A fluorescent peptide substrate facilitates investigation of ghrelin recognition and acylation by ghrelin O-acyltransferase

Article abstract of DOI:10.1016/j.ab.2013.02.013

Ghrelin is a peptide hormone involved in regulation of appetite, glucose homeostasis, and a range of other physiological processes. Ghrelin requires a unique posttranslational modification, octanoylation of a serine side chain, to bind its cognate recepto

Full text of DOI:10.1016/j.ab.2013.02.013

Analytical Biochemistry 437 (2013) 68–76
Contents lists available at SciVerse ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
A fluorescent peptide substrate facilitates investigation of ghrelin recognition
and acylation by ghrelin O-acyltransferase
⇑
Joseph E. Darling, Edward P. Prybolsky, Michelle Sieburg, James L. Hougland
Department of Chemistry, Syracuse University, Syracuse, NY 13244, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Ghrelin is a peptide hormone involved in regulation of appetite, glucose homeostasis, and a range of other
physiological processes. Ghrelin requires a unique posttranslational modification, octanoylation of a ser-
ine side chain, to bind its cognate receptor to activate signaling. The enzyme that catalyzes this modifi-
cation, ghrelin O-acyltransferase (GOAT), is receiving increased interest as a potential drug target for the
treatment of obesity, diabetes, and other diseases proposed to be linked to ghrelin signaling. In this study,
we report the development of a novel fluorescence-based assay for GOAT activity and the use of this assay
to investigate GOAT inhibition and interactions underlying human GOAT (hGOAT) substrate selectivity.
Using a series of mutations and chemical modifications of our fluorescent peptide substrate, we have
identified specific groups on the first two amino acids of ghrelin that potentially contribute to ghrelin rec-
ognition by hGOAT. These data provide the first molecular-level information regarding interactions
within the ghrelin–hGOAT complex. Defining the interactions used by hGOAT to bind and recognize ghre-
lin will provide insight into the structure of the hGOAT active site, aid in the design and optimization of
targeted hGOAT inhibitors, and help to assess the possibility of novel hGOAT substrates beyond ghrelin.
Ó 2013 Elsevier Inc. All rights reserved.
Received 26 December 2012
Received in revised form 12 February 2013
Accepted 16 February 2013
Available online 27 February 2013
Keywords:
Ghrelin
Ghrelin O-acyltransferase
Posttranslational modification
Fluorescent assay
Ghrelin is a 28-amino acid secreted peptide, first discovered in
1999 by Kojima and coworkers, that serves multiple functions in
humans [1,2]. Ghrelin was first recognized to stimulate appetite
[2], but subsequent work also suggests a wide array of additional
physiological roles for ghrelin (for examples and reviews, see Refs.
[3–12]), among which are neurological processes such as learning
and memory, depression, and potentially neuroprotective roles
affecting Parkinson’s and Alzheimer’s diseases [4,13–15]. Ghrelin
signaling has also been linked to control of glucose metabolism
through insulin secretion and sensitivity [16–20].
Ghrelin-dependent pathways present attractive targets for drug
development because ghrelin requires multiple covalent modifica-
tions for biological activity [2]. Following expression as a 117-ami-
no acid precursor named preproghrelin, ghrelin undergoes several
maturation steps prior to secretion into the bloodstream [21,22].
One key maturation step involves a unique posttranslational mod-
ification of the third serine from the N terminus of the ghrelin pre-
cursor des-acyl proghrelin, wherein this serine is acylated by an
octanoyl (C8) fatty acid group (Fig. 1). Although both ghrelin and
des-acyl ghrelin are present in blood serum, only the acylated form
of ghrelin (hereafter referred to simply as ghrelin) can bind its cog-
nate receptor GHSR-1a to activate ghrelin-induced signaling [2].
The unique and essential nature of ghrelin octanoylation makes
blocking this modification an ideal target for inhibiting the biolog-
ical activity of ghrelin.
The enzyme that catalyzes ghrelin acylation, ghrelin O-acyl-
transferase (GOAT),1 was identified independently by two research
groups in 2008 [23,24]. It is the fourth member of a family of mem-
brane-bound O-acyltransferase (MBOAT) enzymes that are proposed
to catalyze acylation of hydroxyl groups in small molecules and a
number of protein substrates [25]. Interestingly, recent studies of
the MBOAT family member hedgehog acyltransferase (Hhat) suggest
that these enzymes may catalyze a broader range of acyl transfer
reactions [26,27]. Ghrelin is the only known substrate of GOAT, mak-
ing GOAT a specific target for inhibiting ghrelin activity. GOAT is pre-
dicted to contain either seven or eight transmembrane helices by
hydropathy plots and computational prediction [23,24,28], with
these predictions and the overall structure of GOAT remaining to
be examined experimentally [29]. Two residues in the C-terminal
1
Abbreviations used: GOAT, ghrelin O-acyltransferase; MBOAT, membrane-bound
O-acyltransferase; Hhat, hedgehog acyltransferase; hGOAT, human ghrelin O-acyl-
transferase; Tris, 2-amino-2-hydroxymethyl-propane-1,3-diol; CoA, coenzyme A; UV,
ultraviolet; HPLC, high-pressure liquid chromatography; MALDI–TOF, matrix-assisted
laser desorption/ionization time-of-flight; DMSO, dimethyl sulfoxide; Heppso, 4-(2-
hydroxyethyl)piperazine-1-(2-hydroxypropanesulfonic acid); TFA, trifluoroacetic
acid; Hepes, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; AcDan, acrylodan;
G1, glycine at the N terminus of the ghrelin-derived GSSFLC peptide; S2, serine at the
second position of the ghrelin-derived GSSFLC peptide; L5, leucine at the fifth position
of ghrelin; F4, phenylalanine at the fourth position of ghrelin.
⇑
Corresponding author. Fax: +1 315 443 4070.
E-mail address: hougland@syr.edu (J.L. Hougland).
0003-2697/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ab.2013.02.013

Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
69
manufacturer’s data sheet). Peptide substrates were either synthe-
sized by Sigma–Genosys (The Woodlands, TX, USA) in the Pep-
screen format or synthesized in-house with amidated C termini,
with purity verified by reverse phase high-pressure liquid chroma-
tography (HPLC) and peptide mass confirmed by matrix-assisted
laser desorption/ionization time-of-flight (MALDI–TOF) mass spec-
trometry (see Fig. S1 in supplementary material). Peptides were
solubilized in absolute ethanol and stored at ꢀ80 °C. Peptide
concentrations were determined spectrophotometrically at
412 nm by reaction of the cysteine thiol with 5,50-dithiobis(2-
nitrobenzoic acid) using an extinction coefficient of 14,150 Mꢀ1
cmꢀ1[37]. Octanoyl-[Dap3]-ghrelin (1–5) (Peptides International)
was solubilized to 1 mM final concentration in dimethyl sulfoxide
(DMSO).
Fig.1. Ghrelin octanoylation by ghrelin O-acyltransferase (GOAT). Using octanoyl-
CoA (coenzyme A) as an acyl donor, GOAT catalyzes the acylation of the serine 3
side chain hydroxyl within proghrelin during ghrelin maturation.
half of GOAT (Asn307 and His338) are proposed to be involved in
catalysis based on conservation across MBOAT family members
and loss of enzyme function on mutation [25,30], but little functional
information is currently available regarding the GOAT catalytic
mechanism and active site architecture.
Acrylodan labeling of peptide substrates
Peptide substrates were labeled with an acrylodan fluorophore
on a cysteine side chain thiol using an adaptation of a previously
reported protocol. Peptide (50 lM) and acrylodan (500 lM) were
Targeted GOAT inhibitors have promise as potential treatments
for a number of diseases such as obesity, diabetes, and Prader–Wil-
li syndrome [17,31–33]. Several research groups have reported
examples of GOAT inhibitors based on mimics of ghrelin/acylated
ghrelin, screening of small-molecule libraries, or bisubstrate ana-
logs [30,34–36]. Cole and coworkers designed a peptide-based
bisubstrate inhibitor (GO-CoA-Tat) capable of crossing mammalian
cell membranes. This inhibitor effectively inhibited GOAT in both
cultured mammalian cells and mice, and treatment of mice with
this inhibitor yielded increased glucose tolerance and reduced
weight gain [34]. Garner and Janda have reported a number of
non-peptide small-molecule GOAT inhibitors with micromolar
IC50 values and suggested that their inhibitors can serve as lead
compounds for developing more potent GOAT inhibitors [35,36].
However, designing novel GOAT inhibitors with increased potency
will be extremely difficult without a molecular-scale model of the
interactions between ghrelin and the GOAT active site.
In this study, we report the development of a novel fluores-
cence-based assay for GOAT activity and the use of this assay to
investigate interactions underlying GOAT substrate selectivity.
Using recombinant human ghrelin O-acyltransferase (hGOAT), we
demonstrate octanoylation of our fluorescent peptide substrate
that mimics the N-terminal sequence of ghrelin. Using structure–
reactivity analysis of a series of peptide substrates, we have
identified specific functional groups within our substrate that
potentially contribute to ghrelin recognition and subsequent
modification by hGOAT. Defining the interactions used by hGOAT
to bind and recognize ghrelin will provide insight into the struc-
ture of the hGOAT active site, aid in the design of targeted hGOAT
inhibitors, and potentially help to assess the possibility of novel
hGOAT substrates beyond ghrelin.
dissolved in 1 ml of 1:1 50 mM Heppso buffer (pH 7.8)/acetonitrile,
followed by incubation at room temperature in the dark overnight
(18 h) with shaking [38]. Acrylodan-labeled peptides were purified
by reverse phase HPLC (Zorbax Eclipse XDB column,
9.4 ꢁ 250 mm) using an isocratic mobile phase of water containing
0.05% trifluoroacetic acid (TFA) (65%) and acetonitrile (35%) flow-
ing at 4.2 ml/min over 21 min; acrylodanylated peptides eluting
at approximately 6 to 10 min were detected by UV absorbance at
360 nM and fluorescence (kex = 360 nm, kem = 485 nm). Peptide
labeling with acrylodan was verified by MALDI–TOF mass spec-
trometry (Bruker Autoflex III, SUNY–ESF) using a matrix of satu-
rated sinapinic acid in 0.1% TFA and 50 mM ammonium
phosphate mixed in a 1:1 volume with the peptide sample; calcu-
lated and observed masses are reported in Table S1 of the supple-
mentary material. Acrylodanylated peptides were solubilized in
ethanol and stored at ꢀ80 °C until use, with peptide concentrations
determined by UV absorbance of the cysteine-conjugated acrylo-
dan group at 360 nm in aqueous solution (
[39].
e )
360 = 13,300 Mꢀ1 cmꢀ1
Expression of hGOAT
A
gene encoding hGOAT (NCBI reference sequence
NP_001094386.1) with a C-terminal His10 tag appended down-
stream of a TEV protease site was commercially synthesized (Gen-
Script, Piscataway, NJ, USA) with EcoRI and XbaI restriction sites at
the 50 and 30 termini, respectively, of the synthetic gene. The hGOAT
insert was cloned into the pFastBacDual vector (Invitrogen) using
the EcoRI and XbaI restriction sites, with the resulting pFastBacDu-
al–hGOAT vector used to produce baculovirus per the Bac-to-Bac
baculovirus expression system protocol (Invitrogen). Sf9 insect
cells (5.0 ꢁ 108 cells in 500 ml total culture volume) were infected
with hGOAT baculovirus at a multiplicity of infection (MOI) of 10
followed by protein expression for 72 h at 27.5 °C in 1-L stirred
flasks (89 rpm). Cells were harvested by centrifugation (500 x g,
5 min), followed by cell pellet resuspension in 25 ml of lysis buffer
[150 mM NaCl, 50 mM Tris–HCl (pH 7.0), 1 mM sodium ethylenedi-
amine tetraacetate (NaEDTA), 1 mM dithiothreitol (DTT), complete
Materials and methods
Miscellaneous methods
All assays were performed at room temperature unless reported
otherwise. All data plotting and curve fitting were performed with
Kaleidagraph (Synergy Software, Reading, PA, USA). Octanoyl coen-
zyme A (octanoyl-CoA) and palmitoyl-CoA (Advent Bio, Downers
Grove, IL, USA) were solubilized to 5 mM final concentration in
10 mM Tris–HCl (pH 7.0) and aliquoted into low-adhesion micro-
centrifuge tubes, with aliquots stored at ꢀ80 °C until use. Acrylo-
dan (Anaspec) was solubilized in acetonitrile, with the stock
concentration determined by ultraviolet (UV) absorbance at
mini-tab protease inhibitor (Roche Pharmaceuticals), 10
lg/ml
pepstatin A, and 100 M bis(4-nitrophenyl)phosphate]. The resus-
l
pended cells were lysed with a dounce homogenizer on ice, fol-
lowed by removal of cell debris by centrifugation (3000 x g,
5 min). The membrane fraction was then isolated by ultracentrifu-
gation of the supernatant (66,912 x g, 30 min). The isolated mem-
brane fraction pellet was resuspended in buffer (50 mM Hepes,
pH 7.0), with protein concentration determined by Bradford assay
393 nm on dilution into methanol
(e
= 18,483 Mꢀ1 cmꢀ1 per

70
Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
(Bio-Rad, Hercules, CA, USA). The resuspended membrane fractions
were aliquoted (50
l) and stored at ꢀ80 °C until use.
the ghrelin peptide upon octanoylation. These labels enable GOAT
activity detection through either autoradiography or chemoselec-
tive immobilization of an enzyme for subsequent fluorescent
detection [30,34,35,40]. We chose to incorporate a fluorescent la-
bel into the peptide substrate itself, thereby allowing detection
and quantitation of both the peptide substrate and octanoylated
peptide product of reaction with hGOAT. To allow attachment of
a thiol-reactive AcDan fluorophore, we extended the minimal ghre-
lin substrate peptide (GSSFL) identified by Yang and coworkers by
the addition of a C-terminal cysteine (GSSFLC) while maintaining
C-terminal amidation shown to be required for peptide activity
with hGOAT [30]. This strategy allows for labeled GOAT substrate
generation using commercially available peptides and fluoro-
phores without the need for synthetic expertise or radioactivity.
As an additional experimental advantage, the thioether linkage
formed on reaction of the cysteine side chain with AcDan should
be resistant to hydrolysis or degradation by thioesterases impli-
cated in the loss of labeled acyl-CoAs in other GOAT assays using
membrane fraction enzyme preparations [30,35]. Peptide sub-
strates with free acid C-termini yielded no octanoylation activity
with hGOAT (data not shown), consistent with previous studies
with mouse GOAT [30]. This cysteine residue provides a thiol side
chain for chemoselective labeling by AcDan, as confirmed by MAL-
DI–TOF mass spectrometry (Fig. 2). In addition to chemoselective
labeling of cysteine thiols [41,42], the environmentally sensitive
nature of the AcDan fluorophore potentially provides a probe for
a shift in local polarity on octanoylation of the serine 3 residue
[38,43–45].
l
hGOAT activity assay
Membrane fractions from Sf9 cells expressing hGOAT were
thawed on ice and passed through an 18-gauge needle 10 times.
Membranes were then centrifuged (1000 x g, 1 min), with the
supernatant collected and added to hGOAT reactions. Unless noted
otherwise, assays were performed with 50
tein, 1.5 M acrylodanylated peptide substrate, 500
CoA, 100 M palmitoyl-CoA, and 50 mM Hepes (pH 7.0) in a total
volume of 50 l. Assays were initiated by the addition of the acry-
lg of membrane pro-
l
l
M octanoyl-
l
l
lodanylated peptide substrate. Assays comparing the reactivity of
different peptide substrates were performed in parallel using the
same stock of membrane protein and other reaction components,
with peptide substrates added last. Assays were incubated at room
temperature and stopped by the addition of 50 ll of 20% acetic acid
in isopropanol; the addition of the stop solution prior to the acry-
lodanylated peptide substrate blocked peptide octanoylation (data
not shown). Assays were analyzed by reverse phase HPLC (Zorbax
Eclipse XDB column, 4.6 ꢁ 150 mm) using a gradient from 30% ace-
tonitrile in water containing 0.05% TFA to 63% acetonitrile in water
containing 0.05% TFA flowing at 1 ml/min over 14 min, followed by
100% acetonitrile for 5 min; acrylodanylated peptides were de-
tected by UV absorbance at 360 nm and fluorescence (kex = 360 -
nm, kem = 485 nm). Peptide substrates typically eluted with a
retention time of 5 to 7 min (ꢂ42–46% acetonitrile), with the octa-
noylated peptide product eluting at approximately 12 min (ꢂ58%
acetonitrile); retention times for all octanoylated peptide products
are reported in Table S2 of the supplementary material. Chromato-
gram analysis and peak integration were performed using Chem-
station for LC (Agilent Technologies), with quantitation reported
in Table S3. For all assays, data reported are the average of three
independent determinations, with error bars representing 1 stan-
dard deviation. The mass of the octanoylated peptide product
([M+Na]+ = 985.50 for octanoyl-GSSFLCAcDan, where AcDan repre-
sents acrylodan) was confirmed by MALDI–TOF mass spectrome-
try, with the octanoylated peptide product purified by HPLC from
octanoylation reactions containing hGOAT membrane fractions.
For inhibition experiments, the inhibitor or DMSO carrier con-
hGOAT octanoylation of GSSFLCAcDan
To initiate our studies of ghrelin acylation by GOAT, we cloned
and expressed hGOAT. Previous studies of GOAT activity have used
the mouse GOAT ortholog, which is 79% identical and 88% homol-
ogous to hGOAT (see Fig. S2 in the supplementary material)
[30,34–36,40]. We expressed hGOAT in Sf9 insect cell membranes
using standard baculoviral methods and isolated membrane pro-
tein fractions per previously reported protocols as described in
Materials and Methods [30].
To determine whether the GSSFLCAcDan fluorescent peptide can
function as a substrate for hGOAT, we incubated the peptide in the
presence of octanoyl-CoA and membrane fractions from Sf9 cells
infected with either a negative control baculovirus encoding b-glu-
curonidase or baculovirus encoding for hGOAT (Fig. 3). Incubation
in the presence of the negative control membrane fraction did not
visibly alter the GSSFLCAcDan peptide, with the HPLC chromatogram
mirroring that for the GSSFLCAcDan peptide alone. In contrast, incu-
bation of GSSFLCAcDan with hGOAT-containing membrane fractions
led to the formation of a new fluorescent peak with a longer reten-
tion time during reverse phase HPLC analysis, consistent with an
increase in GSSFLCAcDan hydrophobicity on octanoylation of serine
3. We isolated the peptide in the new HPLC peak and confirmed
that the mass of the reaction product was consistent with octa-
noyl-GSSFLCAcDan (Fig. 3C, [M+Na]+ = 985.50). We also incubated
a negative control peptide with serine 3 mutated to alanine
(GSAFLCAcDan) in the presence of hGOAT-containing membrane
fractions and octanoyl-CoA and observed no evidence for peptide
modification (data not shown). This lack of reactivity is consistent
with the side chain hydroxyl of serine 3 in the GSSFLCAcDan peptide
serving as the octanoylation site.
trol (ꢂ2
ll) was added to the standard reaction mixture described
above before reaction initiation by the addition of the acrylodany-
lated peptide substrate. Reactions were stopped after 1 h and ana-
lyzed as described above, with percentage activity at each inhibitor
concentration calculated from HPLC fluorescence data using Eqs.
(1) and (2):
% peptide octanoylation in presence of inhibitor
% activity ¼
ð1Þ
% peptide octanoylation in absence of inhibitor
fluorescence of octanoylated peptide
% peptide octanoylation ¼
:
total peptide fluorescenceðoctanoylated and non-octanoylatedÞ
ð2Þ
To determine IC50, the plot of % activity versus [inhibitor] was fit to
Eq. (3), with % activity0 denoting hGOAT activity in the presence of
the DMSO carrier alone:
ꢀ
ꢁ
½inhibitorꢄ
% activity ¼ % activity0 ꢃ 1 ꢀ
:
ð3Þ
½inhibitorꢄ þ IC50
Results
Optimization of GSSFLCAcDan octanoylation
Development of a fluorescently labeled ghrelin mimetic substrate
We explored the potential reaction condition space for
GSSFLCAcDan octanoylation to identify the best assay conditions
for subsequent studies of hGOAT activity and substrate selectivity
Previous assays for detecting GOAT activity have relied on la-
bels within the octanoyl-CoA cosubstrate that are transferred to

Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
71
Fig.2. Generation of the fluorescently labeled GSSFLCAcDan peptide substrate for
hGOAT. (A) Incubation of the GSSFLC peptide with acrylodan leads to conjugation of
the AcDan group to the peptide at the side chain thiol of the C-terminal cysteine. (B)
MALDI–TOF mass spectrum confirming the mass of the GSSFLCAcDan fluorescent
peptide for which the sodium adduct is observed (calculated mass for
[M+Na]+ = 858.98 Da, observed mass = 859.84 Da).
(Fig. 4). In the presence of 50 lg of membrane protein from
hGOAT-infected Sf9 cells, the amount of octanoyl-GSSFLCAcDan in-
creases approximately linearly over 60 min and then reaches a pla-
teau by 120 min. Quantitation of the octanoyl-GSSFLCAcDan
concentration by UV absorbance at 360 nm indicates that the reac-
tion plateaus at approximately 15% to 18% conversion of
GSSFLCAcDan to octanoyl-GSSFLCAcDan (see Table S3 in the supple-
mentary material), with the larger increase in observed yield by
fluorescence (ꢂ30% yield) consistent with solvatochromic fluores-
cence enhancement on octanoylation of GSSFLCAcDan. Although the
change in the local environment of the acrylodan fluorophore
could potentially affect UV absorbance in addition to fluorescence,
polarity effects on the acrylodan extinction coefficient are expected
to be minimal based on studies of the isolated fluorophore [42].
Based on this time course, all subsequent reactions were analyzed
following 60 min of incubation to maximize GSSFLCAcDan octanoy-
lation while minimizing reaction time to aid in identifying substi-
tutions that reduce peptide substrate reactivity with hGOAT.
After selecting 1 h as the standard time point for reaction
monitoring, we optimized enzyme and substrate concentrations.
Fig.3. hGOAT catalyzes octanoylation of GSSFLCAcDan. (A) HPLC chromatogram of
octanoylation reaction with control membrane fractions from cells expressing b-
glucuronidase using fluorescence detection of the AcDan group. One major peak is
observed for unmodified GSSFLCAcDan. (B) HPLC chromatogram of octanoylation
reaction with hGOAT-containing membrane fractions, with a new fluorescent peak
arising at approximately 12.5 min retention time from formation of octanoyl-
GSSFLCAcDan. (C) MALDI–TOF mass spectrum of the peptide corresponding to the
new 12.5-min retention time peak in hGOAT-containing octanoylation reactions,
which confirms the mass of the sodium and potassium adducts of octanoyl-
GSSFLCAcDan (calculated mass for [M+Na]+ = 985.17 Da, observed mass = 985.50 Da;
calculated mass for [M+K]+ = 1001.28 Da, observed mass = 1001.48 Da). An asterisk
(⁄) denotes a peak arising from the sinapinic acid matrix.
Titration of hGOAT membrane protein indicates that adding more
than 50
in GSSFLCAcDan octanoylation, whereas decreasing the amount of
membrane protein below 25 g per reaction reduces the observed
extent of reaction. The best conversion of GSSFLCAcDan to octanoyl-
GSSFLCAcDan occurred in the presence of 1.5 M peptide substrate,
lg of membrane protein yields only a minimal increase
l
l

72
Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
Fig.4. Characterization of hGOAT-catalyzed octanoylation of GSSFLCAcDan. (A) Time dependence of GSSFLCAcDan octanoylation. (B) Dependence of amount of hGOAT-
containing membrane protein fraction. (C) Dependence on GSSFLCAcDan concentration. (D) Dependence on concentration of octanoyl-CoA. Unless noted otherwise, assays
were performed with 50
lg of membrane protein, 1.5 lM acrylodanylated peptide, 500 lM octanoyl-CoA, 100 lM palmitoyl-CoA, and 50 mM Hepes (pH 7.0) in a total
volume of 50 l. Reactions in panels B to D were monitored after 1 h of incubation, with product concentration calculated by integration of the fluorescence in the peak
l
corresponding to octanoyl-GSSFLCAcDan as described in Materials and Methods. Data reported are the averages of three independent determinations, with error bars
representing 1 standard deviation.
with reduced octanoylation observed at either 0.5 or
GSSFLCAcDan. Titration of octanoyl-CoA from 0 to 1000 M revealed
a peak in GSSFLCAcDan octanoylation efficiency between 250 and
500 M octanoyl-CoA, with the possibility of substrate inhibition
5 lM
l
l
at higher concentrations. We do not expect any significant
detergent effects from formation of octanoyl-CoA micelles under
our experimental conditions, with the critical micelle concentra-
tion (CMC) for octanoyl-CoA reported as approximately 30 mM
[46–48]. We also varied the concentration of palmitoyl-CoA in
the reaction (Fig. S3) because inclusion of long-chain acyl-CoAs
has improved GOAT-catalyzed ghrelin acylation in other reported
assays [30,35]. However, in our assay, the addition of palmitoyl-
CoA had a minimal effect on the observed extent of GSSFLCAcDan
octanoylation. Because the presence of palmitoyl-CoA in the other
assays was proposed to block hydrolysis of labeled octanoyl-CoA
analogs by nonspecific thioesterases [30,35], the large excess of
octanoyl-CoA and the non-hydrolyzable thioether bond attaching
the fluorescent label to our peptide substrate presumably ren-
ders our assay less sensitive to interference by endogenous
thioesterases.
Fig.5. Inhibition of hGOAT-catalyzed octanoylation of GSSFLCAcDan by octanoyl-
[Dap3]-ghrelin (1–5). Assays were performed as described in Materials and
Methods, with 50
500 M octanoyl-CoA, 100
total volume of 50 l. The inset shows an expanded view of the data at low inhibitor
lg
of membrane protein, 1.5
lM acrylodanylated peptide,
l
lM palmitoyl-CoA, and 50 mM Hepes (pH 7.0) in a
l
GSSFLCAcDan octanoylation is blocked by a known GOAT inhibitor
concentrations. Percentage (%) activity was calculated as described in Materials and
Methods. Data reported are the averages of three independent determinations, with
error bars representing 1 standard deviation.
To evaluate the potential of our assay for screening for GOAT
inhibitors, we determined whether a known inhibitor of mouse
GOAT, octanoyl-[Dap3]-ghrelin (1–5), blocked ghrelin octanoyla-
tion under our assay conditions [30]. Octanoyl-[Dap3]-ghrelin
(1–5) serves as a very potent inhibitor of GSSFLCAcDan octanoyla-
tion by hGOAT, with an observed IC50 of 0.012 0.001 lM
(Fig. 5). Therefore, our fluorescence-based assay for hGOAT activity
is readily compatible with screening potential hGOAT inhibitors.

Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
73
Structure–activity studies of the GSSFLCAcDan substrate
When investigating the functional groups within GSSFLCAcDan
required for octanoylation by hGOAT, we performed our assays
probing hGOAT selectivity in the presence of the hGOAT mem-
brane protein, peptide substrate, and octanoyl-CoA concentrations
that produced the best yield of octanoyl-GSSFLCAcDan. We deter-
mined the extent of peptide substrate octanoylation after 1 h of
incubation to yield the best sensitivity for detecting changes in
peptide octanoylation velocity relative to GSSFLCAcDan, as discussed
above. All assays comparing modified peptide octanoylation to
GSSFLCAcDan were run in parallel with a reaction with GSSFLCAcDan
using the same reaction stock to eliminate any variability due to
changes in enzyme activity, octanoyl-CoA hydrolysis, or other
factors.
hGOAT recognition of the N-terminal glycine of ghrelin
A study of mouse GOAT indicated that the N-terminal glycine
(G1) of the GSSFL sequence in ghrelin is essential for acylation by
GOAT in the context of a full-length ghrelin peptide [30]. Mouse
GOAT-catalyzed ghrelin octanoylation was blocked by either
replacement of G1 with a serine (SSSFL-) or the addition of a dipep-
tide upstream of G1 (SAGSSFL-). To both confirm that hGOAT
shows similar selectivity as mouse GOAT at G1 and identify the
specific functional groups and interactions involved in hGOAT rec-
ognition of G1, we investigated three potential recognition ele-
ments on this glycine: the
of the glycine, and side chain steric bulk at the
the requirement for the -amino and carbonyl groups, we mea-
sured the reactivity of peptide substrates where G1 is either acet-
ylated ([Ac]GSSFLCAcDan or replaced by an acetyl group
a-amino group, the carbonyl oxygen
a
-carbon. To probe
a
)
([Ac]SSFLCAcDan). G1 acetylation converts the N-terminal amine to
an amide while adding the steric bulk of the acetyl group, whereas
G1 replacement with an acetyl group removes the N-terminal
amine while maintaining the G1
a-carbon and carbonyl group.
Fig.6. hGOAT requires both the N-terminal amine and the absence of a side chain at
The [Ac]GSSFLCAcDan substrate exhibited reactivity approaching
the lower limit of our HPLC detection limit (ꢂ0.05 fluorescence
units), and no octanoylation of [Ac]SSFLCAcDan was detected
(Fig. 6A). The complete loss of activity with [Ac]SSFLCAcDan suggests
that the N-terminal amine plays a much larger role than the gly-
cine carbonyl in ghrelin recognition by hGOAT. The increase in
activity for [Ac]GSSFLCAcDan relative to [Ac]SSFLCAcDan potentially
suggests that maintenance of hydrogen bond donation ability at
the G1 nitrogen atom is beneficial for binding and/or reactivity,
but confirmation of any such role will require further study.
We further probed hGOAT recognition of G1 by measuring the
reactivity of substrates with N-terminal serine (SSSFLCAcDan) and
alanine (ASSFLCAcDan) residues in place of G1 to introduce hydroxy-
G1 for ghrelin recognition. (A) Product formation for the GSSFLCAcDan peptide
compared with a peptide with the N-terminal amine acetylated ([Ac]GSSFLCAcDan
)
or removed ([Ac]SSFLCAcDan). (B) Product formation of the GSSFLCAcDan peptide
compared with peptides with either alanine or serine replacing the G1 glycine. An
asterisk (ꢃ) indicates reactivity below the fluorescence detection limit for our assay
(ꢂ0.05 AU). All reactions were monitored after 1 h of incubation, with product
concentration calculated by integration of the fluorescence in the peak correspond-
ing to octanoylated peptide as described in Materials and Methods. Data reported
are the averages of three independent determinations, with error bars representing
1 standard deviation. Numerical values are reported in Table S3 of the supplemen-
tary material.
work). These two results suggested the potential for a steric con-
straint at S2, with larger amino acids blocking reaction with
hGOAT. To test this model, we designed a series of mutations at
the S2 position to probe the effect of gradually increasing amino
acid volume at this position (Fig. 7). Mutation of S2 to alanine
led to a marked drop in reactivity, in contrast to the lack of effect
of this mutation in the context of full-length ghrelin octanoylation
by mouse GOAT [30]. Moving to larger amino acids, substitution of
threonine for serine at S2 did not decrease reactivity with hGOAT,
whereas the valine, phenylalanine, and tryptophan mutations all
reduced reactivity to the point where octanoylation of the
GWSFLCAcDan substrate was not detectable. The changes in peptide
substrate reactivity are observable when either product fluores-
cence or product UV absorbance is used to monitor peptide sub-
strate octanoylation (Fig. 7 and Table S3), with fluorescence
providing a more sensitive detection limit.
methyl and methyl functional groups at the a-carbon (Fig. 6B). The
SSSFLCAcDan substrate exhibited no activity with hGOAT, in agree-
ment with the loss of activity with mouse GOAT on the same
G1S mutation in ghrelin reported by Yang and coworkers [30].
The more conservative G1A mutation in ASSFLCAcDan did not re-
store detectable reactivity, indicating that hGOAT will not tolerate
any side chain at the
substrate.
a-carbon of the N-terminal amino acid of its
hGOAT recognition of the hydroxyl group at serine 2
We next probed the selectivity, if any, exhibited by hGOAT at
the serine 2 (S2) position of ghrelin. Mutation of S2 to alanine
was reported to have essentially no effect on ghrelin acylation by
mouse GOAT [30], whereas studies in our lab have indicated that
fluorophore attachment at this position completely blocks reactiv-
ity with hGOAT (Darling, Prybolsky, and Hougland, unpublished
The reactivity of these S2-modified substrates suggests that
hGOAT recognizes (at a minimum) two distinct properties of the

74
Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
assays was increased by adding palmitoyl-CoA, which presumably
saturates contaminating thioesterases in the GOAT-containing
membrane fractions and protects octanoyl-CoA from hydrolysis.
Our assay uses a large excess of octanoyl-CoA in comparison with
the fluorescent peptide, which would render the yield of octanoy-
lated peptide less sensitive to octanoyl-CoA hydrolysis; the mini-
mal effect of adding palmitoyl-CoA on octanoylated peptide yield
in our assay further supports this model. Because octanoylated
ghrelin-derived peptides and bisubstrate mimetic inhibitors have
been shown to function as GOAT inhibitors both in vitro and
in vivo [30,34], it is possible, and even likely, that the incomplete
octanoylation of the fluorescent peptide in our assay results from
product inhibition; studies are currently under way to assess this
possibility.
Structure–function analysis of peptide substrate reactivity with
hGOAT allows us to begin building a model for interactions be-
tween our fluorescent peptide substrate and the ghrelin binding
site within hGOAT (Fig. 8). These interactions can potentially serve
to guide the ongoing development of novel small-molecule GOAT
inhibitors [30,34,36]. The N-terminal glycine appears to be recog-
nized primarily through two factors: the N-terminal amino group
Fig.7. hGOAT recognizes both the hydroxyl group and steric bulk at S2 of ghrelin.
Product formation for the GSSFLCAcDan peptide and peptide substrates with
mutations (alanine, threonine, valine, phenylalanine, and tryptophan) at the S2
position is shown. An asterisk (ꢃ) indicates reactivity below the fluorescence
detection limit for our assay (ꢂ0.05 AU). All reactions were monitored after 1 h of
incubation, with product concentration calculated by integration of the fluores-
cence in the peak corresponding to octanoylated peptide as described in Materials
and Methods. Data reported are the averages of three independent determinations,
with error bars representing 1 standard deviation. Numerical values are reported in
Table S3 of the supplementary material.
and the lack of steric bulk/side chain at the
a-carbon itself. The
presence of a hydroxyl group at S2 enhances the activity of the
peptide substrate when compared with a peptide with an isosteric
aliphatic side chain at this position, consistent with the S2 hydro-
xyl group interacting with a partner within the hGOAT active site.
Increasing the volume of the amino acid side chain at S2 results in
loss of peptide reactivity. However, the steric constraint at S2 is
amino acid at this position: presence/absence of a hydroxyl group
and overall amino acid size. Our mutations at S2 allow two pair-
wise comparisons of nearly isosteric amino acids to probe the ef-
fects of removing the hydroxyl group at this position: serine
(89 Å3) versus alanine (88.6 Å3) and threonine (116.1 Å3) versus
valine (140 Å3) [49]. In both cases, replacement of the hydroxyl
substantially less restrictive than that observed at the
a-carbon
of G1 because an increase in aliphatic side chain volume of approx-
imately 60% at S2 (Ala ? Val) does not lead to a reduction in pep-
tide activity. These data support a sterically permissive binding
pocket for the S2 side chain where the hydroxyl group forms an
energetically beneficial contact such as a hydrogen bond to a func-
tional group within the hGOAT active site.
group with either hydrogen (Ser ? Ala) or
a methyl group
(Thr ? Val) reduces peptide reactivity with hGOAT. This suggests
that hGOAT recognizes the hydroxyl group at S2, possibly through
formation of a hydrogen bond to an amino acid within the hGOAT
active site. In addition to recognizing the S2 hydroxyl, the negative
effect of increasing the amino acid volume at S2 on peptide sub-
strate reactivity is consistent with an S2 binding pocket that dis-
criminates against amino acids larger than valine.
We note that our findings for substrate selectivity at S2 differ
from the lack of selectivity at S2 of proghrelin reported by Yang
and coworkers [30]. This difference may arise from one of several
factors such as different isoforms of GOAT (human vs. mouse)
and the use of a full-length recombinant 96-mer proghrelin in
the previous study in contrast to our shorter GSSFLCAcDan ghrelin
mimetic peptide. The latter possibility may suggest that interac-
tions with proghrelin downstream of leucine 5 (L5) may play roles
in defining GOAT selectivity. However, the binding of small molec-
ular inhibitors to hGOAT may be expected to resemble that of our
small peptide mimic more closely than full-length proghrelin,
underscoring the potential importance of the interactions identi-
fied here in the development of therapeutic agents targeting GOAT.
Our proposed model for GSSFLCAcDan binding to hGOAT does not
include any contacts downstream of the phenylalanine in the
fourth position (F4) of ghrelin. This proposal is consistent with
the studies of proghrelin by Yang and workers [30] and the utility
of short ghrelin-mimetic peptide substrates in several different as-
say formats [30,35,40]. Our ability to attach a bulky fluorophore at
the C-terminus of the GSSFLC peptide while maintaining reactivity
with hGOAT provides further support for the lack of peptide sub-
strate selectivity downstream of the first four amino acids of ghre-
lin. In addition to the interactions within the hGOAT active site that
recognize the N-terminal GSSF tetramer sequence of ghrelin, the
enzyme structure must also contain a channel or groove to accom-
modate the remaining amino acids of proghrelin downstream of
the proposed tetrapeptide N-terminal recognition motif. This pro-
posed groove may provide room to accommodate the acrylodan
fluorophore at the C-terminus of our GSSFLCAcDan substrate be-
cause an attachment of the acrylodan fluorophore upstream of
the peptide C-terminus leads to significantly reduced peptide
Discussion
To probe the ghrelin acylation activity of hGOAT, we have
developed a novel fluorescently labeled peptide substrate that
mimics the first five amino acids of human ghrelin [50]. Using this
substrate, we demonstrated in vitro ghrelin peptide octanoylation
using recombinant hGOAT in an HPLC-based assay with fluores-
cence detection. Employing a known GOAT inhibitor, we demon-
strated that our assay can be used to screen for hGOAT
inhibitors. Through amino acid mutation and functional group
modification, we identified several functional groups within the
ghrelin peptide mimic involved in substrate selectivity. These find-
ings constitute an essential step toward defining the interactions
within the hGOAT active site that bind and position ghrelin for
modification, interactions that can be targeted to design novel
hGOAT inhibitors.
In our assay, octanoylation of the fluorescent peptide remains
incomplete and does not proceed beyond approximately 20% con-
version based on quantification of both substrate and product pep-
tides by UV absorbance (Table S3). However, low conversion of
ghrelin to octanoylated ghrelin (<1%) has also been reported in as-
says employing either [3H] octanoyl-CoA or unlabeled octanoyl-
CoA followed by immunodetection [30,40]. Reaction yield in these

Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
75
Fig.8. Proposed model of interactions involved in GSSFLCAcDan recognition within the hGOAT active site. Interactions with the N-terminal amine, the
a-carbon of G1, and the
side chain hydroxyl of S2 identified in this work are shown as dotted lines. Proposed contacts to the serine 3 hydroxyl group and the side chain of F4 (dashed lines) are
included to indicate the potential for catalytic interactions at serine 3 and contact(s) responsible for the substrate selectivity at the F4 position reported in previous studies,
respectively [30,34,40].
[2] M. Kojima, K. Kangawa, Structure and function of ghrelin, Results Probl. Cell
Differ. 46 (2008) 89–115.
[3] V.P. Carlini, M.F. Perez, E. Salde, H.B. Schioth, O.A. Ramirez, S.R. de Barioglio,
reactivity (Darling et al., unpublished work). The addition of steric
bulk at F4 or L5 does not necessarily block peptide binding to
GOAT, however, given that Barnett and coworkers have reported
photo-crosslinking of ghrelin-mimetic inhibitors using benzoyl-
phenylalanine crosslinking groups at F4 and L5 [34]. These findings
illustrate the need for continuing studies to define the size and
nature of the ghrelin binding site within GOAT.
Ghrelin induced memory facilitation implicates nitric oxide synthase
activation and decrease in the threshold to promote LTP in hippocampal
dentate gyrus, Physiol. Behav. 101 (2010) 117–123.
[4] S. Diano, S.A. Farr, S.C. Benoit, E.C. McNay, I. da Silva, B. Horvath, F.S. Gaskin, N.
Nonaka, L.B. Jaeger, W.A. Banks, J.E. Morley, S. Pinto, R.S. Sherwin, L. Xu, K.A.
Yamada, M.W. Sleeman, M.H. Tschop, T.L. Horvath, Ghrelin controls
hippocampal spine synapse density and memory performance, Nat. Neurosci.
9 (2006) 381–388.
[5] N. Nagaya, M. Kojima, M. Uematsu, M. Yamagishi, H. Hosoda, H. Oya, Y.
Hayashi, K. Kangawa, Hemodynamic and hormonal effects of human ghrelin in
healthy volunteers, Am. J. Physiol. Regul. Integr. Comp. Physiol. 280 (2001)
R1483–R1487.
[6] C.D. Murray, N.M. Martin, M. Patterson, S.A. Taylor, M.A. Ghatei, M.A. Kamm, C.
Johnston, S.R. Bloom, A.V. Emmanuel, Ghrelin enhances gastric emptying in
diabetic gastroparesis: a double blind, placebo controlled, crossover study, Gut
54 (2005) 1693–1698.
[7] F. Levin, T. Edholm, P.T. Schmidt, P. Gryback, H. Jacobsson, M. Degerblad, C.
Hoybye, J.J. Holst, J.F. Rehfeld, P.M. Hellstrom, E. Naslund, Ghrelin stimulates
gastric emptying and hunger in normal-weight humans, J. Clin. Endocrinol.
Metab. 91 (2006) 3296–3302.
In summary, we have reported the design and utility of a novel
fluorescent peptide substrate for investigating ghrelin modification
by hGOAT. Using mutations and chemical modifications within this
peptide, we defined specific interactions within the hGOAT active
site that potentially play roles in ghrelin recognition. In addition
to the utility of the HPLC-based assay demonstrated in this work,
the solvatochromic nature of the acrylodan fluorophore may allow
for development of a continuous hGOAT activity assay similar in
concept to those used to monitor other lipidation enzymes
[51,52]. We will also explore the potential of these substrates in
ghrelin–hGOAT binding studies using fluorescence enhancement
and/or anisotropy. By functionally defining the interactions within
the hGOAT active site required to bind and octanoylate ghrelin, we
will gather clues to both the hGOAT catalytic mechanism and the
nature of the ghrelin binding pocket that will aid in designing no-
vel hGOAT inhibitors and optimizing those currently known
[30,34–36]. With ghrelin-based signaling implicated in a number
of diseases, including obesity and diabetes, developing potent
hGOAT inhibitors is essential for evaluating the hGOAT–ghrelin
modification pathway as a therapeutic target.
[8] M. Tschop, D.L. Smiley, M.L. Heiman, Ghrelin induces adiposity in rodents,
Nature 407 (2000) 908–913.
[9] A.M. Wren, L.J. Seal, M.A. Cohen, A.E. Brynes, G.S. Frost, K.G. Murphy, W.S.
Dhillo, M.A. Ghatei, S.R. Bloom, Ghrelin enhances appetite and increases food
intake in humans, J. Clin. Endocrinol. Metab. 86 (2001) 5992–5995.
[10] I. Kishimoto, T. Tokudome, H. Hosoda, M. Miyazato, K. Kangawa, Ghrelin and
cardiovascular diseases, J. Cardiol. 59 (2012) 8–13.
[11] C.T. Lim, B. Kola, M. Korbonits, The ghrelin/GOAT/GHS-R system and energy
metabolism, Rev. Endocr. Metab. Disord. 12 (2011) 173–186.
[12] C.T. Lim, B. Kola, M. Korbonits, A.B. Grossman, Ghrelin’s role as a major
regulator of appetite and its other functions in neuroendocrinology, Prog.
Brain Res. 182 (2010) 189–205.
[13] M. Lutter, I. Sakata, S. Osborne-Lawrence, S.A. Rovinsky, J.G. Anderson, S. Jung,
S. Birnbaum, M. Yanagisawa, J.K. Elmquist, E.J. Nestler, J.M. Zigman, The
orexigenic hormone ghrelin defends against depressive symptoms of chronic
stress, Nat. Neurosci. 11 (2008) 752–753.
Acknowledgments
[14] Z.B. Andrews, D. Erion, R. Beiler, Z.W. Liu, A. Abizaid, J. Zigman, J.D. Elsworth,
J.M. Savitt, R. DiMarchi, M. Tschoep, R.H. Roth, X.B. Gao, T.L. Horvath, Ghrelin
promotes and protects nigrostriatal dopamine function via a UCP2-dependent
mitochondrial mechanism, J. Neurosci. 29 (2009) 14057–14065.
[15] M.D. Gahete, J. Cordoba-Chacon, R.D. Kineman, R.M. Luque, J.P. Castano, Role of
ghrelin system in neuroprotection and cognitive functions: implications in
Alzheimer’s disease, Peptides 32 (2011) 2225–2228.
[16] P.J. Verhulst, I. Depoortere, Ghrelin’s second life: from appetite stimulator to
glucose regulator, World J. Gastroenterol. 18 (2012) 3183–3195.
[17] K.M. Heppner, J. Tong, H. Kirchner, R. Nass, M.H. Tschop, The ghrelin O-
acyltransferase–ghrelin system: a novel regulator of glucose metabolism, Curr.
Opin. Endocrinol. Diabetes Obes. 18 (2011) 50–55.
This work was funded by Syracuse University and a National
Science Foundation (NSF) predoctoral fellowship (DGE-1247399)
to J.E.D. We thank Robert Doyle, Kevin Sweder, and members of
the Hougland laboratory for helpful comments and discussion.
We also thank the Korendovych laboratory (Syracuse University)
for assistance with peptide synthesis.
Appendix A. Supplementary data
[18] P.J. Delhanty, A.J. van der Lely, Ghrelin and glucose homeostasis, Peptides 32
(2011) 2309–2318.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ab.2013.02.013.
[19] Y. Sun, M. Asnicar, R.G. Smith, Central and peripheral roles of ghrelin on
glucose homeostasis, Neuroendocrinology 86 (2007) 215–228.
[20] Y. Sun, M. Asnicar, P.K. Saha, L. Chan, R.G. Smith, Ablation of ghrelin improves
the diabetic but not obese phenotype of ob/ob mice, Cell Metab. 3 (2006) 379–
386.
References
[21] A. Zhou, G. Webb, X. Zhu, D.F. Steiner, Proteolytic processing in the secretory
pathway, J. Biol. Chem. 274 (1999) 20745–20748.
[1] M. Kojima, H. Hosoda, Y. Date, M. Nakazato, H. Matsuo, K. Kangawa, Ghrelin is
a
growth-hormone-releasing acylated peptide from stomach, Nature 402
[22] X. Zhu, Y. Cao, K. Voogd, D.F. Steiner, On the processing of proghrelin to
ghrelin, J. Biol. Chem. 281 (2006) 38867–38870.
(1999) 656–660.

76
Fluorescent GOAT peptide substrate / J.E. Darling et al. / Anal. Biochem. 437 (2013) 68–76
[23] J.A. Gutierrez, P.J. Solenberg, D.R. Perkins, J.A. Willency, M.D. Knierman, Z. Jin,
D.R. Witcher, S. Luo, J.E. Onyia, J.E. Hale, Ghrelin octanoylation mediated by an
orphan lipid transferase, Proc. Natl. Acad. Sci. USA 105 (2008) 6320–6325.
[24] J. Yang, M.S. Brown, G. Liang, N.V. Grishin, J.L. Goldstein, Identification of the
acyltransferase that octanoylates ghrelin, an appetite-stimulating peptide
hormone, Cell 132 (2008) 387–396.
[39] P.L. Post, K.M. Trybus, D.L. Taylor, A genetically engineered, protein-based
optical biosensor of myosin II regulatory light chain phosphorylation, J. Biol.
Chem. 269 (1994) 12880–12887.
[40] H. Ohgusu, K. Shirouzu, Y. Nakamura, Y. Nakashima, T. Ida, T. Sato, M. Kojima,
Ghrelin O-acyltransferase (GOAT) has a preference for n-hexanoyl-CoA over n-
octanoyl-CoA as an acyl donor, Biochem. Biophys. Res. Commun. 386 (2009)
153–158.
[25] K. Hofmann,
A superfamily of membrane-bound O-acyltransferases with
implications for wnt signaling, Trends Biochem. Sci. 25 (2000) 111–112.
[26] J.A. Buglino, M.D. Resh, Hhat is a palmitoylacyltransferase with specificity for
N-palmitoylation of sonic hedgehog, J. Biol. Chem. 283 (2008) 22076–22088.
[27] J.A. Buglino, M.D. Resh, Identification of conserved regions and residues within
hedgehog acyltransferase critical for palmitoylation of sonic hedgehog, PLoS
One 5 (2010) e11195.
[28] A. Krogh, B. Larsson, G. von Heijne, E.L. Sonnhammer, Predicting
transmembrane protein topology with a hidden Markov model: application
to complete genomes, J. Mol. Biol. 305 (2001) 567–580.
[29] M.S. Taylor, Y. Hwang, P.Y. Hsiao, J.D. Boeke, P.A. Cole, Ghrelin O-
acyltransferase assays and inhibition, Methods Enzymol. 514 (2012) 205–228.
[30] J. Yang, T.J. Zhao, J.L. Goldstein, M.S. Brown, Inhibition of ghrelin O-
acyltransferase (GOAT) by octanoylated pentapeptides, Proc. Natl. Acad. Sci.
USA 105 (2008) 10750–10755.
[31] T.R. Castaneda, J. Tong, R. Datta, M. Culler, M.H. Tschop, Ghrelin in the
regulation of body weight and metabolism, Front. Neuroendocrinol. 31 (2010)
44–60.
[41] M.P. Mims, C.B. Sturgis, J.T. Sparrow, J.D. Morrisett, Acrylodan can label
amino as well as sulfhydryl groups: results with low-density lipoprotein,
lipoprotein[a], and lipid-free proteins, Biochemistry 32 (1993) 9215–
9220.
[42] F.G. Prendergast, M. Meyer, G.L. Carlson, S. Iida, J.D. Potter, Synthesis, spectral
properties, and use of 6-acryloyl-2-dimethylaminonaphthalene (acrylodan): a
thiol-selective, polarity-sensitive fluorescent probe, J. Biol. Chem. 258 (1983)
7541–7544.
[43] Y. Sun, L. Breydo, N. Makarava, Q. Yang, O.V. Bocharova, I.V. Baskakov, Site-
specific conformational studies of prion protein (PrP) amyloid fibrils revealed
two cooperative folding domains within amyloid structure, J. Biol. Chem. 282
(2007) 9090–9097.
[44] T. Hiratsuka, ATP-induced opposite changes in the local environments around
Cys697 (SH2) and Cys707 (SH1) of the myosin motor domain revealed by the
prodan fluorescence, J. Biol. Chem. 274 (1999) 29156–29163.
[45] S.W. Reid, E.K. Koepf, L.D. Burtnick, Fluorescent responses of acrylodan-labeled
plasma gelsolin, Arch. Biochem. Biophys. 302 (1993) 31–36.
[46] R.H. Smith, G.L. Powell, The critical micelle concentration of some
[32] J.B. Hillman, J. Tong, M. Tschop, Ghrelin biology and its role in weight-related
disorders, Discov. Med. 11 (2011) 521–528.
[33] C.X. Yi, K. Heppner, M.H. Tschop, Ghrelin in eating disorders, Mol. Cell.
Endocrinol. 340 (2011) 29–34.
physiologically important fatty acyl-coenzyme A’s as
length, Arch. Biochem. Biophys. 244 (1986) 357–360.
a
function of chain
[47] J.G. Boylan, J.A. Hamilton, Interactions of acyl-coenzyme
A
with
[34] B.P. Barnett, Y. Hwang, M.S. Taylor, H. Kirchner, P.T. Pfluger, V. Bernard, Y.Y.
Lin, E.M. Bowers, C. Mukherjee, W.J. Song, P.A. Longo, D.J. Leahy, M.A. Hussain,
M.H. Tschop, J.D. Boeke, P.A. Cole, Glucose and weight control in mice with a
designed ghrelin O-acyltransferase inhibitor, Science 330 (2010) 1689–1692.
[35] A.L. Garner, K.D. Janda, Cat-ELCCA: a robust method to monitor the fatty acid
acyltransferase activity of ghrelin O-acyltransferase (GOAT), Angew. Chem. Int.
Ed. Engl. 49 (2010) 9630–9634.
phosphatidylcholine bilayers and serum albumin, Biochemistry 31 (1992)
557–567.
[48] C.J. Fiol, L.L. Bieber, Effects of octylglucoside and Triton X-100 on the kinetics
and specificity of carnitine palmitoyltransferase, Lipids 23 (1988) 120–125.
[49] A.A. Zamyatin, Protein volume in solution, Prog. Biophys. Mol. Biol. 24 (1972)
107–123.
[50] M. Kojima, K. Kangawa, Ghrelin: structure and function, Physiol. Rev. 85
(2005) 495–522.
[36] A.L. Garner, K.D. Janda,
A small molecule antagonist of ghrelin O-
acyltransferase (GOAT), Chem. Commun. 47 (2011) 7512–7514.
[51] P.B. Cassidy, J.M. Dolence, C.D. Poulter, Continuous fluorescence assay for
protein prenyltransferases, Methods Enzymol. 250 (1995) 30–43.
[52] D.L. Pompliano, R.P. Gomez, N.J. Anthony, Intramolecular fluorescence
enhancement: a continuous assay of Ras farnesyl–protein transferase, J. Am.
Chem. Soc. 114 (1992) 7945–7946.
[37] P.W. Riddles, R.L. Blakeley, B. Zerner, Ellman’s reagent: 5,50-dithiobis(2-
nitrobenzoic acid)—a reexamination, Anal. Biochem. 94 (1979) 75–81.
[38] M.A. Tricerri, A.K. Behling Agree, S.A. Sanchez, A. Jonas, Characterization of
apolipoprotein A–I structure using
a cysteine-specific fluorescence probe,
Biochemistry 39 (2000) 14682–14691.