Sterol C24-methyltransferase: Physio- and stereo-chemical features of the sterol C3 group required for catalytic competence

Article abstract of DOI:10.1016/j.abb.2012.03.002

Sterol C24-methyltransferases (24-SMTs) catalyze the electrophilic alkylation of Δ²⁴-sterols to a variety of sterol side chain constructions, and the C3- moiety is the primary determinant for substrate binding by these enzymes. To determine what specific structural features of the C3-polar group ensure sterol catalysis, a series of structurally related C3-analogs of lanosterol that differed in stereochemistry, bulk and electronic properties were examined against the fungal 24-SMT from Paracoccidioides brasiliensis (Pb) which recognize lanosterol as the natural substrate. Analysis of the magnitude of sterol C24-methylation activity (based on the kinetic constants of V_max/K_m and product distributions determined by GC-MS) resulting from changes at the C3-position in which the 3β-OH was replaced by 3α-OH, 3β-acetyl, 3-oxo, 3-OMe, 3β-F, 3β-NH₂ (protonated species) or 3H group revealed that lanosterol and five substrate analogs were catalyzed and yielded identical side chain products whereas neither the 3H- or 3α-OH lanosterol derivatives were productively bound. Taken together, our results demonstrate a chemical complementarity involving hydrogen bonding formation of specific active site contacts to the nucleophilic C3-group of sterol is required for proper orientation of the substrate C-methyl intermediate in the activated complex.

Full text of DOI:10.1016/j.abb.2012.03.002

Archives of Biochemistry and Biophysics 521 (2012) 43–50
Contents lists available at SciVerse ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Sterol C24-methyltransferase: Physio- and stereo-chemical features of the sterol
C3 group required for catalytic competence q
Alicia L. Howard 1, Jialin Liu 1, Gamal A. Elmegeed 2, Emily K. Collins 3,4, Kalgi S. Ganatra 4,
Chizaram A. Nwogwugwu 4, W. David Nes
⇑
Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, TX 79409, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Sterol C24-methyltransferases (24-SMTs) catalyze the electrophilic alkylation of
D
24-sterols to a variety
Received 27 January 2012
and in revised form 24 February 2012
Available online 14 March 2012
of sterol side chain constructions, and the C3- moiety is the primary determinant for substrate binding by
these enzymes. To determine what specific structural features of the C3-polar group ensure sterol catal-
ysis, a series of structurally related C3-analogs of lanosterol that differed in stereochemistry, bulk and
electronic properties were examined against the fungal 24-SMT from Paracoccidioides brasiliensis (Pb)
which recognize lanosterol as the natural substrate. Analysis of the magnitude of sterol C24-methylation
activity (based on the kinetic constants of Vmax/Km and product distributions determined by GC–MS)
resulting from changes at the C3-position in which the 3b-OH was replaced by 3
3-oxo, 3-OMe, 3b-F, 3b-NH2 (protonated species) or 3H group revealed that lanosterol and five substrate
analogs were catalyzed and yielded identical side chain products whereas neither the 3H- or 3 -OH
lanosterol derivatives were productively bound. Taken together, our results demonstrate a chemical com-
plementarity involving hydrogen bonding formation of specific active site contacts to the nucleophilic
C3-group of sterol is required for proper orientation of the substrate C-methyl intermediate in the
activated complex.
Keywords:
Sterol C24-methyltransferase
Lanosterol
Fluorinated sterol
Sterol catalysis
Hydrogen bonding
Paracoccidioides brasiliensis
a-OH, 3b-acetyl,
a
Ó 2012 Elsevier Inc. All rights reserved.
Introduction
aspects of the role of disordered sterol homeostasis in sterol biosyn-
thesis in animals and plants [7–9].
So far as is known the isoprenoid biosynthesis pathway to 24-al-
kyl sterols is an ancient and ubiquitous property of eukaryotes with
the different sterol side constructions arising from the product
specificities of the sterol C24-methyltransferase enzymes
(24-SMT) [1–3]. An illustrative example of this core pathway for
ergosterol biosynthesis is provided by the concerted C24-methyla-
tion of lanosterol by the fungal Paracoccidioides brasiliensis PbSMT
Based on current concepts [1,11], a steric-electric plug model
for the coupled methylation–deprotonation of
D
24-sterols to
C24-methyl(ene) products has been proposed [10]. This C24-meth-
ylation reaction can be viewed as a nucleophilic attack by the
D
24-double bond of various sterols on the methyl group of the sul-
fonium group of S-adenosyl-L-methionine (SAM). A central feature
of this stereochemical model is the enzyme recognition of sterol
functional groups at either end of the molecule. Notably, a crucial
difference in the anatomy of these flexible substrates targets the
shape of the nucleus and the ring structure’s influence on the tilt
of sterol C3 hydroxyl group and 17(20)-bond [11,12] (Fig. 2).
Comparison of active site substrate requirements directed at the
phyla-specific product specificities for sterol biosynthesis enzymes
has been undertaken with limited success in part due to the ab-
sence of sterol-bound three-dimensional structures of these cata-
lysts. Therefore, a non-structural approach that compares active
sites of related enzymes on the basis of substrate specificities has
proven useful. This specificity for substrates allows 24-SMTs to be
studied for their discrimination between analogs of the natural sub-
strate that differ by a single functional group. Yet, despite several
structure–activity studies that focus on sterol ring A features
[11,13–24], still unclear are what factors in substrate binding gov-
ern catalytic orientation of the reactants in the activated complex.
yielding a single
D
24(28)-methylated sterol product, eburicol [4].
On the other hand, substrate-specific reaction channeling can occur
during the C24-methylation reaction catalyzed by plant and proto-
zoan 24-SMTs to form multiple carbocation intermediates en route
to the biosynthesis of diverse 24b-methyl and 24a-ethyl sterols
(Fig. 1) [1]. Recent metabolite profiling has uncovered complicated
sterol patterns in a range of organisms [1,2,5,6] and helped to define
q
This investigation was supported by a National Science Foundation Grant (MCB
0929212) to W. D. N.
⇑
Corresponding author. Fax: +1 806 742 1289.
E-mail address: wdavid.nes@ttu.edu (W. David Nes).
Authors contributed equally to the work.
Visiting Professor from the Hormones Department, Medical Research Division,
1
2
National Research Center, Dokki, Giza 12622, Egypt.
3
Recipient of Howard Hughes Medical Institute Undergraduate fellowship.
Recipient of National Science Foundation Research Experience for Undergraduate
4
fellowship.
0003-9861/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.abb.2012.03.002

44
A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
A
3
7
HO
4
+
Ergosterol
"CH3
SMT
"
+
H
5
1
2
HO
8
6
Sitosterol
28
B
24
D
C
9
17
B
A
3
HO
HO
HO
HO
10
11
24(28)-Methylene lophenol
Fig. 1. Pathways for ergosterol and sitosterol side chain construction based on the C24-methylation of a
12
Zymosterol
Lanosterol
Cycloartenol
D
24-sterol acceptor molecule 1 (Panel A) and structures of natural
substrates recognized by sterol C24-methyltransferase enzymes from fungi 9 and 12 and plants 10 and 11 (Panel B).
Fig. 2. Conformational perspectives of relevant sterols illustrating the flat structure and tilt of the C3– and C17(20)–bonds in the molecule. Panel A shows the X-ray structures
of lanosterol overlapped on cycloartenol. Panel B shows the X-ray structures of lanosterol overlapped on sitosterol (Adapted from reference [11]).
Previously, we proposed directionality to be a relevant factor of
interaction determined by the crucial non-reacting substrate group
at the proximal end of the acceptor molecule [11]. This direction
can be brought about by a stereospecific complexing of the sub-
strate C3–OH to distinct active site contacts. Correct polarity and
positioning between active site contacts and substrate would then
give favorable entropy of activation leading to efficiency in sterol
methylation. We now report the evaluation of a series of lanosterol
analogs which systematically deviate from the C3b–OH group
against the cloned PbSMT in order to gain further insight into the
nature of the enzyme–substrate complementarity utilized by 24-
SMTs. Collectively, these and our earlier results [4,11,13–17] dem-
onstrate the functional importance of the sterol ring A 3b-OH on
formation of the C24-methyl(ene) intermediate and resulting
product.
samples dissolved in CDCl3 at ambient temperature. Sterols were
analyzed by GC–MS using a Hewlett Packard GC 6890-MSD 5973
(70 eV EI, scan range 50–550 amu) equipped with 0.25 mm i.d.
by 30 m fused silica columns coated with Zebron ZB5
(Phenomenex) with He as carrier at a flow rate of 1.2 mL/min. Cool,
in column injection set for constant flow was used. The oven was
set to be isothermal the first minute, and then ramp to 280 °C at
20 °C/min. The source temperature was 230 °C. Purification by a
combination of flash chromatography on silica gel, Al2O3 and/or
silver nitrate impregnated Al2O3 and TLC on 0.25 m or 1 mm glass
plates and/or HPLC using Whatman or Phenomenex columns pro-
vided pure substrates as previously described [4,25]. Sterols were
identified by their rates of movement in chromatography, ex-
pressed as RF values (TLC) or retention times relative to cholesterol
(RRTc in GLC or ac in HPLC) and by comparison of their mass and
NMR spectra with authentic specimens available in our sterol
collection.
Materials and methods
Analytical methods
Chemicals
1H Spectra were recorded at room temperature on a Varian
Unity Inova 500 MHz spectrometer and chemical shifts reported
in parts per million (ppm) downfield from tetramethyl silane of
SAM as the iodide or chloride salt was purchased from Sigma.
[methyl-3H3]AdoMet, diluted with non-radioactive SAM to a spe-
cific activity of 10 lCi/lmol, was purchased from Perkin-Elmer

A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
45
(Waltham, MA). All other reagents and chemicals were purchased
from Fisher, Aldrich or Sigma unless otherwise noted, and used
without further purification.
HO
F
H
3
O
H
7
4
5
1
Synthetic procedures
+
H3N
H
H
H
HO
Lanosterol and eight C3-modified lanosterol analogs prepared
from lanosterol are shown in Fig. 3. The starting material for the
synthesis of the C3-modified lanosterol derivatives was lanosterol
purified from commercial lanosterol (Sigma) containing lanosterol,
24-dihydrolanosterol, agnosterol and 24-dihydroagnosterol as
described elsewhere [25]. For modification of the 3b-OH group, free
lanosterol was converted to the C3-acetate using acetic anhydride
in pyridine, to the C3–O methyl ether using methyl iodide and
sodium hydride in anhydrous THF or to the 3-oxo derivative using
pyridinium chlorochromate in chloroform by the standard proce-
dures [24]. 3b-Amino lanosterol was prepared from 3-oxo
lanosterol via the 3-oxime intermediate as previously described
[26]. The 3-desoxy lanosterol was prepared from the 3-hydrazine
intermediate in glycol/KOH under reflux followed by extraction
using a water–diethyl ether mixture affording product in 88% yield.
H
8
2
H3CCOO
H
H3CO
H
6
3
Fig. 3. Partial structure (only ring A) of lanosterol and modified lanosterol
derivatives at C3 incubated with PbSMT. In the case of 3-amino lanosterol 5, the
analog is protonated as shown at the buffer pH 8.
Table 1
Chromatographic and special properties for C3 modified analogs of lanosterol.
Substratea
Structureb
RRTc
Rf
MS (M+)e
1H NMRf
c
d
Although we prepared 3a-lanosterol in past studies from the
known route [27], the recovery has been low (0.5–1%). Therefore,
we developed a new procedure as follows: To a stirred solution
of ethylene glycol (20 mL) and KOH (200 mg) was added
3-oxolanosterol (50 mg) and the mixture refluxed for 5 h. After
cooling to room temperature, the sample was poured into iced
water and the precipitate filtered affording a 9:1 mixture of
LA-3b-OH
LA-3a-OH
LA-3b-OMe
LA-3-Oxo
LA-3b-NH2
LA-3b-OAc
LA-3b-F
1
2
3
4
5
6
7
8
1.33
1.30
1.18
1.25
1.25
1.49
1.05
0.93
0.70
0.88
0.97
0.94
0.03
0.98
1.00
1.00
426
426
440
424
425
468
428
408
3.28 dd, 1H
3.43 d, 1H
3.36 m, 1H
ND
2.67 m, 1H
4.12 m, 1H
4.15 m, 1H
ND
3b- to 3a-lanosterol together with significant by-products. 3-Epil-
LA-3-H
anosterol, generated in overall 5–10% yield, was easily separated
from lanosterol and contaminants by a combination of TLC (devel-
oped in benzene/diethyl ether 85/15 (v/v)) and HPLC (reversed
phase C18 Phenomenex semi-preparative column eluted with
methanol at 2.5 m/min) (Table 1).
a
b
c
LA, Lanosterol frame.
Structures are shown in Fig. 3.
Retention time of sterol relative to the retention time of cholesterol in GC; ND,
not determined.
d
Retention factor (Rf) of sterol on silica gel TLC plate developed ꢀ2 with
benzene/ether (85:15, v/v).
In the general procedure for the fluorination of alcohols using
DAST5 [28], 50 mg of 3b-OH–lanosterol was added under nitrogen
to a stirred solution of 50 mL dry heptane and dichloromethane
(5 ml) to facilitate sterol to dissolve in solution, followed by
addition of DAST (1 drop) at room temperature. The reaction mix-
ture was stirred for 10 min and water (1 ml) added carefully to
quench unreacted DAST, then the mixture washed again with
water (100 ml ꢀ 2) and dried over magnesium sulfide. The solvent
e
Molecular weight determined by electron-impact mass spectroscopy.
f
1H NMR of C3–H signal; m, multiplet; d, doublet.
PbSMT assay
Procedures for the heterologous overexpression in Escherichia
coli and isolation of the recombinant PbSMT have been described
[4]. The standard linear range assay procedure for PbSMT involved
incubation in triplicate in 10 mL test tubes containing 20 mM
phosphate buffer (pH 8.0), 5% glycerol, lysate protein (1–2 mg),
sterol (varied from 5 to 150
35 °C for 1 h in total volume of 600
was terminated by brief vortexing and the addition of a solution
(600 L) of 10% methanolic KOH. The C24-methylated sterol prod-
ucts were extracted in hexane and dried; an aliquot of the hexane
extract was analyzed by GC–MS or in the case of radioactive sam-
ples analyzed by scintillation counting to determine the conversion
rate. Control experiments (without sterol) were conducted with
each enzyme preparation to determine solvolytic background,
which generally afforded less than 500 dpm in experiments using
[methyl-3H3]SAM; for optimal sterol methylation the amount of
radioactivity recovered from the organic extract was approxi-
mately 1 ꢀ 106 dpm per assay. Protein concentration was mea-
sured by the Bradford assay [29]. The initial velocity data was
determined using SigmaPlot 2001 with the enzyme kinetics mod-
was removed in vacuo. The product was predominantly 3a-fluorol-
anosterol in poor yield (>1%). MS: 428.3 (M+), 413.3 (M+ –Me),
408.4(M+ –HF), 393.3 (M+ –Me–HF), 365.3329.0, 301.2, 283.1,
261.1, 69.1 (100%). 1H NMR: 3.49 (3b-H, m, 1H). Apparently, syn-
thesis of the desired 3b-fluoro derivative is not readily achieveable
using the convententional procedure. Thus, we developed the fol-
lowing preparation of 3b-fluoro lanosterol 8 from lanosterol 1 in
good yield outlined in Fig. 4. After the mono protection of 1 as
the THP ether, 2 was oxidized to 3 followed by reduction using
sodium hydride, then acetylation affording the C24-acetyl inter-
mediate 4. Preparations of 4 routinely gave a 1:1 mixture of C24-
diastereoisomers in yields of 97–98% over the four steps. Oxidation
of 4 with PCC led to 5. This compound was fluorinated with DAST
l
M) and SAM (fixed at 100
lM) at
l
L. The incubation mixture
l
and the resulting
solution using Adam’s catalyst and hydrogen gas to provide a mix-
ture of D23 24-3b-fluorinated lanosterol (<99% b-isomer). Final
purification of 3b fluoro-
24(25)-lanosterol was achieved by HPLC
D
2-olefin stereospecifically reduced in acidic
/D
D
on TSK gel eluted with methanol. Relevant chromatographic and
spectral properties of the lanosterol analogs are reported in Table
1.
ule software package. Data were fitted to the equation
v = VmaxS/
Km + S using a non-linear least-squares approach. Kinetic con-
stants standard errors estimated from the steady-state kinetic
evaluation for each sterol were never greater than 10% of the
experimental measurement, and R2 values were between 0.90
5
Abbreviations used: DAST, diethylaminosulfur trifluoride; 24-SMT, sterol C24-
methyltransferase ; Pb, Paracoccidiodes brasilienis; SAM, S-adenosyl-L-methionine;
GC-MS, Gas chromatography–mass spectroscopy.

46
A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
O
DHP
CH2Cl2
1. BH3
.
THF
1. NaBH4
ethanol
THF
TsOH (cat.)
100%
2. Ac2 O
2. CH2Cl2
PCC
99%
Pyridine, r.t.
98 %
HO
THPO
THPO
2
5
8
1
3
OAc
OAc
OAc
1. DAST
70-80 oC
1. TsOH
MeOH
1. H2
PtO2
2. PCC
2. POCl3
2. KOH
DMSO, 170 o
C
CH2Cl2
Pyridine
98 - 100%
40%
70 - 76%
THPO
O
F
4
6
Δ23
Δ24
HPLC
F
F
7
Fig. 4. Outline for the synthesis of 3b-fluorolanosterol 8 from lanosterol 1 as described in ‘‘Materials and methods’’.
and 0.99. Kmapp and Vmaxapp values were initially determined by
varying the substrate concentration in the presence of fixed cofac-
tor at 100 lM. To test for non-productive binding of ‘‘inactive sub-
strates’’, in the one case where sufficient analog was available for
Preliminary evaluation of the PbSMT was to determine whether
the C3–OH group is necessary in catalysis. Thus, we tested analogs
in which the molecules’ polarity at C3 was systematically de-
creased from that of C3 in lanosterol. Based on chemical reason-
ableness, polarity ranking of functional groups associated with
biomolecules should follow the order: alcohol > ketone > amine >
ester > ether > alkane or hydrogen atom. In agreement with this
ranking, PbSMT recognition of the analogs ranked in order: C3-hy-
droxyl (100%) > C3-oxo (47%) > C3-amino (21%) > C3-hydrogen
(0%) (Table 2). The binding affinity of the acceptable analogs is
incubation with PbSMT, increasing concentrations of 3-desoxyl-
anosterol was varied (5–150
lanosterol at 25 and 100 M of substrate and fixed concentrations
M) and the kinetics evaluated as described previ-
lM) against two concentrations of
l
of SAM (150
ously [15].
l
comparable to that of the natural substrate (Kmapp = 21 lM). Alter-
Results and discussion
natively, catalytic competence for these analogs differed kinetically
in Vmaxapp showing that not only the presence of the oxygen in the
3-position but also its polarity affects substrate transformation. In
the case of the 3-amino analog, at pH 8 the nitrogen group is fully
protonated, and therefore bears an acidic proton which can H-do-
nor interact in the active site in similar fashion to the interactions
of the C3-hydroxyl to contact residues. However, a surprising re-
sult in the present study is the observation that the C3-methyl
ether derivative (18%), which like the C3-acetyl derivative (15%)
are less reactive substrates, did not fall in the rankings below the
3-amino lanosterol activity to zero as might by expected [11,13].
Catalytic competence of PbSMT in the presence of C3-modified
lanosterol analogs
In attempt to dissect the retention mechanism for substrate
bound to sterol catalysts, several strategies were devised for
anchoring sterol in the active site. A limitation in many of these
earlier studies was a chemical correlation between the substrates,
predicted intermediate of the reaction and active site contacts due
to few analogs tested. In our ongoing research we have evaluated
several C3-variants in different systems and observed that sterol
A-ring has several non-reacting features that can influence enzy-
matic activity [11,17,18]. As a first approximation, the natural sub-
strates recognized by sterol biosynthesis enzymes can possess: (i)
slightly different nuclear conformations affected by the number
and position of double bonds in the molecule, (ii) varied
C4-substitution and/or (iii) a distinct C3-polarity that is associated
with either a 3b-OH or 3-oxo group. Currently, there are few re-
ports that examine the accessibility of the 3b-oxygen in binding.
Thus, to shed further light on the catalytic role of the polar group
at C3, the nature and magnitude of C24-methylation of a series
of synthetic lanosterol derivatives that differed in C3-electronics,
bulk or stereochemistry were examined kinetically with PbSMT
and the specificity constants, Vmax/Km, of the test substrates com-
pared to the catalytic competence of lanosterol normalized to
100%.
Table 2
Catalytic competence of PbSMT tested with substrate analogs.
Substrate
Structurea
Vmax/Km
Competence%b
LA-3b-OH
LA-3a-OH
LA-3b-OMe
LA-3-Oxo
LA-3b-NH2
LA-3b-OAc
LA-3b-F
1
2
3
4
5
6
7
8
53/27 = 1.96
0/0 = 0
100
0
9/25 = 0.36
22/23 = 0.95
13/30 = 0.43
16/28 = 0.57
8/25 = 0.32
0/0 = 0
18
48
22
29
16
0
LA-3-H
a
Structures are shown in Fig. 3.
Vmax(pmol/min/mg protein)/km(lM) ratio normalized to Lanosterol, LA, as
b
100%.

A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
47
These analogs, together with the C3-oxo derivative, have contrast-
ing polarity to the protonated amine and possess a lone pair of
electrons on the oxygen which are available for nucleophilic inter-
actions such that the oxygen group may H-atom accept in H-bond-
ing interactions.
It is thus apparent for PbSMT that the hydrogen atom from the
3-hydroxyl group of lanosterol may not engage in hydrogen bond-
ing solely as a H-donor leading to fixation to the enzyme.
The elution of eburicol relative to lanosterol in GC moves with a
retention factor of 1.07 corresponding to the change in side chain
structure from
D D
24(25)-substrate to 24(28)-product [13]. This reten-
tion factor in product formation was observed routinely in sub-
strate transformation typified in the incubations of lanosterol
and 3b-fluorolanosterol with PbSMT (Fig. 5). The mass spectra of
enzyme-generated products are shown in Fig. 6, except for the
3-amino lanosterol product which was anomalous (M+ 465 peak).
Activity assay of 3-acetyl lanosterol led to 3-acetyl 24(28)-methy-
lene lanosterol (M+ 482). The acetylated methylene sterol was a
minor product in GC analysis and detected only in cases of incom-
plete saponification of the enzyme preparation. Consequently,
lanosterol and eburicol were the main sterols in GC–MS analyses
of 3-acetyl lanosterol transformation. Incubations of 3-desoxy
Anchoring mechanism for sterol-PbSMT interactions
As posed above, although of the analogs incubated with PbSMT
thus far reveals that the 3b-OH group of lanosterol to be the
optimal feature for recognition by PbSMT, several of them were
unexpectedly effective substrates, including those that contained
C3-structures represented by the 3-amino and C3-acetate deriva-
tives never recognized by a sterol catalyst. Analogs that showed
that the C3-oxygen can act as a H-acceptor in H-bonding interac-
tions leads to the possibility that some active site contact other
than histidine observed in homology modeling of 24-SMT [31] is
involved in anchoring sterol. For this reason, we prepared the hal-
ogenated substrate 3b-flurolanosterol. This additional substrate
was chosen for two reasons: it was not expected to perturb the
conformation of the enzyme’s active site since the steric size of a
fluorine and hydroxyl group are similar, and the isosteric replace-
ment of the sterol hydroxyl group with a fluorine atom can hydro-
gen bond accept in a manner similar to the oxygen of a hydroxyl
group and never function as a hydrogen-bond donor [31]. Stea-
dy-state kinetic measurements of lanosterol compared to its 3-flu-
oro analog shows that PbSMT maintains the same binding
efficiencies (Km) for both substrates (Supplementary Fig. S1), but
3b-fluoro lanosterol is transformed much less (16%) efficiently
than 3b-OH lanosterol (100%) (Table 2). These data suggest that
burying of the fluorine atom in the hydrophobic pocket, normally
occupied by a hydroxyl group, angles the C3-F to an acidic hydro-
gen (active site donor) to provide the optimal situation for F–H
bonding. Thus, the 3b-OH group in lanosterol might engage in
hydrogen bonding with active site contacts as a H-bond acceptor,
which agrees and contrasts with other sterol structure–activity
investigations of 24-SMT [11,13]. The experimental data for pro-
ductive substrates, 3b-OH and 3b-F, as well as for 3b-NHþ3 , 3b-
OMe, 3-oxo, and 3b-OAc, showing similar Km values but differences
in Vmax further suggests that substrate specificity is determined
mainly by the transition state structure, rather than by that of
the ground state structure.
(3-H) or 3a-lanosterol sterol derivatives with the recombinant en-
zyme did not lead to any hexane extractable products as judged by
GC–MS analysis, even after prolonged incubation or increasing the
protein to 5 mg per assay tube or increasing cofactor concentra-
tions to 200 lM in the incubation system. Steady state kinetic
analysis of the inactive substrate 3-desoxy lanosterol against
lanosterol showed competitive-type inhibition (Ki3-desoxylanosterol
approximately 67 lM (Supplementary Fig. S2), consistent with
non-productive binding of the analog in the active site. In none
of the GC–MS analyses of the enzyme-generated 24(28)-methylene
sterol was a byproduct detected or hydroxylated sterol intermedi-
ate formed as reported for incubation of suicide substrates with
24-SMTs [1], showing the analogs are stable and that the reaction
progress proceeded to completion. These data suggest that when
the chemical reactivities of the C3 group were altered, the strict re-
gio- and stereo-specificities of the C-24 methylation reaction were
Product characterization
Typically, the methyl–methylene eliminations that form the
exocyclic double bond of the enzyme-bound acceptor molecule
takes place at C28 in the sterol side chain. However, catalysis of
the
protozoan origin can lead to channeling to generate alternate
C24 methyl(ene) products possessing the
23(24)-, 24(25)-,
24(28)- and 25(27)-olefin structure. The different C24-methylation
D
24-sterol substrate from several 24-SMTs of plant, fungal or
D
D
D
D
reaction products are readily distinguished by their chromato-
graphic behaviors in GC [8,21]. Thus, it seemed prudent to consider
whether any of the test substrates can generate products other
than a
D
24(28)-sterol. To assure that our enzyme preparation was
generating the same high yields of 24(28)-methylenelanosterol
(eburicol) from lanosterol as reported previously [4], sterol
C24-methylation of the various lanosterol derivatives was initially
evaluated using lanosterol (100 lM); in agreement with our earlier
findings we detected eburicol as the sole product by GC–MS
Fig. 5. GC analysis (reported as a partial total ion current chromatogram) of the
C24-methylated sterol products, peak 2, generated from PbSMT incubated with
lanosterol peak 1 (Panel A) or 3b-fluorolanosterol peak 1 (Panel B). The retention
time in GC for cholesterol is approximately 13.4 min.
analysis in approximate 40% yield (Fig. 5, Panel A).

48
A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
Fig. 6. Mass spectra (EI, 70 eV) of the C24-methylated products generated by PbSMT incubated with lanosterol (Panel A), 3b-fluorolanosterol (Panel B), 3-methyl ether
lanosterol (Panel C) or 3-oxo lanosterol (Panel D).
still retained consistent with correct anchoring of the test acceptor
molecule in the activated complex. In this model, a non-productive
substrate lacks the necessary structural feature at C3 to induce the
conformational change required for catalytic activity, and thus
does not undergo a reaction.
These new structural findings together with prior structure–
activity studies of the cycloeucalenol-obtusifoliol isomerase [24]
are particularly noteworthy because in both cases bonding to ste-
rol C3–OH is proposed to proceed from the main chain interac-
tions. Since the putative contact 24-SMT histidine residue that
interacts with the sterol C3-hydroxyl group can be replaced with
leucine in the ScSMT (=Erg6p) without abolishing activity [40]
and C3-methyl ether and C3-oxo derivatives are suitable substrates
for ScSMT, it is reasonable to extrapolate to the PbSMT that these
substrate–enzyme interactions are likely more wired than previ-
ously considered. Thus, the anchoring mechanism involved with
sterol C24-methylation in PbSMT catalysis may consist of a topo-
logically constrained side carbonyl and a polar amino acid side
chain which can surround the substrate C3–OH and firmly attach
it in the active site cleft by H-bonds directly or by way of a water
bridge. In this tightly coupled H-bonding ternary complex, the
binding and reaction of the intermediate generated at the active
site are greatly favored over dissociation. This new model, which
envisages the 24-SMT interacts with the sterol C3-hydroxyl group
from two directions (Fig. 7), enables the dynamic conformational
changes that accompany C-24-methylation of acceptor molecule
to play a role in chaperoning product outcome. Since structural dif-
ferences between substrates undergoing C24-methylation are
small and local, the difference in specificity amongst this family
of enzymes is likely caused by only a few amino acid differences
in the substrate-recognition site and/or in subtle differences in
the main frame organization relative to the sterol ring A functional
groups. Indeed, a conserved aromatic residue-phenylalanine, has
General considerations
The results of this investigation have established the sterol
C3–OH group and its 3b-stereochemistry are absolute substrate
requirements to be catalytically functional in PbSMT catalysis.
Although this binding feature is not predictable from the compar-
ative sequence alignments or reaction mechanisms of sterol bio-
synthesis enzymes that form the core pathway, typically these
non-homologous enzymes show narrow substrate selectivity for
the hydrophilic part of the sterol molecule. Several reports indicate
that one or more active site amino acids can interact to form a
hydrogen bond at the 3b-hydroxyl group, including serine, methi-
onine, threonine or aspartate in sterol 14-demethylase [32–35],
histidine in sterol C24-methyltransferase (24-SMT) [31], serine in
sterol C24-reductase (24-SR) [36], threonine in sterol
8,7-isomerase (8-SI) [37] and tyrosine or serine in 3b-hydroster-
oid-dehydrogenase/C4-decarboxylase [3bHSD/4-SD] [38], as sug-
gested by bioinformatic analyses, homology modeling studies
and/or site-directed mutagenesis.
Homology modeling of 24-SMT has shown the sterol aligns in
the active site of 24-SMTs with the A-ring pointed toward a basic
residue, neutral histidine [30]. Thus, one might conclude that the
sterol OH– group functions as the H donor in a H-bond to the lone
pair of electrons in the nitrogen atom which corresponds to the
active site amino acid in the steric-electric plug model [10]. In con-
trast, from the recently obtained X-ray structure of a suicide sub-
been shown to act as a key discriminator between C4
and C4 /b-dimethylated substrates of the sterol 14-demethylase
enzymes [33,41].
a-methylated
a
A key feature in sterol recognition, and therefore specificity, to
distinguish a plant 24-SMT from a fungal 24-SMT is the degree of
C4-substituion in the acceptor molecule. The contribution of one
strate
(14a-methylencyclopropyl-
D
7-24,25-dihydrolanosterol)
complexed to a sterol 14
a
-demethylase from Trypanosoma brucei
or two methyl groups at C4 and associated
D D
7 or 8-bond in rings
which shows a thermodynamically favorable conformation and
orientation that clearly reflects the catalytically relevant binding
mode to the heme, do any of these amino acids interact directly
with the sterol C3-hydroxyl group; rather, the C3b–OH group is
positioned close to the channel entrance of the active site, 3.5A
from a backbone oxygen associated with Met358 residue [39].
B or C for substrate binding can have a positive or negative effective
on 24-SMT catalysis. On the other hand, sterols that contain a mono-
ene
D
5-bond system and lack C4 methyl groups, important features
for sterols to function as membrane components [42–44], are gener-
ally poor substrates for 24-SMTs. It seems to be reasonable to

A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
49
Fig. 7. Proposed aspects of substrate (A) intermediate (B) and product (C) binding to the PbSMT active site. For lanosterol binding in A, the enzyme presents a binding site that
is sterically and electronically complementary, to which the substrate becomes anchored at its C3-hydrophilic group. The sterol C3–OH group interacts in a pre-organized
active site with contacts that form hydrogen bonds against the 3-oxygen atom (from a main frame moiety, M) and hydrogen atom of the 3-oxygen atom (from a basic amino
acid, B1) forming a hydrogen bonded network to stabilize the ground state structure at the proximal end of the acceptor molecule and the side chain assumes a pseudocyclic
conformation. Productive orientation of the substrate side chain affords backside (SN2) addition of ‘‘methyl cation’’ from S-adenosyl-
catalytic sulfur atom, S) to the
24-bond generating the C24b-methyl C25 cation shown in B. Deprotonation of the C28 methyl group from a basic amino acid (B2) can lead to
the C24(28)-methylene product shown in C, followed by disassociation of the methylated sterol from the enzyme (cf. steric-electric model discussed in references [10,11]).
L-methionine (represented by the
D
suppose that for 24-SMTs precise positional control of the substrate
C3-group requires a
8- over 5-substrate for effective catalysis and
that the homoallylic effect from the sterol
5-bond on the C–O bond
at C3 can degrade this control, either by holding the C3–OH group
fixed by donation of electrons through an inductive effect which
should strengthen the H–O bond and therefore weaken H-bonding
of the sterol to the active site contacts, or by allowing multiple alter-
native orientations which are non-productive.
References
D
D
[1] W.D. Nes, Chem. Rev. 111 (2011) 6423–6451.
[2] L.J. Goad, T. Akihisa, Analysis of Sterols, Blackie Academic & Professional, New
York, 1997. pp. 1–40.
[3] R.B. Kolner, R.E. Summons, A. Pearson, N. King, A.H. Knoll, Proc. Natl. Acad. Sci.
U.S.A. 105 (2008) 9897–9902.
[4] M. Pereira, Z. Song, L.K. Santos-Silva, M.H. Richards, T.T.M. Nguyen, J. Liu,
C.M.A. de Almeida Soares, A.H. da Silva Cruz, K. Ganapathy, W.D. Nes, Biochim.
Biophys. Acta 1801 (2010) 1163–1174.
D
[5] W. Zhou, G.A.M. Cross, W.D. Nes, J. Lipid Res. 48 (2007) 665–673.
[6] K. Schrick, C. Cordova, G. Li, L. Murray, S. Fujioka, Phytochemistry 72 (2011)
465–475.
[7] J.X. He, S. Fujioka, T.C. Li, S.G. Kang, H. Seto, S. Takatsuto, S. Yoshida, J.C. Jang,
Plant Physiol. 131 (2003) 1258–1269.
[8] F.D. Porter, G.E. Herman, J. Lipid Res. 52 (2011) 6–34.
[9] P. Benveniste, Annu. Rev. Plant Biol. 55 (2004) 429–457.
[10] W.D. Nes, Biochim. Biophys. Acta 1529 (2000) 63–88.
[11] W.D. Nes, G.G. Janssen, A. Bergenstrahle, J. Biol. Chem. 266 (1991) 15202–
15212.
It is notable that the active site cavity of PbSMT is sufficiently
tight that changes to the active site volume resulting from incuba-
tion of the different analogs failed to generate multiple products
which have been detected in related 24-SMT enzymes catalyzing
the first C1-transfer reaction [1]. Tight conformational control by
the enzyme also implies proximity of the cationic center with
the
D
24-bond of the substrate undergoing C24-methylation, a
charge-stabilizing mechanism. The weaker binding of the 3b-
fluorinated analog did not alter the product outcome, consistent
with a directional component to sterol catalysis independent of
the mechanistic type. Thus, we suggest efficient anchoring of the
C3-hydroxyl group which generates the productive C24-methyla-
tion intermediate in the PbSMT active site affords an element of ki-
netic control over competition between ‘‘slow’’ deprotonation and
‘‘fast’’ methylation affecting product specificity. Although a more
precise formulation of ternary enzyme–sterol–SAM interactions
must await the three-dimensional structure of the 24-SMT, the
present studies in conjunction with previous work [11,13,15], pro-
vide a foundation for substrate binding selectivity amongst this
class of enzyme. Additionally, evaluation of the results, and those
relevant to substrate binding by related catalysts [17–24], allows
effective rational design of mechanism-based inactivators targeted
to inhibit acute enzymes of sterol biosynthesis synthesized in
opportunistic pathogens, efforts currently in progress.
[12] W.D. Nes, K. Koike, Z. Jia, Y. Sakamoto, T. Satou, T. Nikaido, J.F. Griffin, J. Am.
Chem. Soc. 120 (1998) 5970–5980.
[13] M. Venkatramesh, D. Guo, Z. Jia, W.D. Nes, Biochim. Biophys. Acta 1299 (1996)
313–324.
[14] A.T. Mangla, W.D. Nes, Med. Chem. 8 (2000) 925–936.
[15] W.D. Nes, A. Song, A.L. Dennis, W. Zhou, J. Nam, M.B. Miller, J. Biol. Chem. 278
(2003) 34505–34516.
[16] W. Zhou, G.I. Lepesheva, J. Biol. Chem. 281 (2006) 6290–6296.
[17] W.D. Nes, W. Zhou, A.L. Dennis, H. Li, Z. Jia, R.A. Keith, T.M. Piser, S.T. Furlong,
Biochem. J. 387 (2002) 587–599.
[18] A. Bellamine, A.T. Mangla, A.L. Dennis, W.D. Nes, M.R. Waterman, J. Lipid Res.
42 (2001) 128–136.
[19] M. Taton, A. Rahier, Biochem. J. 277 (1991) 483–492.
[20] Y. Aoyama, Y. Yoshida, Y. Sonoda, Y. Sato, Biochim. Biophys. Acta 1006 (1989)
209–213.
[21] Y. Aoyama, Y. Yoshida, Biochem. Biophys. Res. Commun. 183 (1992) 1266–
1272.
[22] G.I. Lepesheva, N.G. Zaitseva, W.D. Nes, W. Zhou, M. Arase, J. Liu, G.C. Hill, M.R.
Waterman, J. Biol. Chem. 281 (2006) 3577–3585.
[23] S. Darnet, A. Rahier, Biochim. Biophys. Acta 1633 (2003) 106–117.
[24] A. Rahier, M. Taton, P. Benveniste, Eur. J. Biochem. 181 (1989) 615–626.
[25] S. Xu, R.A. Norton, F.G. Crumley, W.D. Nes, J. Chromatogr. 452 (1988) 377–398.
[26] D. Guo, J. Hu, J-H. Zheng, S.A. Ross, W.D. Nes, J. Chin. Pharm. Sci. 6 (1997) 133–
137.
[27] J.W. Huffman, J. Org. Chem. 24 (1959) 447–451.
[28] W.J. Middleton, J. Org. Chem. 40 (1975) 574–578.
[29] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[30] B.K. Park, N.R. Kitteringham, P.M. O’Neill, Annu. Rev. Pharmacol. Toxicol. 41
(2001) 443–470.
Acknowledgments
This research was supported in part by grants from the National
Science Foundation (MCB-0929212). Partial support from the
National Science Foundation (REU program) and Howard Hughes
Medical Institute (award to Texas Tech University) to the under-
graduates E.K.C., K.S.G. and C.A.N. is greatly appreciated.
[31] J. Liu, Biochem. J. 439 (2011) 413–422.
[32] D.C. Lamb, D.E. Kelly, J. Biol. Chem. 272 (1997) 5682–5688.
[33] L.M. Podust, L.V. Yermalitskaya, G.I. Lepesheva, V.N. Podust, E.A. Dalmasso,
M.R. Waterman, Structure 12 (2004) 1937–1945.
[34] C. Sheng, Z. Miao, H. Ji, J. Yao, W. Wang, Z. Che, G. Dong, J. Lu, W. Guo, W.
Zhang, Antimicrob. Agents Chemother. 53 (2009) 3487–3495.
[35] N. Strushkevich, S.A. Usanov, J. Mol. Biol. 397 (2010) 1067–1078.
[36] A. Pedretti, E. Bocci, R. Maggi, G. Vistoli, Steroids 73 (2008) 708–719.
[37] A. Rahier, S. Pierre, G. Riveill, F. Karst, Biochem. J. 414 (2008) 247–259.
[38] A. Rahier, M. Bergdoll, G. Genot, F. Bouvier, B. Camara, Plant Physiol. 149
(2009) 1872–1886.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.abb.2012.03.002.

50
A.L. Howard et al. / Archives of Biochemistry and Biophysics 521 (2012) 43–50
[39] T.Y. Hargrove, Z. Wawrzak, J. Liu, M.R. Waterman, W.D. Nes, G.I. Lepesheva, J.
Lipid Res. 53 (2012) 311–320.
[42] W.D. Nes, G.G. Janssen, F.G. Crumley, M. Kalinowska, M. Akihisha, T. Akihisha,
Arch. Biochem. Biophys. 300 (1993) 724–733.
[40] W.D. Nes, P. Jayasimha, W. Zhou, K. Ragu, C. Jin, T.T. Jaradat, R.W. Shaw, J.M.
Bujinicki, Biochemistry 43 (2004) 569–576.
[43] K.E. Bloch, Crit. Rev. Biochem. Mol. Biol. 14 (1983) 47–92.
[44] R. Bitmann, Subcell. Biochem. 28 (1997) 145–171.
[41] G.I. Lepesheva, W.D. Nes, W. Zhou, G.C. Hill, M.R. Waterman, Biochemistry 43
(2004) 10789–10799.