Assemble into Oligomeric Complexes and Are Catalytically Viable—Nondenaturing PAGE of the same E. coli soluble protein samples (see above) showed the variant CSLL and CL40S fusions assembled into oligomeric Rubisco complexes (Fig. 4). By immunodetection, the electrophoretic mobility of the native Rubisco hexadecamer (L8S8) was found to be faster than the oligomeric complexes comprising the CSLL and CL40S fusions (termed fusion-Rubiscos) (Fig. 4B). The observed difference in mobilities are consistent with the fusion peptides assembling predominantly into octameric complexes ((L40S)8 or (CSLL)8), whose mobility decreased as the size of the linker sequence increased. Curiously, in addition to two E. coli proteins, the antibody also recognized a larger, less abundant protein complex in all the fusion peptide samples that corresponded to Rubisco oligomers larger than the octamers and were speculatively assumed to comprise connected octamer pairs (i.e. hexadecamers) that arose from linked L and S coding regions within fusions assembling in separate octameric enzyme complexes, thereby tethering two octamers together. This was examined further by size exclusion chromatography.


Figure 3
View larger version (115K):
[in this window]
[in a new window]

  		FIGURE 3.
Expression of linked and wild-type Synechococcus PCC6301 Rubisco subunits in E. coli. SDS-PAGE analysis of auto-induced Rubisco expression in E. coli BL21(DE3) cells grown at 22 °C (see "Experimental Procedures"). After cell lysis, total cellular protein (containing 10 µg of soluble protein) (A and B) and soluble cellular protein (10 µg) (C and D) were separated by SDS-PAGE. The separated proteins were stained with Coomassie (A and C) or transferred onto nitrocellulose membranes and immunoprobed with antibodies that recognize the Synechococcus large (52.4 kDa)) and small (13.3 kDa) Rubisco subunits and a 55-kDa non-Rubisco E. coli protein (white arrows) (B and D). The molecular mass marker (m) sizes are shown. CL40S/CSLL, Synechococcus Rubisco L and S fusions with different length linker peptides (refer to Fig. 2); CSL{Delta}L, partial fusion peptides that include S, the linker sequence, and varying amounts of the L N-terminal amino acids.

 
The ability of the CSLL and CL40S Rubisco complexes expressed in E. coli to catalyze 14CO2 fixation provided a simple means for their detection during chromatographic purification. Maximal carboxylase activity required preactivation with CO2 and Mg2+ and was dependent on the addition of substrate ribulose-P2. Separation of the E. coli protein by size exclusion chromatography identified two Rubisco activity peaks for the fusion-Rubisco complexes (Fig. 5A). Both activity peaks corresponded to proteins of larger molecular mass than the single activity peak for wild-type L8S8 Rubisco with different elution times for the fusion-Rubisco complexes correlating to the differences in the size of the linker sequences used to fuse the L and S together. Further examination of the chromatography fractions by immunoblot analysis after SDS-PAGE showed both activity peaks were synchronized with the amount of LS fusion peptides detected (data not shown). When the chromatography fractions were separated by nondenaturing PAGE, immunoblot analysis confirmed the second activity peak corresponded to the octameric complex and the earlier peak to the larger LS fusion oligomeric protein complex that was assigned a hexadecameric ((CSLL)16) stoichiometry as it likely comprised two octameric complexes joined by the linker regions of two or more fusions (Fig. 5B). Notably, the linear electrophoretic mobility of the symmetrical L8S8 enzyme in the chromatography fractions (top panel, Fig. 5B) was not emulated by the chromatographically separated (CS40L)8 and (CS40L)16 enzyme fractions that showed considerable variation in their electrophoretic mobility (bottom panel, Fig. 5B). This nonlinear electrophoretic mobility was also observed after non-denaturing PAGE analysis of the other (CSLL)8 and (CSLL)16 complexes after their chromatographic separation (data not shown). Evidently both the octameric and hexadecameric enzymes comprised structurally heterogeneous populations of asymmetrically arranged SL fusions that produced complexes of structurally variable sizes. Irrespective of their structural conformations, however, the CO2/O2 specificity (Sc/o) measured for both the (CS40L)16 and (CS40L)8 peak fractions matched that of the wild-type L8S8 enzyme (Fig. 5A).

