Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2015 Jul 29;169(1):432–441. doi: 10.1104/pp.15.00622

Half-of-the-Sites Reactivity of the Castor Δ9-18:0-Acyl Carrier Protein Desaturase1,[OPEN]

Qin Liu 1,2, Jin Chai 1,2, Martin Moche 1,2, Jodie Guy 1,2, Ylva Lindqvist 1,2, John Shanklin 1,2,*
PMCID: PMC4577402  PMID: 26224800

Heterodimers of a wild-type soluble lipid desaturase and a substrate-binding mutant demonstrate a reactivity that contrasts with membrane desaturases which require two active subunits.

Abstract

Fatty acid desaturases regulate the unsaturation status of cellular lipids. They comprise two distinct evolutionary lineages, a soluble class found in the plastids of higher plants and an integral membrane class found in plants, yeast (Saccharomyces cerevisiae), animals, and bacteria. Both classes exhibit a dimeric quaternary structure. Here, we test the functional significance of dimeric organization of the soluble castor Δ9-18:0-acyl carrier protein desaturase, specifically, the hypothesis that the enzyme uses an alternating subunit half-of-the-sites reactivity mechanism whereby substrate binding to one subunit is coordinated with product release from the other subunit. Using a fluorescence resonance energy transfer assay, we demonstrated that dimers stably associate at concentrations typical of desaturase assays. An active site mutant T104K/S202E, designed to occlude the substrate binding cavity, was expressed, purified, and its properties validated by x-ray crystallography, size exclusion chromatography, and activity assay. Heterodimers comprising distinctly tagged wild-type and inactive mutant subunits were purified at 1:1 stoichiometry. Despite having only one-half the number of active sites, purified heterodimers exhibit equivalent activity to wild-type homodimers, consistent with half-of-the-sites reactivity. However, because multiple rounds of turnover were observed, we conclude that substrate binding to one subunit is not required to facilitate product release from the second subunit. The observed half-of-the-sites reactivity could potentially buffer desaturase activity from oxidative inactivation. That soluble desaturases require only one active subunit per dimer for full activity represents a mechanistic difference from the membrane class of desaturases such as the Δ9-acyl-CoA, Ole1p, from yeast, which requires two catalytically competent subunits for activity.

Fatty acid desaturase enzymes are key regulators of the unsaturation status of storage lipids (Shanklin and Cahoon, 1998). They introduce cis double bonds into fatty acyl chains in a reaction that is dependent on a diiron cofactor, reductant, and molecular oxygen. Fatty acid desaturases can be assigned to two distinct classes based on sequence homology, solubility, and acyl carrier (Shanklin and Somerville, 1991). One class is an acyl-acyl carrier protein (ACP)-dependent, soluble form found only in the plastids of higher plants and mycobacteria (Shanklin and Cahoon, 1998; Dyer et al., 2005). The other class comprises integral membrane proteins that are acyl-CoA or acyl-lipid dependent and is widely dispersed among bacteria, fungi, animals, and plants (Shanklin and Cahoon, 1998). The soluble class of desaturases belongs to a functionally diverse family of four helix bundle diiron containing-enzymes that includes ribonucleotide reductase, hemerythrins, and bacterial multicomponent monooxygenases such as methane monooxygenase (Friedle et al., 2010). A comprehensive understanding of the biochemical mechanisms of these enzymes will contribute to the rational design of desired functionality for applications ranging from oilseed engineering to therapeutics (Cahoon et al., 1997; Whittle and Shanklin, 2001; Guy et al., 2006; Nguyen et al., 2010; Sampath and Ntambi, 2011), and may be useful for efforts to optimize production of unusual fatty acids in production organisms (Napier, 2007).

The castor Δ9 stearoyl-acyl carrier protein desaturase (EC 1.14.19.2, Δ9-18:0-ACP desaturase), which produces oleoyl-acyl carrier protein (18:1cisΔ9-ACP) in plant plastids, is the best studied of the soluble plant fatty acid desaturases (Gomes et al., 2001; Lindqvist, 2001; Fox et al., 2004; Shanklin et al., 2009). In a previous study, we overexpressed the castor desaturase in Escherichia coli and determined the crystal structure of the purified protein (Lindqvist et al., 1996). The three-dimensional structure showed it to be a homodimeric protein, with each subunit comprising a conserved four-helix bundle that coordinates a nonheme diiron active site at the core of each monomer (Lindqvist et al., 1996). The structure also showed each monomer contains a deep, narrow, substrate-binding channel, which passes the diiron center on the same side as the proposed oxygen-binding site. The dimeric organization of the desaturase is consistent with results from size exclusion chromatography (McKeon and Stumpf, 1982) and electrospray ionization mass spectrometry (Guy et al., 2011). The dimer interface of the castor enzyme comprises extensive contacts between helices and interdigitating loops (Lindqvist et al., 1996). The crystal structure of the ivy Δ4-16:0-ACP desaturase also presents as a dimer with a very similar overall structure to that of the castor Δ9-18:0-ACP desaturase (Guy et al., 2007).

