1
Supplementary Information
Direct evidence for a covalent ene adduct intermediate
in NAD(P)H-dependent enzymes
Raoul G. Rosenthal1,*, Marc-Olivier Ebert2,*, Patrick Kiefer1, Dominik M. Peter1, Julia A.
Vorholt1 & Tobias J. Erb1,†
AFFILIATIONS: 1 Institute of Microbiology ETH Zurich, Wolfgang Pauli Str. 10, 8093 Zurich Switzerland.
2 Laboratory of Organic Chemistry, ETH Zurich, Wolfgang Pauli Str. 10, 8093 Zurich Switzerland.
* These authors contributed equally to this work.
† To whom correspondence should be addressed: E-mail: [email protected] (T.J.E.)
Supplementary Results
- Supplementary Figures 1-13 (page 2-25)
- Supplementary Tables 1-3 (page 26-28)
- Supplementary References (page 29)
Nature Chemical Biology: doi:10.1038/nchembio.1385
2
SUPPLEMENTARY RESULTS
Supplementary Figure 1: Fractions of NAD(P)H-dependent enzyme classes and reactions. To
determine the fraction of NAD(P)H-dependent enzyme classes and reactions, (a) the BRENDA
enzyme database (http://www.brenda-enzymes.org/) and (b) the MetaCyc database
(http://www.metacyc.org/) were accessed online on April 20, 2013. 15.8 percent (930/5858 entries) of
the EC classes listed in the BRENDA enzyme database and 19.4 percent (1481/7649 entries) of the
reactions catalyzed by identified enzymes listed in the MetaCyc database require an NADH and/or
NADPH cofactor.
Nature Chemical Biology: doi:10.1038/nchembio.1385
3
Supplementary Figure 2: The superfamily of medium chain dehydrogenases/reductases
(MDRs). Phylogenetic tree of 117 selected MDR superfamily members belonging to 13 of the largest
subfamilies, as defined recently 1. The subfamilies of Ccr (CCR), Etr1p (MCAS), and YhdH (YHDH)
that react through a covalent ene intermediate according to this study are highlighted in purple. Other
subfamilies of Zn2+-independent enzymes (mainly reductases) are labelled in grey. Black branches
show Zn2+-dependent enzyme subfamilies (mainly dehydrogenases) that are thought to react via the
“classical” one-step hydride transfer. Tree topographies and evolutionary distance are given by the
neighbor-joining method. Numbers at nodes represent the percentage bootstrap values for the clades
of this group in 50 replications. Similar trees were obtained by using the minimum evolution and the
maximum likelihood method. For sequences used, please refer to 2.
Nature Chemical Biology: doi:10.1038/nchembio.1385
4
Supplementary Figure 3: 1H-NMR kinetic analysis of the reduction reaction (and the transient intermediate), recorded at 600 Mhz. Conditions: 1 mM
crotonyl-CoA, 1 mM NADPH, 0.67 µM Ccr in 100 mM Na2HPO4 (pH 7.9, 10% D2O) at 4.4 °C. (a) assigned 1H-NMR reference spectra of NADPH (purple),
NADP+ (blue), crotonyl-CoA (green) and butyryl-CoA (yellow), aligned to the 1H-NMR spectrum of the reduction reaction mixture at maximal intermediate
concentration (red). Stars indicate peaks of the intermediate that could not be assigned to any compound of the starting material and that were used to follow
kinetics of the intermediate by 1H-NMR array. (b) 1H-NMR array of the reduction reaction. Characteristic 1H-signals in the reaction mixture are highlighted
as follows: NADPH, blue diamonds; crotonyl-CoA, blue dots; NADP+, gray diamonds; butyryl-CoA, gray dots; intermediate, purple stars. The integrals of
the indicated peaks were used to create Figure 2 (main text). (c) Fits of the data in Figure 2 to determine initial formation and consumption velocities of
compounds in the reaction mixture. Initial kinetics of substrates, products and intermediate were fitted to a first order reaction, as the Michaelis-Menten rate
equation can be simplified to this solution under the initial conditions of the NMR experiment 3. Initial velocities calculated from the fits are listed in
Supplementary Table 1. NADPH, blue diamonds; crotonyl-CoA, blue dots; NADP+, gray diamonds; butyryl-CoA, gray dots; intermediate, purple stars. (d)
Fits of the data in Figure 2 to determine the rate constants for a two-step process of formation and consumption of the transient intermediate. Rate constants
for each compound in the reaction mixture are reported in Supplementary Table 2, labelling according to c. All data represent single experiments.
a
Nature Chemical Biology: doi:10.1038/nchembio.1385
5
b
Nature Chemical Biology: doi:10.1038/nchembio.1385
6
Time (min)
arb
itra
ry N
MR
in
teg
ral u
nit
s
0 50 100 150
0
1000
2000
3000
4000
Time (min)
arb
itra
ry N
MR
in
teg
ral u
nit
s
0 50 100 150
0
1000
2000
3000
4000
crotonyl-CoA
NADPH
intermediate
NADP+butyryl-CoA
intermediate disapperance
crotonyl-CoA
d
c
Nature Chemical Biology: doi:10.1038/nchembio.1385
7
Supplementary Figure 4: HPLC-ESI-MS analysis of the transient intermediate. Conditions: 1 mM
crotonyl-CoA, 1 mM NADPH and 3 µM Ccr in 100 mM Na2HPO4 (pH 7.9) at 5 °C. (a) Kinetics of the
peak area of the characteristic m/z=791.123 ion at different time points. The first time point (t=0 minutes)
was taken before addition of Ccr. (b) m/z spectrum of t=55 minutes sample, the insert shows the isotopic
distribution. (c) Full measured mass spectrum of the isolated ene adduct used to make figure 3d. The
NADP+ and butyryl-CoA peaks can be attributed to in source fragmentation of the ene adduct –both
peaks perfectly co-elute with the ene adduct around 12.5min. The unlabeled 870.154 m/z peak did not co-
elute with the ene adduct. Shown data are representative of at least two independent experiments.
reaction time (min)
Peak
are
a (
ions x
min
)
0 50 100 150 200 250
0
1100 6
2100 6
3100 6
4100 6
m/z
Ion c
ount
600 700 800 900 1000
0
5100 4
1100 5
2100 5
2100 5
3100 5
744.0
83
791.1
26
802.1
15
838.1
66
810.1
00
m/z
To
tal
ion
co
un
t
791 792 793
0
5100 4
1100 5
2100 5
2100 5
3100 5
791.1
255
791.6
258
792.1
259
792.6
270
793.1
273
m/z
Ion
co
un
t
600 700 800 900 1000
0
5100 5
1100 6
2100 6
2100 6
3100 6
744.0
83
791.1
26
802.1
15
838.1
66
87
0.1
54
a
b
c
Nature Chemical Biology: doi:10.1038/nchembio.1385
8
Supplementary Figure 5: Two-dimensional NMR analysis of the transient intermediate (ene
adduct) in the reaction mixture, recorded at 600 MHz. Conditions: 15 mM NADPH, 15 mM
crotonyl-CoA, and 7.5 μg Ccr were mixed in 100 mM NaHPO4 (pH 7.9, 10% D20) were reacted at 25°C
to generate the intermediate. Then, the reaction was stopped by raising the pH to 11, and NMR spectra
were acquired at 4.4 ˚C. (a) DQF-COSY spectrum of the intermediate (b) HSQC spectrum of the
intermediate in the reaction mixture. For corresponding reference spectra of NADPH, NADP+, butyryl-
CoA, and crotonyl-CoA, see Supplementary Figure 13.
a
Nature Chemical Biology: doi:10.1038/nchembio.1385
9
b
Nature Chemical Biology: doi:10.1038/nchembio.1385
10
Supplementary Figure 6: (In)-stability of the ene adduct in solution as followed by HPLC analysis.
(a) HPLC analysis of subsequent injections of 0.5 mM ene adduct dissolved in 100 mM Na2HPO4 (pH
7.9) incubated at ambient temperature (21°C) that shows the spontaneous decay of the ene addduct into
NADP+ and butyryl-CoA. The times indicated show the absolute time difference between dissolving the
isolated, lyophilized ene adduct in buffer and the time by which the ene adduct passed the UV/Vis
detector. (b) Fraction of the total injection integral at 260 nm for each component in the solution, fitted to
a first order decay. Rate constants (K) and half-times (t1/2, 21 °C) for each component are given.
Extrapolation to t=0 shows that the intermediate is pure, after lyophilization and before it is dissolved.
Shown data represents single experiments.
a
bRetention time (min)
Ab
so
rpti
on
@ 2
60
nm
(m
AU
)
0 5 10 15 200
100
200
300
400
t=13 min
t=43 min
t=73 min
t=103 min
t=133 min
NADP+
intermediatebutyryl-CoA
Time (min)
fra
cti
on
of
tota
l
0 50 100 150
0.0
0.5
1.0
intermediate-decay
t1/2 = 24.50.3min
K=0.029 0.001 min -1
butyryl-CoA-formation
t1/2 = 25.84.4 min
K=0.027 0.001 min -1
NADP+-formation
t1/2 = 23.43.1 min
K=0.028 0.000 min -1
Nature Chemical Biology: doi:10.1038/nchembio.1385
11
Supplementary Figure 7: NMR analysis of the isolated ene adduct. Selected correlations of the ene
adduct are highlighted as follows: bold lines, COSY correlations; arrows, 1H to 13C HMBC correlations.
For full peak assignment, see Supplementary Table 3. The stereochemistry at the Cα-C2 bond of the ene
adduct was derived from the orientation of NADP+ and CoA ester at the active site of CinF, a close
homolog of Ccr that has been recently co-crystallized with both compounds (PDB-code 4a0s; see also
Figure 4a).
Nature Chemical Biology: doi:10.1038/nchembio.1385
12
Supplementary Figure 8: HPLC-ESI-MS analysis of reaction products when using isolated ene
intermediate as substrate for (inactivated) Ccr or Etr1p in the presence or absence of CO2. Products
were confirmed by mixing (“spiking”) with 13C-uniformally labelled metabolite standards. (a-f) Extracted
ion chromatograms of different reactions at m/z = 838.162 (butyryl-CoA, [M+H]+) and m/z = 882.154
(ethylmalonyl-CoA, [M+H]+). (g-l) Extracted mass spectra of the reactions (a-f) at 20.5 min confirming
formation of butyryl-CoA (m/z13C[U]-standard = 860.114) in reactions a, b, d, e, and f, or 15.3 min
confirming formation of ethylmalonyl-CoA (m/z13C[U]-standard = 908.241) in reaction c. Shown data
represent single experiments.
Nature Chemical Biology: doi:10.1038/nchembio.1385
13
Supplementary Figure 9: Reaction of isolated ene intermediate with Ccr, Etr1p YhdH. Time
dependent wavelength scans of reactions of 100 μM ene adduct in 100 mM Na2HPO4 (pH 7.9) at 30
°C with different enzymes. (a) Sample reaction with 0.11 µM Ccr containing additionally 50 mM
NaHCO3 and 10 µg carbonic anhydrase in the reaction mixture (“carboxylation reaction”), (b) Sample
reaction with 0.60 µM Ccr (“reduction reaction”). Note the partial conversion (“back reaction”) of the
ene adduct into NADPH (and crotonyl-CoA). (c) Sample reaction with 0.06 µM Etr1p. (d) Sample
reaction with 0.60 µM YhdH. (e) Control incubation of ene intermediate in the absence of Ccr, Etr1p or
YhdH. In the absence of enzymes, the ene intermediate is slowly decomposed into NADP+ and butyryl-
CoA (see Supplementary Figure 6 for details). Detailed analysis of reaction products from experiments
a-c are shown in Supplementary Figure 8, kinetic parameters for the reactions a-e are listed in Table 1.
Shown data are representative of at least two independent experiments.
a b
c dYhdH
Wavelength (nm)
Ab
so
rpti
on
(A
U)
300 350 400 4500.0
0.2
0.4
0.6
0.8
1.00 s
6 s
11 s
17 s
23 s
28 s
34 s
Ccr +CO2
Wavelength (nm)
Ab
so
rpti
on
(A
U)
300 350 400 4500.0
0.2
0.4
0.6
0.8
1.0 0
11 s
28 s
56 s
112 s
224 s
548 s
Ccr -CO2
Wavelength (nm)
Ab
so
rpti
on
(A
U)
300 350 400 4500.0
0.2
0.4
0.6
0.8
1.0
6 s
11 s
8931 s
0 s
39 s
123 s
2200 s
Etr1p
Wavelength (nm)
Ab
so
rpti
on
(A
U)
300 350 400 4500.0
0.2
0.4
0.6
0.8
1.0 0 s
11 s
28 s
56 s
112 s
224 s
550 s
e uncatalyzed
wavelength (nm)
Ab
so
rpti
on
(A
U)
300 350 400 4500.0
0.2
0.4
0.6
0.8
1.0 0 s
120 s
360 s
840 s
1800 s
3720 s
2760 s
Nature Chemical Biology: doi:10.1038/nchembio.1385
14
Supplementary Figure 10: Detection of the ene adduct by stopped-flow spectroscopy at 385 nm.
Stopped flow measurements of the reaction of (a) 40 μM Ccr or (b) 110 μM YhdH in 25 mM Na2HPO4
(pH 7.9) and 200 mM NaCl, with 250 μM NADPH and 1 mM crotonyl-CoA. The Ccr reaction contained
additionally 100 mM NaHCO3 in the reaction mixture. Formation of the intermediate in the assay was
monitored at 385 nm (purple traces) and compared to control reactions that contained no crotonyl-CoA
(black traces). At 385 nm wavelength, the difference in absorption between NADPH (ε385 = 0.83 mM-1
cm-1) and ene intermediate (ε385 = 5.9 mM-1 cm-1) is maximal (Δε385 = 5.1 mM-1 cm-1). Approximately 0.4
μM ene adduct builds up during carboxylation reaction of Ccr (calculated from the increase of 2 mAU at
385 nm, purple trace in panel a). During reaction of Etr1p, approximately 4 μM ene adduct builds up
(calculated from the increase of 20 mAU at 385 nm, purple trace in panel b). Control spectra of enzyme
with NADPH alone did not show an increase in absorbance at 385 nm (black traces in panel a and b).
Shown data are representative of three technical replicates.
Nature Chemical Biology: doi:10.1038/nchembio.1385
15
Supplementary Figure 11: Apparent steady-state kinetic parameters for the reactions of Ccr, Etr1p
and YhdH with crotonyl-CoA, NADPH, and isolated ene adduct. The data are summarized in Table 1
(main text). (a-c) Characterization of the carboxylation reaction of Ccr in the presence of 50 mM
NaHCO3 and 0.7 µM carbonic anhydrase. (d-f) Characterization of the reduction reaction of Ccr. (g-i)
Characterization of the reduction reaction of Etr1p. (j-l) Characterization of the reduction reaction of
YhdH. Error bars in panels a-e, g and h represent mean standard deviation of three technical replicates.
In panels f, i, and j-l, each data point represents one measurement, all graphs were fitted to at least 10
data points.
intermediate (M)
Rate
(S
-1)
0 50 100 150 2000
1
2
3
[NADPH] (M)
Rate
(S
-1)
0 100 200 300 4000
1
2
3
[crotonyl-CoA] (M)
Rate
(S
-1)
0 2000 4000 60000
1
2
3
[intermediate] (M)
Rate
(S
-1)
0 10 20 300.0
0.5
1.0
1.5
[crotonyl-CoA] (M)
Rate
(S
-1)
0 50 100 1500
20
40
60
80
100
[NADPH] (M)
Rate
(S
-1)
0 20 40 60 800
20
40
60
80
[crotonyl-CoA] (M)
Rate
(S
-1)
0 500 1000 1500 20000
5
10
15
20
[NADPH] (M)
Rate
(S
-1)
0 1000 2000 3000 40000
10
20
30
[intermediate] (M)
Rate
(S
-1)
0 200 400 6000
10
20
30
40
a b c
d e f
g h i
[NADPH] (M)
Rate
(S
-1)
0 20 40 60 800.0
0.5
1.0
1.5
[crotonyl-CoA] (M)
Rate
(S
-1)
0 1000 2000 3000 4000 50000.0
0.2
0.4
0.6
0.8
1.0
[intermediate] (M)
Rate
(S
-1)
0 10 20 30 400.0
0.5
1.0
1.5
2.0
2.5j k l
Nature Chemical Biology: doi:10.1038/nchembio.1385
16
Supplementary Figure 12: Denaturing polyacrylamide gel electrophoretic (SDS-PAGE) analysis of
recombinant enzymes used in this study. (a) Purification of Ccr. Lane 1, Whole cell extract before
purification; Lane 2, soluble protein extract; Lane 3, Ccr after Ni-NTA purification, expected molecular
weight 49.0 kDa. (b) Etr1p after Ni-NTA purification, expected molecular weight 42.0 kDa (c) YhdH
after Ni-NTA purification, expected molecular weight 36.6 kDa. Molecular weight markers (PageRuler
Plus Prestained Protein Ladder, Thermo Scientific) are indicated for each gel.
a b c
Nature Chemical Biology: doi:10.1038/nchembio.1385
17
Supplementary Figure 13: Reference NMR spectra, recorded at 600 MHz in 100 mM Na2HPO4 (pH
7.9) at 4.4 ˚C (a) COSY of butyryl-CoA (b) HSQC of butyryl-CoA (c) COSY of crotonyl-CoA (d)
HSQC of crotonyl-CoA (e) HMBC of crotonyl-CoA (f) COSY of NADPH (g) HSQC of NADPH (h)
COSY of NADP+ (i) HSQC of NADP+.
a COSY of butyryl-CoA
Nature Chemical Biology: doi:10.1038/nchembio.1385
18
b HSQC of butyryl-CoA
Nature Chemical Biology: doi:10.1038/nchembio.1385
19
c COSY of crotonyl-CoA
Nature Chemical Biology: doi:10.1038/nchembio.1385
20
d HSQC of crotonyl-CoA
Nature Chemical Biology: doi:10.1038/nchembio.1385
21
e HMBC of crotonyl-CoA
Nature Chemical Biology: doi:10.1038/nchembio.1385
22
f COSY of NADPH
Nature Chemical Biology: doi:10.1038/nchembio.1385
23
g HSQC of NADPH
Nature Chemical Biology: doi:10.1038/nchembio.1385
24
h COSY of NADP+
Nature Chemical Biology: doi:10.1038/nchembio.1385
25
i HSQC of NADP+
Nature Chemical Biology: doi:10.1038/nchembio.1385
26
SUPPLEMENTARY TABLES 1-3
Supplementary Table 1: Initial velocities of formation or consumption for individual components
in the reaction mixture of Figure 2. Initial velocities were derived by fitting the data in Figure 2
according to a first order decay (see Supplementary Figure 3c for fits) and extrapolating the first order
derivative to time point t=0 min. For details on the fitting procedure, please refer to Supplementary
Figure 3 and Online Methods. Data represent mean 95% confidence intervals of fitted data.
Compound Initial velocity (arbitrary
NMR integrals∙ min-1)
Initial velocity of substrate consumption NADPH crotonyl-CoA
-262 ± 14 -248 ± 4
Initial velocity of product formation NADP+ butyryl-CoA
61 ± 2 44 ± 2
Initial velocity of intermediate formation ene adduct 217 ± 0
Nature Chemical Biology: doi:10.1038/nchembio.1385
27
Supplementary Table 2: Rate constants for the formation and consumption of reaction products,
substrates and the transient intermediate in Figure 2. Rate constants were derived from the data
presented in Figure 2 by fitting the kinetics according to a two-step process →
→ ; with
→
corresponding to formation of the ene adduct at a rate of k1, and → corresponding to consumption of
the ene adduct at a rate of k2 (see Supplementary Figure 3d for fits). Please refer to Supplementary
Figure 3d and Online Methods for details on the fitting procedure. Data represent mean 95%
confidence intervals of fitted data.
compound rate constant k1 (min-1) rate constant k2 (min-1)
Substrates NADPH crotonyl-CoA
0.060 ± 0.03 0.056 ± 0.002
N/A1
N/A
Intermediate ene adduct 0.079 ± 0.019 0.019 ± 0.005
Products NADP+
butyryl-CoA 0.060 ± 0.008 0.074 ± 0.011
0.017 ± 0.001 0.013 ± 0.001
Average 0.066 ± 0.011 0.017 ± 0.003
1 N/A, not applicable
Nature Chemical Biology: doi:10.1038/nchembio.1385
28
Supplementary Table 3: NMR analysis of the isolated ene adduct. Peak assignment for the isolated
ene adduct recorded at 600 MHz in D2O, 100 mM Na2DPO4 (pH 7.9) at 4.4 °C. The corresponding
COSY NMR spectrum and important HMBC and COSY correlations are shown in Supplementary
Figure 7. See Supplementary Figure 5 for the atom labeling used.
Position1 1H-shift2 13C-shift HMBC COSY
α 2.41 61 - Hβa, H2
Βa 1.01 (m, 1H) 213 - Hγ, Hα, Hβb
Βb 1.14 (m, 1H) 213 - Hβa
γ 0.29 (m, 3H) 11 Cα, Cβ Hβa,
2 4.38 56 Cα, C3, C4, C6, 1' Hα
3 - 1113 - -
4 6.40 (d, 1H) 132 C2, C6 H5
5 4.77 (t, 1H) 100 C4 H4, H5
6 6.23 (d, 1H) 137 C2, C4, C(ONH2), 1' H5
1' 4.50 (d, 1H) 97 C2', C2, H2'
2' 3.96 (dd, 1H) 70 1' H1'
3' 3.87 70 1' -4
4' 3.87 65 - -4
5a' 3.59 66 - 5b"
5b' 3.65 66 - 5a"
1a" 3.17 (dd) 72 - H1b"
1b" 3.47 (dd, 1H) 72 H1a"
2" - 38 - -
3" 5 0.37 (s, 3H) 18 C1", C2", C4", C5" -
4" 5 0.51 (s, 3H) 21 C1", C2", C3", C5" -
5" 3.65 (s) 74 C1", C2", C3", C4", C6" -
6" - 175 - -
7" - - - -
8" 3.07 (m, 2H) 36 C6", C9" H9"
9" 2.05 (t, 2H) 36 C8", C10" H8"
10" - 174 - -
11" - - - -
12a" 2.81 (m, 1H) 38 C13", C10" H13"
12b" 2.90 (m, 1H) 38 C13", C10" H13"
13a" 2.50 (m, 1H) 28 C12" H12"
13b" 2.44 (m) 28 C12" H12"
1 Diastereotopic protons are indicated with subscript 'a' and 'b', the upfield proton being assigned as proton 'a'. No absolute stereochemistry was measured. 2 Where possible multiplicity and integral are given in brackets. 3 Chemical shift assigned from HMBC spectrum. 4 COSY signals overlap, therefore no unambiguous assignment was possible. 5 No absolute stereochemistry was determined in this study, however the upfield protons were
assigned according to the literature as belonging to the pro-S methyl 4.
Nature Chemical Biology: doi:10.1038/nchembio.1385
29
REFERENCES
1. Persson, B., Hedlund, J. & Jornvall, H. The MDR superfamily. Cellular and Molecular Life Sciences 65, 3879-3894 (2008).
2. Schada von Borzyskowski, L., Rosenthal, R.G. & Erb, T.J. Evolutionary history and biotechnological future of carboxylases. Journal of biotechnology, in press (2013).
3. Golicnik, M. Exact and Approximate Solutions for the Decades-old Michaelis-Menten Equation: Progress-curve Analysis Through Integrated Rate Equations. Biochemistry and Molecular Biology Education 39, 117-125 (2011).
4. Dordine, R.L., Paneth, P. & Anderson, V.E. 13C NMR and 1H-1H NOEs of Coenzyme-A - Conformation of the Pantoic Acid Moiety. Bioorganic Chemistry 23, 169-181 (1995).
Nature Chemical Biology: doi:10.1038/nchembio.1385