The Poly-Cis Pathway of Carotene Desaturation: Enzymology, Herbicide Action and Retrograde Signaling Inaugural Dissertation to obtain the Doctoral Degree Faculty of Biology Albert-Ludwigs-Universität Freiburg im Breisgau Presented by Julian Koschmieder born in Bühl (Baden), Germany July 2017
171
Embed
The Poly-Cis Pathway of Carotene Desaturation: Enzymology ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Poly-Cis Pathway of Carotene Desaturation:
Enzymology, Herbicide Action and Retrograde Signaling
Inaugural Dissertation to obtain the Doctoral Degree
Faculty of Biology
Albert-Ludwigs-Universität
Freiburg im Breisgau
Presented by
Julian Koschmieder
born in Bühl (Baden), Germany
July 2017
Dean of the Faculty of Biology: Prof. Dr. Bettina Warscheid
Promotion chairman: Prof. Dr. Andreas Hiltbrunner
Supervisor of work: Prof. Dr. Peter Beyer
1st Reviewer: Prof. Dr. Peter Beyer
2nd Reviewer: Prof. Dr. Ralf Reski
3rd Reviewer: Prof. Dr. Jens Timmer
Date of defense: 12th of September 2017
This work was carried out from April 2014 to July 2017 under the supervision of
Prof. Dr. Peter Beyer, Institute of Biology II, Cell Biology.
Major parts of this thesis have been published in:
Bruno M*, Koschmieder J*, Wüst F, Schaub P, Fehling-Kaschek M, Timmer J,
Beyer P, Al-Babili S (2016) Enzymatic study on AtCCD4 and AtCCD7 and their
potential to form acyclic regulatory metabolites. J Exp Bot 67: 5993–6005
* equal contribution
Brausemann A, Gemmecker S, Koschmieder J, Ghisla S, Beyer P, Einsle O
(2017) Structure of Rice (O. sativa) Phytoene Desaturase Provides Insights Into
Herbicide Binding and Reaction Mechanisms Involved In Carotene
Desaturation. Structure, in press, DOI: 10.1016/j.str.2017.06.002
Gemmecker S, Schaub P, Koschmieder J, Brausemann A, Drepper F,
Rodriguez-Franco M, Ghisla S, Warscheid B, Einsle O, Beyer P (2015) Phytoene
Desaturase from Oryza sativa: Oligomeric Assembly, Membrane Association
and Preliminary 3D-Analysis. PLoS One 10: e0131717
Schaub P, Wüst F, Koschmieder J, Yu Q, Virk P, Tohme J, Beyer P (2017) Non-
Enzymatic β-Carotene Degradation in (Provitamin A-Biofortified) Crop Plants.
J Agr Food Chem, in press
Submitted for publication:
Koschmieder J, Fehling-Kaschek M, Schaub P, Ghisla S, Brausemann A, Timmer
J, Beyer P (2017) Rice (O. sativa) Phytoene desaturase: a Functional
All organic solvents were purchased as pro analysi quality from common suppliers, if used for HPLC analysis as HPLC / LC-MS grade. All other chemicals were purchased as pro analysi quality from common suppliers.
PCR products and other DNA fragments were purified from agarose gels and enzymatic reactions using the Illustra GFXTM PCR DNA / Gel Band Purification Kit (GE Healthcare) according to manufacturer instructions. Plasmids were isolated from E.
coli using the kits PureYieldTM Plasmid Miniprep / Midiprep System (Promega) according to manufacturer instructions.
3.1.2 Separation of nucleic acids
PCR products and plasmids were separated by agarose gelelectrophoresis. Gels with an agarose concentration of 1.0 – 1.2 % (w/v) were prepared with TAE buffer (0.484 % (w/v) Tris, 0.114 % (v/v) glacial acetic acid, 0.037 % (w/v) disodium-EDTA, pH 8.0 adjusted with acetic acid). 0.002 % (v/v) of Midori Green was added for DNA visualization. Samples were mixed with 6x DNA loading buffer (30 % (v/v) glycerol, 0.25 % (w/v) xylencyanol, Orange G and bromophenol blue each). Gels were run in TAE buffer at 7 V cm-1. GeneRulerTM 1 kb DNA ladder (Fermentas) served as a size marker.
3.1.3 Sequencing of nucleic acids
DNA sequencing was carried out by GATC Biotech (Constance, Germany) with the appropriate primers given in 2.2.
3.1.4 Polymerase chain reaction
All reactions were carried out with Phusion ® High Fidelity DNA Polymerase (NEB) or AccuPrime GC-rich DNA polymerase (ThermoFisher) for OsZISO amplification according to manufacturer instructions. Melting temperatures were calculated with the Geneious 8.0.2 software.
3.1.5 Site-directed mutagenesis
Site-directed point mutations were introduced into OsPDS-His6 using overlap extension PCR (Heckman and Pease, 2007). For amplification of the PDS fragment 5’ of the mutated site, primers P1 and P2x were used and primers P4 and P3x for the 3’ fragment. “x” stands for the letter assigned to the mutation (Table 2).
Table 2 Primers used for site-directed mutagenesis of OsPDS and resulting mutations
For mutations mutG / mutH / mutI, a mutated fragment of the PDS cDNA was synthesized and cloned into puC57 by GenScript (New Jersey, USA), excised directly 5’ / 3’ of the fragment with either HindIII and BglII (mut H) or NdeI and SphI (mutG / I) and cloned into the pRice expression backbone previously digested with the required restriction enzymes to reconstitute a point-mutated full length PDS cDNA. Besaid primers were also used to verify all sequences and point mutations in all pRiceOsPDS-His6 plasmids.
3.1.6 Gene cloning
3.1.6.1 OsPDS-His6 and OsZDS-His6
The plasmid pRiceOsPDS-His6 was kindly provided by Dr. Sandra Gemmecker and Dr. Patrick Schaub (University of Freiburg). Briefly, the OsPDS (Gen Bank AF049356) sequence was deprived of the 87 amino acid transit peptide sequence, equipped with a 5´ NdeI site and 3´ His6 coding sequence followed by a HindIII site and synthesized by GenScript®. The vector pCrtI-His6, used previously to express the bacterial carotene desaturase CrtI (Schaub et al., 2012), was digested with NdeI and HindIII to remove the CrtI-His6 cassette and replace it by the OsPDS-His6 coding sequence, resulting in the vector pRicePDS-His6. To obtain a vector pRiceOsZDS-His6, the same procedure as described above was performed with rice ZDS (Gen Bank Acc. AP004273.2), based on a sequence which was codon-optimized, synthesized and sequence-verified by GenScript®.
3.1.6.2 (Mistic-)OsZISO-His6
The coding sequence of OsZISO (Gen Bank Acc. AK066126.1) was deprived of the N-terminal 46 amino acid transit peptide sequence and equipped with a 5’ His6 tag. The sequence was synthesized and cloned into the pUC57 plasmid by GenScript®. The OsZISO-His6 sequence was amplified from pUC57 by PCR using the primers linker-ZISO FW and pCDF-ZISO RV (see 2.2). The sequence for Mistic M110 (Gen Bank Acc. AY874162.1) was amplified from Bacillus subtilis genomic DNA (kindly provided by Dr. Hervé Joel Soufo, University of Freiburg) using the primers pCDF-Mistic FW and linker-Mistic RV (see 2.2), thus depriving it from its stop codon. The coding sequences for Mistic and OsZISO-His6 were cloned into NcoI / BamHI – digested pCDFDuetTM-1 plasmid (Novagen) by Gibson assembly (Gibson et al., 2009). The resulting plasmid pCDFDuet-MisticOsZISO-His6 encoded a chimeric ZISO protein fused to Mistic at its N-terminus via a 20 amino acid flexible linker with a TEV protease cleavage site and rich in glycine, serine and proline. The OsZISO-His6 coding sequence was cloned into pCDFDuetTM-1 stand alone by the same procedure using primers pCDF-ZISO FW and pCDF-ZISO RV (see 2.2), resulting in the plasmid pCDFDuet-OsZISO-His6 encoding a ZISO-His6 protein.
Methods
36
3.1.7 Transformation of E. coli with plasmid DNA
E. coli cells were made competent with the Z-competent E. coli Transformation buffer set (Zymo Research, USA) according to manufacturer instructions and were stored at – 80 °C. 50 µl aliquots of cells were thawed on ice for 5 min, 100 ng of plasmid DNA was added and cells were kept on ice for 30 min. Cells were grown at 37 °C for 1 h after the addition of 1 ml LB medium. They were selected on LB agar plates (1 % (w/v) agar in LB) with the required antibiotic over night at 37 °C (ampicillin, 100 µg ml-1; kanamycin, 50 µg ml-1; chloramphenicol, 20 µg ml-1; streptomycin, 20 µg ml-1).
3.2 Protein methods
3.2.1 Protein expression and purification
3.2.1.1 Expression and IMAC purification of OsPDS-His6
Tuner (DE3) E. coli cells were transformed with pRice-PDS-His6 and grown in 2YT-medium containing ampicillin (100 µg ml-1) under agitation (120 rpm) at 37 °C using baffled Erlenmeyer flasks. The expression of PDS-His6 was induced at OD600nm of 0.5 – 0.7 with 0.5 mM IPTG and took place at 15 °C over night. Cells were harvested by centrifugation, frozen in liquid nitrogen and stored at - 20 °C. IMAC purification was carried out keeping samples on ice. 15 g of cells were resuspended in 20 ml lysis buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl, 1 mM TCEP) and catalytic amounts of DNase I were added. Cells were disintegrated by using a French Pressure Cell at 1380 bar twice. After 10 min of centrifugation at 20,000 x g, 90 ml of lysis buffer were added to the supernatant and the cell lysate was supplemented with 0.25 % (w/v) CHAPS (0.7 CMC). 4.5 ml of TALON® Metal Affinity Resin (Takara Bio Europe / Clontech) was equilibrated in lysis buffer and added. The mixture was incubated for 30 min at 10 °C under over head agitation (10 rpm). The resin was collected by centrifugation for 5 min at 800 x g. The resin pellet was washed twice – with 40 ml of wash buffer 1 (lysis buffer containing 500 mM NaCl and 2 % glycerol) and 40 ml of wash buffer 2 (wash buffer 1 containing 10 mM imidazole-HCl (pH 8.0)). After reequilibration of the resin with 40 ml of lysis buffer, PDS was eluted from the resin on gravity flow columns with 5 ml of elution buffer (lysis buffer containing 150 mM imidazole-HCl (pH 8.0)). This was followed by dialysis at 10 °C against dialysis buffer (lysis buffer containing 10 % (v/v) glycerol) for imidazole removal. The protein could then be stored at - 80 °C for several months in an active state.
3.2.1.2 Expression and IMAC purification of OsZDS-His6
Rosetta (DE3) E. coli cells were transformed with pRiceZDS-His6 and grown in 2YT-medium containing ampicillin (100 µg ml-1) and chloramphenicol (10 µg ml-1) under agitation (120 rpm) at 37 °C using baffled Erlenmeyer flasks. ZDS-His6 expression was induced at OD600nm 0.5 – 0.7 with 0.2 mM IPTG and protein was expressed at 16 °C over night. Cells were harvested by centrifugation, frozen in liquid nitrogen and stored at - 20 °C. IMAC purification was carried out keeping samples on ice. 15 g of cells were resuspended in 20 ml lysis buffer (20 mM Tris-HCl pH 8.0, 100 mM NaCl, 2 % (v/v) glycerol, 10mM MgCl2, 0.05 % (v/v) Triton X-100 (2.5 CMC), spatula tip of DNase I and lysozyme). Cells were disintegrated by using a French Pressure Cell at 1380 bar twice. After 10 min of centrifugation at 20,000 x g, 90 ml of lysis buffer were added to the supernatant. 4.5 ml of TALON® Metal Affinity Resin (Takara Bio Europe / Clontech) was equilibrated in lysis buffer and added. The mixture was incubated for 30 min at
Methods
37
10 °C under over head agitation (10 rpm). The resin was collected by centrifugation for 5 min at 800 x g. The resin pellet was washed with 40 ml of wash buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl, 10mM MgCl2,5 % (v/v) glycerol, 15 mM imidazole-HCl (pH 8.0)). After reequilibration of the resin with 40 ml of lysis buffer, ZDS was eluted from the resin on gravity flow columns with 5 ml of elution buffer (20 mM Tris-HCl (pH 8.0), 100 mM NaCl, 10mM MgCl2, 5 % (v/v) glycerol and 100 mM imidazole-HCl (pH 8.0)). This was followed by dialysis at 10 °C against dialysis buffer (lysis buffer containing 10 % glycerol) for imidazole removal. The protein could then be stored at - 80 °C for several weeks in an active state.
3.2.1.3 Expression and crude preparation of Mistic-OsZISO-His6
C43 (DE3) E. coli cells were transformed with pCDFDuet-MisticOsZISO-His6 and grown in 2YT-medium containing streptomycin (50 µg ml-1) under agitation at 37 °C using non-baffled Erlenmeyer flasks. The expression of Mistic-OsZISO-His6 was induced at OD600nm 0.5 – 0.7 with 1 mM IPTG and took place at 20 °C over night. Cells were harvested by centrifugation, frozen in liquid nitrogen and stored at - 20 °C. For crude protein preparation samples were kept on ice. 0.25 g of cells were resuspend in 2 ml of lysis buffer (50 mM Tris-HCl (pH 8.0), 100 mM NaCl, spatula tip of DNase I and lysozyme) and disintegrated by two passages through a French Pressure cell at 1380 bar. The cell lysate was clarified by centrifugation at 20000 x g for 1 min and the supernatant as a crude enzyme source was stored on ice and used for in vitro assays within 2 h.
3.2.1.4 Expression and IMAC purification of LeCRTISO-His6
The plasmid pHLT-CRTISO was kindly provided by Dr. Qiuju Yu (University of Freiburg) in order to express LeCRTISO as a HLT fusion protein with His6 tag and to purify it according to Yu et al. (2011) with minor changes. Briefly, pHLT-LeCRTISO was transformed into BL21 (DE3) E. coli cells harboring the plasmid pGro7 (Takara Bio Inc.), grown in 2YT-medium containing ampicillin (100 µg ml-1) and chloramphenicol (10 µg ml-1) in baffled flasks at 37 °C to an OD600nm of 0.5 – 0.8, induced by adding arabinose (0.2 % (w/v)) and IPTG (0.2 mM) and protein expression took place over night at 16 °C. Cells were harvested and stored at - 20 °C. For IMAC purification, samples were kept on ice or at 4 °C. Cells were disintegrated in 1.5 ml lysis buffer per gram of cell mass (25 mM sodium phosphate buffer (pH 7.5), 2.5 mM MgCl2, 300 mM NaCl, 15 % (v/v) glycerol) by two passages through a French Press Cell at 1380 bar. The crude lysate was clarified by centrifugation at 20,000 × g for 10 min and could either be used as a crude protein preparation directly or for IMAC purification. For further purification, Tween 20 was added to the supernatant to a final concentration of 10 x CMC (0.067 % (v/v)) for membrane solubilization. After 30 min of incubation on ice, TALON® was added and binding was allowed for 45 min at 10 rpm in an overhead shaker. The resin was pelleted by centrifugation at 800 x g for 5 min, washed three times with wash buffer (25 mM sodium phosphate buffer (pH 7.5), 2.5 mM MgCl2, 300 mM NaCl, 15 % (v/v) glycerol, 0.02 % (v/v) Tween 20 and 15 mM imidazole (pH 8.0)). Elution of LeCRTISO-His6 was accomplished on small gravity flow columns by the addition of 100 mM imidazole (pH 8) to the wash buffer. The protein was dialyzed against dialysis buffer (25 mM sodium phosphate buffer (pH 6.2), 2.5 mM MgCl2, 300mM NaCl, 15 % (v/v) glycerol) in a 30 kDa MWCO dialysis tube in order to remove imidazole. The protein was then stored at - 80 °C.
Methods
38
3.2.1.5 Expression and crude preparation for AtCCDs
All four AtCCD enzymes were expressed as thioredoxin fusion proteins using the pThio-vector system. The vectors pThio-Dan1-AtCCD1 and pThio-Dan1-AtCCD4 were kindly provided by Dr. Mark Bruno (University of Freiburg; Bruno et al., 2016), the vectors pThio-Dan2-AtCCD7 and pThio-AtCCD8 were kindly provided by Dr. Adrian Alder (University of Freiburg; Alder et al., 2012). pThio-AtCCD plasmids were transformed into BL21 (DE3) E. coli cells harboring the plasmid pGro7 (Takara Bio Inc.). Cells were grown at 37 °C in 50 ml 2YT growth medium supplemented with ampicillin (100 µg ml-1) and chloramphenicol (10 µg ml-1). Protein expression was induced with 0.2 % (w/v) arabinose at an OD600nm of 0.5. Cells were grown for 4 h at 28 °C, harvested and stored at - 20 °C. For crude protein preparation, cell pellets were resuspended in 1 ml modified LEW buffer (50 mM NaH2PO4, 300 mM NaCl, 1 mg ml-1 lysozyme, 1 mM dithiothreitol (pH 8.0)), passed twice through a French pressure cell at 690 bar and centrifuged at 20,000 x g for 5 min. Protein concentration was quantified using the Quick StartTM Bradford Protein Assay (Bio-Rad Laboratories) and adjusted to 20 µg µl-1.
3.2.1.6 Gel permeation chromatography for OsPDS-His6
For gel permeation chromatography (GPC) analysis of PDS, the GPC system Äkta Explorer 10 consisting of Box-900, pH/C-900, UV-900 und P-900 (Pharma Biotech, Uppsala, Sweden) was used. Samples were separated on a Superose 6 10/300 GL column (GE Healthcare) preequilibrated in GPC buffer (50 mM Tris-HCl (pH 8.0), 100 mM NaCl, 150 mM imidazole-HCl (pH 8.0), 5 mM TCEP) at an isocratic flow rate of 0.8 – 1 ml min-1. Protein elution was monitored via absorption at 280 nm, fluorescence of free and protein-bound FAD at 535 nm upon excitation at 450 nm was monitored with a WatersTM 474 Scanning Fluorescence Detector (Waters GmbH). GPC control and data analysis was carried out using the Unicorn 5.0 software (GE Healthcare). Peak area was quantified by multiplying height of the peak and its width at 50 % height.
3.2.1.7 Gel permeation chromatography for OsZDS-His6
For gel permeation chromatography analysis of ZDS, see 3.2.1.6 with the following changes. Samples were separated in GPC buffer (20 mM Tris-HCl (pH 8.0), 100 mM NaCl, 100 mM imidazole-HCl (pH 8.0), 0.5mM TCEP if not stated otherwise) at an isocratic flow rate of 0.8 – 1ml min-1. For determination of native protein size and potential oligomerization, samples were separated on the columns HiLoad 16/60 (fractionation range 10 – 600 kDa), Superose 6 10/300 (fractionation range 5 – 5000 kDa), Superdex 75 10/300GL (fractionation range 3-75 kDa) (all GE Healthcare) that were calibrated with the following proteins / organic compounds: Blue dextran, carboanhydrase from bovine erythrocytes, alcohol dehydrogenase from S.
cerevisiae, β-amylase from sweet potato, cytochrome c from equine heart, apoferritin from equine spleen, albumin from bovine serum (all Sigma) and ovalbumin (GE Healthcare).
Methods
39
3.2.2 Enzymatic assays
3.2.2.1 OsPDS-His6 in vitro
Under standard conditions, 100 µl of liposomes containing 5 nmol of 15-cis-phytoene (prepared according to 3.2.8) were vortexed after the addition of 0,6 µl of methanolic decylplastoquinone (30 mM). Assay buffer (50 mM MES-KOH (pH 6.0), 100 mM NaCl) was added to a final volume of 700 µl. Assays were started by the addition of 25 µg OsPDS-His6 (ca. 10 µl; see 3.2.1.1). The incubation was carried out at 37 °C in the dark for 15 min. Assays were stopped and carotenoids extracted by mixing them with 300 µl of CHCl3/methanol (2:1, v/v) with α-tocopherol acetate (0.1 mg ml-1) as internal standard. After 5 min of centrifugation at 20000 x g, the carotenoid-containing hypophase was dried, resuspended in 40 µl of CHCl3/methanol (2:1, v/v) and 5 µl were analyzed by HPLC (system 1, see 3.3.1). For inhibitor studies, 2 µl of methanolic inhibitor solution was added to the assay buffer and control assays contained 2 µl of methanol.
3.2.2.2 OsZDS-His6 in vitro
Under standard conditions, 100 µl of liposomes containing 5 nmol of 9,9’-di-cis-ζ-carotene (prepared according to 3.2.8) and were vortexed after the addition of 0.6 µl of methanolic decylplastoquinone (6 mM). Assay buffer (50 mM MES-KOH (pH 6.5), 100 mM NaCl) was added to a final volume of 700 µl. Assays were started by the addition of 25 µg OsZDS-His6 (ca. 10 µl; see 3.2.1.2). The incubation was carried out at 36 °C in the dark for 30 min. Assays were stopped and carotenoids extracted by mixing them with 300 µl of CHCl3/methanol (2:1, v/v) with α-tocopherol acetate (0.1 mg ml-1) as internal standard. After 5 min of centrifugation at 20000 x g, the carotenoid-containing hypophase was dried, resuspended in 40 µl of CHCl3/methanol (2:1, v/v) and 5 µl were analyzed by HPLC (system 2, see 3.3.1). For inhibitor studies, 2 µL of methanolic inhibitor solution was added to the assay buffer and control assays contained 2 µl of methanol. When determining the pH dependency of OsZDS-His6 activity, a combined MES-Tris buffer system (50 mM MES-Tris, 100 mM NaCl, pH adjusted by MES-Tris titration) was used.
3.2.2.3 Mistic-OsZISO-His6 in vitro For in vitro activity measurements, 100 µl of liposomes containing 5 nmol of 9,15,9’-tri-cis-ζ-carotene (prepared according to 3.2.8) were mixed with 100 µl of Mistic-OsZISO-His6–containing cell lysate (see 3.2.1.3) and assay buffer (50mM MES-KOH (pH 6.0), 100 mM NaCl) was added to a final volume of 700 µl. Assays were incubated for 2 h at 30° C in the dark and carotenoids were extracted by mixing samples with 1 volume of CHCl3/methanol (2:1, v/v) with α-tocopherol acetate (0.1 mg ml-1) as internal standard. After 5 min of centrifugation at 20000 x g, the carotenoid-containing hypophase was dried, resuspended in 40 µl of CHCl3/methanol (2:1, v/v) and 5 µl were analyzed by HPLC (system 1, see 3.3.1).
3.2.2.4 LeCRTISO-His6 in vitro
In vitro assays were based on crude enzyme preparation (3.2.1.4). 100 µl of substrate-containing liposomes (prepared according to 3.2.8) were mixed with 191 µl of crude protein preparation and 3 µl of FAD (5 mM). Anaerobic conditions were achieved by adding 3 µl of NADH (5 mM) and freshly prepared sodium dithionite (100 mM) each, resulting in a final volume of 300 µl. Assays were incubated at 37 °C for 1 – 3 h in the dark, depending on the carotenoid substrate. Carotenoid extraction was achieved
Methods
40
by the addition of 900 µl acetone, 300 µl of petroleum ether/diethyl ether (2:1, v/v) with α-tocopherol acetate (0.1 mg ml-1) as internal standard and sonication (Branson Digital Sonifier, USA). After 5 min of centrifugation at 20000 x g, the carotenoid-containing hypophase was dried, resuspended in 40 µl of CHCl3/methanol (2:1, v/v) and 5 µl were analyzed by HPLC (system 6, see 3.3.1). Prolycopene served as a positive activity control.
3.2.2.5 AtCCDs in vitro
Assays were performed in a total volume of 200 µl. Purified substrates were dissolved in CHCl3 (equaling 30 µM as final concentration if not stated otherwise) and mixed with 20 µl ethanolic Triton X-100 (2 % (v/v)) in the case of AtCCD 4 / 7 / 8 and with 50 µl ethanolic ß-octylglucoside (4 % (v/v)) in the case of AtCCD1 dried using a vacuum evaporator and dissolved in 100 µl of 2x incubation buffer (2 mM TCEP, 0.4 mM FeSO4, 200 mM HEPES-NaOH (pH 7.8)and 2 mg ml-1 catalase (Sigma). For 9-cis-lycopene micelles, the substrate was mixed with 10 µl of ethanolic Triton X165 (3 % (v/v)). Assays were started by the addition of each, 50 µl lysate and H2O and then incubated for 1 h, if not stated otherwise, under shaking (200 rpm) at 30 °C in darkness. For extraction, 400 µl acetone were added, followed by short sonication (Branson Digital Sonifier, USA) and the addition of 600 µl petroleum ether/diethyl ether (2:1, v/v) using α-tocopherol acetate (0.1 mg ml-1) as an internal standard. After centrifugation, the epiphase was dried and redissolved in 40 µl CHCl3. 5 µl of the extract were subjected to HPLC (system 7 if not stated otherwise, see 3.3.1). The following substrates served as positive activity controls: β-carotene (AtCCD4), β-apo-8-carotenal (AtCCD1), 9-cis-β-carotene (AtCCD7) and 9-cis-β-apo-10’-carotenal (AtCCD8).
3.2.2.6 OsZISO-His6 in vivo
DH5α E. coli cells were transformed with the plasmid pζ-carotene (see 2.3; kanamycin) to produce tri-cis-ζ-carotene and di-cis-phytofluene by the expression of, inter alia, phytoene desaturase from Synechococcus sp. PCC 6803. For exclusive phytoene production, E. coli was transformed with the plasmid pPhytoene_PS (see 2.3; kanamycin). Cells were cotransformed with pCDFDuet-MisticOsZISO-His6 or pCDFDuet as a control, grown in LB to an OD600nm of 0.5 – 0.7 at 28 °C and ZISO expression was induced by the addition of 0.2 mM IPTG. Cells were further grown in the dark for 18 h at 28 °C, harvested by centrifugation at 8000 x g for 5 min and carotenoids were extracted in the dark by the addition of 3 times 2 ml acetone and sonication (Branson Digital Sonifier, USA). 2 ml of petroleum ether / diethyl ether (2:1, v/v) were added to the combined acetone extract, filled up to 14 ml with water, partitioned and centrifuged at 4000 x g for 5 min. The epiphase was dried and carotenoid extracts were subjected to HPLC to investigate ζ-carotene isomerization (system 3, see 3.3.1). In order to reconstitute the entire poly-cis desaturation pathway and investigate how well light can substitute for ZISO activity in vivo, cells were cotransformed with the plasmids pRiceOsZDS-His6 (ampicillin) and pCDFDuet-MisticOsZISO-His6 (streptomycin) or pCDFDuet-empty as a control. Cells were grown at 32 °C for 8 h after IPTG induction at OD 0.5 – 0.7, either in the dark or under light (800 µmol m-2 sec-1). Carotenoids were extracted and analyzed as given above.
3.2.3 SDS-PAGE
For protein separation, discontinuous sodium dodecyl – polyacrylamide gelelectrophoresis (SDS-PAGE) was applied. Self-casted gels (8.5 x 8.2 x 0.1 cm)
Methods
41
consisted of a 4 % collecting gel and a 8 % / 12 % separation gel composed as follows:
Separation gel 12 % Separation gel 8 % Collecting gel
acrylamide 30 18 ml 12 ml 2,4 ml
separation buffer (1.88 M Tris-HCl; pH 8.8)
9 ml 9 ml x
collecting buffer (1.25 M Tris-HCl; pH 6.8)
x x 1,5 ml
aqua dd 17,4 ml 23,4 ml 10,8 ml
10 % (w/v) SDS 450 µl 450 µl 150 µl
TEMED 22,5 µl 22,5 µl 15 µl
10 % (w/v) APS 210 µl 210 µl 150 µl
Protein samples were mixed with 3 x protein loading buffer (65 mM Tris-HCl (pH 6.75), 20 % (v/v) glycerol, 10 % (v/v) β-mercaptoethanol, 4 % (w/v) SDS, spatula tip of Coomassie Brilliant Blue (CBB) G250), solubilized at 90 °C for 5 min and separated for 1.5 – 2 h at 230 V and 30 mA. Gels were run in SDS-PAGE running buffer (0.25 M Tris, 2 M glycine, 1 % (w/v) SDS). PageRulerTM Prestained-Protein Ladder (Fermentas, St.Leon-Rot) was used as a size marker if not stated otherwise. Gels were stained using Coomassie staining solution (2.5 % (w/v) CBB G250 in ethanol/H2O/acetic acid (227:227:46; v/v/v)) and destaining solution (ethanol/H2O/acetic acid (30:60:10; v/v/v) and scanned.
3.2.4 Protein precipitation
3.2.4.1 Ammonium sulfate precipitation
31 % (w/v) ammonium sulfate was added to PDS or ZDS samples originating from GPC analyses (see 3.2.1.6 and 3.2.1.7), equaling 60 % saturation at 4 °C. Samples were incubated on ice on a rotary shaker (50 rpm) for 30 min and then centrifuged at 20000 x g for 20 min at 4 °C. Protein pellets were resuspended in appropriate volumes of GPC buffer and centrifuged at 20000 x g for 10 min at 4 °C to remove aggregates and denatured protein.
3.2.4.2 Chloroform-methanol precipitation
Proteins samples were denatured and precipitated according to Wessel and Flügge (1984) and resolubilized at 90 °C in 3 x SDS loading buffer for SDS-PAGE (see 3.2.3).
3.2.5 Protein quantification
3.2.5.1 Bradford assay
Protein concentrations were determined using the Quick StartTM Bradford Protein Assay (Bio-Rad Laboratories) according to the manufacturer’s protocol.
Methods
42
3.2.5.2 Nanodrop
IMAC-purified protein was quantified in its dialysis buffer with a Nanodrop photometer (Implen, Munich) using ε280nm = 72,400 l mol-1 cm-1 for OsPDS-His6 and ε280nm = 68,300 l mol-1 cm-1 for OsZDS-His6, both extinction coefficients calculated for the fully reduced proteins with the ExPASy protein parameter online tool.
3.2.6 Membrane association assay
Standard PDS in vitro assays with 50 µg OsPDS-His6 each (see 3.2.2.1) were incubated for 15 min at 37 °C, layered on top of a 30 % (w/v) sucrose in PDS assay buffer (50 mM MES-KOH (pH 6.0), 100 mM NaCl) and centrifuged for 30 min at 110000 x g. The liposomes were isolated from the density boundary, PDS was precipitated by CHCl3/methanol precipitation (see 3.2.4.2) and SDS-PAGE was carried out (see 3.2.3). To test if PDS interaction with liposomes is based on ionic interactions, isolated liposomes were washed with PDS assay buffer supplemented with 0.5 M KCl, incubated for 15 min and PDS liposomes were isolated as described above.
3.2.7 Preparation of carotenoid-containing liposomes
For liposome preparation for in vitro assay with OsPDS-His6, OsZDS-His6 or MIstic-OsZISO-His6 , 5 mg phosphatidylcholine were dissolved in CHCl3 and added to 50 nmol of the required carotenoid substrate. After vortexing, the lipid-carotene mixture was dried under N2 and 1 ml liposome buffer (50 mM Tris-HC (pH 8.0), 200 mM NaCl) was added, followed by 15 min incubation on ice. Liposomes were formed by gentle sonication (Branson Digital Sonifier, USA). Small unilamellar vesicles were formed by a passage through a French pressure cell at 1380 bar. For liposome preparation for CRTISO in vitro assays, 10 mg phosphatidylcholine ml-1 were used as final concentration and liposome buffer consisted of 25 mM sodium phosphate buffer (pH 7.5), 2.5 mM MgCl2, 300mM NaCl and 15 % (v/v) glycerol.
3.2.8 Photometric quantification of protein-bound FAD
OsPDS-His6 or OsZDS-His6 were denatured in dialysis buffer (see 3.2.1.1 and 3.2.1.2) for 10 min at 80 °C in order to release non-covalently bound FAD from the flavoenzymes. Protein was pelleted by centrifugation at 20000 x g for 10 min. FAD in the supernatant was quantified photometrically at 450 nm after establishing a FAD calibration curve. The percentage of flavinylation on a monomeric protein base was calculated from the FAD concentration in the supernatant and the initial protein concentration determined according to 3.2.5.2.
3.3 Chromatography
3.3.1 High performance liquid chromatography
Analytical carotenoid samples were dissolved in 40 µl CHCl3:methanol (2:1, v/v) and 2 – 5 µl were analyzed using a Prominence UFLC XR separation module equipped with a SPD-M20A photodiode array detector (Shimadzu). The column temperature was 40 °C. Peak area integration and data analysis was carried out using the Shimadzu software LabSolutions. For quantification of carotenoids, the HPLC system was calibrated with β-carotene. For PDS assays, molar extinctions coefficients for the all-trans species were used (Britton et al., 1995): 15-cis-phytoene (285 nm) = 68125 l mol-1 cm-1; 9,15-di-cis-phytofluene (350 nm) = 73300 l mol-1 cm-1; 9,15,9’-tri-cis-ζ-
Methods
43
carotene: 400nm = 138000 l mol-1 cm-1. For ZDS assays, values were estimated based on the coefficients published for the trans-carotenoid and the influence of cis-configuration in lycopene isomers according to Hengartner et al. (1992): 9,9’-di-cis-ζ-carotene (400 nm) = 97000 l mol-1 cm-1; 7,9,9’-tri-cis-neurosporene (433 nm) = 108,000 l mol-1 cm-1; 7,9,7’,9’-tetra-cis-lycopene (440 nm) = 100840 l mol-1 cm-1. All other carotenoids were quantified based on molar extinction coefficients published for the trans-configured species (Britton et al., 1995).
HPLC system 1 for PDS assays and ZISO in vitro and in vivo assays: Samples were separated on a YMC carotenoid C30 column (150 mm x 3 mm, 5 µm; YMC) with the solvent system A: methanol / tert-butyl methyl ether (TBME) (1:1, v/v) and B: methanol / TBME / water (5:1:1, v/v/v). The flow rate was 0.7 ml min-1. Separation started at 60 % A and increased linearly to 86 % A within 6.5 min, followed by 2 min of reequilibration.
HPLC system 2 for ZDS assays: Samples were separated on an Accucore C30 column (150 mm x 3.0 mm, 2.6 µm; Thermo Scientific) with the solvent system A: methanol / TBME (1:1, v/v) and B: methanol / TBME / water (5:1:1, v/v/v). The flow rate was 0.7 ml min-1. Separation started at 40 % A and increased linearly to 80 % A within 8 min, followed by 2 min of reequilibration.
HPLC system 3 for ZISO in vivo assays upon co-expression with PDS and ZDS: Samples were separated on a Nucleosil C18 100-5 (150 mm x 4.6 mm, 5 µm; Macherey-Nagel) with an isocratic flow of acetonitrile at 1.4 ml min-1.
HPLC system 4 for purification of ζ-carotene isomers and 9,15-di-cis-phytofluene from
tangerine tomato: Samples were separated on a YMC Carotenoid C30 column (250 mm x 10 mm, 5 µm; YMC) with an isocratic flow of methanol / TBME (4:1, v/v) at 2.0 ml min-1.
HPLC system 5 for purification of prolycopene and proneurosporene from tangerine
tomato: Samples were separated on a Nucleosil C18 100-10 (250 mm x 4.6 mm, 10 µm; Macherey-Nagel) with an isocratic flow of acetonitrile / water (98:2, v/v) at 1.2 ml min-1.
HPLC system 6 for CRTISO in vitro assays and all-trans- and 9-cis-neurosporene
purification: Samples were separated on a YMC carotenoid C30 column (150 mm x 3 mm, 5 µm; YMC) with the solvent system A: methanol / TBME (1:1, v/v) and B: methanol / TBME / water (5:1:1, v/v/v). The flow rate was 0.75 ml min-1. Separation started at 0 % A and increased linearly to 100 % A within 20 min. Final conditions were maintained for 4 min, followed by 2 min of reequilibration.
HPLC system 7 for analysis of CCD assays with phytofluene, ζ-carotene, prolycopene,
proneurosporene: Samples were separated on a YMC carotenoid C30 column (150 mm x 3 mm, 5 µm; YMC) with the solvent system A: methanol / TBME (4:1, v/v) and B: methanol / TBME / water (30:1:10, v/v/v). The flow rate was 0.6 ml min-1. Separation started at 0 % A and increased linearly to 60 % A within 20 min and to 0 % A within 5 min. Conditions were maintained for 9 min, followed by 6 min of reequilibration.
HPLC system 8 for CCD assays with 9-cis-lycopene and all-trans-lycopene: Samples were separated on a YMC carotenoid C30 column (150 mm x 3 mm, 5 µm; YMC) with the solvent system A: methanol / TBME (4:1, v/v) and B: methanol / TBME / water (5:1:1, v/v/v). The flow rate was 0.6 ml min-1. Separation started at 0 % A and
Methods
44
increased linearly to100 % A within 24 min. Conditions were maintained for 4 min, followed by 6 min of reequilibration.
HPLC system 9 for CCD assays with 9-cis-neurosporene: Samples were separated on a YMC carotenoid C30 column (150 mm x 3 mm, 5 µm; YMC) with the solvent system A: methanol / TBME (4:1, v/v) and B: methanol / TBME / water (30:1:10, v/v/v). The flow rate was 0.6 ml min-1. Separation started at 0 % A and increased linearly to100 % A within 20 min. Conditions were maintained for 4 min, followed by 6 min of reequilibration.
3.3.2 Extraction and purification of carotenoids from tangerine tomato and
carotenoid-containing bacteria
Total carotenoid extraction from from bacteria or plants: Either bacterial cell pellets or tomato fruit tissue that was previously homogenized in a blender was extracted in the dark with acetone under sonication (Branson Digital Sonifier, USA) until sufficient decoloration was achieved. The acetone extract was mixed with 1/10 volume of petroleum ether / diethyl ether (2:1, v/v) in a separating funnel. Water was added for partitioning of the carotenoids into the epiphase until a clear phase separation was achieved. The epiphase was collected and dried under vacuum, dissolved in an appropriate volume of CHCl3/methanol (2:1, v/v) and subjected to chromatography required for isolation of the desired carotenoid as follows.
Purification of phytoene from transgenic E. coli containing the plasmid pPhytoene_PS
(see 2.3): The phytoene extract from E. coli was applied on a thin layer silica gel plate 60 F254 (Merck) and developed sufficiently with petroleum ether / diethyl ether (4:1, v/v). The running front containing the phytoene was scrapped and phytoene was eluted with five times 2ml acetone. The eluate was dried under vacuum and phytoene was stored in petroleum ether at - 20 °C in the dark.
Purification of ζ-carotene isomers from transgenic E. coli containing the plasmid pz-
carotene (see 2.3): The extract from ζ-carotene producing E. coli was applied on a silica TLC plate and developed sufficiently with petrolether / diethylether / acetone (4:1:1, v/v/v). The running front, containing the ζ-carotene (and its precursors phytoene and phytofluene), was scrapped and ζ-carotene was eluted from the silica with five times 2 ml acetone. The eluate was dried under vacuum and resuspended in CHCl3/methanol (2:1, v/v). ζ-carotene isomers were isolated with HPLC system 4 (see 3.3.1).
Purification of all-trans-neurosporene from Rhodovulum sulfidophilum (Hagemann et
al. (1996); kindly provided by Nasser Gad’on, University of Freiburg, Microbiology): The carotenoid extract was separated with HPLC system 6 (see 3.3.1) and neurosporene was collected.
Purification of carotenes from tangerine tomato: Total carotene extracts from tangerine tomato (kindly provided by Prof. Dr. Joseph Hirschberg, Hebrew University Jerusalem, Israel) fruits were applied on a silica TLC plate and developed sufficiently in pure petroleum ether. Three fractions were scrapped and eluted with acetone, listed by weaker adsorption: 1) a colorless 9,15-di-cis-phytofluene fraction identified by green fluorescence upon excitation with 350 nm UV light, 2) a bright yellow fraction of 9,9’-di-cis-ζ-carotene below fraction 1 and 3) a dark orange proneurosporene / prolycopene fraction close to the starting point. 9,15-di-cis-phytofluene and 9,9’-di-cis-ζ-carotene were further purified using HPLC system 4 (see
Methods
45
3.3.1). Proneurosporene and prolycopene were further purified using HPLC system 5 (see 3.3.1). The isomeric state of these carotenoids in tangerine tomato fruit was reported by Clough and Pattenden (1979) using nuclear magnetic resonance analysis.
3.3.3 Effective liposomal concentrations of carotenes, decylplastoquinone
and norflurazon
In the PDS assay established by Gemmecker et al. (2015), liposomes are used to accommodate the highly hydrophobic substates phytoene during liposome formation. The second substrate decylplastoquinone and the inhibitor NFZ are added after liposome formation from organic solvent solutions (see 3.2.2.1), both being hydrophobic and partially partitioning into the lipid layer. It is assumed that PDS as a monotopic membrane protein only gains access to its substates and NFZ via the lipid bilayer. Therefore, effective concentrations of besaid compounds – meaning their concentration in the lipid bilayer volume of liposomes (ceff) – are relevant. Therefore, in the framework of the dissertation of Dr. S. Gemmecker (University of Freiburg, 2015) incorporation efficiencies into liposomes were determined according to procedures of Degli Esposti et al. (1983). The incorporation efficiencies were: 100 % for carotenes, 55 % for DPQ and 86 % for NFZ. Given the partial volume of phosphatidylcholine (PC) of 0.997 ml g-1 (Greenwood et al., 2006) and the presence of 0.5 mg PC per assay (see 3.2.2.1), the compounds are present in a lipid bilayer volume of 0.5 µl. Accordingly, ceff is calculated as follows: ceff [mM] = (compound [nmol] / 0.5 µl) * x, with “n” being the amount of carotene, DPQ or NFZ added into the assay and “x” being the incorporation efficiency of the compound given above.
3.3.4 Liquid chromatography – mass spectrometry
3.3.4.1 Identification of apocarotenoids and carotene deuteration analysis
Non-volatile cleavage products produced by CCDs were identified by LC-MS using a Dionex UltiMate 3000 UPLC coupled to a Q-Exactive mass spectrometer (Thermo Fisher Scientific). Sample separation was achieved with a Hypersil Gold C18 UPLC-column (150 x 2.1 mm i.d., 1.9 µm) and the solvent system A, 0.05 % (v/v) formic acid in H2O and B, 0.05 % (v/v) formic acid in acetonitrile. Initial conditions were 70 % B for one minute, followed by a gradient to 100 % B within four minutes. The final conditions were maintained for ten minutes, all at a flow-rate of 0.5 ml min-1. Ionization of apocarotenoids was achieved with atmospheric pressure chemical ionization (APCI) and analyzed in the positive mode. Nitrogen was used as sheath and auxiliary gas, set to 20 and 10 arbitrary units, respectively. The vaporizer temperature was set to 350 °C and the capillary temperature was 320 °C. The spray voltage was set to 5 kV and the normalized collision energy (NCE) to 35 arbitrary units. For data analysis the TraceFinder (3.1) software and authentic apocarotenoid standards were used.
3.3.4.2 Identification of C35 carotenoids
PDS desaturation products originating from 15-cis-nor-phytoene (15-cis-1´,2´,3´,16´,17´-penta-nor-phytoene) were identified by LC-MS using a Dionex UltiMate 3000 UPLC coupled to a Q-Exactive mass spectrometer (Thermo Fisher Scientific). Sample separation was achieved with a YMC carotenoid C30 column (150 mm x 3 mm, 5 µm; YMC) with the solvent system A: methanol / TBME / water (5:1:1, v/v/v) in 0.1 % (v/v) formic acid and B: methanol / TBME (1:1, v/v) in 0.1 % (v/v) formic
Methods
46
acid. Conditions started at 50 % B, increased linearly to 60 % B within 15 min and to 100 % B within further 5 min. Final conditions were maintained for 10min, all at a flow-rate of 0.6 ml min-1. Ionization of apocarotenoids was achieved with atmospheric pressure chemical ionization (APCI) and analyzed in the positive mode. Nitrogen was used as sheath and auxiliary gas, set to 20 and 10 arbitrary units, respectively. The vaporizer temperature was set to 350 °C and the capillary temperature was 320 °C. The spray voltage was set to 5 kV and the normalized collision energy (NCE) to 35 arbitrary units. For data analysis the TraceFinder (3.1) software and authentic apocarotenoid standards were used.
3.3.4.3 Identification of protein-bound nucleotide cofactors
Sample separation was achieved with a Hypersil Gold C18 UPLC column (150 x 2.1 mm i.d., 1.9 µm) and the solvent system A: 50 mM ammonium acetate in 1 % (v/v) formic acid in H2O and B: 1.7 mM ammonium acetate in H2O. The flow rate was maintained at 0.5 ml min-1. Conditions started at 100 % A, decreased to 50 % A within 10 min and finale conditions were maintained for 5 min. Ionization of mono-/dinucleotide cofactors was achieved with electrospray ionization (ESI) and analyzed in the positive mode. Nitrogen was used as sheath and auxiliary gas, set to 40 and 10 arbitrary units, respectively. The vaporizer temperature was set to 300 °C and the capillary temperature was 350 °C. The spray voltage was set to 3.5 kV and the normalized collision energy (NCE) to 25 / 40 arbitrary units. For analysis of the full MS and data dependant MS2 data the TraceFinder 3.1 software was used. For cofactor identification, the following combinations of precursor ions [M+H]+ in MS1 and fragment ions in MS2 (in brackets) were established with authentic, purified standards: FAD, MS1 m/z 786.16441 (348.1, 439); FMN, MS1 m/z 457.11689 (359.2, 439.1); NAD+, MS1 m/z 664.11640 (524, 542.1); NADH, MS1 m/z 666.13205 (348.2, 649.2); NADPH, MS1 m/z 745.09055 (428.1, 729.1).
3.3.5 Gas chromatography – mass spectrometry
Volatile cleavage products such as 6-methyl-5-hepten-2-one and geranylacetone were collected by solid phase micro extraction (SPME; PDMS, 100 µm; Supelco). The fiber was exposed to the in vitro assay head space for 15 min and thermodesorbed in the injector of the Trace GC coupled to a Trace DSQ II mass spectrometer (Thermo Fisher Scientific). Separation was achieved on a 30 m Zebron ZB-5 column 0.25 mm i.d., 0.25 µm film thickness (Phenomenex). The initial temperature of 50 °C was maintained constant for five minutes, followed by a ramp of 25 °C min-1 to a final temperature of 280 °C which was maintained for five minutes. The helium carrier gas flow rate was 1 ml min-1 and the injector temperature was set to 220 °C. Electron impact ionization (EI) was used at an ion source potential of 70 eV and a source temperature of 200 °C. Spectra were matched to the NIST (2.0) database using the Excalibur software. Additionally, standards of geranylacetone and 6-methyl-5-hepten-2-one were used (Sigma).
3.4 Cryo scanning electron microscopy
The liposomes with bound OsPDS-His6 were isolated from membrane binding assays (see 3.2.6) and concentrated by ultracentrifugation at 150000 x g for 30 min. The enzymatic activity of PDS was verified by the observed yellow color of the liposomes upon conversion of phytoene to ζ-carotene. The isolated liposomes of four PDS in
vitro assays (see 3.2.2.1) with 50 µg PDS each were combined to 40 µl of PDS-
Methods
47
liposome suspension (in 12.5 mM MES-KOH, pH 6.0). After the addition of 30 % (v/v) glycerol, suspension was pipetted into the 50 µm cavity of two 3 mm aluminum specimen carriers before sandwiching them. The assembly was frozen using the HPM 100 (Leica) freezer, transferred into the Freeze Fracture System EM BAF060 (Leica) and fractured. Samples were visualized directly in a Zeiss Auriga SEM system (- 115 °C, 5 kV acceleration voltage, 20 µm aperture using the inlens SE detector) or after 5 min of sublimation at - 105 °C in order to display liposomal surfaces. Sublimated as well as untreated samples were coated with 2.5 nm of platinum/carbon and backed with 4 nm of carbon at a gun angle of 45° and under steady stage rotation with 40 rpm.
3.5 Homology modeling and in silico docking of OsZDS-His6
Homology modeling of Oryza sativa ZDS (XP_015646524.1), deprived of its N-terminal plastid target peptide and equipped with an N-terminal methionine, was carried out using the SWISS-MODEL protein structure homology modeling server in automated mode with the PDS structure as template (Brausemann et al., 2017). The obtained ZDS structure represented the apoenzyme. Riboflavin as a precursor of the ZDS redox cofactor FAD was docked into the ZDS structure using the SwissDock ligand docking web service. The thermodynamically most favorable holoenzyme structure, i.e. the structure with the lowest free enthalpy ΔG, obtained from in silico docking was chosen for structural analysis of OsZDS.
3.6 Mathematical modeling of PDS time courses and kinetics
General procedures. The model consists of a set of ordinary differential equations (ODEs) for the contributing processes following mass action kinetics. The maximum likelihood method is used to estimate parameters such that the model prediction optimally describes the observed PDS time courses. Setting up the likelihood,
normally distributed noise is assumed. The cost function χ�(θ) = ∑ (��(�, ))���� needs to
be minimized in order to maximize the likelihood. Here, θ denotes the model parameters, the index i runs over the data points taken at time ti with value xi and uncertainty σ� and x(t�, θ) is the model prediction at time t� . The nonlinear minimization of the cost function is performed by a trust region optimizer (Nocedal and Wright, 1999). Derivatives of the cost function, upon which the optimizer relies, are provided by sensitivity equations. Prior knowledge about parameter values, e.g. values of the initial states, are incorporated by either fixing the parameter value or adding a penalty to the cost function via a quadratic prior function. In general, the cost function can have several local optima, besides the global optimum. In order to find the global optimum a multistart approach is performed by seeding the optimization in different points of the parameter space. The ODEs and sensitivity equations are integrated with the lsodes solver (Soetaert et al., 2010). Identifiability of the parameters and their confidence intervals are determined by the profile likelihood method (Raue et al., 2009). The model was implemented using the dMod package for dynamic modeling in R (Kaschek et al., 2016). Data preprocessing. For PDS time courses of the conversion of phytoene and phytofluene, the amounts of phytoene, phytofluene and ζ-carotene were measured over time. The experiments were conducted in triplicate. Uncertainties for the computed mean values were first estimated by a maximum likelihood method
Methods
48
combining the empirical mean values and variances with an error model. However, additional fluctuations between neighboring time points, larger than those represented by the replicates, were observed. They cannot be captured by the error model described above, but would lead to an underestimation of the derived parameter profiles and uncertainties. Therefore, the uncertainty parameters of the error model were estimated together with the other model parameters, including the log(σ�)-term originally contained in the log-likelihood, leading to a new cost function:
−2 log L (θ) = ��x�(θ) −x��σ�(θ) ��
�+ log (σ�(θ)�)
The uncertainty parameters σ� include a relative and an absolute contribution for each observable, e.g. σ[!] =σ[!]#$% ∙ [p] + σ[!]()* and may vary between the different
reaction time courses. The relative normalizations of phytoene, phytofluene and ζ-carotene measurements were investigated by a preceding optimization. It is based on conservation of mass, i.e. the total sum of carotenes is conserved during reaction time courses. Such normalization is needed because of inaccuracies during carotene quantification. The molar extinction coefficient is known for 15-cis-phytoene but not for 9,15-di-cis-phytofluene and 9,15,9’-tri-cis-ζ-carotene. Therefore, the molar extinction coefficients for the all-trans species of phytofluene and ζ-carotene are used as approximation. Scaling parameters s,!, s,!- and s,. for phytoene, phytofluene and ζ-carotene,
respectively, were estimated by minimizing the discrepancy s,! ∙ [p]�/� +s,!- ∙ [pf]�/� +s,. ∙ [z]�/� − c
at all time points t� for an arbitrary constant c. Since the absolute scale incorporated
by the constant c is unknown, the ratios l3 = *,4*,5 and ratios l� = *,�
*,5 including their
confidence intervals are estimated by a least squares approach. The scaling parameters s! , s!- and s. used for phytoene, phytofluene and ζ-carotene in the
model prediction are related to the ratios via s! = l� ∙ s!- and s. = %4%� s!- and the
constraints on l3 and l� are added via a quadratic prior to the cost function. For additional information about data preprocessing, see supplemental data.
3.7 Deduction of non-covalent interactions from changes in electron
density gradients
A single point ab initio quantum mechanical calculation (density functional theory,
6-31G* model chemistry) was performed on the complex of interest, the PDS – NFZ
cocrystal structure (Brausemann et al., 2017). Non-covalent interactions were
deduced from changes in the gradient of the electron density between atoms
according to the methods described by Johnson et al. (2010). The system for
calculations consisted of only 16 flavin and PDS residues with a total of 313 atoms
that are in direct contact with norflurazon (irrelevant parts of each residue
disconsidered), representing a pared down representation of the PDS – NFZ binding
site.
Results
49
4 Results
4.1 Biochemical characterization of phytoene desaturase PDS
4.1.1 Association and oligomerization of OsPDS-His6 at liposomal surfaces
PDS associates with plastid membranes to access its hydrophobic,
membrane-soluble carotene substrates and plastoquinone. As suggested by
hydropathy plots, the protein has no transmembrane helices. Moreover, it can
be natively purified as soluble in the absence of detergents and associates
spontaneously to liposomes to convert phytoene to ζ-carotene (Gemmecker
et al., 2015). Accordingly, PDS is assumed to represent a monotopic
membrane protein, interacting with only one lipid leaflet. Furthermore,
experiments pointed towards homotetrameric assembly of OsPDS-His6 in
solution (Fig. 1-3 B). It remained unclear whether homotetramers represent the
catalytically active form at membrane surfaces.
4.1.1.1 Monotopic association of OsPDS-His6 with liposomes
In order to better characterize the mode of membrane association, the
interaction of OsPDS-His6 with phosphatidylcholine liposomes was examined.
First, it was tested whether membrane interaction required only the
phospholipid bilayer or whether the simultaneous presence of substrates (DPQ
and phytoene) was required (Fig. 4-1 A). As depicted in lane “1 L”, OsPDS-His6
readily attached to phosphatidylcholine liposomes containing the substrates.
A small portion of the protein formed aggregates that appeared in the pellet
upon centrifugation (lane 1 P). In the absence of liposomes, OsPDS-His6
formed aggregates (lane 2 P) since the assay buffer contained neither
imidazole nor glycerol, both promoting protein solubility in the absence of
membranes (Gemmecker et al., 2015). Lane “3 L” shows that OsPDS-His6
readily attached to membranes in the absence of substrates. Second, it was
tested whether PDS membrane association mainly relies on electrostatic
interactions or rather on hydrophobic interactions. Washing proteoliposomes
with a high-salt buffer did not detach PDS, as the amount of protein
recovered was very similar as upon washing with a low-salt buffer (Fig. 4-1 B,
Results
50
compare lanes 1 and 2). This points towards hydrophobic interaction of the
protein with the membrane and accordingly, PDS could only be released
from liposomes by solubilization with detergents (not shown). The OsPDS-His6
interaction with membranes requires only phospholipid bilayers and is mainly
mediated by hydrophobic interactions with the membrane core.
Fig. 4-1 SDS-PAGE analysis of OsPDS-His6 liposome binding assays.
(A) Dependency of OsPDS-His6 membrane association on substrate-containing
liposomes (lane 1) and substrate-free phosphatidylcholine liposomes (lane 3).
Fraction L represents liposome-bound OsPDS-His6 and fraction P represents pelletable
aggregates of unbound enzyme. Conducting the assay in the absence of liposomes
led to complete OsPDS-His6 precipitation (lane 2). (B) Characterization of the mode
of interaction between OsPDS-His6 and liposomes. Lane 1 represents protein bound
to liposomes washed with 100 mM KCl, lane 2 washed with 500 mM KCl.
4.1.1.2 Homooligomerization of OsPDS-His6 at liposomal membrane surfaces
Cryo scanning electron microscopy (cryo-SEM) was used to address the
question whether the PDS homotetramers found in solution (Gemmecker et
al., 2015; Fig. 1-3 B) are artifactual or rather represent the biologically active
form at membrane surfaces. The size of OsPDS-His6 in its active state at
membrane surfaces should be compared to the size of soluble OsPDS-His6
homotetramers previously investigated by negative staining. Additionally,
liposome fracture faces should reveal the presumed absence of membrane
spans.
21 3
L P L P L P
1 2A BL L
Results
51
Supporting monotopic membrane interaction, fracture faces of PDS
proteoliposomes capable of performing phytoene desaturation (Fig. 4-2 A, B,
black arrow in Fig. 4-2 B) did not support the presence of membrane spans.
Sublimation revealed the liposomal surface with OsPDS-His6 particles (Fig. 4-2
B, red arrows) that were 14.5 ± 1.9 nm in diameter (n = 30), whereof 2 nm are
due to Pt/C-coating. Accordingly, active OsPDS-His6 at liposomal surfaces is
12.5 nm ± 1.9 nm in size. This corresponds well with the reported size of soluble
homotetramers of 11.8 ± 1.3 nm (Gemmecker et al., 2015).
Fig. 4-2 Freeze fracture electron microscopy of OsPDS-His6 proteoliposomes.
(A) Membrane fracture face of OsPDS-His6 proteoliposome visualized by cryo
scanning electron microscopy. The outer fracture face boundary, surrounded by ice,
is indicated by a black arrow. (B) Membrane fracture face (surface/fracture face
boundary indicated by black arrow) and membrane surface with associated OsPDS-
His6 particles (red arrows) exposed upon ice sublimation.
In summary, cryo-SEM analysis corroborated the notion that PDS is a
monotopic enzyme being present as homotetramer in its active state on
membrane surfaces. In accordance with this, PDS monomers were shown to
be unflavinylated and enzymatically inactive (Gemmecker et al., 2015).
4.1.2 Regio-specificity of carotene desaturation by OsPDS-His6
PDS catalyzes a highly regio-specific reaction. No carotene desaturation at
positions other than C11-C12 and C11’-C12’ has ever been reported. It
remains elusive how the correct positioning of the reaction sites relative to the
PDSA B
200 nm
Results
52
active center is achieved. In VP14, a 9-cis-epoxy-carotenoid cleavage
dioxygenase, the 9-cis-configured double bond acts as a restrictor arresting
the carotenoid in the correct position relative to the active site (Messing et al.,
2010). Alternatively, the long carotenoid substrate might be introduced into
the substrate cavity until reaching the cavity end that acts as restrictor. In this
case, the substrate length might mediate correct positioning.
The modeling of 15-cis-phytoene in its extended conformation into the
substrate cavity of OsPDS-His6 supports both scenarios (Fig. 4-3). Charged
amino acid residues at the back end of the hydrophobic substrate cavity
to end product formation (Fig. 4-14 A). Thus, higher affinity for PQ than for
carotene substrates can be considered as a third mechanism in PDS to favor
end product formation. It is conceivable that rapid reoxidation of the subunits
within homotetramers assists substrate channeling of the nascent phytofluene
that still resides within the PDS microdomain.
Discussion
132
The substrate channeling model and the rate constants deduced from
reaction time courses could in principle simulate and reflect the observed
substrate concentration-dependent relations of intermediate and end
product formation by OsPDS-His6 (Fig. 4-14). This supports the validity of the
model. It needs to be stated that it also systematically overestimates
Michaelis-Menten parameters, used as a quantitative measure to allow
comparisons, by factors of 1.1 to 4.1 (Table 5). The error is likely due to gradual
alterations of the structure and physicochemical properties of liposomal
membranes when increasing concentrations of the hydrocarbon substrates
are incorporated that conceivably impair OsPDS-His6 activity (Sikkema et al.,
1993). The model was established at low substrate concentrations and cannot
consider this structural circumstance upon extrapolation. Moreover, the
production of enzyme and liposomes batches with identical specific activities
showed to be notoriously difficult. The preparations used in reaction time
course experiments to develop the model were different from the ones used
to investigate substrate concentration dependencies. This contributes to the
quantitative deviations from the model despite qualitative similarities, too.
5.1.6 Characterization of norflurazon - OsPDS-His6 interactions
5.1.6.1 Quinone-competitive inhibition by norflurazon is mediated by several
non-covalent interactions
Phytoene desaturase is a prominent target for a structurally heterogeneous
group of bleaching herbicides such as norflurazon (NFZ), fluridone and
diflufenican sharing a meta-trifluoromethylphenyl (m-CF3-phenyl) moiety often
fused to a second (heterocyclic) carbonyl-substituted ring structure (see
1.3.1.3). The lack of structural information has long hampered a detailed
molecular characterization of the inhibitory mode of these herbicidal
compounds. With emerging resistances of weeds against commonly used
herbicides, knowledge of structure-function relationship can pave the way
towards the rational design of optimized herbicides and the development of
herbicide resistant PDS versions in crops. The importance of such investigations
Discussion
133
is reflected by the fact that they have met the interest of the agrochemical
company Syngenta and collaboration has been initiated. In accordance
with the corresponding confidentiality agreement, not all results can be
presented at this point. The refinement of the OsPDS-His6 structure in a
complex with NFZ was key to the investigations concerning PDS-targeting
herbicides (Brausemann et al., 2017).
The data presented reveal that NFZ exhibits competitive inhibition towards PQ
(Fig. 4-15) which is in accordance with results obtained with complex cell-free
assay systems (Breitenbach et al., 2001). Moreover, as shown in a preceding
publication (Gemmecker et al., 2015), NFZ is tightly bound to OsPDS-His6
during enzyme purification but can be displaced by DPQ. This suits the
structure-derived notion that PQ and NFZ share the same binding site. In
contrast, NFZ exhibits either uncompetitive or non-competitive inhibition
towards phytoene (Fig. 4-15), occupying the same substrate cavity. Non-
competitive inhibition is supported by the fact that the crystallized PDS-NFZ
complex represents an enzyme-inhibitor complex (Brausemann et al., 2017)
with no substrate bound. This situation can only occur for competitive or non-
competitive inhibitors whereas uncompetitive inhibitors sensu stricto require
the presence of the enzyme-substrate complex (Copeland, 2000). Moreover,
non-competitive inhibition towards phytoene has been suggested earlier
(Sandmann et al., 1989; Mayer et al., 1989).
This differential inhibition mode regarding the two substrates further supports
the prevalence of an ordered ping pong bi bi mechanism (see 5.1.3), with
phytoene or phytofluene binding to the FADox state of the enzyme and DPQ
and NFZ binding to the FADred state. In line with this, Gemmecker et al. (2015)
reported that NFZ binds and – being redox-inactive - stabilizes the reduced
enzyme state after carotene desaturation.
NFZ as quinone-competitive inhibitor lowers the apparent PQ availability,
thereby preventing FADred reoxidation and the execution of continuous
desaturation cycles (see 5.1.5.2). Thus, NFZ inhibition diminishes enzymatic
Discussion
134
activity, i.e. carotene flux from phytoene via phytofluene to ζ-carotene. This
favors phytofluene formation relative to ζ-carotene formation, as the latter
depends on the former in this two-step reaction. This fits with the observation
that low PQ supply increases the phytofluene proportion (Fig. 4-14). Previously,
this circumstance has been misinterpreted in terms of a differential NFZ
susceptibility of the two desaturation reactions.
Despite significant structural differences compared to NFZ, the m-CF3-phenyl
containing herbicides diflufenican and fluridone compete with PQ binding, as
shown here (Fig. 4-15) and as hypothesized earlier (Laber et al., 1999). This
emphasizes the importance of the common structural features, namely the
m-CF3-phenyl moiety and the presence of a carbonyl group in a second
rather hydrophobic ring structure. However, in silico docking of diflufenican
and fluridone into the PQ binding site was not successful, suggesting an
induced fit upon inhibitor binding. Further investigation would require
crystallography of OsPDS-His6 containing complexed diflufenican or fluridone.
These structures would provide valuable models for the calculation and
prediction of non-covalent interactions between PDS and inhibitors using the
density functional theory-based method of Johnson et al. (2010).
Such calculations were used during this thesis to elucidate the molecular basis
of the long-known importance of the m-CF3 substituent in NFZ (Mayer et al.,
1989; Sandmann et al., 1989; Sandmann and Böger, 1982), fluridone
(Sandmann et al., 1992) and diflufenican (Cramp et al., 1987). The role of this
functionality has mainly been attributed to its lipophilicity and
electronegativity so far. The quantum mechanical calculations presented
reveal that the m-CF3 substituent provides five weakly attractive non-covalent
interactions with hydrogen or sulphur atoms of Met188, Met277, Ala280 and
Phe423 (Fig. 4-16), all representing conserved PDS residues of the PQ binding
site (Fig. 4-20). Far less attractive interactions are expected when CF3 is
replaced by CH3, Cl or H and thus, the “magic” of CF3 indeed involves a
combination of lipophilicity, electronegativity and fluorine effects rather than
Discussion
135
a single “all-or-none” interaction. Such computational approaches when
applied to additional compounds could substantially facilitate the rational
design of new PDS-targeting herbicides – as opposed to cumbersome
combinatorial chemistry approaches.
Interestingly, despite the lack of structural information, early studies have led
to a fairly accurate prediction of the binding site of m-CF3-phenyl herbicides
by Laber et al. (1999). Indeed, the m-CF3-phenyl moiety is surrounded by a
conserved pocket of lipophilic residues such as Met188, Met277, Ala280, Phe423,
Phe162 and Met310 (Fig. 4-16). The conserved Arg300 represents the postulated
residue for hydrogen bonding of the oxygen functionality in vicinity to the m-
CF3-phenyl group.
5.1.6.2 Site-directed mutagenesis confers norflurazon resistance at the
expense of catalytic activity
The effective use of PDS-targeting herbicides in agriculture would benefit from
resistant crops. This can nowadays be achieved by applying state of the art
genome-editing methods since PDS mutations conferring resistance to NFZ
and fluridone have been identified in cyanobacteria, algae and in one plant
(see 1.3.1.3). However, it remained elusive how these mutations confer
resistance and in which way enzyme characteristics might be affected.
In order to address these questions, such PDS mutant enzymes were
generated and biochemically characterized (see 4.3.2). In summary,
individual mutation of the residues Phe162, Arg300 and Leu538 resulted in
enzymes that suffered a substantial loss of enzymatic activity in vitro
compared to the wild type enzyme, despite retaining a native-like tertiary
structure. In depth analysis of the most active mutant enzyme, Arg300Ser PDS,
revealed that homooligomerization was not affected either (Fig. 4-18).
However, the kinetic parameters were drastically altered (Table 6) as
witnessed by higher affinity for DPQ (KM lower by factor 3) and phytoene (KM
lower by factor 12) and lower Vmax regarding both substrates, the latter
reflecting the enzyme’s loss of activity. Moreover, as expected from reports
Discussion
136
about this mutation (Arias et al., 2006; Martinez-Férez et al., 1994), the affinity
for NFZ was found to be lowered as witnessed by a 7 fold higher Ki (Table 6).
Thus, the point mutation Arg300Ser led to global kinetic changes.
Removing a formal charge from an active center as in Arg300Ser PDS
represents a major change with possible long-range conformational changes
exerting multiple effects. More specifically, the increased active center
lipophilicity in the mutant might result in enhanced affinity for the lipophilic
substrates phytoene and DPQ (lower KM for both) and the decreased affinity
for the smaller, more hydrophilic inhibitor NFZ (higher Ki). These parameter
changes contradict the notion of a simple analogy in NFZ and DPQ binding
via hydrogen bonding of their keto functions (see 5.1.6.1) but plead for a
contribution of the hydrocarbon moiety in the case of PQ. Interpreting
lowered KM in terms of affinities, the increased affinity for substrates might as
well be accompanied by increased affinity for the similarly lipophilic products.
It may therefore be conceivable that product release and the rapid
sequential order of carotene and PQ binding are hindered in the mutant,
leading to diminished catalytic activity. In terms of herbicide resistance, the
relative affinities for DPQ and the quinone-competitive NFZ need to be
considered. Despite increased DPQ affinity, the enzyme exhibits increased
NFZ resistance (higher Ki). This emphasizes the importance of hydrogen
bonding of NFZ by Arg300 – besides other non-covalent interactions, e.g. with
the m-CF3-phenyl moiety (see 4.3.3). This supports the above given notion that
affinity for the more hydrophobic PQ does not solely rely on hydrogen
bonding of its keto function but strongly relies on additional non-covalent
interactions with its hydrocarbon moiety.
Substrate concentration dependencies for Arg300Ser PDS showed that
increasing phytoene concentrations favored the formation of the
intermediate phytofluene in relation to the end product ζ-carotene, like with
wild type OsPDS-His6 (compare Fig. 4-19 and Fig. 4-14). However, the mutant
preferentially released phytofluene and not ζ-carotene at any given
Discussion
137
phytoene concentration – in contrast to the wild type enzyme (compare Fig.
4-19 and Fig. 4-14). This was mainly due to strongly impaired ζ-carotene
formation, while phytofluene was present in wild type-like quantities. In the
light of the mathematical model for OsPDS-His6, the simplest explanation
would be that the lowered enzymatic activity and carotene flux through the
mutant enzyme mainly diminishes ζ-carotene formation as it additionally
depends on previous phytofluene synthesis. Mathematical modeling of
reaction time courses with Arg300Ser PDS was tried but did not allow further
disentangling the sub-processes affected in the mutant. Impaired substrate
channeling appears unlikely as a cause since homooligomerization as the
prerequisite remained unaffected (Fig. 4-18).
As a conclusion, all PDS mutant enzymes generated to confer NFZ resistance
(Arg300Ser, Arg300Thr, Leu538Phe, Leu538Arg and Phe162Val) in principle allowed
for ζ-carotene formation in E. coli, engineered to produce 15-cis-phytoene,
and in vitro – but at much lower rates than the wild type enzyme. In
consideration of the results obtained for Arg300Ser PDS, engineering NFZ
resistance trades-in lowered catalytic activity of the enzyme. However, this
might suffice for carotenogenesis in vivo, as witnessed by the wild type-like
carotenoid levels of NFZ-resistant mutants (see 1.3.1.3). It is conceivable that
the loss of activity is compensated by regulatory mechanisms in planta.
Genome editing would be the method of choice to investigate whether
engineering of PDS can produce agriculturally useful herbicide-tolerant plants
with uncompromised performance.
Discussion
138
5.2 ζ-carotene desaturase ZDS: a comparison to PDS
5.2.1 OsZDS-His6 utilizes an ordered ping pong bi bi mechanism and a
“flavin only” mechanism
In recent years and during this thesis, substantial progress has been made in
the understanding of PDS (Gemmecker et al., 2015; Brausemann et al., 2017;
see 4.1, 4.2, 4.3, 5.1). Little is known about ZDS, although many similarities are
anticipated because of homology and a similar function. Thus, investigations
on ZDS were initiated as this enzyme is crucial to an understanding of the
poly-cis pathway of carotene desaturation (for details, see 4.4).
OsZDS-His6 was natively purified in the absence of detergents to near
homogeneity as a yellow enzyme. LC-MS analysis revealed FAD as sole
cofactor bound by a conserved Rossman fold (Fig. 4-22). OsZDS-His6 rapidly
attached to substrate containing liposomes and is active in the absence of
supplemented FAD, yielding proneurosporene (PN) and prolycopene (PL)
from 9,9’-di-cis-ζ-carotene and exhibiting high regio-specificity (Fig. 4-21). Thus,
it represents a monotopic membrane flavoprotein. Its redox cofactor FAD
requires reoxidation by PQ, the latter serving as diffusible, terminal electron
acceptor. The occurrence of carotene desaturation in the absence of PQ
and the observed release of the intermediate from the enzyme revealed that
both carotene desaturation reactions and PQ reduction are per se
independent of each other (see 4.4.1.3). Consequently, OsZDS-His6 is
expected to follow an ordered ping pong bi bi mechanism and represents a
bifunctional ζ-carotene – neurosporene desaturase, catalyzing a two step-
desaturation cascade with formally identical reactions on the substrate half
sides. PQ is only required for FADred reoxidation to allow repeated reaction
cycles and exerts kinetic control over the overall reaction towards
prolycopene. Thus, regarding these basic mechanisms, similarities with PDS
are very significant.
In line with this, homology modeling (carried out because crystallization
attempts failed) revealed high structural similarity with OsPDS (Fig. 4-24). A
Discussion
139
single hydrophobic substrate cavity was present, capable of entirely
accommodating both carotene substrates and PQ but not simultaneously.
The FAD-containing active center of OsZDS was found to lack acidic and
basic amino acids that could mediate acid-base catalysis of carotene
desaturation. Thus, ZDS seems to employ a “flavin only” mechanism with the
isoalloxazine as the sole catalyst, similar to the reaction mechanism proposed
for PDS (for details, see 4.4.2).
Characterization of the substrate concentration-dependent kinetics of OsZDS-
His6 revealed high similarities with OsPDS-His6, supporting that ZDS utilizes similar
mechanisms to favor end product formation in its two-step carotene
desaturation. First, OSZDS-His6 exhibits higher affinity for the intermediate PN
than for the initial substrate ζ-carotene. The KM for PN is 13.9 ± 2.5 mM and,
following the rationale detailed for PDS that KM for the substrate should be
determined via formation of the intermediate (see 5.1.5.2), KM could not be
determined experimentally for ζ-carotene and is higher than 50 mM (see
4.4.3.3). Accordingly, increasing concentrations of the initial substrate ζ-
carotene favored formation of the intermediate over the end product, as it
competes with PN (Fig. 4-28 A). Thus, low carotene fluxes through ZDS are
essential to favor end product formation. Second, OSZDS-His6 exhibits higher
affinity for PQ than for its carotene substrates with a KM of only 0.3 ± 0.1 mM for
DPQ. This ensures rapid enzyme reoxidation for repeated carotene
desaturation cycles and conceivably facilitates rapid conversion of the
recently expelled intermediate PN to PL. In line with this, high PQ supply
favored end product formation in relation to intermediate formation (Fig. 4-28
C). Thus, similarities with PDS are significant in terms of kinetic properties.
5.2.2 OsZDS-His6 might not employ homooligomerization and substrate
channeling
A difference between OsZDS-His6 and OsPDS-His6 came to light upon GPC
analysis. OsZDS-His6 remained flavinylated and is most likely active as
monomer (see 4.4.2.3) whereas OsPDS-His6 is only flavinylated and active as
Discussion
140
homotetramer (see 4.2.2 and 5.1.5). It remains to be investigated whether
OsZDS-His6 forms homooligomers at membrane surfaces to employ substrate
channeling of its intermediate PN. In analogy to OsPDS-His6, the size and
oligomeric state of OsZDS-His6 in its active state on liposomal surfaces could
be determined by cryo scanning electron microscopy. Notably, certain
amounts of higher-order oligomers observed during GPC only under non-
reducing conditions might point towards unstable homooligomerization (see
4.4.2.3). Breitenbach et al. (1999) observed homodimerization for Capsicum
ZDS, stating that this might be due to hydrophobic aggregation.
Mathematical modeling of OsZDS-His6 reaction time courses could be applied
to further investigate oligomerization and substrate channeling. Given the
strong similarity of PDS and ZDS, the “monomeric model” and “substrate
channeling model” for OsPDS-His6 might be applicable to OsZDS-His6.
Should OsZDS-His6 not assemble as homooligomer, the question arises whether
ZDS is rather a constituent of a heterooligomeric metabolon, e.g. with CRTISO
that is located downstream of ZDS, as suggested by Lundqvist et al. (2017).
Native expression of OsZDS and LeCRTISO (Yu et al., 2011) in E. coli has
already been achieved and bimolecular fluorescence complementation
(BiFC) assays could help revealing structural interactions. Alternatively, a
different organization might facilitate intermediate release from ZDS to serve
other functions. It has been proposed that a cis-neurosporene might serve as
precursor for a signaling compound involved in the feedback regulation of
carotenogenesis (Kachanovsky et al., 2012). In fact, as will be discussed in
5.4, 9-cis-neurosporene that is formed by CRTISO from proneurosporene (Yu et
al., 2011) can be converted by AtCCD7 to yield 9-cis-ζ-apo-10’-carotenal as
a potential signal precursor.
Discussion
141
5.3 ζ-Carotene Isomerase ZISO: a bona fide enzyme with the
potential for additional functions
In the framework of this thesis, functional expression in E. coli and a liposomal
in vitro assay were established for OsZISO-His6 (see 4.5.1). N-terminal fusion of
the Mistic protein from Bacillus subtilis, a fusion partner mediating membrane
targeting and insertion in E. coli (Roosild et al., 2005), was crucial for obtaining
active OsZISO-His6. This is in accordance with ZISO being a transmembrane
protein, as predicted (Chen et al., 2010). Cell lysates containing Mistic-OsZISO-
His6 isomerized 9,15,9‘-tri-cis-ζ-carotene to 9,9‘-di-cis-ζ-carotene in the
absence of all other carotenogenic enzymes (Fig. 4-31) showing that ZISO is
indeed an enzyme rather than a modifier of PDS specificity, the latter being
capable of isomerizing the C9-C10 double bond during phytoene
desaturation (Fig. 1-3 A). Enzyme function was then shown by Beltrán et al.
(2015) who reported its first native purification and in vitro isomerization of tri-
cis-ζ-carotene shortly afterwards. Beltran et al. (2015) furthermore reported on
the cofactor requirements and mechanistics of ZISO, involving a ferrous heme
b cofactor. Their findings are not further detailed here and prompted a
discontinuation of OsZISO-His6 purification attempts during this thesis.
Nevertheless, the data presented here allow first insights into substrate
recognition by ZISO. Apart from 9,15,9‘-tri-cis-ζ-carotene, 9,15-di-cis-
phytofluene was shown to be readily isomerized by Mistic-OsZISO-His6, in
contrast to 15-cis-phytoene (see 4.5.1). It can be concluded that a petaene
and a 9,15-di-cis configuration are the minimal substrate requirements. This
would correspond to recognition of only one half side of the symmetric,
canonical substrate 9,15,9’-tri-cis-ζ-carotene. But further experimentation is
needed. Carotene isomerization can be reversible as shown for the cis-trans-
ß-carotene isomerase D27 (Bruno and Al-Babili, 2016). Reversibility from 9,9’-di-
cis- to 9,15,9’-tri-cis-ζ-carotene would inter alia imply that 9-cis configuration
suffices for recognition by ZISO. Accordingly, 9-cis-isomers of other linear
carotenes (ζ-carotene, neurosporene and lycopene) could be potential ZISO
Discussion
142
substrates as well, depending on the contribution of the polyene system to
ZISO substrate specificity. These questions surely need to be addressed in vitro.
Future research should focus on ZISO topology and protein-protein
interactions in the light of suggestions in favor of a carotenogenic
supercomplex achieving metabolite channeling (Cunningham and Gantt,
1998; Shumskaya et al., 2013; Nisar et al., 2015). ZISO is the only
transmembrane protein of the poly-cis carotene desaturation pathway while
all other enzymes interact monotopically with membranes (Gemmecker et
al., 2015; Albrecht et al., 1995; Yu et al., 2011; Yu and Beyer, 2012). ZISO would
be a predestined adaptor to organize the metabolon by protein-protein
interaction. Such a role was for instance found for the transmembrane protein
Erg28P in sterol biosynthesis (Mo and Bard, 2005). First observations may point
towards such a role: Lundqvist et al. (2017) used a BN-PAGE – mass
spectrometry approach and reported that ZISO and PDS as well as CRTISO
and ZDS might be arranged in two heterocomplexes in Arabidopsis.
Additionally, the data presented point towards substrate channeling
between PDS and ZISO in vitro (see 4.5.3). In the absence of OsPDS-His6, when
9,15,9’-tri-cis-ζ-carotene was deposited in liposomes as substrate, Mistic-
OsZISO-His6 activity was low. In contrast, the identical amount of enzyme
mediated much more isomerization (on a relative and absolute scale) when
the substrate was simultaneously produced by OsPDS-His6. This suggests that
ZISO gains more readily access to its substrate in the presence of OsPDS-His6,
conceivably due to substrate channeling. Mass spectrometry approaches
represent a viable option for the identification and characterization of the
suggested membrane-bound carotenogenic metabolon, as demonstrated
for the peroxisomal importomer (Oeljeklaus et al., 2012).
Discussion
143
5.4 The poly-cis pathway of carotene desaturation and retrograde
signaling
5.4.1 AtCCD7 forms linear 9-cis-apocarotenoids as potential signaling
molecule precursors
In recent years, apocarotenoids originating from CCD-mediated cleavage of
poly-cis-configured linear carotenes have been suggested to serve regulatory
functions in plants (see 1.4). Such ideas are of special interest in the light of a
significant conundrum in carotenoid research: why has the highly intricate
poly-cis pathway of carotene desaturation evolved in plants although its
overall reaction can be achieved with only one enzyme, as in bacteria and
fungi (Fig. 1-2)? Sensing desaturation intermediate levels and translating these
into regulatory functions could provide a clue. These lines of thinking have
been fostered by the discovery that strigolactones and abscisic acid, both
important phytohormones, are derived from the cleavage of cis-configured
carotenoids. Avendaño-Vázquez et al. (2014) postulated cis-ζ-carotene or cis-
phytofluene to be cleaved by CCD4 to yield an apocarotenoid participating
in plastid retrograde signaling that regulates seedling development in
Arabidopsis. Kachanovsky et al. (2012) suggested prolycopene,
proneurosporene or a cis-neurosporene to give rise to an apocarotenoid
mediating feedback regulation of fruit-specific PSY1 in tomato.
The lack of evidence for the existence of such linear apocarotenoids,
prompted the here presented research on CCD-mediated cleavage of
canonical intermediates of the poly-cis pathway of carotene desaturation
and non-canonical isomers derived thereof (see 4.6). The carotenoid
cleavage dioxygenases AtCCD4 and AtCCD7 were investigated for primary
cleavage since they are plastid localized (the site of carotene desaturation)
and they are known to cleave cyclic C40 carotenes (Bruno et al., 2016; Bruno
et al., 2014). In contrast, CCD8 does not cleave C40 carotenoids (Alder et al.,
2012) and CCD1 is localized in the cytosol (Floss and Walter, 2009).
Discussion
144
Contrary to expectation (Avendaño-Vázquez et al., 2014), AtCCD4, did not
cleave of any cis- or trans-configured linear carotene (see 4.6.1). This is in
accordance with the reported specificity of AtCCD4 for trans-configured
cyclic carotenoids (Bruno et al., 2016). Thus, AtCCD4 seems to require the
cyclic ionone moieties for substrate recognition. The data presented in this
work rule out that AtCCD4 cleaves cis-ζ-carotenes. The only possibility for
AtCCD4 involvement would therefore relate to a secondary cleavage of an
apocarotenoid originating from a desaturation intermediate.
In contrast, AtCCD7 was found to specifically cleave the trans-configured C9-
C10 double bond in 9-cis-ζ-carotene, 9’-cis-neurosporene and 9-cis-lycopene
whereas poly-cis- and all-trans-configured species were not converted (see
4.6.2). This is in accordance with the 9-cis-specificity of AtCCD7 for bicyclic
carotenoids (Bruno et al., 2014) that appears to persist with acyclic carotenes.
The apocarotenoids formed by AtCCD7 were identified as 9-cis-ζ-apo-10’-
carotenal (from ζ-carotene and neurosporene) and 9-cis-apo-10’-lycopenal
(from lycopene) by LC-MS and the stereo-configuration of the lycopenal was
verified by its chromatographic behavior differing from authentic all-trans-
apo-10’-lycopenal. Cleavage of the 9-cis isomers of neurosporene and
lycopene was comparably weaker than cleavage of 9-cis-ζ-carotene. This is
somewhat surprising considering the fact that they resemble the canonical
CCD7 substrate 9-cis-β-carotene much more with respect to the number of
double bonds present. The explanation might be that chromophore
elongation hinders single bond rotation within the polyene. The resulting
enhanced rigidity is known to cause problems upon solubilization with
detergents so that more desaturated substrates like neurosporene and
lycopene are not optimally presented to the CCD enzymes.
The findings of this work are reminiscent of SL biosynthesis where CCD7
cleaves 9-cis-β-carotene at the trans-configured C9-C10 double bond to
yield a 9-cis-C27-apocarotenoid (9-cis-β-apo-10’-carotenal) (Al-Babili and
Bouwmeester, 2015). Moreover, ABA as the second carotenoid-derived
Discussion
145
phytohormone has a 9-cis-carotenoid precursor, too (Marion-Poll and
Nambara, 2005). There seems to be the recurrent motif of a 9-cis-configured
carotenoid representing precursors of regulatory molecules. Following this
idea, the data presented in this work support an involvement of AtCCD7 in
the biosynthesis of regulatory molecule derived directly or indirectly from the
poly-cis pathway of carotene desaturation.
Interestingly, 9-cis configuration is widespread amongst carotenes of the poly-
cis pathway: 9-cis double bonds are introduced at both C9-C10 and C9’-C10’
by PDS (Fig. 1-3 A) and persist throughout the desaturation pathway
intermediates until being isomerized to trans by CRTISO (see Fig. 1-2). Thus, all
of the poly-cis pathway intermediates might be predestined as precursors of
regulatory molecules. Moreover, 9-mono-cis-isomers of linear carotenes might
indeed be formed in planta. 9’-cis-neurosporene and 9-cis-lycopene can be
enzymatically formed in vitro by the carotene isomerase CRTISO (Isaacson et
al., 2004; Yu et al., 2011), a property that has been exploited in this work to
generate 9’-cis-neurosporene as substrate. In contrast, the potential origin of
9-cis-ζ-carotene in planta remains elusive. Taking the resemblance with SL
biosynthesis into account, the canonical 9,9’-di-cis-ζ-carotene should be
assayed in vitro as substrate with the 9-cis-β-carotene isomerase D27. Further
candidates could be the yet uncharacterized CRTISO-like enzymes (Fantini et
al., 2013) that might represent carotene isomerases with the required
specificity. Moreover, thermo- and photoisomerization should be considered.
9-cis-ζ-carotene and other isomers were regularly found in E. coli (see 4.5.1), in
the absence of specific carotene isomerases.
5.4.2 Linear 9-cis-apocarotenoids are not converted into strigolactone-like
metabolites by AtCCD8
Signaling molecules need to move between aqueous cell compartments in
order to exert their functions, a feature that does not apply to the rather
hydrophobic C27 apocarotenoids 9-cis-ζ-apo-10’-carotenal and 9-cis-apo-10’-
lycopenal. Further modification or cleavage by CCDs should be considered.
Discussion
146
Given a potential resemblance to SL biosynthesis, CCD8 would be the prime
candidate to potentially yield strigolactone-like compounds from the linear
apocarotenoids (Alder et al., 2012). However, no conversion was observed
(Fig. 4-36) which is somewhat surprising: The postulated CCD8 reaction
mechanism for carlactone formation from 9-cis-β-apo-10’-carotenal (Alder et
al., 2012) involves only C11=C12 and C13=C14 to participate directly in the
intramolecular rearrangement being surrounded by C15=C15’ and C9=C10.
All of these double bonds as well as 9-cis-configuration are present in 9-cis-ζ-
apo-10’-carotenal and 9-cis-apo-10’-lycopenal, the latter being equipped
with a polyene that strongly resembles β-apo-10-carotenal. It remains to be
assayed whether the lycopenal can be converted by AtCCD8. However, the
absence of the ionone moiety might hinder recognition by AtCCD8.
Consequently, the newly identified apocarotenoids 9-cis-ζ-apo-10’-carotenal
and 9-cis-apo-10’-lycopenal might rather undergo modification by hitherto
unknown enzyme activities. Accordingly, conversion of 9-cis-ζ-apo-10’-
carotenal and 9-cis-apo-10’-lycopenal by AtCCD1 and AtCCD4, both known
to convert apocarotenoids (Bruno et al., 2016; Floss and Walter, 2009) should
be considered. However, AtCCD4 did not cleave the ζ-carotenal and
cleavage by AtCCD1 did not yield a cleavage product that was visible in the
UV/VIS wavelength range from 250 nm to 400 nm (Fig. 4-36). Considering the
broad cleavage site specificity of CCD1 enzymes yielding diverse products
from one substrate (Floss and Walter, 2009), it seems unlikely that a CCD1-
mediated reaction might yield a regulatory molecule.
The history of low molecular mass signaling molecule identification shows that
plant mutant characterization might help to further elucidate the fate of
these two apocarotenoids. Alternatively, the classic approach of synthesizing
radioactivity-labeled apocarotenoids, track their fates upon application to
plant cells and identify their derivatives by LC-MS and NMR analyses remains
a viable option.
Summary
147
6 Summary
Plant carotenoid biosynthesis in plastid membranes proceeds via the so-
called poly-cis pathway of carotene desaturation, converting 15-cis-
phytoene into all-trans-lycopene via poly-cis-configured intermediates. Its four
constituent enzymes are poorly understood in terms of mechanisms and
structure. In addition, the reason for the prevalence of these intricate
reactions remains enigmatic considering that the bacterial carotene
desaturation pathway requires only one enzyme.
In this work, the crystal structure of phytoene desaturase (PDS) from O. sativa,
the first desaturation pathway enzyme, in complex with its inhibitor norflurazon
(NFZ) was refined and structural implications were functionally evaluated by a
kinetic characterization. The kinetics of the downstream ζ-carotene
desaturase (ZDS), representing a PDS homolog, were investigated for
comparison and revealed pronounced similarities. The data presented
support an ordered ping pong bi bi kinetic mechanism for PDS and ZDS.
Desaturation of their substrates occurs via two formally identical,
mechanistically independent reactions on half sides of the symmetric
substrate. The intermediate is released into the membrane after the first
desaturation. The enzyme-bound FADred produced by carotene desaturation
is reoxidized by plastoquinone (PQ) to allow repeated rounds of catalysis. As
inferred from the active site structure, data support an unprecedented “flavin
only” mechanism characterized by FADox as sole catalyst of carotene
desaturation. PDS and ZDS interact monotopically with membranes to access
their substrates. Mathematical modelling implies substrate channeling
between subunits within PDS homotetramers whereas ZDS is active as
monomer. Notably, PDS does not produce ζ-carotene in the
stereoconfiguration required by ZDS. In this work, a cell-free activity assay
demonstrated that the membrane-integral ζ-carotene isomerase (ZISO) is a
bona fide isomerase representing the link between PDS and ZDS.
Summary
148
An additional aspect of this work relates to the mode of action of bleaching
herbicides targeting PDS and containing meta-CF3-phenyl moieties, such as
NFZ. It is demonstrated that they compete with PQ and NFZ, occupying the
PQ binding site as suggested in crystallo. The meta-CF3 substituent provides
unique non-covalent interactions with conserved residues and is crucial for
inhibitory effectiveness. Mutagenesis of the conserved Arg300, forming a
hydrogen bond with NFZ, allows engineering NFZ-resistant PDS but at the
expense of altered enzyme kinetics and partial loss of enzymatic activity.
Evidence has been presented that carotenoid cleavage dioxygenases
(CCDs) might form retrograde signals from cis-configured carotene
desaturation intermediates, regulating plastid development and carotenoid
biosynthesis. Investigating intermediate cleavage by CCDs, Arabidopsis CCD7
was found to cleave 9-cis-ζ-carotene yielding 9-cis-ζ-apo-10’-carotenal (C27)
which, in contrast to strigolactone biosynthesis, is not converted by AtCCD8.
These data suggest this apocarotenoid as a candidate signal precursor
requiring further modification by hitherto unidentified enzyme activities.
References
149
7 References
Al-Babili S, Hugueney P, Schledz M, Welsch R, Frohnmeyer H, Laule O, Beyer P (2000)
Identification of a novel gene coding for neoxanthin synthase from Solanum tuberosum. FEBS
Lett 485: 168–72
Al-Babili S, von Lintig J, Haubruck H, Beyer P (1996) A novel, soluble form of phytoene
desaturase from Narcissus pseudonarcissus chromoplasts is Hsp70-complexed and
competent for flavinylation, membrane association and enzymatic activation. Plant J 9: 601–
612
Albrecht M, Klein A, Hugueney P, Sandmann G, Kuntz M (1995) Molecular cloning and
functional expression in E. coli of a novel plant enzyme mediating ζ-carotene desaturation.
FEBS Lett 372: 199–202
Alder A, Jamil M, Marzorati M, Bruno M, Vermathen M, Bigler P, Ghisla S, Bouwmeester H, Beyer P, Al-Babili S (2012) The Path from β-carotene to Carlactone, a Strigolactone-Like Plant
Hormone. Science 335: 1348–1351
Álvarez D, Voß B, Maass D, Wüst F, Schaub P, Beyer P, Welsch R (2016) Carotenogenesis Is
Regulated by 5′UTR-Mediated Translation of Phytoene Synthase Splice Variants. Plant Physiol.
172: 2314-2326
Arango J, Jourdan M, Geoffriau E, Beyer P, Welsch R (2014) Carotene Hydroxylase Activity
Determines the Levels of Both β-Carotene and Total Carotenoids in Orange Carrots. Plant Cell
26: 2223–2233
Arias RS, Dayan FE, Michel A, Howell J, Scheffler BE (2006) Characterization of a higher plant
herbicide-resistant phytoene desaturase and its use as a selectable marker. Plant Biotechnol
J 4: 263–273
Arnoux P, Morosinotto T, Saga G, Bassi R, Pignol D (2009) A structural basis for the pH-
dependent xanthophyll cycle in Arabidopsis thaliana. Plant Cell 21: 2036–2044
Avendaño-Vázquez A-O, Cordoba E, Llamas E, San Román C, Nisar N, De la Torre S, Ramos-Vega M, Gutiérrez-Nava MDLL, Cazzonelli CI, Pogson BJ, León P (2014) An Uncharacterized
Apocarotenoid-Derived Signal Generated in ζ-Carotene Desaturase Mutants Regulates Leaf
Development and the Expression of Chloroplast and Nuclear Genes in Arabidopsis. Plant Cell
Tetrahydropyrimidinones: A New Herbicidal Class of Compounds Inducing Chlorosis by
Inhibition of Phytoene Desaturation. Pestic Biochem Physiol 52: 33–44
Bai L, Kim E-H, DellaPenna D, Brutnell TP (2009) Novel lycopene ε-cyclase activities in maize
revealed through perturbation of carotenoid biosynthesis. Plant J 59: 588–599
Bartley GE, Scolnik PA, Beyer P (1999) Two Arabidopsis thaliana carotene desaturases,
phytoene desaturase and ζ-carotene desaturase, expressed in Escherichia coli, catalyze a
poly-cis pathway to yield pro-lycopene. Eur J Biochem 259: 396–403
Bartnikas TB, Tosques IE, Laratta WP, Shi J, Shapleigh JP (1997) Characterization of the nitric
oxide reductase-encoding region in Rhodobacter sphaeroides 2.4.3. J Bacteriol 179: 3534–
3540
Beltrán J, Kloss B, Hosler JP, Geng J, Liu A, Modi A, Dawson JH, Sono M, Shumskaya M, Ampomah-Dwamena C, Love JD, Wurtzel ET (2015) Control of carotenoid biosynthesis through
a heme-based cis-trans isomerase. Nat Chem Biol 11: 598–605
Beyer P, Al-Babili S, Ye X, Lucca P, Schaub P, Welsch R, Potrykus I (2002) Golden Rice:
introducing the β-carotene biosynthesis pathway into rice endosperm by genetic engineering
to defeat vitamin A deficiency. J Nutr 132: 506S–510S
References
150
Beyer P, Mayer M, Kleinig H (1989) Molecular oxygen and the state of geometric isomerism of
intermediates are essential in the carotene desaturation and cyclization reactions in daffodil
chromoplasts. Eur J Biochem 184: 141–150
Bonk M, Hoffmann B, Lintig J, Schledz M, Al-Babili S, Hobeika E, Kleinig H, Beyer P (1997)
Chloroplast Import of Four Carotenoid Biosynthetic Enzymes In Vitro Reveals Differential Fates
Prior to Membrane Binding and Oligomeric Assembly. Eur J Biochem 247: 942–950
Booker J, Auldridge M, Wills S, McCarty D, Klee H, Leyser O (2004) MAX3/CCD7 Is a
Carotenoid Cleavage Dioxygenase Required for the Synthesis of a Novel Plant Signaling
Molecule. Curr Biol 14: 1232–1238
Bouvier F, D’Harlingue A, Backhaus RA, Kumagai MH, Camara B (2000) Identification of
neoxanthin synthase as a carotenoid cyclase paralog. Eur J Biochem 267: 6346–6352
Bouvier F, D’Harlingue A, Hugueney P, Marin E, Marion-Poll A, Camara B (1996) Xanthophyll
Biosynthesis. J Biol Chem 271: 28861–28867
Bramley PM (2002) Regulation of carotenoid formation during tomato fruit ripening and
development. J Exp Bot 53: 2107–2113
Brausemann A, Gemmecker S, Koschmieder J, Ghisla S, Beyer P, Einsle O (2017) Structure of
Rice (O. sativa) Phytoene Desaturase Provides Insights Into Herbicide Binding and Reaction
Mechanisms Involved In Carotene Desaturation, Structure, in press
Breitenbach J, Bruns M, Sandmann G (2013) Contribution of two ζ-carotene desaturases to
the poly-cis desaturation pathway in the cyanobacterium Nostoc PCC 7120. Arch Microbiol
195: 491–498
Breitenbach J, Fernández-González B, Vioque A, Sandmann G (1998) A higher-plant type ζ-
carotene desaturase in the cyanobacterium Synechocystis PCC6803. Plant Mol Biol 36: 725–
732
Breitenbach J, Kuntz M, Takaichi S, Sandmann G (1999) Catalytic properties of an expressed
and purified higher plant type ζ-carotene desaturase from Capsicum annuum. Eur J Biochem
265: 376–383
Breitenbach J, Sandmann G (2005) ζ-Carotene cis isomers as products and substrates in the
plant poly-cis carotenoid biosynthetic pathway to lycopene. Planta 220: 785–793
Bruno M, Al-Babili S (2016) On the substrate specificity of the rice strigolactone biosynthesis
enzyme DWARF27. Planta 243: 1429–1440
Bruno M, Hofmann M, Vermathen M, Alder A, Beyer P, Al-Babili S (2014) On the substrate- and
stereospecificity of the plant carotenoid cleavage dioxygenase 7. FEBS Lett 588: 1802–1807
Bruno M, Koschmieder J, Wüst F, Schaub P, Fehling-Kaschek M, Timmer J, Beyer P, Al-Babili S
(2016) Enzymatic study on AtCCD4 and AtCCD7 and their potential to form acyclic
regulatory metabolites. J Exp Bot 67: 5993–6005
Bruno M, Vermathen M, Alder A, Wüst F, Schaub P, van der Steen R, Beyer P, Ghisla S, Al-Babili S (2017) Insights into the formation of carlactone from in-depth analysis of the CCD8-
catalyzed reactions. FEBS Lett 591: 792–800
Bugos RC, Hieber AD, Yamamoto HY (1998) Xanthophyll cycle enzymes are members of the
lipocalin family, the first identified from plants. J Biol Chem 273: 15321–15324
Carol P, Kuntz M (2001) A plastid terminal oxidase comes to light: Implications for carotenoid
References
151
biosynthesis and chlororespiration. Trends Plant Sci 6: 31–36
Cazzonelli CI (2011) Carotenoids in nature: Insights from plants and beyond. Funct Plant Biol
38: 833–847
Cazzonelli CI, Pogson BJ (2010) Source to sink: regulation of carotenoid biosynthesis in plants.
Sandmann G, Schmidt A, Linden H, Böger P (1991) Phytoene desaturase, the essential target
for bleaching herbicides. Source Weed Sci 39: 474–479
Sandmann G, Ward CE, Lo WC, Nagy JO, Böger P (1990) Bleaching herbicide flurtamone
interferes with phytoene desaturase. Plant Physiol 94: 476–478
Santabarbara S, Casazza AP, Ali K, Economou CK, Wannathong T, Zito F, Redding KE, Rappaport F, Purton S (2013) The requirement for carotenoids in the assembly and function of
the photosynthetic complexes in Chlamydomonas reinhardtii. Plant Physiol 161: 535–546
Schaub P, Yu Q, Gemmecker S, Poussin-Courmontagne P, Mailliot J, McEwen AG, Ghisla S, Al-Babili S, Cavarelli J, Beyer P (2012) On the structure and function of the phytoene desaturase
CRTI from Pantoea ananatis, a membrane-peripheral and FAD-dependent
References
156
oxidase/isomerase. PLoS One (7): e39550
Schledz M, Al-Babili S, Lintig J v., Haubruck H, Rabbani S, Kleinig H, Beyer P (1996) Phytoene
synthase from Narcissus pseudonarcissus: functional expression, galactolipid requirement,
topological distribution in chromoplasts and induction during flowering. Plant J 10: 781–792
Serganov A, Nudler E (2013) A Decade of Riboswitches. Cell 152: 17–24
Shumskaya M, Wurtzel ET (2013) The carotenoid biosynthetic pathway: Thinking in all
dimensions. Plant Sci 208: 58–63
Sikkema J, De Bont JAM, Poolman B (1994) Interactions of cyclic hydrocarbons with biological
membranes. J Biol Chem 269: 8022–8028
Soetaert K, Petzoldt T, Setzer RW (2010) Solving Differential Equations in R: Package deSolve. J.
Stat. Softw. 33: 1–25
Steinbrenner J, Sandmann G (2006) Transformation of the green alga Haematococcus
pluvialis with a phytoene desaturase for accelerated astaxanthin biosynthesis. Appl Environ
Microbiol 72: 7477–7484
Suarez J V., Banks S, Thomas PG, Day A (2014) A new F131V mutation in Chlamydomonas
phytoene desaturase locates a cluster of norflurazon resistance mutations near the FAD-
binding site in 3D protein models. PLoS One 9: e99894
Takabayashi A, Kadoya R, Kuwano M, Kurihara K, Ito H, Tanaka R, Tanaka A (2013) Protein co-
migration database (PCoM-DB) for Arabidopsis thylakoids and Synechocystis cells.
Springerplus 2: 148
Tian L (2015) Recent advances in understanding carotenoid-derived signaling molecules in
regulating plant growth and development. Front Plant Sci 6: 1–4
Tian L, Dellapenna D (2004) Progress in understanding the origin and functions of carotenoid
hydroxylases in plants. Arch Biochem Biophys 430: 22–29
Wagner T, Windhövel U, Römer S (2002) Transformation of tobacco with a mutated
cyanobacterial phytoene desaturase gene confers resistance to bleaching herbicides. Z
Naturforsch 57c: 671-679
Walter MH, Strack D (2011) Carotenoids and their cleavage products: biosynthesis and
functions. Nat Prod Rep 28: 663–692
Welsch R, Beyer P, Hugueney P, Kleinig H, von Lintig J (2000) Regulation and activation of
phytoene synthase, a key enzyme in carotenoid biosynthesis, during photomorphogenesis.
Planta 211: 846–854
Wessel D, Flügge UI (1984) A Method for the Quantitaive Recovery of Protein in Dilute Solution
in the Presence of Detergents and Lipds. Anal Biochem 138: 141–143
Wheeldon I, Minteer SD, Banta S, Barton SC, Atanassov P, Sigman M (2016) Substrate
channelling as an approach to cascade reactions. Nat Chem 8: 299–309
Yu Q, Beyer P (2012) Reaction specificities of the ε-ionone-forming lycopene cyclase from rice
(Oryza sativa) elucidated in vitro. FEBS Lett 586: 3415–3420
Yu Q, Ghisla S, Hirschberg J, Mann V, Beyer P (2011) Plant carotene cis-trans isomerase
CRTISO: A new member of the FADred-dependent flavoproteins catalyzing non-redox
reactions. J Biol Chem 286: 8666–8676
Yu Q, Schaub P, Ghisla S, Al-Babili S, Krieger-Liszkay A, Beyer P (2010) The lycopene cyclase
CrtY from Pantoea ananatis (formerly Erwinia uredovora) catalyzes an FADred-dependent
non-redox reaction. J Biol Chem 285: 12109–12120
Acknowledgements
157
8 Acknowledgements
I would like to thank Prof. Dr. Peter Beyer for supervising my dissertation,
offering guidance and encouraging me to work and think independently.
I’m indebted to Dr. Patrick Schaub, Dr. Ralf Welsch and Dr. Florian Wüst, for
teaching me the knowledge and skills that were crucial for the success of this
work and for their technical assistance and useful advice.
I’m glad to have conducted this dissertation alongside Dr. Mark Bruno and Dr.
Daniel Álvarez and would like to thank them for their constant support. Having
spent a lot of time together, inside and outside the laboratory, I consider you
not only colleagues but good friends.
I would like to thank Carmen Schubert for her help whenever it was needed.
Thank you to Prof. Dr. Sandro Ghisla for helpful advice and providing a better
understanding of enzyme kinetics and mechanisms over the past years.
I would like to show appreciation for the work of my cooperation partners
Anton Brausemann and Dr. Mirjam Fehling-Kaschek who provided invaluable
structural and functional insight into phytoene desaturase.
A big thanks goes to Domi, Frede and Steve for sharing my passion for sports
on all our skiing and biking trips and for giving me an understanding of a new
take on things that helped me during this thesis.
Lastly but most importantly, I would like to express my deep gratitude to my
parents, my sister and my grandparents. I know you won’t agree but all I am
and have achieved, I owe to you. Thank you for always being there!