Quaternary structures of proteins can impact structural and regulatory properties that can be critical for our understanding of enzyme mechanism (Ali and Imperiali, 2005). Experimental evidence from the yeast (Saccharomyces cerevisiae) membrane-bound Δ9 acyl-CoA desaturase, Ole1p, which forms a dimer in vivo, showed that both protomers of the dimer must be catalytically active to produce a functional desaturase dimer, providing evidence for a subunit-subunit interaction (Lou and Shanklin, 2010). Evidence for interaction between subunits of the castor desaturase comes from studies comparing its reactivity upon reduction by the native redox partners compared with chemically reduced desaturase (Broadwater et al., 1998). Whereas interaction with the native redox partners facilitated desaturation, the 4e− chemically reduced desaturase formed a quasi-stable 2:2 enzyme:substrate complex in which O2 was slowly reduced to H2O without desaturation of the substrate acyl chain. These results suggest that desaturation may proceed by an alternating subunit half-of-the-sites reactivity mechanism in which 18:0-ACP binding to one subunit could be coordinated with release of the 18:1-ACP product from the other subunit (Broadwater et al., 1998), as illustrated in Figure 1A. Under such a model, only one of the subunits can be active at a time. Half-of-the-sites reactivity has been previously reported for a related diiron protein, the R2 component of ribonucleotide reductase (Friedle et al., 2010). Key experiments on R2 were performed by expressing and purifying wild-type or mutant subunits, mixing them, and following reaction outcomes from the resulting heterodimers (Sjöberg et al., 1987). Our approach to testing the half-of-the-sites reactivity hypothesis for the soluble desaturase involves measuring the activity of heterodimers comprising one wild-type subunit paired with one inactive mutant subunit. The mixing approach used for R2 was successful because monomers readily disassociate and reassociate in vitro, but this approach is likely not useful for the soluble castor desaturase, which is a homodimeric diiron protein with an extensive dimer interface (of 5,826 Å2) containing multiple hydrophobic interactions, suggesting the dimer is not dissociable (Lindqvist et al., 1996). Consistent with this view, no monomers were detected upon size exclusion chromatography during protein purification or by electrospray ionization mass spectrometry analysis of the purified enzyme (Guy et al., 2011). Thus, to analyze desaturase heterodimers, we developed a coexpression strategy in which separately epitope-tagged wild-type and mutant desaturases were coexpressed in E. coli, which facilitated heterodimer purification via tandem affinity chromatography. A mutant T104K/S202E was designed in which we engineered a salt bridge to block the substrate-binding cavity to preclude substrate binding. X-ray crystallographic analysis confirmed the occlusion of the binding cavity, and no activity was detected for the mutant when expressed alone. Comparison of the activities of purified wild-type-wild-type homodimers with that of wild-type-inactive subunit heterodimers was used to distinguish between the three potential mechanisms shown in Figure 1. The observation of multiple turnovers of the heterodimer is inconsistent with the alternating subunit half-of-the-sites mechanism shown in Figure 1A. Rather, the activities presented are consistent with the half-of-the-sites mechanism shown in Figure 1C. That a single functional active site is sufficient for full activity of the soluble desaturase dimer represents a stark mechanistic contrast to that reported for membrane-bound desaturases in which two functional subunits are required to enable desaturation (Lou and Shanklin, 2010; Lou et al., 2014).

Figure 1.

Figure 1.

Open in a new tab

Scheme illustrating three possible mechanistic models for the catalytic cycle of for acyl-ACP desaturase. The proposed alternating subunit half-of-the-sites reactivity model (A) in which substrate (represented as a straight acyl chain, shown in red, esterified to ACP) binds to the first subunit and is desaturated to form the unsaturated product (represented by the kinked acyl chain); the energy gained upon binding of substrate to the second subunit is transferred to the product-containing subunit, resulting in product release. In this model, the reaction cycle would produce a single product molecule. B, Model shows the two subunits and their respective active sites acting independently. In this model, each reaction cycle would produce two product molecules. C, Model shows negative cooperativity (represented by blue bars) upon binding of substrate that prevents a second substrate from binding, the result being a half-of-the-sites reactivity that differs from A in that substrate binding is not required to facilitate product release. In this model, each reaction cycle would produce one product molecule. Quantitation of the activity of wild-type-nonfunctional mutant design (shown in Fig. 6) can distinguish between models in A and C.

RESULTS

The correct interpretation of our experimental results is dependent on the formation of stable desaturase dimers. We therefore sought to experimentally validate that heterodimers do not reassociate under assay conditions. To evaluate potential exchange of subunits between dimers, we conducted fluorescence resonance energy transfer (FRET)-based experiments. In FRET experiments, a donor chromophore is excited with a specific wavelength of light, and if the acceptor chromophore is sufficiently close to the donor, the energy is transferred from donor to acceptor, which consequently emits light at a characteristic wavelength and is quantitated (Minor, 2006). The experiments presented here were designed to evaluate exchange of monomers between desaturase dimers by mixing populations of dimers labeled exclusively with either donor or acceptor chromophores and measuring FRET signal arising from reassociation of monomers to create heterodimers labeled with donor and acceptor fluorophores. A convenient way to label target proteins is with maleimide-linked fluorescent dyes designed to covalently bind to Cys residues. Analysis of such experiments is simplified by creating uniform singly labeled proteins; however, the wild-type castor desaturase contains three endogenous cysteines (Cys-61, Cys-222, and Cys-328), two of which, Cys-61 and Cys-328, are surface accessible based on analysis of the crystal structure (Fig. 2). Therefore, we mutated the desaturase gene to convert each of these cysteines to Ala residues and added a C-terminal Cys residue (Cys-364) to create a desaturase we refer to as CTC, which has only a single Cys accessible for labeling. CTC expressed in E. coli BL21-Gold (DE3) cells and purified to near homogeneity retained approximately 60% activity relative to that of the wild type. This reduction in activity suggests that stability of the protein may be slightly compromised by the mutations.

Figure 2.

Figure 2.

Open in a new tab

Design and characterization of desaturase mutant C-terminal Cys (CTC). Surface accessibility of Cys in the castor wild-type desaturase. Monomers comprising the desaturase dimer are shown in black and gray, respectively, with surface Cys colored in yellow and the C-terminal residue, Leu-363, shown in cyan. Cys-222 is not visible because it is not surface accessible.

The commercially available thiol-specific dyes Alexa Fluor 488 C5-maleimide and Alexa Fluor 594 C5-maleimide (Invitrogen) were used as the donor and acceptor dyes, respectively. CTC was labeled with either donor or acceptor (Fig. 3). As a positive control, CTC was labeled with an equimolar mixture of donor and acceptor fluorophores (Fig. 3, lane 4). Unlabeled 6×His-tagged wild-type protein is shown as a negative control for acceptor emission signal (Fig. 3, lane 1). SDS-PAGE of these same samples analyzed by Coomassie staining and fluoroimaging revealed that only the unlabeled control protein cannot be visualized by fluoroimaging (Fig. 3), indicating that the labeling is specific.

Figure 3.

Figure 3.

Open in a new tab

Fluorescence labeling of desaturase mutant CTC. SDS-PAGE of the CTC mutant and its corresponding fluorescent conjugates (top, Coomassie; bottom, fluorescence). Purified CTC was labeled with donor (lane 2), acceptor (lane 3), or both dyes simultaneously (dual, lane 4) as detailed under “Materials and Methods.” Unlabeled protein served as a negative control (lane 1). All protein samples were subjected to SDS-PAGE followed by either Coomassie staining or fluorescence imaging.