The catalytic stability of the (CS40L)16 Rubisco complexes were found to be less stable than the (CS40L)8 enzyme. Equal amounts of protein from E. coli-expressing CS40L fusion-Rubiscos were incubated at different temperatures and centrifuged before separation by size exclusion chromatography (Fig. 5C). After incubating the E. coli extract at 25 °C for 30 min, 92% of the carboxylase activity measured in a parallel sample immediately chromatographed after extraction (0 min at 0 °C) was recovered. As shown in Fig. 5C, this loss was almost exclusively attributed to a reduction in the activity recovered in the (CS40L)16 Rubisco peak. The reduced activity was not due to a dissociation of the (CS40L)16 complex into insoluble CS40L aggregates as immunoblot analyses after SDS-PAGE found no discernable difference in the amount of S40L fusion peptides in the chromatography fractions between these two samples or any insoluble CS40L in the centrifugal pellet of the 25 °C-treated sample before loading onto the column (data not shown). Notably, the reduced activity of the (CS40L)16 fusion-Rubisco complexes could be avoided by maintaining the enzyme at 0 °C as, even after 4 h, almost all (>98%) of the activity was recovered after chromatography.


Figure 4
View larger version (86K):
[in this window]
[in a new window]

  		FIGURE 4.
Quaternary structures of the Synechococcus PCC6301 fusion-Rubiscos and wild-type L8S8 enzyme. Nondenaturing PAGE comparison of Rubisco complexes assembled in the soluble protein extract of E. coli BL21(DE3) cells. The separated proteins were stained with Coomassie (A) or transferred onto nitrocellulose membranes and immunoprobed with antibodies that recognize Synechococcus PCC6301 Rubisco and non-Rubisco E. coli proteins (white arrows) (B). Molecular mass markers (m) are thyroglobulin (669 kDa), ferritin (440 kDa), and catalase (232 kDa). (CL40S)8 and (CSLL)8, Rubisco octamers of PCC6301 L and S fusions tethered together with different linker peptides (refer to Fig. 2); (CL40S)16 and (CSLL)16, hexadecamers of the L and S fusions.

 
Rubisco Fusion Complexes Bind Carboxyarabinitol-P2 Tightly—The active sites in the fusion-Rubiscos composed of CL40S or CSLL peptides bound the Rubisco inhibitor, carboxyarabinitol-P2, tightly. When CO2-Mg2+-preactivated sample was incubated with saturating levels (13 or 39 µM) of [14C]carboxypentitol-P2 for 15 or 30 min at 25 °C, there was less than a 7% difference in the amount of Rubisco-[14C]carboxyarabinitol-P2 complexes recovered after gel filtration. The return of radioactivity to base line between the Rubisco-bound and -unbound [14C]carboxyarabinitol-P2 peaks confirmed tight binding of the inhibitor. Therefore, the amount of the Rubisco complexes expressed in the E. coli was quantified by assuming stoichiometric binding by the inhibitor to Rubisco active sites (Table 1, expressed as a percentage of cellular soluble protein) and are analogous to the relative amounts identified by immunoblot analyses of the same protein samples (Fig. 3D). Compared with wild-type L8S8, the amount of functional fusion-Rubisco enzyme complexes produced were reduced by >2-fold, indicating functional assembly of the larger CL40S and CLLS fusions in E. coli was impaired, in particular the fusions containing histidine tags or the 20-amino acid linker.


View this table:
[in this window]
[in a new window]

  		TABLE 1
Expression and kinetic properties of Synechococcus PCC6301 wild-type (CLS) and fusion-Rubiscos expressed in E. coli

 
Carboxylation rate (Vmaxc) and Km(RuBP) were measured in non-purified soluble protein extract from E. coli cells, and the active site content was measured by [2-14C]carboxyarabinitol-P2 binding. The Formula of the CSLL Rubiscos were generally the same or slightly improved compared with the wild-type enzyme except for the Rubiscos comprising the C-terminal His6-tagged (CS40LH6) and CL40S fusions whose carboxylation rates were reduced 3–4-fold (Table 1). Km(RuBP) measurements for the wild-type enzyme matched those measured previously for purified enzyme (34). For the fusion-Rubiscos, modest improvements in ribulose-P2 affinity relative to the wild-type enzyme were found for those containing the S20L, S40L, S60L, and S60H4L fusions, no change was found for the CS40LH6 fusion, and ribulose-P2 affinity was reduced for enzymes composed of CL40S fusions (Table 1).


