188
M.L. Díaz Chávez et al. / Archives of Biochemistry and Biophysics 507 (2011) 186–193
sites of the transfer vector pVL1392 (BD Bioscience, Franklin Lakes,
NJ, USA). Recombinant baculovirus was generated by cotransfec-
tion with linearized baculovirus DNA using the BaculoGold Trans-
fection Kit (BD Biosciences) and Spodoptera frugiperda Sf9 cells
according to the manufacturer’s instructions. Viral amplification
proceeded according to Ref. [13]. Sf9 cells were propagated in
TC-100 medium (Sigma–Aldrich, St. Louis, MO, USA) supplemented
with 10% fetal bovine serum (Hyclone, South Logan, UT, USA) and
were regularly maintained as previously described [15]. Full length
berberine bridge enzyme cDNA from E. californica (in bacterial vec-
tor pUC19) was generated by PCR with the following primers that
incorporated Pst I and Bam HI restriction sites into the 50- and 30-
ends (respectively): forward (50-GGGCTGCAGATGGAAAACAAAA
CTCCCATC-30) and reverse (50-GGGGGATCCCTATATTACAACTTCTCC
ACC-30). Amplification was achieved with Pfu Hotstart polymerase
using the following cycle parameters: 2 min at 95 °C, succeeded by
30 cycles of amplification (30 s at 95 °C, 30 s at 52 °C, 2 min at
72 °C), and a final extension of 5 min at 72 °C in a total reaction
microsomal P450 protein content using an extinction coefficient of
91 cmꢁ1 mMꢁ1 (data not shown). Cytochrome P450 protein con-
tent in whole cells used for enzyme assays was calculated from a
comparison of enzyme activity obtained with hypotonically lysed
whole cells and microsomes for which the cytochrome P450 con-
tent had been measured.
Enzyme assays
Triple infections that resulted in the production of CYP719A13,
CYP719A14 and CPR were done to initially screen large numbers of
substrates for enzymatic conversion by either cytochrome P450.
The screen was done in this manner in order to conserve on sub-
strate used, since many of these chemicals are not commercially
available. Double infections were then carried out to produce
CYP719A13 and CPR, and CYP719A14 and CPR for kinetic analysis.
The quadruple infection was a readily achievable test for channel-
ing in the pathway. Standard assay conditions contained 30 mM
volume of 50 ll. Recombination in baculovirus and viral amplifica-
tion was identical to CYP719A13 and CYP719A14.
potassium phosphate buffer (pH 8.0), 1.25 mM NADPH, 5
lM sub-
strate, and 70 l of Sf9 cell suspension in a final volume of 200
l
ll.
Control assays were run alongside each reaction without NADPH
and without enzyme. Reactions incubated for 2 h at 30 °C and were
Heterologous expression in S. frugiperda Sf9 cells
terminated by the addition of 400
ll sodium carbonate buffer (pH
Sf9 cells were grown in 50 ml of suspension media (TC-100 sup-
plemented with 10% fetal bovine serum and 0.1% pluronic) to a
density of 2 ꢀ 106 cells/ml (27 °C, 140 rpm). Cells were collected
by centrifugation (900g for 10 min at room temperature), resus-
pended in 7.5 ml of suspension media and moved to a sterile
50 ml flask with foam plug. For double infections, 1.25 ml (MOI
5) of virus containing CYP719A13 or CYP719A14 and 1.25 ml (MOI
5) of virus containing full-length E. californica cytochrome P450
reductase (CPR)3 were added. Each P450 was also coexpressed with
petunia CPR and Arabidopsis thaliana CPR as previously described
[13]. For triple infections, the 7.5 ml of cells were supplemented with
0.83 ml (MOI 3.3) of each recombinant virus (CYP719A13,
CYP719A14, and CPR). Quadruple infections were carried out by the
addition of 0.625 ml (MOI 2.5) of virus containing BBE1 (berberine
bridge enzyme), CYP719A13, CYP719A14, and CPR. Protein collection
for each infection proceeded as previously described [13]. For large
scale infection of cells coexpressing CYP719A13 or CYP719A14 and
CPR, Sf9 cells were grown in 750 ml of suspension media in a 1 L
flask to a density of 2 ꢀ 106 cells/ml (27 °C, 140 rpm). Infection pro-
ceeded as described above with viral volumes proportional to the
culture volume.
9.5) and 400 l chloroform followed by rapid vortexing for 1 min
l
and centrifugation (2 min at 13,000 rpm). Extraction was repeated
on the organic layer once and dried with N2. The final product was
dissolved in 200 ll of 80% methanol.
Kinetic constants for CYP719A13 were determined for (S)-chei-
lanthifoline, (S)-scoulerine, and (S)-tetrahydrocolumbamine and
(S)-scoulerine for CYP719A14 using similar assay conditions as
above with the following amendments. The reaction mixtures with
CYP719A14 contained 70
25.7 pmol cytochrome P450 protein) and were stopped after
15 min. Assays with CYP719A13 contained 35 l of cell suspension
ll of cell suspension (containing 6.6–
l
(containing 1.9–6.0 pmol cytochrome P450 protein) and were
stopped after 10 min. All reactions contained increasing substrate
concentrations (0, 0.5, 1, 5, 10, 15, 20, 30, and 50 lM), proceeded
at 30 °C and were terminated by freezing in liquid nitrogen fol-
lowed by chloroform extraction (described above). The enzymatic
products were quantified by LC–MS/MS analysis and the kinetic
parameters (Km and Vmax) were estimated by non-linear regression
with GraphPad Prism in three independent experiments. Due to
high variation in protein content between different viral infections,
kinetic data were normalized using protein concentration deter-
mined from CO difference spectrum assays. Conversion rates of
product formation for CYP719A13 were determined for (S)-scoul-
erine, (S)-tetrahydrocolumbamine, and (S)-coreximine using the
conversion of (S)-cheilanthifoline to (S)-stylopine as 100% with
reaction conditions stated above.
Microsome preparation and reduced carbon monoxide difference
spectrum
Cells coexpressing CYP719A13 or CYP719A14 and CPR were har-
vested from a large scale infection by centrifugation (1000g for
5 min at room temperature) and washed twice with 40 ml of 1ꢀ
PBS buffer (130 mM NaCl, 3 mM NaH2PO4, 7 mM Na2HPO4). The
pellet was resuspended in 40 ml of ice-cold suspension buffer. Half
of the cells were immediately frozen in liquid nitrogen and placed
in ꢁ80 °C and the remaining 20 ml were subjected to sonication
(Fisher Scientific Sonic Dismembrator Model 100; four times on
ice, 10ꢀ pulsed, 50 W). The remainder of each microsome prepara-
tion and reduced CO difference spectrum assays were carried out
exactly according to Refs. [13,16] using a Cary 300 UV–Visible
spectrophotometer. The resulting spectrum was used to determine
LC–MS/MS analysis of enzyme assays
Product identification and alkaloid quantitation was performed
using LC–MS/MS with Analyst 1.4.2 on a 4000 QTrap (AB Sciex
Instruments) for mass spectroscopic analysis with chromato-
graphic separation on an LC-20AD (Shimadzu). Program parame-
ters included a TurboIonSpray ionization source temperature of
500 °C and low resolution for Q1 and Q3 done with EPI and MRM
scans in the positive ion mode. Fragmentation patterns for (S)-
scoulerine, (S)-cheilanthifoline, and (S)-stylopine were identified
with EPI scans for m/z 328, m/z 326, and m/z 324 ions, respectively.
Similar programs for (S)-coreximine, (S)-coreximine product, (S)-
tetrahydrocolumbamine, (S)-canadine, and (S)-nandinine were
identified with EPI scans for m/z 328, m/z 326, m/z 342, m/z 340,
and m/z 326, respectively. Specific parameters for each method
3
Abbreviations used: CPR, cytochrome P450 reductase; RT-PCR, reverse transcrip-
tase-polymerase chain reaction; PCR, polymerase chain reaction; LC–MS/MS, liquid
chromatography–tandem mass spectrometry; MRM, multiple reaction monitoring;
EPI, enhanced product ion; BBE, berberine bridge enzyme; MOI, multiplicity of
infection.