Assays are typically performed with 1.6 nm desaturase, so we examined the subunit reassociation between two dimer populations uniformly labeled with either Alexa488 (donor) or Alexa594 (acceptor) upon mixing to a final concentration of 1 nm and incubation for a time period exceeding that of an assay (for scheme, see Fig. 4A). The emission wavelength (617 nm) was monitored to measure the rate of reassociation of monomers (Fig. 4B). One-half of the dimers in the dual-labeled sample should be composed of one donor and one acceptor fluorophore (the remainder being composed of either two donors or two acceptors). As expected, upon excitation, the positive-control sample displayed significant fluorescence intensity (Fig. 4B), whereas low levels of FRET signal were emitted by the donor-only or the acceptor-only samples. Similarly, when uniformly labeled donor and acceptor preparations were mixed and incubated prior to evaluating for exchange of monomers between dimers (Fig. 4B), no increase in FRET signal relative to the donor-only or acceptor-only samples was observed, demonstrating that, under assay conditions, no significant reassociation of subunits had occurred. That no exchange of dimers was observed in the potentially mildly destabilized CTC protein suggests that the more stable wild-type dimers would be even less likely to undergo exchange.

Figure 4.

Figure 4.

Open in a new tab

FRET analysis shows no detectable reassociation of monomers between dimers. A, Schematic of the FRET-based assay strategy. Desaturase is indicated by two large white circles, smaller white circles represent C-terminal Cys, and colored circles represent donor fluorophore (yellow) or acceptor fluorophore (purple). B, Fluorescence intensities of four different protein samples at the emission wavelength (617 nm) of the acceptor (Alexa Fluor 594). Labeled CTC were diluted to 1 nm and excited at the excitation wavelength (494 nm) of the donor (Alexa 488). Spectra were recorded as detailed under “Materials and Methods.” a.u., Absorbance units.

Coexpression and Purification of Recombinant Heterodimers

Having demonstrated that the castor desaturase exists as a stable dimer, we proceeded to produce heterodimers composed of one active wild-type subunit and one inactive mutant subunit. To engineer a mutant desaturase that is unable to bind substrate, we examined the substrate-binding cavity of the wild-type desaturase for amino acid locations suitable for substitution with amino acid side chains that could form a salt bridge designed to occlude the cavity, thereby blocking substrate binding. Such a double mutant, T104K/S202E, was therefore constructed to create a salt bridge between K104 and E202. The mutant desaturase was expressed and purified to near homogeneity (Fig. 5A). To assess whether the designed salt bridge between residues K104 and E202 had formed, we solved a crystal structure of the T104K/S202E mutant at 2.9-Å resolution (from which we had also deleted the disordered 31-amino acid N-terminal section of the desaturase to facilitate crystallization). The C-α traces of the wild-type and T104K/S202E mutant desaturases superimpose well, showing that there are no global folding differences (Fig. 5B). The side chains of K104 and E202 are displayed along with a model of the stearic acid substrate in the active site cavity in Figure 5C. The geometry of side chains K104 and E202 is consistent with the formation of a salt bridge that intersects the cavity. Although 2.9-Å resolution is not particularly high, we display the electron density along with the crystallographic model in Figure 5D to show that the correspondence between the data and model is very good, and the positions of these side chains are unambiguously defined.

Figure 5.

Figure 5.

Open in a new tab

Analysis of mutant T104K/S202E (K/E-H). A, SDS-PAGE analysis of purified fractions of wild-type (WT) desaturase tagged with 6×His or strep tags along with mutant T104K/S202E-H. Proteins were separated by 11% SDS-PAGE and subjected to Coomassie Brilliant Blue staining. B, Overall folds of the castor desaturase wild type (green) and T104K/S202E mutant (yellow) in cartoon representation reveals no overall changes in their structures. C, Structure of the substrate binding cavity of mutant K/E with substrate modeled in place to illustrate the steric blockage of the cavity. D, Superimposed electron density map with the structure to show that there is good fit to the observed density in the region of the T104K/S202E substitutions. E, Size exclusion chromatography of purified wild-type (blue lines, top) and K/E mutant (red lines, bottom) desaturases in the presence or absence of substrate.

As a direct test of substrate binding to the T104K/S202E mutant, size exclusion chromatography experiments were performed on samples containing the wild-type (Fig. 5E, top) or T104K/S202E mutant (Fig. 5E, bottom) enzymes in the presence or absence of 18:0-ACP substrate and, as an additional negative control, in the presence of holo-ACP (i.e. ACP lacking the acyl chain). As expected, the wild-type enzyme plus substrate eluted significantly earlier than either desaturase alone or desaturase plus holo-ACP. In contrast, the T104K/S202E mutant showed no change in mobility upon incubation of substrate relative to its mobility in the absence of substrate or upon incubation with holo-ACP, directly demonstrating the mutant does not bind substrate. Consistent with a lack of substrate binding, enzyme assays confirmed that T104K/S202E mutant homodimers have no detectable activity (data not shown).

Isolation of Wild-Type*-T104K/S202E Heterodimers

Our goal was to isolate the wild type-mutant heterodimers with 1:1 stoichiometry by in vivo coexpression of wild-type and mutant subunits. The approach requires approximately equal expression of the two proteins. We therefore used the pETDuet expression vector, which has two separate T7 lac promoters to drive the expression of the target genes. To avoid homologous recombination between the open reading frames encoding wild-type and mutant proteins, we redesigned a wild-type DNA coding sequence (referred to as wild-type*), which shares only 62% DNA identity with the native sequence while maintaining the same translated amino acid sequence. To facilitate affinity purification, the wild type was tagged with a 2×strep tag, whereas the mutant protein was 6×His tagged. See Figure 6A for a schematic of the experimental heterodimer samples. Purified wild-type*-S+wild-type-H heterodimers were then identified by SDS-PAGE (Fig. 6B); under these conditions, the desaturases exhibit different mobilities due to the presence of the different affinity tags. The purified protein species behave as a heterodimer of approximately 1:1 stoichiometry (Fig. 6B).

Figure 6.

Figure 6.

Open in a new tab

Heterodimer strategy purification and analysis. A, Schematic of heterodimers tested in this work. Desaturase monomers are represented by ovals with a deep substrate binding channel. Black circles represent diiron clusters. S and H represent strep and His affinity tags, respectively. Left, Wild-type (WT) homodimer control: wild-type*-S+wild-type-H; right, heterodimer of wild-type*-S+T104K/S202E-H (i.e. one monomer defective in substrate binding, indicated by an x). B, Portion of Coomassie-stained PAGE gels showing purified wild-type*-S+wild- type-H and wild-type*-S+T104K/S202E-H. The relative intensities of two subunits in each dimer were examined by densitometry analysis using ImageQuant TL (GE Healthcare). Intensities of wild-type*-S were set to 1. The results are shown as the mean ± sd of three replicates.

The Wild Type-Mutant Heterodimers Have Equivalent Activity to Wild-Type Homodimers

To distinguish between the models presented in Figure 1 (i.e. the alternating subunit half-of-the-sites mechanism in which substrate binding is required for product release [Fig. 1A], independent active sites [Fig. 1B], and half-of-the-sites mechanism in which substrate binding is not required for product release [Fig. 1C]), the activities of heterodimers in which a wild-type subunit was paired with the T104K/S202E mutant that is unable to bind substrate were determined along with that of wild-type homodimers. Preliminary experiments were first performed to validate the linearity of the desaturase assays for the enzyme concentrations and times used in this study. In three independent experiments, the heterodimer (wild-type*-S+K/E-H) exhibited no significant difference in activity from that of equivalently tagged wild-type homodimers (wild-type*-S+wild-type-H; Fig. 7).

Figure 7.

Figure 7.

Open in a new tab

Relative activities of purified heterodimers. Relative activities of purified wild-type (WT)*-S+wild-type-H and wild-type*-S+T104K/S202E-H dimers are presented as the mean ± sd of three replicates. One hundred percent indicates a specific activity of 416 nm min−1 mg−1.

DISCUSSION

Enzyme oligomerization can lead to increased protein stability and permits active sites to be shared among protomers. In addition, subunit-subunit interactions can result in biochemical activation or inhibition, and dynamic reversible oligomerization plays an important role in signal transduction (Ali and Imperiali, 2005). Previous work on the castor desaturase invoked subunit-subunit communication to explain why the 4e−-reduced desaturase-substrate complex forms a long-lived 2:2 substrate:desaturase complex that performs oxidase rather than desaturase chemistry (Broadwater et al., 1998). In the same work, the authors propose that the desaturase may operate by a half-of-the-sites reactivity mechanism in which binding of substrate to one subunit is coordinated with product release from the other subunit as diagrammed in Figure 1A. Under this scenario, only one subunit would be catalytically active at a time, and the active subunit would alternate between the two protomers with each successive reaction cycle. A prediction of the coordination of substrate-binding and product-release hypothesis is that heterodimers composed of one wild type and one subunit deficient in substrate binding would be able to perform only a single turnover, at which time the product would be retained in the active site cavity because substrate is unable to bind to the defective subunit to facilitate release of product from the wild-type subunit. However, multiple turnovers were observed in assays of the desaturase heterodimer comprising one wild-type monomer and one monomer deficient in substrate binding. These data are inconsistent with an alternating subunit half-of-the-sites mechanism. Because a wild-type homodimer has 2× the number of active sites relative to a heterodimer composed of one wild-type and one inactive subunit, if the subunits act independently, one would expect the wild-type homodimer to exhibit twice the activity of the heterodimer, as diagrammed in Figure 1B. The activity of the heterodimer exceeds one-half that of the wild-type homodimer, suggesting negative cooperativity between the subunits upon binding of the first substrate. Indeed, that the heterodimer does not show any significant decrease in activity relative to the wild-type homodimer indicates that only one monomer of the dimer is active at a time (i.e. that the desaturase operates by a half-of-the-sites reactivity mechanism, as diagrammed in Fig. 1C).

Whether in desaturase wild-type homodimers one subunit is always active and the other always inactive, or whether the subunits can alternate roles in a manner similar to that described for the E1 component of the bacterial pyruvate dehydrogenase complex is unknown (Frank et al., 2004; Seifert et al., 2006). A conformational signal transmitted from the active subunit to the other subunit, which prevents simultaneous productive complex formation in both subunits, seems the most likely mechanism to explain the half-of-the-sites reactivity. An example of such a mechanism comes from Asp β-semialdehyde dehydrogenase, in which substrate binding induces conformational changes that induce a catalytically competent geometry in one active site of the dimer while distorting it in the other (Nichols et al., 2004). The nature of the structural signal that inactivates the second subunit of the desaturase is currently unknown. Crystal structures of several acyl-ACP desaturase homodimers solved in the absence of substrate have shown the subunits to be structurally equivalent (Lindqvist et al., 1996; Guy et al., 2007); however, several biochemical studies have presented experimental evidence to support conformational changes in response to substrate binding. For instance, binding studies using fluorescence anisotropy showed that initial substrate binding to produce a 1:2 substrate:enzyme complex is 30-fold tighter than that of a second substrate binding that produces a 2:2 substrate:enzyme complex (Haas and Fox, 2002). The negative cooperativity for substrate binding was attributed to changes in conformation of the desaturase upon binding of the first substrate. In a second study, conformational changes of the desaturase resulting from substrate binding were proposed to explain the +104-mV increase in redox potential that favors reduction of substrate-bound desaturase relative to unbound enzyme by a process referred to as redox gating (Reipa et al., 2004). Solving a crystal structure of the heterodimer with a 1:2 substrate:enzyme complex could provide valuable insights into the structural nature of the communication between the subunits that underlies the half-of-the-sites mechanism. This could be accomplished by cocrystallizing the heterodimer with substrate, but to facilitate this, the yield of purified heterodimer would have to increase from its current <1% level typical of complexes purified by tandem affinity purification (Kerrigan et al., 2011). Future studies will focus on optimizing heterodimer yield to facilitate detailed structural and kinetic analysis. In addition, the heterodimer approach developed in this work is currently being exploited to define the electron transport pathway to the diiron centers of the desaturase.

Similar half-of-the-sites reactivity to that described here has been demonstrated for another structurally related dimeric diiron enzyme, the B2 component of ribonucleotide reductase (Sjöberg et al., 1987). In this case, experiments were based on heterodimers comprising one wild-type subunit and one catalytically inactive mutant, Tyr-122-Phe. The heterodimer containing only one Tyr-122 has the same enzymatic activity as that of the wild-type dimer that contains a Tyr-122 in both of its subunits (Larsson et al., 1988). More recently, two diiron-containing bacterial multicomponent monooxygenases, phenol hydroxylase and toluene/o-xylene monooxygenases, were shown to have maximal product yields of 50% of their diiron centers, consistent with a half-of-the-sites mechanism (Tinberg et al., 2011). In contrast, a related diiron multicomponent monooxygenase protein, the methane monooxygenase hydroxylase component, was shown to have a full-sites mechanism (Liu et al., 1997).

The data presented here raise the question as to why enzymes would have evolved as dimers with two active sites but have a mechanism that allows only one to operate at a time. A possible explanation is that such a mechanism could buffer against enzyme inactivation by oxidative damage resulting from poor coordination of events leading to oxygen activation (Tinberg et al., 2011). Diiron enzymes are potentially vulnerable to such damage because they activate molecular oxygen to produce highly oxidizing species necessary to activate C-H bonds. A heterodimer composed of one active and one inactivated subunit would maintain the same activity as a wild-type homodimer, thereby buffering damage to one of its active sites. It has been noted that newly evolved enzymes tend to exhibit lower catalytic activity and stability than the archetypes from which they evolved (Shanklin, 2000). These properties have hindered attempts to accumulate high levels of unusual fatty acids in crop plants upon the expression of variant acyl-ACP desaturases (Nguyen et al., 2010). It is possible that coexpression of a variant desaturase along with its progenitor (or a T104K/S202E mutant thereof) from the same species might facilitate heterodimer formation that could stabilize the variant and thereby increase unusual fatty acid production.

Previously, we demonstrated that the Δ9-acyl-CoA membrane-bound desaturase Ole1p from bakers’ yeast and a series of higher plant membrane desaturases, fatty acid desaturase2 (FAD2), FAD3, FAD6, FAD7, and FAD8, like the evolutionarily unrelated soluble desaturases, form dimers in vivo (Lou and Shanklin, 2010; Lou et al., 2014). For both the yeast Ole1 (Lou and Shanklin, 2010) and plant FAD2 desaturases (Chapman et al., 2001, 2008), coexpression of inactive mutant subunits results in the inactivation of the endogenous wild-type desaturases, presumably by the formation of heterodimers. Thus, for the membrane class of desaturase enzymes, two catalytically competent subunits are required for catalysis. That only one functional monomer is required for full activity of the soluble acyl-ACP desaturase dimer, whereas two active subunits are required for activity of membrane desaturases, underscores a fundamental mechanistic difference between the convergently evolved soluble and membrane classes of diiron desaturases.

MATERIALS AND METHODS

Materials

The pET28b and pETDuet-1 vectors were purchased from Novagen. All restriction enzymes and T4 DNA ligase were from New England BioLabs. Primers were synthesized by Integrated DNA Technologies. The QIAquick spin plasmid kit and QIAquick gel extraction kit were from Qiagen. The Escherichia coli strain BL21-Gold (DE3) was purchased from Stratagene. The Econo-Pac 10DG desalting columns are from Bio-Rad. Alexa Fluor 488 C5-maleimide and Alexa Fluor 594 C5-maleimide were purchased from Invitrogen.

Plasmid Construction

The construction of recombinant plasmids used in this study is detailed in Table I. Manipulation of nucleic acids was performed using standard protocols (Sambrook et al., 2003). The coding region for the mature wild-type castor Ricinus communis ∆9-18:0-ACP was PCR amplified and ligated into pET24b plasmid (Novagen) with a C-terminal 6×His or strep tag (referred to as wild-type-H or wild-type-S). The mutant T104K/S202E was constructed by PCR-driven overlap extension according to the method of Heckman and Pease (Ho et al., 1989) using wild-type-H as a template. A Cys-less mutant was also generated using pET24b with an exogenous Cys at the C terminus of the sequence (referred to as CTC). For the dual-tagged heterodimer expression, the T104K/S202E mutant (and codon-modified wild-type sequence, which we refer to as wild-type*) was ligated into multiple cloning site MCS1 and MCS2 of the pETDuet-1 vector (Novagen) with 6×His and 2×strep tags, respectively (referred to as T104K/S202E-H+wild-type*-S).

Table I. Summary of recombinant plasmids and expressed proteins.

Plasmid Promoter Protein Expressed or Coexpressed Protein Tag
pET24b T7 Wild-type-H 6×His
pET24b T7 Wild-type-S 1×Strep
pET24b T7 T104K/S202E-H 6×His
pET24b T7 CTC
pETDuet-1 T7lac Wild-type-H, Wild-type*-S 6×His and 2×Strep
pETDuet-1 T7lac Wild-type*-S, T104K/S202E-H 6×His and 2×Strep
Open in a new tab

The wild-type castor desaturase and codon-modified wild-type sequence were ligated into pETDuet-1 as well to serve as a control (referred to as wild-type-H+wild-type*-S). All constructs were sequenced to verify their integrity and that the genes were cloned in frame with the appropriate epitope tags.

Protein Expression and Purification

All pET expression plasmids were transformed into electrocompetent E. coli BL21-Gold (DE3) cells to obtain expression under the control or T7 or T7 lac promoter. Typically, cells containing the expression plasmid were grown in 2.8-L baffled Fernbock flasks in 1 L of Lauria-Bertellini broth at 37°C with shaking at 275 rpm to an optical density at 600 nm of 0.4 to approximately 0.6, and the expression was induced at 30°C for 4 h (or 20°C overnight, in the case of the heterodimer proteins) by the addition of isopropylthio-β-galactoside to a final concentration of 0.2 mm.

For two-step affinity purification of T104K/S202E-H+wild-type*-S heterodimers, cells were harvested by centrifugation at 6000g for 8 min. Cell paste was resuspended in five volumes of lysis buffer A (6 mm HEPES, 6 mm MES, 6 mm NaOAc, 5 mm MgCl2, pH 7.5) supplemented with protease inhibitor cocktail (EDTA free, Sigma-Aldrich). One hundred micrograms of DNAse I was added per gram of cells, and the slurry was then passed through a French pressure cell at 12,000 psi. The supernatant was collected by centrifugation at 20,000g for 30 min at 4°C followed by fractionation with ammonium sulfate. Solid ammonium sulfate was slowly added to the supernatant to reach 25% saturation at 4°C. After stirring for 15 min, the mixture was centrifuged at 20,000g for 15 min to separate the supernatant and pellet fractions. The supernatant was added with ammonium sulfate to a final concentration of 90% (w/v) and stirred at 4°C for another 15 min. After a second centrifugation, the pellet was resuspended in lysis buffer B (50 mm NaH2PO4, 300 mm NaCl, 10 mm imidazole, pH 8.0) and applied to a Ni-NTA column (Qiagen). The column was washed with lysis buffer plus 20 mm imidazole and eluted in lysis buffer plus 250 mm imidazole. Eluted fractions containing a mixture of T104K/S202E-H homodimers and T104K/S202E-H+wild-type*-S heterodimers were pooled and applied to a Strep-Tactin column (Qiagen). The column was washed with wash buffer (100 mm Tris-HCl, 150 mm NaCl, pH 8.0) and eluted with wash buffer plus 2.5 mm d-desthiobiotin (Sigma-Aldrich). The eluted fractions were pooled and the protein concentrations were measured with the Bradford assay (Sigma-Aldrich) using bovine serum albumin as a standard.

For one-step cation exchange purification of other proteins, the cells were lysed as described above. The separated supernatant was applied to a 20CM column (Applied Biosystems) and purified with an NaCl gradient elution (0–1.5 m). The purity of the samples was assayed by SDS-PAGE and Coomassie staining.

SDS-PAGE

Protein samples were separated using 11% (w/v) SDS-PAGE followed by Coomassie Brilliant Blue staining. Densitometry analysis of the stained gel was performed using ImageQuant TL software (GE Healthcare).

Fluorescence Labeling of Cys Mutant with Alexa Dye

Purified CTC mutant of the castor ∆9-18:0 desaturase was labeled simultaneously with both donor and acceptor fluorescent probes using procedures from Life Technologies for labeling Cys-containing proteins, with modifications. One milligram of purified protein was reduced by 5 mm dithiothreitol at 4°C for 1 h. The residual dithiothreitol was then removed using Micro Bio-Spin 6 columns (Bio-Rad). The protein was then added to an Eppendorf microcentrifuge tube containing a 2.5-fold molar excess of Alexa Fluor 488 C5-maleimide (Alexa 488, donor) and Alexa Fluor 594 C5-maleimide (Alexa 594, acceptor) dyes (Life Technologies) dissolved in a ratio of 1:1 in dimethyl sulfoxide. The tube was purged with nitrogen and the reaction was carried out in darkness for 16 h at 4°C. The reaction mixture was then loaded to a HiLoad 16/60 Superdex 75 column (Amersham-Pharmacia Biotech) equilibrated in HEPES-MES buffer to separate the unreacted dye from the labeled protein. Alternatively, CTC was used for donor- or acceptor-only labeling similar to what is described above.

Fluorescence Measurements

All the spectral measurements were performed at room temperature. The standard buffer was 6 mm HEPES, 6 mm MES, and 6 mm NaOAc (pH 7.5). Fluorescence spectroscopic measurements used an ISS model PC-1 photon-counting spectrofluorometer equipped with a Xenon arc lamp. Dual-labeled and donor-only-labeled samples were excited at 494 nm (excitation slit, 8 nm) with an emission spectrum from 454 to 750 nm. For acceptor-only-labeled samples, the excitation wavelength was set to 590 nm with an emission spectrum from 590 to 750 nm and the same setting for the excitation slit.

Size Exclusion Chromatography

Desaturase samples in 20 mm HEPES and 300 mm NaCl (pH 7.0) were incubated alone, with holo-ACP, or with 18:0-ACP before separation using a 30- × 0.46-cm Tosoh Super SW3000 HPLC column (Tosoh Biosciences) developed with a flow rate of 0.3 mL min−1. Typical protein samples contained 0.3 mm desaturase with 0.5 mm of holo-ACP or 18:0-ACP.

Determination of Crystal Structure of the T104K/S202E Mutant

The hanging drop vapor diffusion method was used for crystallization. The protein solution was mixed in equal amounts (2.7 + 2.7 μL) and equilibrated against a reservoir solution in Falcon 24-well tissue culture plates (Corning). The protein solution consisted of 9.3 mg mL−1 protein in 20 mm HEPES and 70 mm NaCl, pH 7.0. The reservoir solution consisted of 20% (w/v) polyethylene glycol 4000, 80 mm Tris hydrochloride buffer (pH 8.5), 120 mm magnesium chloride, and 20% (w/v) glycerol.

Data Collection and Refinement

Protein crystals were frozen under a stream of liquid nitrogen, and cryo temperature diffraction data to 2.8-Å resolution were collected at National Synchrotron Light Source (Brookhaven National Laboratory, NY) beamline X25 and processed with HKL2000 (Otwinowski and Minor, 1997) and CCP4 suite software packages (Winn et al., 2011). The protein crystallized in space group P31 with cell parameters a = b = 139.7 and c = 86.87 and six molecules belonging to three dimers in the asymmetric unit. Molecular replacement was performed using MOLREP (Vagin and Teplyakov, 1997) and refinement performed in autoBUSTER (Smart et al., 2012) and REFMAC5 (Murshudov et al., 1997), both supporting automatic local noncrystallographic symmetry restraints definitions for convenient use. Manual model building was performed in COOT (Emsley and Cowtan, 2004) where the option to make a noncrystallographic symmetry averaged map was particularly useful. The 18:0-ACP substrate was modeled into the active site cavity as previously described (Lindqvist et al., 1996).

Desaturase Assays

Typically castor 18:0-ACP desaturase preparations were assayed at a concentration of 1.6 nm in assays containing 1.5 μm [1-14C]18:0-ACP substrate using recombinant spinach (Spinacia oleracea) ACP-I as previously described (Rock and Garwin, 1979; Beremand et al., 1987; Cahoon et al., 1997; Cahoon et al., 1998). Assays are typically performed with desaturase and 1.5 μm [1-14C]18:0-ACP substrate for between 4 and 15 min. Labeled methyl esters of fatty acids that include the assay products and unreacted substrates were analyzed by argentation thin-layer chromatography, and radioactivity in products was quantified via phosphorimaging.

Sequence data from this article can be found in the GenBank/EMBL data libraries under accession number 4V0J.

Acknowledgments

We thank Drs. William Studier, Paul Freimuth, Allen Orville, Walter Mangel, and William McGrath for helpful discussion; and Michael Blewitt, Vito Graziano, Dr. Lin Bai, Ed Whittle, and John Trunk for technical assistance.

Glossary

ACP

acyl carrier protein

FRET

fluorescence resonance energy transfer

CTC

C-terminal Cys

Footnotes

1

This work was supported by the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences of the U.S. Department of Energy (grant no. DOE KC0304000) and the Swedish Research Council. Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences (contract no. DE–AC02–98CH10886).

J.S. conceived the study; Q.L. and J.S. designed the experimental plan; Q.L. performed the biochemical experiments; M.M., J.G., and Y.L. performed the crystallography experiments and analysis; J.C. provided technical assistance with protein purification; Q.L. and J.S. analyzed the data and wrote the article with assistance of all coauthors.

[OPEN]

Articles can be viewed without a subscription.

References

  1. Ali MH, Imperiali B (2005) Protein oligomerization: how and why. Bioorg Med Chem 13: 5013–5020 [DOI] [PubMed] [Google Scholar]
  2. Beremand PD, Hannapel DJ, Guerra DJ, Kuhn DN, Ohlrogge JB (1987) Synthesis, cloning, and expression in Escherichia coli of a spinach acyl carrier protein-I gene. Arch Biochem Biophys 256: 90–100 [DOI] [PubMed] [Google Scholar]
  3. Broadwater JA, Ai J, Loehr TM, Sanders-Loehr J, Fox BG (1998) Peroxodiferric intermediate of stearoyl-acyl carrier protein delta 9 desaturase: oxidase reactivity during single turnover and implications for the mechanism of desaturation. Biochemistry 37: 14664–14671 [DOI] [PubMed] [Google Scholar]
  4. Cahoon EB, Lindqvist Y, Schneider G, Shanklin J (1997) Redesign of soluble fatty acid desaturases from plants for altered substrate specificity and double bond position. Proc Natl Acad Sci USA 94: 4872–4877 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cahoon EB, Shah S, Shanklin J, Browse J (1998) A determinant of substrate specificity predicted from the acyl-acyl carrier protein desaturase of developing cat’s claw seed. Plant Physiol 117: 593–598 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chapman KD, Austin-Brown S, Sparace SA, Kinney A, Ripp KG, Pirtle IL, Rirtle RM (2001) Transgenic cotton plants with increased seed oleic acid content. J Am Oil Chem Soc 78: 941–947 [Google Scholar]
  7. Chapman KD, Neogi PB, Hake KD, Stawska AA, Speed TR, Cotter MQ, Garrett DC, Kerby T, Richardson CD, Ayre BG. , et al. (2008) Reduced oil accumulation in cottonseeds transformed with a Brassica nonfunctional allele of a delta-12 fatty acid desaturase (FAD2). Crop Sci 48: 1470–1481 [Google Scholar]
  8. Dyer DH, Lyle KS, Rayment I, Fox BG (2005) X-ray structure of putative acyl-ACP desaturase DesA2 from Mycobacterium tuberculosis H37Rv. Protein Sci 14: 1508–1517 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132 [DOI] [PubMed] [Google Scholar]
  10. Fox BG, Lyle KS, Rogge CE (2004) Reactions of the diiron enzyme stearoyl-acyl carrier protein desaturase. Acc Chem Res 37: 421–429 [DOI] [PubMed] [Google Scholar]
  11. Frank RAW, Titman CM, Pratap JV, Luisi BF, Perham RN (2004) A molecular switch and proton wire synchronize the active sites in thiamine enzymes. Science 306: 872–876 [DOI] [PubMed] [Google Scholar]
  12. Friedle S, Reisner E, Lippard SJ (2010) Current challenges of modeling diiron enzyme active sites for dioxygen activation by biomimetic synthetic complexes. Chem Soc Rev 39: 2768–2779 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gomes CM, Le Gall J, Xavier AV, Teixeira M (2001) Could a diiron-containing four-helix-bundle protein have been a primitive oxygen reductase? ChemBioChem 2: 583–587 [DOI] [PubMed] [Google Scholar]
  14. Guy JE, Abreu IA, Moche M, Lindqvist Y, Whittle E, Shanklin J (2006) A single mutation in the castor Δ9-18:0-desaturase changes reaction partitioning from desaturation to oxidase chemistry. Proc Natl Acad Sci USA 103: 17220–17224 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Guy JE, Whittle E, Kumaran D, Lindqvist Y, Shanklin J (2007) The crystal structure of the ivy Delta4-16:0-ACP desaturase reveals structural details of the oxidized active site and potential determinants of regioselectivity. J Biol Chem 282: 19863–19871 [DOI] [PubMed] [Google Scholar]
  16. Guy JE, Whittle E, Moche M, Lengqvist J, Lindqvist Y, Shanklin J (2011) Remote control of regioselectivity in acyl-acyl carrier protein-desaturases. Proc Natl Acad Sci USA 108: 16594–16599 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Haas JA, Fox BG (2002) Fluorescence anisotropy studies of enzyme-substrate complex formation in stearoyl-ACP desaturase. Biochemistry 41: 14472–14481 [DOI] [PubMed] [Google Scholar]
  18. Ho SN, Hunt HD, Horton RM, Pullen JK, Pease LR (1989) Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77: 51–59 [DOI] [PubMed] [Google Scholar]
  19. Kerrigan JJ, Xie Q, Ames RS, Lu Q (2011) Production of protein complexes via co-expression. Protein Expr Purif 75: 1–14 [DOI] [PubMed] [Google Scholar]
  20. Larsson A, Karlsson M, Sahlin M, Sjöberg BM (1988) Radical formation in the dimeric small subunit of ribonucleotide reductase requires only one tyrosine 122. J Biol Chem 263: 17780–17784 [PubMed] [Google Scholar]
  21. Lindqvist Y. (2001) Delta nine stearoyl-acyl carrier protein desaturase. John Wiley & Sons, Chichester UK [Google Scholar]