Figure 5
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 5.
Size exclusion chromatography separation of Synechococcus PCC6301 Rubisco complexes. A, 1–1.5 mg of total cell soluble protein (in 0.2 ml) from E. coli cells expressing native PCC6301 Rubisco (cLS, gray line) and fusion-Rubisco complexes (cL40S, {circ}; cS20L, solid line; cS40L, dotted line; cS60L, dashed line) were separated on a Superdex 200HR 10/30 column (see "Experimental Procedures"). Fractions (0.3 ml) were collected and assayed for ribulose-P2-dependent carboxylase activity. Activity peaks corresponding to fusion-Rubisco hexadecamers ((CS40L)16) and octamers ((CS40L)8) and the wild-type L8S8 Rubisco are indicated (arrows). Sc/o, the CO2/O2 specificity of peak fractions measured as described in Kane et al. (32). B, immunoblot analysis of cLS and cS40L Superdex fractions (adjacent fractions from panel A, pooled as indicated by arrows) separated by nondenaturing PAGE and probed with an antibody that recognizes Synechococcus Rubisco and another E. coli protein (white arrow). C, recovery of ribulose-P2-dependent carboxylase activity by Superdex 200HR chromatography of replica soluble protein E. coli extract (1.3 mg) expressing the cS40L fusion. The extract was incubated 0 min at 25 °C (bar graph), 30 min at 25 °C, then 0 °C for 60 min (•), or 0 °C for 3 h before (solid line) before chromatography.

 
Purification—Sucrose gradient ultracentrifugation of E. coli cell extracts was used to purify adequate amounts of non-Histagged Rubiscos and confirmed the CL40S and CSLL peptides assemble into functional fusion-Rubisco complexes that were larger than the native Rubisco hexadecamer (supplemental Fig. 2). For the native L8S8 complex, 14CO2 carboxylase and immunoblot assays identified a single symmetrical peak corresponding to the sedimentation between fractions 9 and 12. Consistent with the non-denaturing PAGE (Fig. 4) and chromatography analyses (Fig. 5), the CL40S and various CSLL peptides all assembled into larger octameric ((CSLL)8) and hexadecameric ((CSLL)16) complexes that sedimented further down the gradient than native L8S8 cyanobacterial Rubisco. Unlike that obtained by size exclusion chromatography, the octameric and hexadecameric fusion-Rubisco complexes did not resolve into two separate activity peaks; however, the sedimentation behavior of the Rubisco complexes predicted from activity measurements matched the amount of Rubisco subunits identified by immunodetection (supplemental Fig. 2B). SDS-PAGE analyses of Rubiscos purified using sucrose gradients showed they were ~50% pure except for the native L8S8 enzyme that was >90% pure (not shown). Attempts to purify the fusion-Rubiscos from E. coli extracts by anion exchange chromatography using the procedure of (35) yielded dilute enzyme preparations of low purity (Rubisco constituting <30% of protein, data not shown). By this method, the wild-type L8S8 Rubisco resolved separate from other proteins at 0.25 M potassium phosphate at pH 7.2. In contrast, the Rubisco complexes made of S20L, S40L, and S60L fusion peptides eluted with other E. coli proteins earlier in the potassium phosphate gradient at ~0.23, 0.21, and 0.2 M, respectively.

For the His-tagged enzymes, almost complete recovery of enzyme activity (>90% of initial activity) was obtained by immobilized metal affinity chromatography of the soluble cellular extract. However, because the level of Rubisco expression in these extracts was low (Table 1), other E. coli proteins were co-purified, resulting in final purities of ~60% as judged by SDS-PAGE (data not shown).

Unlinked S Peptides Do Not Assemble in CSLL Rubisco Complexes—Immunoblot analysis of sucrose gradient fractions showed that unlinked S and CS60{Delta}L peptides did not assemble with CS60L peptides into functional enzyme complexes. Sucrose gradient fractionation of protein from E. coli cells co-expressing CS60L and unfused S (pETCS60LS, Fig. 2B) found the 13.3-kDa S and the 22.7- and 24.5-kDa CS60{Delta}L peptides sedimented toward the top of the gradient and did not co-purify with the large functional octameric and hexadecameric complexes that sedimented toward the bottom of the gradient (supplemental Fig. 2B). This indicates that, in the absence of L, available S are excluded from the assembly of CSLL Rubisco complexes. Moreover, it demonstrated that if the CS60{Delta}L peptides arise from proteolysis, then the CSLL peptides appear immune to proteolysis once assembled in an oligomeric complex.


Figure 6
View larger version (65K):
[in this window]
[in a new window]

  		FIGURE 6.
Expression of linked tobacco Rubisco subunits in E. coli. SDS-PAGE analysis of autoinduced Rubisco expression in BL21(DE3) E. coli cells grown at 22 °C. After cells lysis, total (soluble and insoluble) and soluble cellular protein were separated by SDS-PAGE and either stained with Coomassie (A) or transferred onto nitrocellulose membranes and the upper region of the blot probed with an antibody to spinach Rubisco that recognizes the tobacco large subunit (52.9 kDa)), the TS40L (70.4 kDa) and TS60L (71.9 kDa) fusions (refer to Fig. 2) and two E. coli proteins (black triangles) (B). C, the lower section of the membrane was probed with an antibody to tobacco Rubisco that strongly recognizes the tobacco small subunit (14.6 kDa) and two other E. coli proteins (white arrows). m, molecular mass marker; TSL{Delta}L, uncharacterized 22–25-kDa peptides containing S, the linker sequence and N-terminal sequence of the L; Leaf, soluble tobacco leaf protein extracted as described in Whitney et al. (16); pET30, control plasmid directing expression of a 14-kDa non-Rubisco soluble protein.

 
Substrate Specificity and CO2 Affinity—Purified enzyme preparations were used to measure the CO2/O2 specificity (Sc/o) and the Michaelis constant for CO2 (Km(CO2)) and, for the wild-type enzyme, are similar to that reported previously (35). Irrespective of the linker sequence length or the organization of L and S in the fusion, the CO2/O2 specificity (Sc/o) for all the fusion-Rubiscos was comparable with wild-type (Table 1). In contrast, the Km(CO2) for Rubiscos comprising cS40L and cS60L fusions were ~10% lower than wild-type, whereas Rubiscos comprising the CL40S or CS60LH6 fusions were 40 and 60% higher than wild-type, respectively.

Fused Tobacco L and S Peptides Do Not Assemble in E. coli— For tobacco Rubisco, two genes coding for peptides TS40L and TS60L were cloned by merging the C-terminal Tyr-123 of S to the N-terminal Met-1 of L and joined by the 40- and 60-amino acid linker sequences, respectively. Like the cyanobacterial enzyme, both tobacco Rubisco fusions expressed well in E. coli; however, they were unable to fold and assemble correctly into functional enzyme. SDS-PAGE and immunoblot analyses showed that the large amount of TS40L and TS60L peptides produced were entirely insoluble (Fig. 6). This was confirmed by [2-14C]carboxyarabinitol-P2 binding and carboxylase assays (using high specific activities of 14CO2 and 10-min assay periods) that were unable to detect any assembled or active enzyme in the soluble cell protein extracts (data not shown). This is consistent with previous findings that higher plant L and S are unable to assemble into L8S8 enzyme in E. coli due to problems with chaperone incompatibility (24, 36). Notably, as with the cyanobacterial fusion peptides, the immunoblot analyses of the soluble E. coli protein identified two smaller TS{Delta}L peptides (~20- and 21.5-kDa species in TS40L, ~21.5- and 23-kDa species in TS60L) that were consistent with proteolysis within the N-terminal region of the fused L around Lys-14 and Lys-32. The remaining portion of a proteolyzed tobacco L could not be detected in the total cellular protein extract by immuno-detection, indicating the L was either rapidly degraded or the TS{Delta}L peptides are the products of premature translational processing of the rbcTS40/60L mRNAs toward the 5' end of rbcL.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Linked Cyanobacterial L and S Peptides Assemble into Functional Rubisco Oligomers—Our interest in applying a peptide fusion strategy to link the Rubisco L and S subunits arises from a desire to simultaneously engineer the genes for both native and foreign Rubisco subunits within the plastids of higher plants. Using Synechococcus Rubisco for proof of principal, this study has identified suitable linker sequences useful for generating L-S and S-L fusions that maintain the capacity to assemble into functional cyanobacterial fusion-Rubisco complexes in E. coli. The ability of all the fusion-Rubiscos to catalyze ribulose-P2 carboxylation and tightly bind the reaction intermediate carboxyarabinitol-P2 indicated the integrity of the active site geometry remained relatively unperturbed, particularly for the enzymes comprising the CSLL fusions whose kinetics closely mirrored, and in some instances showed an improvement on, those of the wild-type L8S8 enzyme.

The capacity of the different cyanobacterial Rubisco fusions to functionally assemble in E. coli varied depending on the size of the linker, the order of S and L, and whether or not histidine tags were incorporated. By utilizing the T7 promoter pETbased expression system, large amounts of the different Rubisco peptides were expressed in E. coli BL21 cells, particularly when stimulated by autoinduction. Comparable to previous findings, the majority (>98%) of the wild-type cyanobacteria L produced was insoluble, presumably forming inclusion body aggregates, with a pool of the more soluble S still available for assembling with L8 cores (22–24, 37). As well, the CLLS and S40L fusions were produced in amounts that rivaled that of wild-type L; however, their productive folding and assembly by E. coli chaperones appeared additionally compromised. Possibly these processes are constrained by the large size of the fusion peptides or by additional complexities in assembling the asymmetrical enzyme complexes (see below). Whether overexpressing the GroES-GroEL chaperonins can improve the yield of Rubisco-fusion complexes produced in E. coli, as it can for cyanobacterial L8S8 Rubisco (23, 37), remains to be examined. Curiously the incorporation of successive histidine residues appeared to hamper proper folding and/or assembly of the cS60H4L and cS40LH6 peptides. As well, the fusion peptides containing the 40- or 60-amino linker sequences appeared more adept to functional assembly than the fusion peptides containing the 20-amino acid linker (Table 1). Based on the predicted distances between the L and S termini in L8S8 structures (Fig. 1), it is possible the shorter linker sequence (~65 Å) may have impeded functional assembly of the cS20L fusion by imposing a constraint on where the linked L and S subunits could be positioned within the holoenzyme complexes.

Non-denaturing PAGE and size exclusion chromatography clearly showed the fusion peptides assembled into two structurally distinct catalytic complexes that were larger than the wild-type L8S8. It is likely these represented octameric structures ((CL40S)8 or (CSLL)8) and unprecedented hexadecameric configurations that presumably consisted of adjoined octamers. Unlike the L8S8 enzyme, the assembly of the fusion peptides into hexadecamers through the integration of adjoined L and S sequences into different octamers clearly indicated they are asymmetrical. Similarly, structural asymmetry would be expected for the octameric complexes, particularly those comprising fusion peptides with larger linker sequences where there might be less constraint as to where the connected L and S subunits could be positioned within the fusion-Rubisco complex. Consistent with this, the Rubisco-fusion complexes containing larger linker sequences displayed more diffuse separation through non-denaturing PAGE (Fig. 4), broader elution profiles after size exclusion chromatography (Fig. 5), and increased asymmetry in their sedimentation behavior in sucrose gradients after ultracentrifugation (supplemental Fig. 2).

Linking to or Modifying the C Terminus of L Impairs Catalytic Performance—Although all the Synechococcus PCC6301 fusion-Rubisco complexes, irrespective of linker length and subunit order, were appropriately assembled to form apparently normal active sites, it was clear that engineering changes to the C-terminal end of L was detrimental to catalytic viability. Normally during catalysis the binding of substrate within a Rubisco active site initiates the folding over ("closing") of loop 6 residues in L that subsequently retracts ("opens") after catalysis (7, 8, 38, 39). The cycling between the open and closed confirmation during catalysis is thought to be assisted by the flexible C-terminal "tail" of L that stretches across the top of loop 6 when it is closed and then retracts to an {alpha}-helical structure upon loop 6 opening, allowing product release. The importance of the C-terminal tail on catalysis has been highlighted by the observed reduction in catalytic performance by Rubiscos whose L subunits have had their C termini shortened by proteolysis (40, 41) or increased by cloning extensions to Synechococcus L (42). At 25 °C, the substrate affinities and catalytic turnover for the mutant Synechococcus Rubiscos with C-terminal extensions were impaired, whereas their ability to distinguish between CO2 and O2 as substrates was not (42). Comparable differences were observed here in the kinetic properties of enzymes comprising cS40LH6 and cL40S fusions, supporting the concept that engineering changes to the C terminus of L (such as linking it to the N terminus of S) invariably lead to reduced catalytic performance. In contrast, there was no compromise in catalytic viability when adjoining the C terminus of S to the N terminus of L using flexible linker peptides (Table 1), indicating this is the favored configuration for linking these Rubisco subunits.

Future Prospects for Rubisco Engineering in Higher Plants—Future work will assess the applicability of this linking strategy for assembling functional Rubisco complexes of linked higher plant Rubisco L and S in chloroplasts. This study successfully showed that TSLL fusions comprising linked tobacco S and L subunits could be abundantly produced in E. coli; however, linking of the subunits was unable to circumvent the assembly restriction faced by all eukaryotic L8S8 Rubiscos in this prokaryotic host that are proposed to arise from chaperone incompatibilities (12, 24, 36). Further studies as to whether the chaperone complexes of higher plant plastids are capable of assembling these tobacco TSLL fusions into functional Rubisco complexes are obviously warranted. Successful replacement of the rbcL copy in the tobacco plastome with the comparable gene from sunflower (18) and the bacterium Rhodospirillum rubrum (15, 17) by homologous replacement indicates that a comparable strategy for introducing the rbcTS40L or rbcTS60L genes is feasible. Although transgenic manipulation by this means is a protracted process, it does provide a specific means by which to examine comprehensively the molecular, cellular, and physiological consequences of the transplanted fusion-Rubisco into plant plastids. That is, the kinetics of functional Rubisco fusions could be examined both in vitro after enzyme purification and in vivo using whole leaf gas exchange (15–17, 20, 21). Moreover, such a transplantation would provide a means to assess whether the finding in this study that unassembled cyanobacterial S are excluded from assembling with CS60L fusions can be emulated in higher plant plastids. Notably, in transplastomic tobacco expressing the Rubisco dimer (L2) from R. rubrum, which does not require any S, the unassembled tobacco S was rapidly degraded without apparent detriment to the plants (15).


   FOOTNOTES
 
* This work was funded by Australian Research Council Discovery Grant DP0450564. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

Formula The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. 1 and 2. Back

1 To whom correspondence should be addressed. Tel.: 61-2-6125-5073; Fax: 61-2-6125-5075; E-mail: spencer.whitney{at}anu.edu.au.

2 The abbreviations used are: S, Rubisco small subunit; L, Rubisco large subunit; carboxyarabinitol-P2,2'-carboxyarabinitol-1,5-bisphosphate; carboxypentitol-P2, isomeric mixture of carboxyarabinitol-P2 and 2'-carboxyribitol-1,5-bisphosphate; CLLS, cyanobacterial (Synechococcus PCC6301) Rubisco L-linker-S fusion peptide; EPPS, 4-([2-hydroxyethyl)-1-piperazinepropanesulfonic acid; RuBP, ribulose-P2; MES, 2-[N-morpholino]ethanesulfonic acid; ribulose-P2, D-ribulose-1,5-bisphosphate; Rubisco, ribulose-P2 carboxylase/oxygenase; TLLS, tobacco Rubisco L-linker-S fusion peptide; Bis-Tris, 2-[bis(2-hydroxyethyl) amino]-2-(hydroxymethyl)propane-1,3-diol. Back


   ACKNOWLEDGMENTS
 
We thank John Andrews for valuable contribution to this research and Heather Kane for comments on the manuscript.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Andrews, T. J., and Whitney, S. M. (2003) Arch. Biochem. Biophys. 414, 159-169[Medline] [Order article via Infotrieve]
  2. Zhu, X.-G., Portis, A. R., and Long, S. P. (2004) Plant Cell Environ. 27, 155-165[CrossRef]
  3. Long, S. P., Zhu, X., Naidu, S. L., and Ort, D. R. (2006) Plant Cell Environ. 29, 315-330[CrossRef][Medline] [Order article via Infotrieve]
  4. Morell, M. K., Paul, K., Kane, H. J., and Andrews, T. J. (1992) Aust. J. Bot. 40, 431-441[CrossRef]
  5. Whitney, S. M., Baldett, P., Hudson, G. S., and Andrews, T. J. (2001) Plant J. 26, 535-547[CrossRef][Medline] [Order article via Infotrieve]
  6. Tcherkez, G. G. B., Farquhar, G. D., and Andrews, T. J. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 7246-7251[Abstract/Free Full Text]
  7. Spreitzer, R. J., and Salvucci, M. E. (2002) Annu. Rev. Plant Biol. 53, 449-475[CrossRef][Medline] [Order article via Infotrieve]
  8. Parry, M. A., Andralojc, P. J., Mitchell, R. A., Madgwick, P. J., and Keys, A. J. (2003) J. Exp. Bot. 54, 1321-1333[Abstract/Free Full Text]
  9. Raines, C. A. (2006) Plant Cell Environ. 29, 331-339[Medline] [Order article via Infotrieve]
  10. Spreitzer, R. J., Peddi, S. R., and Satagopan, S. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 17225-17230[Abstract/Free Full Text]
  11. Whitney, S. M., and Andrews, T. J. (2001) Plant Cell 13, 193-205[Abstract/Free Full Text]
  12. Roy, H., and Andrews, T. J. (2000) in Photosynthesis: Physiology and Metabolism (Leegood, R. C., Sharkey, T. D., and von Caemmerer, S., eds) pp. 53-83, Kluwer Academic Publishers, Dordrecht, The Netherlands
  13. Houtz, R. L., and Portis, A. R., Jr. (2003) Arch. Biochem. Biophys. 414, 150-158[Medline] [Order article via Infotrieve]
  14. Spreitzer, R. J. (2003) Arch. Biochem. Biophys. 414, 141-149[CrossRef][Medline] [Order article via Infotrieve]
  15. Whitney, S. M., and Andrews, T. J. (2003) Plant Physiol. 133, 287-294[Abstract/Free Full Text]
  16. Whitney, S. M., von Caemmerer, S., Hudson, G. S., and Andrews, T. J. (1999) Plant Physiol. 121, 579-588[Abstract/Free Full Text]
  17. Whitney, S. M., and Andrews, T. J. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 14738-14743[Abstract/Free Full Text]
  18. Kanevski, I., Maliga, P., Rhoades, D. F., and Gutteridge, S. (1999) Plant Physiol. 119, 133-141[Abstract/Free Full Text]
  19. Kanevski, I., and Maliga, P. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 1969-1973[Abstract/Free Full Text]
  20. Dhingra, A., Portis, A. R., and Daniell, H. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 6315-6320[Abstract/Free Full Text]
  21. Zhang, X. H., Ewy, R. G., Widholm, J. M., and Portis, A. R., Jr. (2002) Plant Cell Physiol. 43, 1302-1313[Abstract/Free Full Text]
  22. Emlyn-Jones, D., Woodger, F. J., Price, G. D., and Whitney, S. M. (2006) Plant Cell Physiol. 47, 1630-1640[Abstract/Free Full Text]
  23. Goloubinoff, P., Gatenby, A. A., and Lorimer, G. H. (1989) Nature 337, 44-47[CrossRef][Medline] [Order article via Infotrieve]
  24. Kettleborough, C. A., Parry, M. A. J., Keys, A. J., and Phillips, A. L. (1990) J. Exp. Bot. 41, 1287-1292[Abstract/Free Full Text]
  25. Gatenby, A. A. (1992) Plant Mol. Biol. 19, 677-687[CrossRef][Medline] [Order article via Infotrieve]
  26. Pierce, J., Tolbert, N. E., and Barker, R. (1980) 19, 934-942
  27. Kane, H. J., Wilkin, J. M., Portis, A. R., and Andrews, T. J. (1998) Plant Physiol. 117, 1059-1069[Abstract/Free Full Text]
  28. Andrews, T. J. (1988) J. Biol. Chem. 263, 12213-12219[Abstract/Free Full Text]
  29. Studier, F. W. (2005) Protein Expression Purif. 41, 207-234[CrossRef][Medline] [Order article via Infotrieve]
  30. Paul, K., Morell, M. K., and Andrews, T. J. (1991) Biochemistry 30, 10019-10026[CrossRef][Medline] [Order article via Infotrieve]
  31. Ruuska, S., Andrews, T. J., Badger, M. R., Hudson, G. S., Laisk, A., Price, G. D., von, and Caemmerer, S. (1998) Aust. J. Plant Physiol. 25, 859-870
  32. Kane, H. J., Viil, J., Entsch, B., Paul, K., Morell, M. K., and Andrews, T. J. (1994) Aust. J. Plant Physiol. 21, 449-461
  33. Farr, G. W., Fenton, W. A., Chaudhuri, T. K., Clare, D. K., Saibil, H. R., and Horwich, A. L. (2003) EMBO J. 22, 3220-3230[CrossRef][Medline] [Order article via Infotrieve]
  34. Pearce, F. G. (2006) Biochem. J. 399, 525-534[CrossRef][Medline] [Order article via Infotrieve]
  35. Morell, M. K., Paul, K., O'Shea, N. J., Kane, H. J., and Andrews, T. J. (1994) J. Biol. Chem. 269, 8091-8098[Abstract/Free Full Text]
  36. Gatenby, A. A. (1984) Eur. J. Biochem. 144, 361-366[Medline] [Order article via Infotrieve]
  37. Larimer, F. W., and Soper, T. S. (1993) Gene (Amst.) 126, 85-92[CrossRef][Medline] [Order article via Infotrieve]
  38. Duff, A. P., Andrews, T. J., and Curmi, P. M. G. (2000) J. Mol. Biol. 298, 903-916[CrossRef][Medline] [Order article via Infotrieve]
  39. Taylor, T. C., and Andersson, I. (1997) J. Mol. Biol. 265, 432-444[CrossRef][Medline] [Order article via Infotrieve]
  40. Portis, A. R., Jr. (1990) Arch. Biochem. Biophys. 283, 397-400[CrossRef][Medline] [Order article via Infotrieve]
  41. Gutteridge, S. (1986) Biochem. Soc. Trans. 14, 28-29[Medline] [Order article via Infotrieve]
  42. Zhu, G. H., Jensen, R. G., Bohnert, H. J., Wildner, G. F., and Schlitter, J. (1998) Photosynth. Res. 57, 71-79

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Plant Physiol.Home page
S. M. Whitney, H. J. Kane, R. L. Houtz, and R. E. Sharwood
Rubisco Oligomers Composed of Linked Small and Large Subunits Assemble in Tobacco Plastids and Have Higher Affinities for CO2 and O2
Plant Physiology, April 1, 2009; 149(4): 1887 - 1895.
[Abstract] [Full Text] [PDF]

Home page
 J Exp BotHome page
S. M. Whitney and R. E. Sharwood
Construction of a tobacco master line to improve Rubisco engineering in chloroplasts
J. Exp. Bot., May 1, 2008; 59(7): 1909 - 1921.
[Abstract] [Full Text] [PDF]

Home page
 Plant Physiol.Home page
R. E. Sharwood, S. von Caemmerer, P. Maliga, and S. M. Whitney
The Catalytic Properties of Hybrid Rubisco Comprising Tobacco Small and Sunflower Large Subunits Mirror the Kinetically Equivalent Source Rubiscos and Can Support Tobacco Growth
Plant Physiology, January 1, 2008; 146(1): 83 - 96.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
282/6/3809    most recent
M610479200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Whitney, S. M.
Right arrow Articles by Sharwood, R. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Whitney, S. M.
Right arrow Articles by Sharwood, R. E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2007 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement