Synthesis of α,β-unsaturated aldehydes as potential substrates for bacterial luciferases

Article abstract of DOI:10.1016/j.bmc.2017.01.013

Bacterial luciferase catalyzes the monooxygenation of long-chain aldehydes such as tetradecanal to the corresponding acid accompanied by light emission with a maximum at 490?nm. In this study even numbered aldehydes with eight, ten, twelve and fourteen carbon atoms were compared with analogs having a double bond at the α,β-position. These α,β-unsaturated aldehydes were synthesized in three steps and were examined as potential substrates in vitro. The luciferase of Photobacterium leiognathi was found to convert these analogs and showed a reduced but significant bioluminescence activity compared to tetradecanal. This study showed the trend that aldehydes, both saturated and unsaturated, with longer chain lengths had higher activity in terms of bioluminescence than shorter chain lengths. The maximal light intensity of (E)-tetradec-2-enal was approximately half with luciferase of P. leiognathi, compared to tetradecanal. Luciferases of Vibrio harveyi and Aliivibrio fisheri accepted these newly synthesized substrates but light emission dropped drastically compared to saturated aldehydes. The onset and the decay rate of bioluminescence were much slower, when using unsaturated substrates, indicating a kinetic effect. As a result the duration of the light emission is doubled. These results suggest that the substrate scope of bacterial luciferases is broader than previously reported.

Full text of DOI:10.1016/j.bmc.2017.01.013

Accepted Manuscript
Synthesis of α,β-unsaturated aldehydes as potential substrates for bacterial lu-
ciferases
Eveline Brodl, Jakov Ivkovic, Chaitanya R. Tabib, Rolf Breinbauer, Peter
Macheroux
PII:
S0968-0896(16)31416-X
http://dx.doi.org/10.1016/j.bmc.2017.01.013
BMC 13492
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
13 December 2016
9 January 2017
11 January 2017
Please cite this article as: Brodl, E., Ivkovic, J., Tabib, C.R., Breinbauer, R., Macheroux, P., Synthesis of α,β-
unsaturated aldehydes as potential substrates for bacterial luciferases, Bioorganic & Medicinal Chemistry (2017),
doi: http://dx.doi.org/10.1016/j.bmc.2017.01.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis of α, β-unsaturated aldehydes as potential substrates for
bacterial luciferases
Eveline Brodla, Jakov Ivkovicb,1, Chaitanya R. Tabiba, Rolf Breinbauerb and Peter Macherouxa,*
aGraz University of Technology, Institute of Biochemistry, Petersgasse 12/2, 8010 Graz,
Austria
bGraz University of Technology, Institute of Organic Chemistry, Stremayrgasse 9, 8010 Graz,
Austria
1current affiliation: University of Lund, Kemicentrum, Center for Analysis and Synthesis,
Naturvetarvägen 14, 22100 Lund, Sweden.
*Corresponding author at present address: Graz University of Technology, Petersgasse 12/2,
A-8010 Graz, Austria. Tel.: +43 3168736450. Email address: peter.macheroux@tugraz.at
Keywords: Keywords: bacterial bioluminescence; luciferase; molecular docking; α, β-
unsaturated aldehydes;
Abbreviations: DIBALH: diisobutylaluminum hydride; DMAP: 4-(dimethylamino)-pyridine;
DMF: N,N-dimethylformamide; DCM: dichloromethane; THF: tetrahydrofuran
1

Abstract
Bacterial luciferase catalyzes the monooxygenation of long-chain aldehydes such as
tetradecanal to the corresponding acid accompanied by light emission with a maximum at
490 nm. In this study even numbered aldehydes with eight, ten, twelve and fourteen carbon
atoms were compared with analogs having a double bond at the α, β-position. These α, β-
unsaturated aldehydes were synthesized in three steps and were examined as potential
substrates in vitro. The luciferase of Photobacterium leiognathi was found to convert these
analogs and showed a reduced but significant bioluminescence activity compared to
tetradecanal. This study showed the trend that aldehydes, both saturated and unsaturated,
with longer chain lengths had higher activity in terms of bioluminescence than shorter chain
lengths. The maximal light intensity of (E)-tetradec-2-enal was approximately half with
luciferase of P. leiognathi, compared to tetradecanal. Luciferases of Vibrio harveyi and
Aliivibrio fisheri accepted these newly synthesized substrates but light emission dropped
drastically compared to saturated aldehydes. The onset and the decay rate of
bioluminescence were much slower, when using unsaturated substrates, indicating a kinetic
effect. As a result the duration of the light emission is doubled. These results suggest that
the substrate scope of bacterial luciferases is broader than previously reported.
1. Introduction
The “cold-light” phenomenon - the enzymatic production of light commonly known as
bioluminescence - can be found in prokaryotes and eukaryotes.1 The involvement of long-
chain aliphatic aldehydes as substrates in bacterial bioluminescence has been known since
1953 when various chain lengths of the substrates were investigated by Strehler and
Cormier.2–4 In the 1960s, the role of these potential substrates was analyzed concerning the
reaction velocity, the initial maximal intensity and the decay of luminescence.5,6 In 1963
Spudich and Hastings tested the first unsaturated aldehyde, 2-decenal, and showed
complete inactivity with this substrate.7 Cormier et al. were the first to prove that long-chain
aldehydes were definitely required for light production.8,9 Further investigations focused on
the determination of the stoichiometry, the quantum yield and the product of this
bioluminescent reaction.10–13 It took some more years to identify tetradecanal as the
“natural” substrate for bacterial bioluminescence in 1974 by Shimomura.14 By now it is
known, that the luciferase catalyzes the monooxygenation of long-chain aliphatic aldehydes
to the corresponding acids employing reduced flavin mononucleotide (FMNH2) as redox
cofactor (Scheme 1).1
Scheme 1: General reaction of bacterial bioluminescence.
2

The initial step of the reaction is the binding of reduced FMN to luciferase. The enzyme-
FMNH2 complex reacts with molecular oxygen to form flavin-4a-hydroperoxide. This
relatively stable enzyme-FMNHOOH complex subsequently reacts with a long chain aliphatic
aldehyde to form a flavin-4a-peroxyhemiacetal intermediate.15 Its slow decay results in the
oxidation of the aldehyde to the corresponding acid and the free energy released during this
reaction populates an excited state flavin-4a-hydroxide, which in turn serves as the light
emitting molecule.16
Bacterial luciferases are heterodimeric enzymes consisting of an α-subunit and a β-subunit.
The two subunits have a sequence identity of approximately 32 % and have evolved from a
common ancestor.17 The active site of the enzyme is exclusively on the α-subunit and also
distant from the subunit interface. The exact role of the β-subunit is not clear, but deletion
or mutation of this subunit reveals less or complete loss of activity. A mutation of βTyr151,
for instance, has a negative effect on FMNH2 binding. It seems that the β-subunit is
responsible for high quantum yield and protein stability.17,18
Only two crystal structures of bacterial luciferases have been reported, where one of them
17,19
elucidated the structure of the apo-LuxAB of Vibrio harveyi
and the other one revealed
the apo-LuxAB of Vibrio harveyi soaked with FMN.18 The isoalloxazine ring of the flavin
shows a planar conformation and is held in place by mainly backbone contacts. The amino
acids involved in the binding of the 5’ phosphate are Arg107, Arg125, Glu175, Ser176 and
Thr179 (Figure 1).18,20 Both structures designate a TIM barrel fold (βα)8 for the enzyme. Both
subunits have a loop between the β-strand 7 and α-helix 7. The α-subunit, in contrast to the
β-subunit, has 29 additional amino acids and a stretch of disordered residues from Lys283 to
Arg290. This loop region is the most conserved region of the luciferase sequence. It is highly
protease-labile, but binding of FMN or polyvalent anions can prevent proteolytic
inactivation.17,18 Complete deletion of the loop results in reduction of total quantum yield by
two orders of magnitude. It was hypothesized that the mobile loop has a lid-gating
mechanism similar to other TIM-barrel enzymes.21 This loop is in close proximity to the
active center and seems to undergo conformational changes from an open or semi open
state to a closed state after flavin binding and before reaction with oxygen.17,18,21
3

Figure 1: Graphical representation of FMN bound to luciferase of V. harveyi. Based on the crystal
structure18, FMN and a few key residues discussed in the text are depicted in stick confirmation with
according labelling. General color code is used for the atoms. Residues Arg107, Arg125, Glu175,
Ser176 and Thr179 are responsible for binding of the 5’ phosphate group of FMN. His44, Ser227 and
Trp250 might have a role in substrate binding and interaction.
Currently, structural information on the spatial arrangement of FMN and the aldehyde
substrate in the active site is lacking, however various mutagenesis and modelling studies
were performed in the last years. It was assumed that the flavin binding pocket is large
enough to accommodate FMNH2, O2 and long chain aldehydes.22 In particular, two amino
acids, Cys106 and Ser227, have attracted interest, because their exchange substantially
affected enzymatic activity. The former amino acid apparently plays an important catalytic
role as alkylation of its reactive thiol led to inactivation of the luciferase (Figure 1).17,18 In
addition it was found that the Cys106Val variant exhibits decreased aldehyde utilization and
reduced stability of the flavin-4a-hydroperoxide intermediate.20 On the other hand,
replacement of Ser227 to phenylalanine in the α-subunit led to a steric effect in the well-
characterized mutant AK-20 (Figure 1). Generally, replacement of Ser227 by large aromatic
amino acids led to a 10-fold decreased binding affinity for aldehyde, smaller amino acids,
e.g. alanine had no influence.17,23 Modelling studies suggest that the bioluminescent reaction
occurs on the si-face of the isoalloxazine ring facing the amino acid His44 (Figure 1). The
distance of the C4a atom of flavin and the Nδ atom of His44 is approximately 7 Å. This
distance leaves enough space for functional groups of the intermediates (peroxide,
hydroxide) and aldehyde binding. Additionally, a spacious cavity is formed in the active site,
which is surrounded by hydrophobic residues. Among those residues is Trp250, which was
suggested to interact with the aldehyde substrate (Figure 1).20,24 Despite these structural
analyses, the exact structure of bacterial luciferases in complex with FMNH2 and aldehyde
4

substrate is still unknown. A 13C and 15N NMR study by Vervoort et al. analyzed the
differences of bound FMN and bound FMNH2 to luciferase of Vibrio harveyi. There is clearly a
change in electron densities, hydrogen bonding and planarity of the oxidized and reduced
state of FMN. The N10 atom of FMNH2 seems to be slightly out of the molecular plane.25 All
these indications prompt speculations about the reaction mechanism as well as the
substrate scope.
In this study, α, β-unsaturated aldehydes with chain lengths of eight, ten, twelve and
fourteen carbon atoms were synthesized. To investigate the mechanism of bacterial
bioluminescence, the recombinant luciferases from Photobacterium leiognathi, Vibrio
harveyi and Aliivibrio fisheri were chosen as model systems to test these potential substrates
and analyze the substrate specificity of bacterial luciferases.
2. Materials and Methods
2.1. General experimental information
All commercially available reagents and solvents were purchased from Sigma-Aldrich, Alfa
Aesar, Fisher Scientific, Acros Organics, Roth or VWR, and were used without further
purification if not stated otherwise. When it was required, non-dry solvents were distilled
before use. If reactions were performed under inert conditions, e.g. exclusion of water,
oxygen or both, all experiments were carried out using established Schlenk techniques.
Herein solvents were dried and/or degassed with common methods and afterwards stored
under inert gas atmosphere (argon or N2) over molecular sieves. In some cases, when
explicitly mentioned, dry solvents were received from the listed suppliers. DCM (EtOH
stabilized) was distilled first over P4O10 to remove the stabilizer and then over CaH2 under
argon atmosphere and stored over 4 Å molecular sieves in an amber-colored 1000 mL
Schlenk bottle.
All reactions were stirred with Teflon-coated magnetic stirring bars. Molecular sieves (Sigma
Aldrich, beads with 8-12 mesh) were activated in a round-bottom flask with a gas-inlet
adapter by heating them carefully in a heating mantle for approximately 12 h under high
vacuum until complete dryness was obtained. These activated molecular sieves were stored
at room temperature under argon atmosphere.
Temperatures were measured externally if not otherwise stated. Reactions that were carried
out at -78 °C were cooled by keeping the reaction vessel immersed in a properly sized Dewar
vessel containing acetone/dry ice.
Analytical thin layer chromatography (TLC) was carried out on Merck TLC silica gel 60 F254
aluminum sheets and spots were visualized by UV light (λ = 254 and/or 366 nm) or by
staining with iodide, cerium ammonium molybdate (2.0 g Ce(SO4)2, 50.0 g (NH4)6Mo7O24 and
50 mL conc. H2SO4 in 400 mL water) (CAM) or potassium permanganate (0.3 g KMnO4, 20 g
5

K2CO3, 5 mL 5 % aqueous NaOH in 300 mL H2O) followed by the development of the stains in
the heat. Flash column chromatography was performed on silica gel 0.035-0.070 mm, 60 Å
(Acros Organics). A 30 to 100 fold excess of silica gel was used with respect to the amount of
dry crude product, depending on the separation problem. The dimensions of the column
were selected in such a way that the required amount of silica gel formed a pad between
10 cm and 25 cm. The column was equilibrated first with the solvent or solvent mixture, and
the crude product diluted with the eluent was applied onto the top of the silica pad. In case
when the crude product was insoluble in the eluent, the sample was dissolved in an
appropriate solvent (EtOAc or DCM), and the equal amount of diatomaceous earth was
added, followed by removal of the solvent under reduced pressure and drying the sample in
vacuum, which was then directly loaded onto the top of the silica pad. The mobile phase was
forced through the column using a rubber bulb pump.
2.2. General procedure GP-1 (Synthesis of α, β-unsaturated ethyl esters)
In a 100 mL single neck round-bottom flask equipped with a magnetic stir bar, 4-
(dimethylamino)pyridine (122 mg, 1.00 mmol, 0.10 eq), mono-ethyl malonate (2.36 mL,
20.0 mmol, 2.0 eq) and saturated alkyl aldehyde (10.0 mmol, 1.0 eq) were dissolved in DMF
(50 mL). The reaction mixture was stirred at room temperature for 42 h. Subsequently, the
mixture was diluted with diethyl ether (50 mL), washed with saturated aqueous NH4Cl
(50 mL), saturated aqueous NaHCO3 (50 mL), water (50 mL), and concentrated under
reduced pressure. Flash chromatography (SiO2, 5 % EtOAc in cyclohexane) afforded the
desired unsaturated ethyl ester as a colorless liquid.
2.3. General procedure GP-2 (Synthesis of allyl alcohols)
In a nitrogen-purged 100 mL Schlenk tube equipped with a magnetic stir bar, unsaturated
ethyl ester (4.5 mmol, 1.0 eq) was dissolved in dry dichloromethane (18 mL), the vessel was
sealed with a glass stopper and cooled to -78 °C in an acetone/dry ice bath. 1.0 M solution of
diisobutylaluminum hydride in hexanes (10.8 mL, 10.8 mmol, 2.4 eq) was added dropwise
via a syringe and a septum throughout 10 min. The reaction was stirred at -78 °C until TLC
indicated quantitative conversion (3 h). The reaction mixture was quenched by the dropwise
addition of MeOH (1 mL). Subsequently, the cooling bath was removed, saturated aqueous
potassium sodium tartrate solution (18 mL) was added, and the mixture was stirred
vigorously for 2 h. After phase separation the aqueous layer was extracted with
dichloromethane (10 mL). The combined organic layers were dried over anhydrous Na2SO4
and concentrated under reduced pressure. Flash chromatography (SiO2, 20 % EtOAc in
cyclohexane) afforded the desired allyl alcohol as a colorless liquid.
2.4. General procedure GP-3 (Synthesis of α, β-unsaturated aldehydes)
In a nitrogen-purged 10 mL Schlenk tube equipped with a magnetic stir bar, manganese(IV)
oxide (494 mg, 5.0 mmol, 5.0 eq) and activated 4 Å molecular sieves were suspended in dry
dichloromethane (4 mL). Allyl alcohol (1.0 mmol, 1.0 eq) was dissolved in dry
6

dichloromethane (3.3 mL), added to the mixture in the Schlenk tube, which was sealed with
a glass stopper. After stirring the mixture overnight at room temperature the dark brown
reaction mixture was filtered through a compressed pad of diatomaceous earth. The pad
was washed with dichloromethane (2 mL), and the filtrate was concentrated under reduced
pressure. Flash chromatography (SiO2, 10 % EtOAc in cyclohexane) afforded the desired
unsaturated aldehyde as a pale yellow liquid.
2.5. Ethyl (E)-oct-2-enoate (3d)
Unsaturated ester 3d was synthesized and isolated according to the general procedure GP-1
and its stated stoichiometry.
Yield: 940 mg (5.52 mmol, 55 %), colorless liquid.
Rf = 0.35 (cyclohexane/EtOAc 40:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 6.97 (dt, 3J = 15.6 Hz, 3J = 7.0 Hz, 1H), 5.81 (d, 3J = 15.6 Hz, 1H),
4.18 (q, 3J = 7.1 Hz, 2H), 2.25–2.14 (m, 2H), 1.52–1.40 (m, 2H), 1.36–1-25 (m, 7H), 0.82 (t, 3J =
6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 166.9 (s, 1C), 149.6 (s, 1C), 121.4 (s, 1C), 60.2 (s, 1C), 32.3 (s,
1C), 31.4 (s, 1C), 27.8 (s, 1C), 22.6 (s, 1C), 14.4 (s, 1C), 14.1 (s, 1C) ppm.
2.6. Ethyl (E)-dec-2-enoate (3c)
Unsaturated ester 3c was synthesized and isolated according to the general procedure GP-1
and its stated stoichiometry.
Yield: 1.480 g (7.46 mmol, 75 %), colorless liquid.
Rf = 0.35 (cyclohexane/EtOAc 40:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 6.96 (dt, 3J = 15.6 Hz, 3J = 7.0 Hz, 1H), 5.80 (d, 3J = 15.6 Hz, 1H),
4.18 (q, 3J = 7.1 Hz, 2H), 2.26–2.12 (m, 2H), 1.51–1.39 (m, 2H), 1.35–1-21 (m, 11H), 0.88 (t, 3J
= 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 166.9 (s, 1C), 149.6 (s, 1C), 121.4 (s, 1C), 60.3 (s, 1C), 32.3 (s,
1C), 31.9 (s, 1C), 29.3 (s, 1C), 29.2 (s, 1C), 28.2 (s, 1C), 22.8 (s, 1C), 14.4 (s, 1C), 14.2 (s, 1C)
ppm.
7

2.7. Ethyl (E)-dodec-2-enoate (3b)
Unsaturated ester 3b was synthesized and isolated according to the general procedure GP-1
and its stated stoichiometry.
Yield: 1.235 g (5.46 mmol, 55 %), colorless liquid.
Rf = 0.36 (cyclohexane/EtOAc 40:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 6.96 (dt, 3J = 15.6 Hz, 3J = 7.0 Hz, 1H), 5.80 (d, 3J = 15.6 Hz, 1H),
4.18 (q, 3J = 7.1 Hz, 2H), 2.25–2.13 (m, 2H), 1.51–1.39 (m, 2H), 1.35–1-21 (m, 15H), 0.88 (t, 3J
= 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 166.9 (s, 1C), 149.6 (s, 1C), 121.3 (s, 1C), 60.2 (s, 1C), 32.3 (s,
1C), 32.0 (s, 1C), 29.6 (s, 1C), 29.5 (s, 1C), 29.4 (s, 1C), 29.3 (s, 1C), 28.2 (s, 1C), 22.8 (s, 1C),
14.4 (s, 1C), 14.2 (s, 1C) ppm.
2.8. Ethyl (E)-tetradec-2-enoate (3a)
Unsaturated ester 3a was synthesized and isolated according to the general procedure GP-1
and its stated stoichiometry.
Yield: 1.733 g (6.81 mmol, 68 %), colorless liquid.
Rf = 0.38 (cyclohexane/EtOAc 40:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 6.96 (dt, 3J = 15.6 Hz, 3J = 7.0 Hz, 1H), 5.80 (d, 3J = 15.6 Hz, 1H),
4.17 (q, 3J = 7.1 Hz, 2H), 2.25–2.12 (m, 2H), 1.51–1.38 (m, 2H), 1.35–1-21 (m, 19H), 0.87 (t, 3J
= 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 166.9 (s, 1C), 149.6 (s, 1C), 121.3 (s, 1C), 60.2 (s, 1C), 32.3 (s,
1C), 32.0 (s, 1C), 29.8 (m, 2C), 29.7 (s, 1C), 29.5 (s, 1C), 29.4 (s, 1C), 29.3 (s, 1C), 28.2 (s, 1C),
22.8 (s, 1C), 14.4 (s, 1C), 14.2 (s, 1C) ppm.
2.9. (E)-Oct-2-en-1-ol (4d)
Allyl alcohol 4d was synthesized and isolated according to the general procedure GP-2 and
its stated stoichiometry.
8

Yield: 449 mg (3.50 mmol, 78 %), colorless liquid.
Rf = 0.57 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 5.77–5.54 (m, 2H), 4.08 (br s, 2H), 2.03 (dt, 3J = 6.9 Hz, 3J = 6.6
Hz, 2H), 1.45–1.19 (m, 7H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 133.8 (s, 1C), 129.0 (s, 1C), 64.0 (s, 1C), 32.3 (s, 1C), 31.5 (s, 1C),
29.0 (s, 1C), 22.7 (s, 1C), 14.2 (s, 1C) ppm.
2.10. (E)-Dec-2-en-1-ol (4c)
Allyl alcohol 4c was synthesized and isolated according to the general procedure GP-2 and its
stated stoichiometry.
Yield: 605 mg (3.87 mmol, 86 %), colorless liquid.
Rf = 0.57 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
3
1H NMR (300 MHz, CDCl3) δ = 5.77–5.55 (m, 2H), 4.07 (br s, 2H), 2.03 (dt, J = 6.9 Hz, 3J = 6.5
Hz, 2H), 1.46–1.17 (m, 11H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 133.7 (s, 1C), 129.0 (s, 1C), 64.0 (s, 1C), 32.3 (s, 1C), 32.0 (s, 1C),
29.3 (s, 3C), 22.8 (s, 1C), 14.2 (s, 1C) ppm.
2.11. (E)-Dodec-2-en-1-ol (4b)
Allyl alcohol 4b was synthesized and isolated according to the general procedure GP-2 and
its stated stoichiometry.
Yield: 617 mg (3.35 mmol, 74 %), colorless liquid.
Rf = 0.57 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
3
1H NMR (300 MHz, CDCl3) δ = 5.77–5.56 (m, 2H), 4.08 (br s, 2H), 2.03 (dt, J = 6.9 Hz, 3J = 6.6
Hz, 2H), 1.41–1.18 (m, 15H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 133.8 (s, 1C), 129.0 (s, 1C), 64.0 (s, 1C), 32.4 (s, 1C), 32.1 (s, 1C),
29.8–29.6 (m, 2C), 29.5 (s, 1C), 29.4–29.2 (m, 2C), 22.8 (s, 1C), 14.3 (s, 1C) ppm.
2.12. (E)-Tetradec-2-en-1-ol (4a)
9

Allyl alcohol 4a was synthesized and isolated according to the general procedure GP-2 and
its stated stoichiometry.
Yield: 641 mg (3.02 mmol, 67 %), colorless liquid.
Rf = 0.57 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
3
1H NMR (300 MHz, CDCl3) δ = 5.77–5.56 (m, 2H), 4.08 (br s, 2H), 2.03 (dt, J = 6.9 Hz, 3J = 6.6
Hz, 2H), 1.36–1.17 (m, 19H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 133.8 (s, 1C), 129.0 (s, 1C), 64.0 (s, 1C), 32.4 (s, 1C), 32.1 (s, 1C),
29.9–29.7 (m, 3C), 29.7 (s, 1C), 29.5 (s, 1C), 29.4–29.2 (m, 2C), 22.8 (s, 1C), 14.3 (s, 1C) ppm.
2.13. (E)-Oct-2-enal (5d)
Unsaturated aldehyde 5d was synthesized and isolated according to the general procedure
GP-3 and its stated stoichiometry.
Yield: 103 mg (0.816 mmol, 82 %), pale yellow liquid.
Rf = 0.37 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 9.50 (d, 3J = 7.9 Hz, 1H), 6.86 (dt, 3J = 15.6 Hz, 3J = 6.9 Hz, 1H),
6.11 (dd, 3J = 15.6 Hz, 3J = 7.9 Hz, 1H), 2.33 (dt, 3J = 7.2 Hz, 3J = 7.2 Hz, 2H), 1.58–1.42 (m, 2H),
1.37–1.26 (m, 4H), 0.98–0.85 (m, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 194.3 (s, 1C), 159.2 (s, 1C), 133.1 (s, 1C), 32.8 (s, 1C), 31.4 (s,
1C), 27.7 (s, 1C), 22.5 (s, 1C), 14.1 (s, 1C) ppm.
2.14. (E)-Dec-2-enal (5c)
Unsaturated aldehyde 5c was synthesized and isolated according to the general procedure
GP-3 and its stated stoichiometry.
Yield: 122 mg (0.791 mmol, 79 %), pale yellow liquid.
Rf = 0.37 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 9.51 (d, 3J = 7.9 Hz, 1H), 6.85 (dt, 3J = 15.6 Hz, 3J = 6.9 Hz, 1H),
6.11 (dd, 3J = 15.6 Hz, 3J = 7.9 Hz, 1H), 2.33 (dt, 3J = 7.2 Hz, 3J = 7.2 Hz, 2H), 1.57–1.44 (m, 2H),
1.40–1.19 (m, 8H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
10

13C NMR (75 MHz, CDCl3) δ = 194.2 (s, 1C), 159.2 (s, 1C), 133.1 (s, 1C), 32.9 (s, 1C), 31.8 (s,
1C), 29.2 (s, 1C), 29.1 (s, 1C), 28.0 (s, 1C), 22.8 (s, 1C), 14.2 (s, 1C) ppm.
2.15. (E)-Dodec-2-enal (5b)
Unsaturated aldehyde 5b was synthesized and isolated according to the general procedure
GP-3 and its stated stoichiometry.
Yield: 136 mg (0.746 mmol, 75 %), pale yellow liquid.
Rf = 0.38 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 9.50 (d, 3J = 7.9 Hz, 1H), 6.85 (dt, 3J = 15.6 Hz, 3J = 6.9 Hz, 1H),
6.11 (dd, 3J = 15.6 Hz, 3J = 7.9 Hz, 1H), 2.33 (dt, 3J = 7.2 Hz, 3J = 7.2 Hz, 2H), 1.57–1.43 (m, 2H),
1.39–1.18 (m, 12H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 194.3 (s, 1C), 159.2 (s, 1C), 133.1 (s, 1C), 32.9 (s, 1C), 32.0 (s,
1C), 29.6 (s, 1C), 29.5 (s, 1C), 29.4 (s, 1C), 29.3 (s, 1C), 28.0 (s, 1C), 22.8 (s, 1C), 14.2 (s, 1C)
ppm.
2.16. (E)-tetradec-2-enal (5a)
Unsaturated aldehyde 5a was synthesized and isolated according to the general procedure
GP-3 and its stated stoichiometry.
Yield: 159 mg (0.756 mmol, 76 %), pale yellow liquid.
Rf = 0.40 (cyclohexane/EtOAc 9:1 (v/v); staining: KMnO4).
1H NMR (300 MHz, CDCl3) δ = 9.50 (d, 3J = 7.8 Hz, 1H), 6.85 (dt, 3J = 15.6 Hz, 3J = 6.9 Hz, 1H),
6.11 (dd, 3J = 15.6 Hz, 3J = 7.9 Hz, 1H), 2.33 (dt, 3J = 7.2 Hz, 3J = 7.2 Hz, 2H), 1.57–1.44 (m, 2H),
1.38–1.20 (m, 16H), 0.88 (t, 3J = 6.7 Hz, 3H) ppm.
13C NMR (75 MHz, CDCl3) δ = 194.3 (s, 1C), 159.2 (s, 1C), 133.1 (s, 1C), 32.9 (s, 1C), 32.0 (s,
1C), 29.8 (s, 2C), 29.6 (s, 1C), 29.5–29.4 (m, 2C), 29.3 (s, 1C), 28.0 (s, 1C), 22.8 (s, 1C), 14.2 (s,
1C) ppm.
2.17. Instrumentation
UV/Vis absorption spectra were recorded with a Specord 210 spectrophotometer (Analytic
Jena, Jena, Germany). The light emission was measured by a Berthold Technologies Centro
LB 960 microplate Luminometer with Mikro Win version 4.16. Gel filtration was performed
11

using a Superdex-200 column (prep grade XK 16/100; GE Healthcare) with an Äktaexplorer
100 Pharmacia Biotech (GE Healthcare).
1H-, 13C-NMR spectra were recorded on a Bruker AVANCE III 300 spectrometer (1H:
13
300.36 MHz; C: 75.53 MHz). Chemical shifts were referenced to the residual proton and
carbon signal of the deuterated solvent, respectively (CDCl3: δ = 7.26 ppm (1H), 77.16 ppm
(13C)). Signal multiplicities are abbreviated as s (singlet), bs (broad singlet), d (doublet), dd
(doublet of doublet), t (triplet), q (quadruplet), p (pentet) and m (multiplet). Deuterated
solvents for nuclear resonance spectroscopy were purchased from Euriso-top®.
2.18. Design, expression and purification of recombinant His-tagged proteins
LuxAB from Photobacterium leiognathi (ATCC 27561; PL_LuxAB) and YcnD from Bacillus
subtilis were cloned into pET21a vector and transformed into E. coli BL21 (DE3) strain for
expression as described previously.26,27 LuxAB from Vibrio harveyi (ATCC 14126; VH_LuxAB)
and Aliivibrio fisheri (ATCC 7744; AF_LuxAB) were cloned similarly. The genes for VH_LuxAB
and AF_LuxAB were integrated into pET24b vector and transformed into E. coli Rosetta
strain. Both constructs had an additional C-terminal octa-histidine tag. Heterologous
expression cultures were grown at 37 °C in LB media containing kanamycin (50 µg/mL) and
chloramphenicol (20 µg/mL) as selection markers until an OD (600 nm) of 0.6 was reached.
The expression was induced with 0.1 mM IPTG and cells were further grown at 20 °C for
16 h. Cells were harvested by centrifugation (4400 g, 10 min, 4 °C) and the wet cell pellets
were stored at -20 °C. The cell pellets were resuspended in lysis buffer (50 mM NaH2PO4,
300 mM NaCl, 10 mM imidazole, pH 8), lysed by addition of lysozyme and sonication and
after centrifugation the clear supernatant was loaded on 5 mL HisTrap FF/HP columns (GE
Healthcare) for purification via Ni-NTA affinity chromatography. The columns were washed
with wash buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole, pH 8) and the purified
protein fractions were gained with elution buffer (50 mM NaH2PO4, 300 mM NaCl, 300 mM
imidazole, pH 8). After concentrating and buffer exchange to 45 mM Tris-buffer containing
40 mM MES and 20 mM L-malic acid pH 8, a subsequent gel filtration using a Superdex-200
column was performed. The concentration of the various proteins was determined
spectrophotometrically at 280 nm using the extinction coefficients of 82,335 M-1 cm-1
(PL_LuxAB), 84,230 M-1 cm-1 (VH_LuxAB), 83,200 M-1 cm-1 (AF_LuxAB). In the case of YcnD
the concentration was determined using an extinction coefficient of 12,190 M-1 cm-1 at 450
nm.
2.19. In vitro assay
The in vitro assay was performed in 96 well white assay plates. For the assay all enzymes and
substrates were prepared and/or diluted in 100 mM potassium phosphate buffer pH 7. The
reaction mixture contained the enzyme luciferase with either 50 nM for P. leiognathi or
200 nM for V. harveyi or A. fischeri, respectively. YcnD and FMN were adjusted to 1.5 fold of
the luciferase concentration, respectively. Additionally 500 nM NADPH and the substrate-
buffer suspension were added to make up the final volume of 250 µL. The tested substrates
12

include even chain length (8 to 14 carbon atoms) saturated and unsaturated aldehydes,
respectively. Due to the relatively low solubility of aldehydes in water, concentrated
aldehyde suspensions were obtained by adding 5 µL of the substrate to 10 mL of the
reaction buffer, respectively (concentration range of 1,94-3,30 mM; Supplementary Data).5
The reaction was started by injecting NADPH to the reaction mixture (after a delay of 5
seconds) and the readings were subsequently taken every 0.01th of a second for a total of 90
seconds. The light was measured using the luminometer. The light emission was recorded as
emission counts. The area under the curve was taken for calculation of the percentage and
the total light emission of luciferase with tetradecanal (6a) as substrate was considered as
100 %. The data was calculated to 100 nM of luciferase concentration to compare the results
with each other. The values for the conversion of luciferase of P. leiognathi are shown by
means ± SD of seven individual measurements for the saturated aldehydes and four
individual measurements for the unsaturated aldehydes, respectively. The values for the
conversion of luciferase of V. harveyi and A. fischeri are shown by means ± SD of four
individual measurements, respectively.
2.20. Molecular docking
In silico molecular docking studies were performed using Yasara Structure 13.9.8.28 The
crystal structure of the luciferase/flavin complex of Vibrio harveyi was retrieved from the
Protein Data Bank (PDB entry: 3FGC). Structure preparation and all following experiments
were performed within Yasara Structure 13.9.8. All the crystallographic water molecules and
the β-subunit of the luciferase were removed before molecular docking. Missing hydrogens
were added to the molecules by using the clean mode of Yasara.
For the docking experiments the oxidized flavin structure was modified to the flavin-4a-
hydroperoxide intermediate. Therefore the additional hydrogens and the two oxygens were
attached to the molecule and refined by energy minimization using AMBER99 force field,
while fixing Lys283 and Arg290 which connect the luciferase backbone with an unstructured
and therefore missing loop.18 The resulting crystal structure was utilized for docking the
substrate molecules (5a-d, 6a-d) in flexible mode into the rigid receptor using the plugin
Autodock Vina in Yasara Structure 13.9.8.29,30 The docking simulation cell was set to 15 Å
around the flavin-4a-hydroperoxide intermediate and 500 docking runs with an RMSD cutoff
of 2 Å were performed. The docked conformations for each substrate (5a-d, 6a-d) were
ranked according to the distance between the C1 atom of the substrate molecule and the
terminal oxygen atom of the hydroperoxide functional group. The best-ranked docking pose
for each substrate (5a-d, 6a-d) was analyzed in Yasara Structure 13.9.8.
3. Results
To obtain new insights into the activity and selectivity of luciferases, unsaturated aldehydes
with various chain lengths were synthesized and analyzed. The substrate synthesis was
13

carried out in three steps (Scheme 2). The α, β-unsaturated aldehydes were synthesized
starting from commercially available saturated aldehydes with two carbons less in chain
length. The starting aldehydes 1a-d were subjected to a DMAP-catalyzed Knoevenagel
condensation with mono-ethyl malonate (2) to obtain the corresponding unsaturated esters
3a-d.31 3a-d were reduced to the corresponding allyl alcohols 4a-d using 1.2 equivalents
diisobutylaluminum hydride (DIBALH) in DCM at -78 °C and were subsequently oxidized with
manganese(IV) oxide (MnO2) to afford the desired α, β-unsaturated aldehydes 5a-d.
Scheme 2: Synthesis route for aliphatic, unsaturated aldehydes with different chain lengths (C8-
C14). Using DMAP-catalyzed Knoevenagel condensation the starting aldehydes 1a-d were converted
to the unsaturated ethyl esters 3a-d, which were reduced to the corresponding allyl alcohols 4a-d
and were finally oxidized to the α, β-unsaturated aldehydes 5a-d. The exact equivalents, solvents and
temperature conditions are given and the yield for each step is given as percentage.
To test the newly synthesized substrates, an in vitro assay was developed (Scheme 3).
Briefly, 50 nM recombinant luciferase of P. leiognathi (LuxAB) was used in a reaction mixture
with 100 mM potassium phosphate buffer pH 7, 75 nM FMN, 75 nM YcnD, 500 nM NADPH
and substrate-buffer suspensions of 5a-d and 6a-d (see Materials and Methods). YcnD, an
NADPH-dependent oxidoreductase from Bacillus subtilis, reduces FMN to provide the
cosubstrate FMNH2.27 The luciferase then oxidizes the various substrates to their
corresponding acids using the enzyme-bound flavin-4a-hydroperoxide with concomitant
emission of light. We assume that saturated aldehydes 6a-d and unsaturated aldehydes 5a-d
are accepted in a similar way as substrates and converted to the corresponding acids. The
photons, emitted during this reaction, were collected by a luminometer for 90 seconds and
the areas under the light emission curve were compared to tetradecanal (6a).
14

Scheme 3: Schematic representation of the in vitro assay. The synthesized unsaturated aldehydes
5a-d, as well as the four saturated aldehydes 6a-d, were investigated in an in vitro assay. The
oxidation reaction catalyzed by 50 nM luciferase (LuxAB), employing molecular oxygen (O2) and
reduced FMN (FMNH2), results in long-chain aliphatic acids 7a-d and 8a-d and the emission of light
(hν). For the reduced FMN a recycling system was established using the NADPH-dependent
oxidoreductase YcnD from Bacillus subtilis. The light emission is measured by a luminometer and
subsequently converted to total light emission in per cent for comparison and analysis.
As expected, tetradecanal (6a) showed the highest light emission and was set to 100 %. The
other aldehydes showed lower activity with the luciferase (Figure 2). In the case of
dodecanal (6b), decanal (6c) and (E)-tetradec-2-enal (5a), light emission was greater than
50 % in comparison to tetradecanal (6a). Octanal (6d) exhibited the lowest yield of the
saturated aldehydes. The unsaturated aldehyde substrates 5b-d, however, resulted in yields
below 10 %. Comparing saturated aldehydes with each other, substrates with longer chain
length emit more light than those with shorter chain length and therefore are apparently
better substrates for luciferase. This tendency was already observed earlier2,6, however a
clear comparison and definite values were not reported. The same tendency was found for
the unsaturated aldehydes, where total light emission decreases with shorter chain length.
Thus, saturated and unsaturated aldehydes exhibit a similar chain length dependency but
are clearly accepted as substrates for bacterial luciferase.
15

Figure 2: Comparison of saturated and unsaturated aldehydes as potential substrates for the
luciferase of P. leiognathi. The conversion of unsaturated aldehydes 5a-d (light grey) and saturated
aldehydes 6a-d (dark grey) by the luciferase during the in vitro assay (Scheme 3) can be determined
by the total light emission of the reaction measured by the luminometer. The counts of the light
emission were converted to percent. The total light emission (as percentage) was plotted against the
substrates (here differentiated by their chain lengths). The production of light by the conversion of
tetradecanal (6a) was set to 100 %. The values are shown by means ± SD of seven individual
measurements for the saturated aldehydes and four individual measurements for the unsaturated
aldehydes, respectively.
Next, we analyzed the time course of light emission for unsaturated and saturated
aldehydes. As an example, the kinetics of light emission with tetradecanal (6a) and (E)-
tetradec-2-enal (5a) as substrates are shown in Figure 3. Generally, the onset as well as the
decay of the light emission is faster with saturated aldehydes. In the case of 6a a maximum
light emission is reached after ca. 7 s whereas light emission with 5a peaks at ca. 10 s. On the
other hand, light emission lasted much longer for the unsaturated aldehyde 5a than for the
saturated aldehyde 6a. A similar kinetic behavior was observed for all other
saturated/unsaturated aldehyde pairs. This result indicates that the rate-limiting step (or
steps) leading to the population of the excited state luciferin (presumably the flavin-4a-
hydroxide32) is slowed down when unsaturated aldehydes are used as substrates.
16

Figure 3: Time course of the total light emission of tetradecanal (6a) and (E)-tetradec-2-enal (5a).
The light emission during the conversion of the substrates tetradecanal (6a, solid line) and (E)-
tetradec-2-enal (5a, dashed line) was measured as a function of time by the luminometer. The counts
of light emission were plotted as per cent, setting the maximal bioluminescence intensity of
tetradecanal (6a) as 100 %, against the time (in seconds). This is a representative figure of a single
measurement.
Tetradecanal (6a) and (E)-tetradec-2-enal (5a) were chosen as substrates for the in vitro
assay with luciferases from different genera (P. leiognathi, V. harveyi, A. fisheri) as depicted
in Figure 4. The assay conditions were adopted for V. harveyi and A. fisheri. For the latter, a
luciferase concentration of 200 nM was used and the concentrations for FMN and YcnD
were set to 300 nM, respectively. Because light emission was highest with the luciferase
from P. leiognathi it was used as a reference point, i.e. set to 100 %. V. harveyi and A. fisheri
accepted both substrates but showed a much lower activity than P. leiognathi. Comparison
of these two substrates with various luciferases confirms the previous results by depicting a
decline of light emission with unsaturated aldehydes. Nevertheless, (E)-tetradec-2-enal (5a)
is a substrate for various bacterial luciferases.
17

Figure 4: Comparison of the total light emission for luciferases from P. leiognathi, V. harveyi and A.
fisheri. Three luciferases of different genera were compared with each other by analyzing the total
light emission (as percentage) during conversion of the two potential substrates tetradecanal (6a,
dark grey) and (E)-tetradec-2-enal (5a, light grey) in in vitro assays. The values are shown by means ±
SD of four individual measurements, respectively.
To evaluate whether the binding of the substrate within the active site might influence the
enzyme activity and maximal light intensity, a preliminary docking study was performed.
Based on the crystal structure with bound FMN18, the intermediate state of flavin-4a-
hydroperoxide was predicted and the various substrates were docked into the active site.
After energy minimization, the structure with the flavin-4a-hydroperoxide in the active site
was used for further docking studies with the substrates used in this study, i. e. 5a-d and 6a-
d. The docking results obtained with the saturated and unsaturated aldehydes indicate
similar distances of the C1 atom of the respective aldehyde and the distal oxygen atom of
the flavin-4a-hydroperoxide (Supplementary Data).
In Figure 5, an overlay of the two docking results with the substrates tetradecanal (6a) and
(E)-tetradec-2-enal (5a) is depicted. It appears that the two substrates 5a and 6a adopt
similar conformations and orientations, except for the position of the oxygen atom of the
aldehyde that points in opposite directions. The distance between the C1 of the aldehyde
and the distal oxygen of the flavin intermediate is around 3.6-3.7 Å (Supplementary Data).
18

Figure 5: Docking of tetradecanal (6a) and (E)-tetradec-2-enal (5a) into the active site of the
luciferase of V. harveyi with bound flavin-4a-hydroperoxide intermediate. A: Crystal structure of
the α-subunit of the luciferase of V. harveyi with modelled flavin-4a-hydroperoxide intermediate is
depicted in yellow. The unstructured loop is displayed as pink dots. The two substrates tetradecanal
(6a, cyan) and (E)-tetradec-2-enal (5a, blue) are docked into the active site of the luciferase. B: Zoom
into the active site and overlay of tetradecanal (6a, cyan) and (E)-tetradec-2-enal (5a, blue). The
flavin-4a-hydroperoxide intermediate (yellow) is in close proximity to the substrates 6a and 5a,
having a distance between the distal oxygen of flavin-4a-hydroperoxide to C1 of 3,637 Å and 3,740 Å,
respectively.
Thus, our docking results support our experimental findings that unsaturated aldehydes are
accepted substrates and indicate that luciferases have a broader substrate range as
previously assumed. The different orientation observed for the aldehyde function may be a
first hint why unsaturated aldehydes show a substantial difference in kinetics as compared
to their saturated counterparts. The hydrophobic pocket, lined for example by Trp250
(Figure 5) within the active site does not allow binding of bulkier or larger substrates,
however, replacement of amino acids in the active site of luciferase may help to engineer
the putative substrate binding pocket for other aldehyde substrates. Clearly, further
structural and computational methods need to be applied to enhance our understanding of
the mechanism and substrate scope of bacterial luciferases.
4. Discussion
Four different compounds, namely (E)-tetradec-2-enal (5a), (E)-dodec-2-enal (5b), (E)-dec-2-
enal (5c) and (E)-oct-2-enal (5d), were successfully synthesized and analyzed as potential
substrates for recombinant luciferases from three different genera. Spudich and Hastings
showed in 1963 that 2-decenal (referring to (E)-dec-2-enal (5c)) is completely inactive in the
production of light with the luciferase of Achromobacter fisheri. On the contrary, this
compound was found to be a potent competitive inhibitor in bioluminescence with decanal
in the reaction inhibited by decenal. Strangely enough, they have reported similar quantum
yields for the reaction with saturated and unsaturated substrate.7 Additionally, Lei and
coworkers proposed an inhibitory effect of a postulated “luciferase-aldehyde dead end
19

complex”. In this case the sequence of substrate binding (aldehyde and FMNH2) to the
luciferase seems to be essential for activity or inhibition.33 Thus, it was assumed that all α, β-
unsaturated aldehydes exert an inhibitory effect and were therefore not considered as
possible substrates. In contrast to that, we show here that unsaturated aldehydes are
accepted as substrates by various recombinant luciferases from the genera Photobacterium,
Vibrio and Aliivibrio, although the light emission yield was lower with the unsaturated
aldehydes.
In light of our observations, the previously observed inhibitory effect7 seems to be a kinetic
one. The time course of the light emission is strongly influenced by the various substrates.
Unfortunately, substrate-buffer suspensions had to be used for the assays; therefore, it was
not possible to conduct more detailed kinetic measurements. Studies with organic co-
solvents were attempted, but led to denaturation of the luciferases (data not shown). The
solubility of the aldehydes corresponds to the aldehyde chain length according to molar
solubility values (Supplementary Data). Octanal (6d), for instance, should presumably give
better results than tetradecanal (6a), as its solubility in aqueous buffer is higher. However,
the reverse dependency was observed, as aldehydes with longer chain length are more
efficient in light emission in our in vitro assay system.
As mentioned in the introduction, structural information on the active site of luciferase is
still scarce in particular in regard of the positioning of the aldehyde substrate. Current
mechanistic considerations are based on the crystal structure of luciferase of V. harveyi with
bound FMN.18 Modelling studies suggested several amino acids that may play an important
role for binding or interacting with the aldehyde substrate, as for example His44 and Trp250.
Additionally, a spacious hydrophobic cavity was postulated as potential substrate binding
position.20 Nevertheless, a structure of the ternary complex of luciferase, FMNH2 and
aldehyde is still lacking leading to speculations concerning substrate binding and the reaction
mechanism.
5. Conclusion
Bacterial bioluminescence is a fascinating phenomenon and the structure-function
relationships responsible for the population of an excited state remains a scientific challenge
even after decades of research. Here, we have demonstrated that the scope of substrates
utilized by bacterial luciferases is not as limited as previously thought. In this study, the α, β-
unsaturated aldehydes with chain length of 8, 10, 12 and 14 carbon atoms were synthesized
in a three step synthesis approach. To elucidate the conversion of these potential substrates,
an in vitro assay was developed. The four synthesized, unsaturated aldehydes 5a-d, as well
as their saturated analogs 6a-d were analyzed with luciferases from three different genera
(P. leiognathi, V. harveyi, A. fisheri). The results indicate, that all of them are accepted by the
enzymes and show reasonable to low light emission. Comparing the different potential
substrates, tetradecanal (6a) exhibits the highest light emission yield, while three other
20

substrates (5a, 6b, 6c) reached only about 54-75 % of the best performing tetradecanal (6a).
This study comprises the first comparison of these eight aldehydes (5a-d, 6a-d) as substrates
for bacterial luciferases. Having a closer look at the time course of light emission, the
different kinetics in the onset as well as decay of light emission for tetradecanal (6a) and (E)-
tetradec-2-enal (5a) were evident. Also, we have shown that luciferases from other
bioluminescent bacteria show a similar pattern with regard to yield and kinetics of light
emission. In summary, all eight substrates 5a-d and 6a-d were accepted by the luciferase
leading to the conclusion that further investigations on substrate specificity and
compatibility will lead to new insights in to bacterial bioluminescence.
Acknowledgments
This work was supported by the Austrian “Fonds zur Förderung der wissenschaftlichen
Forschung” (FWF) to PM (P24189) and the PhD program “DK Molecular Enzymology” to PM
and RB (W901). We would like to thank Johannes Niederhauser and Dr. Grit Straganz (Graz
University of Technology) for their help with molecular docking experiments using Yasara.
References
(1) Dunlap, P. In Encyclopedia of Microbiology; Schaechter, M., Ed.; Elsevier: Oxford,
2009; pp 45–61.
(2) Strehler, B. L.; Cormier, M. J. J. Biol. Chem. 1954, 211, 213–225.
(3) Strehler, B. L. J. Am. Chem. Soc. 1953, 75, 1264.
(4) Strehler, B. L.; Cormier, M. J. Arch. Biochem. Biophys. 1953, 47, 16–33.
(5) Hastings, J. W.; Gibson, Q. H. J. Biol. Chem. 1963, 238 (7), 2537–2554.
(6) Hastings, J. W.; Spudich, J.; Malnic, G. J. Biol. Chem. 1963, 238 (9), 3100–3105.
(7) Spudich, J.; Hastings, J. W. J. Biol. Chem. 1963, 238 (9), 3106–3108.
(8) Cormier, M. J.; Totter, J. R. Biochim. Biophys. Acta 1957, 25, 229–237.
(9) Cormier, M. J.; Eley, M.; Abe, S.; Nakano, Y. Photochem. Photobiol. 1969, 9, 351–358.
(10) Lee, J. Biochemistry 1972, 11 (18), 3350–3359.
(11) Shimomura, O.; Johnson, F. H.; Kohama, Y. Proc. Natl. Acad. Sci. USA 1972, 69 (8),
2086–2089.
(12) McCapra, F.; Hysert, D. W. Biochem. Biophys. Res. Commun. 1973, 52 (1), 298–304.
(13) Lee, J.; Murphy, C. L. Biochemistry 1975, 14 (10), 2259–2268.
(14) Shimomura, O.; Johnson, F. H.; Morise, H. Proc. Natl. Acad. Sci. USA 1974, 71 (12),
4666–4669.
(15) Macheroux, P.; Ghisla, S.; Hastings, J. W. Biochemistry 1993, 32, 14183–14186.
(16) Dunlap, P. In Bioluminescence: Fundamentals and Applications in Biotechnology -
Volume 1; Thouand, G., Marks, R., Eds.; Springer Berlin Heidelberg: Berlin, Heidelberg,
2014; Vol. 144, pp 37–64.
(17) Fisher, A. J.; Raushel, F. M.; Baldwin, T. O.; Rayment, I. Biochemistry 1995, 34, 6581–
6586.
21

(18) Campbell, Z. T.; Weichsel, A.; Montfort, W. R.; Baldwin, T. O. Biochemistry 2009, 48
(26), 6085–6094.
(19) Fisher, A. J.; Thompson, T. B.; Thoden, J. B.; Baldwin, T. O.; Rayment, I. J. Biol. Chem.
1996, 271 (36), 21956–21968.
(20) Lin, L. Y.-C.; Sulea, T.; Szittner, R.; Vassilyev, V.; Purisima, E. O.; Meighen, E. A. Protein
Sci. 2001, 10 (8), 1563–1571.
(21) Campbell, Z. T.; Baldwin, T. O.; Miyashita, O. Biophys. J. 2010, 99 (12), 4012–4019.
(22) Baldwin, T. O.; Christopher, J. A.; Raushel, F. M.; Sinclair, J. F.; Ziegler, M. M.; Fisher, A.
J.; Rayment, I. Curr. Opin. Struct. Biol. 1995, 5, 798–809.
(23) Chen, L. H.; Baldwin, T. O. Biochemistry 1989, 28, 2684–2689.
(24) Li, Z.; Meighen, E. A. Biochemistry 1995, 34, 15084–15090.
(25) Vervoort, J.; Muller, F.; O’Kane, D. J.; Lee, J. Biochemistry 1986, 25, 8067–8075.
(26) Bergner, T.; Tabib, C. R.; Winkler, A.; Stipsits, S.; Kayer, H.; Lee, J.; Malthouse, J. P.;
Mayhew, S.; Müller, F.; Gruber, K.; Macheroux, P. Biochim. Biophys. Acta BBA -
Proteins Proteomics 2015, 1854 (10), 1466–1475.
(27) Morokutti, A.; Lyskowski, A.; Sollner, S.; Pointner, E.; Fitzpatrick, T. B.; Kratky, C.;
Gruber, K.; Macheroux, P. Biochemistry 2005, 44 (42), 13724–13733.
(28) de Groot, B. L.; van Aalten, D. M. F.; Scheek, R. M.; Amadei, A.; Vriend, G.; Berendsen,
H. J. C. Proteins Struct. Funct. Genet. 1997, 29, 240–251.
(29) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.; Olson,
A. J. J. Comput. Chem. 1998, 19 (14), 1639–1662.
(30) Trott, O.; Olson, A. J. J. Comput. Chem. 2009, 455–461.
(31) List, B.; Doehring, A.; Fonseca, M. T. H.; Wobser, K.; van Thienen, H.; Torres, R. R.;
Galilea, P. L. Adv. Synth. Catal. 2005, 347 (11-13), 1558–1560.
(32) Kurfürst, M.; Ghisla, S.; Hastings, J. W. Proc. Natl. Acad. Sci. USA 1984, 81, 2990–2994.
(33) Lei, B.; Cho, K. W.; Tu, S.-C. J. Biol. Chem. 1994, 269 (8), 5612–5618.
22

Graphical abstract
23

Highlights

α, β-unsaturated aldehydes were synthesized in a three-step approach
 (un-)saturated aldehydes were investigated as potential substrates for luciferases
 saturated aldehydes showed higher light emission than unsaturated aldehydes
 light emission increases with longer aldehyde chain length, i.e. C14>C12>C10>C8
 unsaturated aldehydes display slower kinetics of light emission
24