  22. Lindqvist Y, Huang W, Schneider G, Shanklin J (1996) Crystal structure of delta9 stearoyl-acyl carrier protein desaturase from castor seed and its relationship to other di-iron proteins. EMBO J 15: 4081–4092 [PMC free article] [PubMed] [Google Scholar]
  23. Liu Y, Nesheim JC, Paulsen KE, Stankovich MT, Lipscomb JD (1997) Roles of the methane monooxygenase reductase component in the regulation of catalysis. Biochemistry 36: 5223–5233 [DOI] [PubMed] [Google Scholar]
  24. Lou Y, Schwender J, Shanklin J (2014) FAD2 and FAD3 desaturases form heterodimers that facilitate metabolic channeling in vivo. J Biol Chem 289: 17996–18007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lou Y, Shanklin J (2010) Evidence that the yeast desaturase Ole1p exists as a dimer in vivo. J Biol Chem 285: 19384–19390 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. McKeon TA, Stumpf PK (1982) Purification and characterization of the stearoyl-acyl carrier protein desaturase and the acyl-acyl carrier protein thioesterase from maturing seeds of safflower. J Biol Chem 257: 12141–12147 [PubMed] [Google Scholar]
  27. Minor L. (2006) Handbook of assay development in drug discovery. CRC/Taylor & Francis, Boca Raton, FL [Google Scholar]
  28. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53: 240–255 [DOI] [PubMed] [Google Scholar]
  29. Napier JA. (2007) The production of unusual fatty acids in transgenic plants. Annu Rev Plant Biol 58: 295–319 [DOI] [PubMed] [Google Scholar]
  30. Nguyen HT, Mishra G, Whittle E, Pidkowich MS, Bevan SA, Merlo AO, Walsh TA, Shanklin J (2010) Metabolic engineering of seeds can achieve levels of omega-7 fatty acids comparable with the highest levels found in natural plant sources. Plant Physiol 154: 1897–1904; erratum Nguyen M, Whittle P, Bevan M, Walsh S (2011) Plant Physiol 155: 1473 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nichols CE, Dhaliwal B, Lockyer M, Hawkins AR, Stammers DK (2004) High-resolution structures reveal details of domain closure and “half-of-sites-reactivity” in Escherichia coli aspartate beta-semialdehyde dehydrogenase. J Mol Biol 341: 797–806 [DOI] [PubMed] [Google Scholar]
  32. Otwinowski Z, Minor W (1997) Processing of X-ray Diffraction Data Collected in Oscillation Mode, Vol 276 Academic Press, Waltham, MA: [DOI] [PubMed] [Google Scholar]
  33. Reipa V, Shanklin J, Vilker V (2004) Substrate binding and the presence of ferredoxin affect the redox properties of the soluble plant Delta9-18:0-acyl carrier protein desaturase. Chem Commun (Camb) (21): 2406–2407 [DOI] [PubMed] [Google Scholar]
  34. Rock CO, Garwin JL (1979) Preparative enzymatic synthesis and hydrophobic chromatography of acyl-acyl carrier protein. J Biol Chem 254: 7123–7128 [PubMed] [Google Scholar]
  35. Sambrook J, Fritsch EF, Maniatis T (2003) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. [Google Scholar]
  36. Sampath H, Ntambi JM (2011) The role of stearoyl-CoA desaturase in obesity, insulin resistance, and inflammation. Ann N Y Acad Sci 1243: 47–53 [DOI] [PubMed] [Google Scholar]
  37. Seifert F, Golbik R, Brauer J, Lilie H, Schröder-Tittmann K, Hinze E, Korotchkina LG, Patel MS, Tittmann K (2006) Direct kinetic evidence for half-of-the-sites reactivity in the E1 component of the human pyruvate dehydrogenase multienzyme complex through alternating sites cofactor activation. Biochemistry 45: 12775–12785 [DOI] [PubMed] [Google Scholar]
  38. Shanklin J. (2000) Exploring the possibilities presented by protein engineering. Curr Opin Plant Biol 3: 243–248 [PubMed] [Google Scholar]
  39. Shanklin J, Cahoon EB (1998) Desaturation and related modifications of fatty acids. Annu Rev Plant Physiol Plant Mol Biol 49: 611–641 [DOI] [PubMed] [Google Scholar]
  40. Shanklin J, Guy JE, Mishra G, Lindqvist Y (2009) Desaturases: emerging models for understanding functional diversification of diiron-containing enzymes. J Biol Chem 284: 18559–18563 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shanklin J, Somerville C (1991) Stearoyl-acyl-carrier-protein desaturase from higher plants is structurally unrelated to the animal and fungal homologs. Proc Natl Acad Sci USA 88: 2510–2514 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Sjöberg BM, Karlsson M, Jörnvall H (1987) Half-site reactivity of the tyrosyl radical of ribonucleotide reductase from Escherichia coli. J Biol Chem 262: 9736–9743 [PubMed] [Google Scholar]
  43. Smart OS, Womack TO, Flensburg C, Keller P, Paciorek W, Sharff A, Vonrhein C, Bricogne G (2012) Exploiting structure similarity in refinement: automated NCS and target-structure restraints in BUSTER. Acta Crystallogr D Biol Crystallogr 68: 368–380 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Tinberg CE, Song WJ, Izzo V, Lippard SJ (2011) Multiple roles of component proteins in bacterial multicomponent monooxygenases: phenol hydroxylase and toluene/o-xylene monooxygenase from Pseudomonas sp. OX1. Biochemistry 50: 1788–1798 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vagin A, Teplyakov A (1997) Molrep: An automated program for molecular replacement. J Appl Cryst 30: 1022–1025 [Google Scholar]
  46. Whittle E, Shanklin J (2001) Engineering delta 9-16:0-acyl carrier protein (ACP) desaturase specificity based on combinatorial saturation mutagenesis and logical redesign of the castor delta 9-18:0-ACP desaturase. J Biol Chem 276: 21500–21505 [DOI] [PubMed] [Google Scholar]
  47. Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR, Keegan RM, Krissinel EB, Leslie AG, McCoy A, et al. (2011) Overview of the CCP4 suite and current developments. Acta Crystallogr D Biol Crystallogr 67: 235–242 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES