PROTEIN EXPRESSION AND PURIFICATION 10, 175–179 (1997)ARTICLE NO. PT970730

Phytoene Desaturase: Heterologous Expression in anActive State, Purification, and Biochemical Properties

Christian Schneider, Peter Boger, and Gerhard Sandmann*,1

Lehrstuhl Physiologie und Biochemie der Pflanzen, Universitat Konstanz, 78464 Konstanz, Germany; and*Botanisches Institut, FB Biologie, J. W. Goethe Universitat, P.O. Box 111932, 60054 Frankfurt, Germany

Received October 1, 1996, and in revised form January 23, 1997

tion product differs structurally and functionally fromConditions were developed for heterologous expres- another type found in heterotrophic bacteria and fungi.

sion in Escherichia coli of the membrane-bound cyano- This bacterial enzyme in general carries out a four-bacterial/plant-type phytoene desaturase (PDS) from step desaturation of phytoene to lycopene (4).Synechococcus in an active form. Decrease of growth Several different phytoene desaturase genes have re-temperature for the transformant to 287C resulted in cently been overexpressed in Escherichia coli (5,6), in-an increase of proteins in a supernatant fraction ob- cluding the gene from Synechococcus (7). The condi-tained after pressure disruption (20 MPa) of cells and tions used resulted in high-level production of thiscentrifugation. This supernatant in which the highest cyanobacterial/algal/higher plant-type phytoene desa-PDS activity was found was used for purification to a turase in an inactive state. The enzyme was seques-homogenous protein by ammonium sulfate precipita- tered in inclusion bodies from which it had to be solubi-tion and DEAE chromatography. The purified PDS was lized by urea. After optimization of reactivation proce-employed to determine substrate specificity and cofac- dures only a low level of enzyme activity could betor requirement. Substrates in addition to phytoene detected. Therefore, it was necessary to use radioac-were phytofluene and 1,2-epoxy phytoene which were

tively labeled phytoene in order to detect z-carotene asconverted to z-carotene and the corresponding 1,2-ep-a reaction product.oxide. The reaction was stimulated by NAD, NADP, and

We now report conditions which allow the expressionoxygen. The Km values determined for phytoene andof substantial amounts of the phytoene desaturaseNADP were 3.5 mM and 14.3 mM, respectively. q 1997(PDS)2 in an active and soluble form which facilitatesAcademic Pressthe purification of this enzyme. The activity obtained ishigh enough to establish an enzymatic assay involvingnonradioactive substrates. This makes it possible to

Cyclic and acyclic carotenoids are indispensible for carry out enzyme kinetic studies and also to analyzea functional photosynthesis. They protect the photosyn- phytoene-related carotenoids as potential substrates ofthetic apparatus from excess light preventing photody- this PDS.namic damage of chlorophylls and destruction of otherchloroplast components (1). In all organisms with oxy- MATERIALS AND METHODSgenic photosynthesis, including cyanobacteria, algae,

Plasmids and growth of cultures. The E. coliand higher plants, cyclic carotenoids are essential forstrain JM101 was the host for the PDS-overex-the photosynthetic process (2). All accumulating cyclicpressing plasmid pPDSdel35 (7,8) and plasmidcarotenoids are derived from the maximally desatur-pACCRT-EB which mediates the formation of phy-ated lycopene which possesses 11 conjugated doubletoene (6). Both transformants were routinely grownbonds. These double bonds are introduced into phy-at 287C, if not stated otherwise, in Luria–Bertanitoene by two subsequently acting enzymes, phytoenebroth containing either ampicillin (50 mg/ml) ordesaturase and z-carotene desaturase (3). This phy-chloramphenicol (30 mg/ml) as described by Maniatistoene desaturase which forms z-carotene as the reac-

1 To whom correspondence and reprint requests should be ad- 2 Abbreviations used: DEAE, diethylaminoethyl cellulose; PDS,plant-type phytoene desaturase; SDS, sodium dodecyl sulfate.dressed. Fax: /49 69 7982 4822.

1751046-5928/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID PEP 0730 / 6q10$$$101 06-04-97 23:58:35 pepa AP: PEP

SCHNEIDER, BOGER, AND SANDMANN176

TABLE 1et al. (9). The protein was expressed by inducing anovernight culture with 0.25 mM b-D-isopropylthioga- Temperature-Dependent Accumulation of Expressed PDS

from Synechococcus in the Soluble Fraction afterlactopyranoside for another 3 h. The phytoene-pro-Breaking of Escherichia coli Cellsducing culture was grown for 48 h. Phycomyces

blakesleeanus mutant S442 (10) was grown as aSoluble protein PDS activitysource for phytofluene in a medium according to

Growth from 100 ml (ng z-caroteneThan et al. (11) for 3 days at 287C under constanttemperature cells ml01 sup h01)

shaking. 1,2-Epoxyphytoene was from a former prep-aration (12). 237C 17.2 mg 70

257C 14.4 mg 74Isolation and purification of PDS. Cells were har-287C 32.9 mg 106

vested by centrifugation at 16,000g for 15 min and the 337C 24.5 mg 61pellet was resuspended in 0.1 M Tris–HCl buffer, pH 377C 12.8 mg 758, containing 5 mM dithiothreitol in 1/10 of the original

Note. Sup, supernatant.culture volume. The cells were disrupted in a frenchpress at a pressure of 20 MPa. The resulting homoge-nate was incubated with DNase (2 mg/ml) for 20 minand centrifugated for 2 h at 10,500g. The resulting su- (85/10/5, by volume) with a flow of 1 ml/min was usedpernatant was precipitated with ammonium sulfate (7). Absorbance and spectra were recorded with a Kon-(0–40% saturation) and the proteins were collected by tron 440 diode array detector.centrifugation (23,000g, 15 min). The pellet was resus-pended in the buffer mentioned above and then de- RESULTSsalted on a Sephadex G25M column of 5-cm length and

Expression and purification of PDS. The activity ofa diameter of 1.5 cm. Proteins were absorbed on aPDS is strongly destroyed during treatment with vari-DEAE-cellulose column (length 5 cm, diameter 2 cm)ous detergents (data not shown). Thus, solubilizationand separated by a gradient of 0 to 0.3 M NaCl in 25of this membrane-bound protein in an active state ismM Tris–HCl buffer, pH 8. The purification processvery crucial. Therefore, it was attempted to expresswas monitored by SDS–polyacrylamide gel electropho-PDS in a soluble form retaining its activity. For thisresis on 12.5% gels (13). Proteins were stained withpurpose, the E. coli transformant was cultivated at dif-Coomassie brilliant blue and scanned with a flatbedferent temperatures in a range from 23 to 377C. Then,scanner and the relative amount of PDS was estimatedthe amount of total soluble protein after disruption ofby a densitometric software (7). Protein concentrations the harvested cells and also the amount of soluble PDSwere determined using the method of Bradford (14). activity were determined (Table 1). Activity of soluble

Substrate preparation and enzyme assay. Phytoene PDS was based on the volume of cell suspension. Thiswas extracted from JM101/pACCRT-EB cells by heat- value indicates under which growth condition the max-ing at 607C for 20 min in methanol and partitioning imum amount of active, soluble PDS can be providedinto 10% diethyl ether in petrol. For application to the for subsequent purification of this enzyme. The highestenzyme assay a lipid emulsion with soybean L-a-phos- amount of soluble protein was found in cultures grownphatidylcholine and the carotenes was made as pre- at 287C and the PDS activity per aliquot of supernatantviously described (15). also was highest therein. Consequently, the superna-

The incubation mixtures contained, in a total volume tant from broken E. coli cells grown at this temperatureof 1 ml of 0.1 M Tris–HCl buffer, pH 8, proteins from was used for purification of PDS.the different purification steps equivalent to about 20 PDS is the prominent protein band on SDS–poly-mg of PDS, 3.5 mg of substrate carotene in the lipid acrylamide gels at 53 kDa found in whole cells grownsuspension described above, and 1 mmol of NADP. at 287C, in the homogenate, and also in the supernatantWhen anaerobic conditions were investigated, they after centrifugation (Fig. 1, lanes 2 and 3). About one-were established by adding glucose (2 mM), glucose oxi- fifth of total PDS was recovered in the supernatantdase (20 U/ml), and catalase (20,000 U/ml) before the (Table 2). Nevertheless, specific activity in the superna-reaction vessels were tightly sealed. The assays were tant was only slightly lower than in the homogenate.carried out at 377C for 90 min and shaking at a fre- The purification steps involved ammonium sulfate pre-quency of 200 rpm. After termination by addition of 0.5 cipitation and subsequent DEAE chromatography ofml of methanol the carotenoids were partitioned into the 40% ammonium sulfate fraction. A steady increase10% diethylether in petrol, evaporated, redissolved in of specific activity was observed during this purificationacetone, and determined by HPLC. For HPLC separa- of PDS from the supernatant indicating that activity istion, a Sherisorb ODS-1, 5-mm column and a solvent basically retained. The combination of both purification

steps was sufficient to obtain a 0.12 mM NaCl fractionsystem consisting of acetonitrile/methanol/2-propanol

AID PEP 0730 / 6q10$$$101 06-04-97 23:58:35 pepa AP: PEP

HETEROLOGOUS PHYTOENE DESATURASE 177

intermediate phytofluene could also serve as a sub-strate and was desaturated to z-carotene (Fig. 2B). Inaddition to phytoene, its 1,2-epoxyde was also con-verted by PDS (Fig. 2C). The product formed was 1,2-epoxy-z-carotene. In the experiments of Fig. 2, anamount of 3.5 mg of substrate carotene was routinelyadded to an assay for PDS activity. However, the mini-mum amount for detection and reliable quantitation ofreaction products was 1 mg of substrate carotene.

Cofactor requirement of PDS was studied by deter-mination of enzyme activity in the presence of NAD,NADP, FAD, or ATP and under a reduced oxygen atmo-

FIG. 1. SDS–PAGE of fractions obtained during purification of sphere (Table 3). The activity showed a basic valuePDS. Lane 1, molecular mass markers; lane 2, cell homogenate; lane without any cofactor which could be decreased by mak-3, supernatant; lane 4, 40% ammonium sulfate precipitate; lane 5, ing the incubation mixture anaerobic. Either in thefraction from DEAE-cellulose column eluted with 0.12 M NaCl.

presence or in the absence of oxygen activity was stimu-lated equally well by NAD or NADP. This positive effectwas much stronger under anaerobic conditions. FADfrom the DEAE column in which PDS exhibited a pu- decreased the reaction rate to the lowest values,rity of ú99% (Table 2; Fig. 1, lane 5). From 1 liter of whether with or without oxygen or in the presence ofan E. coli culture 5 mg of purified PDS was obtained NAD. Addition of ATP did not significantly influencewhich represents a yield of 5%. PDS activity.

In vitro reaction and enzyme kinetic studies. The With both the substrate phytoene and the cofactorpurified PDS is active enough to convert nonradioactive NADP enzyme kinetic studies were carried out to ob-phytoene at a rate which is high enough to allow the tain their Km values. A series of assays with varieddetection of the reaction product(s) by HPLC. A corre- phytoene amounts was performed. The double recipro-sponding separation of products from residual sub- cal Lineweaver–Burk plot of the reaction productstrate in such an assay is shown in Fig. 2. Different formed versus added phytoene resulted in a straightsubstrates were applied to be converted by PDS. In line with r Å 99% (Fig. 3A). From the intercept withthe individual HPLC traces of Fig. 2, the peaks of the the abscissa the Km value for phytoene as substrate ofsubstrates are marked by dotted lines. For the experi- PDS was calculated to be 3.5 mM. A similar kineticment exhibited in part A, an isomeric mixture of phy- analyis was carried out for NADP interaction withtoene with a cis isomer (peak 3) and a trans isomer PDS. The double-reciprocal plot again resulted in a(peak 3*) was used as substrate. Alternatively, the straight line with r Å 97% (Fig. 3B) which allowed theother substrates were phytofluene in the trans (peak calculation of the Km value for NADP of 14.3 mM.2) and in the cis (peak 2*) form in trace B, and 1,2-epoxyphytoene (peak 5) in trace C of Fig. 2. The substrate

DISCUSSIONphytoene was converted into z-carotene as the endproduct of a two-step desaturation reaction via phy- The most successful strategy for isolation of caroten-

ogenic enzymes is to start with heterologous expressiontofluene as an intermediate. The typical isomeric mix-ture of all-trans z-carotene (peak 1) and two domi- (3). In E. coli overexpressed enzymes are usually se-

questered in inclusion bodies which makes them insol-nating cis isomers (peaks 1* and 19) was found. The

TABLE 2

Purification of PDSa

Volume Protein PDS Purity Yield Sp actStep (ml) (mg) (mg) (%) (%) (mg h01 mg prot01)

Homogenate 100 490 108 22 100 0.9Supernatant 100 155 20 13 11 0.8(NH4)2SO4

b 10 41 19 45 8 4.4DEAE fractionc 5 5 5 ú99 1 13.5

a Purification started from 1 liter of E. coli cell suspension.b Reextracted pellet of a 40% precipitation.c The homogenous protein eluted with 0.12 mM NaCl.

AID PEP 0730 / 6q10$$$101 06-04-97 23:58:35 pepa AP: PEP

SCHNEIDER, BOGER, AND SANDMANN178

FIG. 2. Conversion of phytoene (A), phytofluene (B), and 1,2-epoxyphytoene (C) by the Synechococcus PDS. These substrates are indi- FIG. 3. Double reciprocal plots of PDS activity versus the concen-cated by dotted lines. The identified carotenes were labeled with 1 tration of phytoene (A) or NADP (B) to determin the Km values.for z-carotene, 2 for phytofluene, 3 for phytoene, 4 for 1,2-epoxy-z-carotene, and 5 for 1,2-epoxy phytoene.

the optimum growth temperature to obtain solublePDS in an active form (Table 1).uble and inactive. This is also the case for PDS from

In the supernatant PDS was already highly enrichedSynechococcus (8). Temperature dependence of solubil-over all other proteins (Fig. 1). This made it possibleity of proteins as a result of inclusion body formationto purify this enzyme by only two steps (Table 2). Thewas reported for several recombinant enzymes (16).purified enzyme is highly active and its specific activityCultivating our E. coli transformant with the pds geneis in the same range as other desaturases (6,17). This isat 377C and lower temperatures showed that 287C isalso the case for its Km value for the carotene substrate.With our new assay it was possible to investigate the

TABLE 3 structural requirements of a carotenoid substrate to beconverted by PDS. Structurally the substrate moleculeStimulation of Enzyme Activity by Various Cofactorsmust resemble at least half the molecule of phytoene.

Specific activity Obviously, a 1,2-epoxy group does not negatively affectAdditions (ng z-carotene mg01 prot h01) desaturation of phytoene (Fig. 2C).

The mechanism of desaturation involves proton andAerobic control 5.4

electron transfer to an acceptor (3). In case of the bacte-NAD 7.6rial crtI-type phytoene desaturases, FAD is involved asNADP 8.8

FAD 2.7 coenzyme (5,6). With the reactivated PDS isolated fromAnaerobic control 2.7 urea-solubilized inclusion bodies, it was already dem-Anaerobic/NAD 13.8 onstrated that either NAD or NADP is the cofactor forAnaerobic/NADP 14.8

this reaction (7). In addition, it was found with our newAnaerobic/FAD 2.7PDS preparation that oxygen can to a certain extentAnaerobic/NAD

/ FAD 3.2 also stimulate desaturation (Table 3). Nevertheless, ox-/ ATP 12.7 ygen seems to interfere to a certain degree with the

NAD- or NADP-dependent activity of isolated PDSNote. The cofactor concentration was 1 mM in each case. For anaer-which may indicate a competition at the cofactor bind-obic conditions the oxygen in the incubation mixture was depleted

with glucose/glucose oxidase. ing. The direct oxygen interaction with PDS found here

AID PEP 0730 / 6q10$$$101 06-04-97 23:58:35 pepa AP: PEP

HETEROLOGOUS PHYTOENE DESATURASE 179

6. Linden, H., Misawa, N., Chamovitz, D., Pecker, I., Hirschberg,may have been caused by the purification procedureJ., and Sandmann, G. (1991) Functional complementation inand the question remains open whether oxygen is alsoEscherichia coli of different phytoene desaturase genes and anal-a direct electron acceptor for PDS in the thylakoid ysis of accumulated carotenes. Z. Naturforsch. 46c, 1045–1051.

membrane of its natural host organism. It should be 7. Fraser, P. D., Linden, H., and Sandmann, G. (1993) Purificationmentioned here that the only electron acceptor found and reactivation of recombinant Synechococcus phytoene desa-

turase from an overexpressing strain of Escherichia coli. Bio-for a purified crtI-type z-carotene was oxygen (17).chem. J. 291, 687–692.The established expression conditions and purifi-

8. Raisig, A., Bartley, G., Scolnik, P., and Sandmann, G. (1996)cation procedures described here result in about 5 mgPurification in an active state and properties of the 3-step phy-

of purified PDS from 1 liter of E. coli culture (Table 2). toene desaturase from Rhodobacter capsulatus overexpressed inThis opens the possibility of obtaining enough protein Escherichia coli. J. Biochem. 119, 559–564.for future NMR and crystallographic studies in order 9. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) ‘‘Molecular

Cloning: A Laboratory Manual,’’ 2nd ed., Cold Spring Harborto identify the different domains in the PDS proteinLaboratory Press, Cold Spring Harbor, NY.responsible for binding of carotenes, the cofactor

10. Bejarano, E. R., Govind, N. S., and Cerda-Olmedo, E. (1987) z-NAD(P), and herbicidal inhibitors which interactCarotene and other carotenes in a Phycomyces mutant. Phyto-

with PDS. chemistry 26, 2251–2254.11. Than, A., Bramley, P. M., Davies, B. H., and Rees, A. F. (1972)

Stereochemistry of phytoene. Phytochemistry 11, 3187–3192.REFERENCES

12. Albrecht, M., Sandmann, G., Musker, D., and Britton, G. (1991)Identification of epoxy- and hydroxyphytoene from norflurazon1. Young, A., and Britton, G. (1990) Carotenoids and stress, intreated Scenedesmus. J. Agric. Food Chem. 39, 566–569.‘‘Stress Response in Plants: Adaptation and Acclimation Mecha-

13. Laemmli, U. K. (1970) Cleavage of structural protein during thenisms’’ (Alscher, R. G., and Allen, N. S., Eds.), pp. 87–112, Wiley-assembly of the head of bacteriophage T4. Nature 227, 680–688.Liss, New York.

14. Bradford, M. (1976) A rapid and sensitive method for the quanti-2. Siefermann-Harms, D. (1987) The light harvesting and protec-tation of microgram quantities of proteins using the principle oftive function of carotenoids in photosynthetic membranes. Phys-protein-dye binding. Anal. Biochem. 72, 248–254.iol. Plant. 69, 561–568.

15. Schnurr, G., Misawa, N., and Sandmann, G. (1996) Expression,3. Sandmann, G. (1994) Carotenoid biosynthesis in microorgan- purification and properties of lycopene cyclase from Erwinia ure-

isms and plants. Eur. J. Biochem. 223, 7–24. dovora. Biochem. J. 315, 869–874.4. Sandmann, G. (1995) Phytoene desaturase: Genes, enzymes, and 16. Schein, C. H. (1989) Production of soluble recombinant proteins

phylogenetic aspects. J. Plant Physiol. 143, 444–447. in bacteria. Bio/Technology 7, 1141–1149.17. Albrecht, M., Linden, H., and Sandmann, G. (1996) Biochemical5. Fraser, P. D., Misawa, N., Linden, H., Yamano, S., Kobayashi,

K., and Sandmann, G. (1992) Expression in Escherichia coli, characterization of purified z-carotene desaturase from Ana-baena PCC 7120 after expression in Escherichia coli. Eur. J.purification, and reactivation of the recombinant Erwinia ure-

dovora phytoene desaturase. J. Biol. Chem. 267, 19891–19895. Biochem. 236, 115–120.

AID PEP 0730 / 6q10$$$101 06-04-97 23:58:35 pepa AP: PEP