Article

pubs.acs.org/ac

Metabolism of Growth Hormone Releasing Peptides

Andreas Thomas,*, Philippe Delahaut, Oliver Krug, Wilhelm Schan zer, and Mario Thevis

Center for Preventive Doping Research/Institute of Biochemistry, German Sport University, Am Sportpark Mungersdorf 6, Cologne
50933, Germany

Departement Sante, CER Groupe, Rue du Point du Jour, 8, Marloie, Belgium

S Supporting Information
*

ABSTRACT: New, potentially performance enhancing compounds have frequently been

introduced to licit and illicit markets and rapidly distributed via worldwide operating
Internet platforms. Developing fast analytical strategies to follow these new trends is one
the most challenging issues for modern doping control analysis. Even if reference
compounds for the active drugs are readily obtained, their unknown metabolism
complicates eective testing strategies. Recently, a new class of small C-terminally
amidated peptides comprising four to seven amino acid residues received considerable
attention of sports drug testing authorities due to their ability to stimulate growth
hormone release from the pituitary. The most promising candidates are the growth
hormone releasing peptide (GHRP)-1, -2, -4, -5, -6, hexarelin, alexamorelin, and
ipamorelin. With the exemption of GHRP-2, the entity of these peptides represents
nonapproved pharmaceuticals; however, via Internet providers, all compounds are readily
available. To date, only limited information on the metabolism of these substances is
available and merely one metabolite for GHRP-2 is established. Therefore, a comprehensive in vivo (po and iv administration in
rats) and in vitro (with human serum and recombinant amidase) study was performed in order to generate information on
urinary metabolites potentially useful for routine doping controls. The urine samples from the in vivo experiments were puried
by mixed-mode cation-exchange solid-phase extraction and analyzed by ultrahigh-performance liquid chromatography (UHPLC)
separation followed by high-resolution/high-accuracy mass spectrometry. Combining the high resolution power of a benchtop
Orbitrap mass analyzer for the rst metabolite screening and the speed of a quadrupole/time-of-ight (Q-TOF) instrument for
identication, urinary metabolites were screened by means of a sensitive full scan analysis and subsequently conrmed by highaccuracy product ion scan experiments. Two deuterium-labeled internal standards (triply deuterated GHRP-4 and GHRP-2
metabolite) were used to optimize the extraction and analysis procedure. Overall, 28 metabolites (at least three for each GHRP)
were identied from the in vivo samples and main metabolites were conrmed by the human in vitro model. All identied
metabolites were formed due to exopeptidase- (amino- or carboxy-), amidase-, or endopeptidase activity.

the search for unmodied target analytes in urine was described

to be insucient.3
Known and well-investigated representatives of GHRPs are
GHRP-1, GHRP-2, GHRP-4, GHRP-5, GHRP-6, ipamorelin,
alexamorelin, and hexarelin.12 The sequences for these most
promising candidates comprising four to seven amino acids are
summarized in Table 1. Their commonality comprising the
amidated C-terminus presumably represents the bioactive site
of these peptides. In order to prove this hypothesis, theories on
structural pharmacophores based on molecular modeling and
computational design were described in former studies.1,9,12 In
contrast to GHRH, the releasing peptides deemed to follow a
counteract of somatostatinergic activity at hypothalamatic as
well as at pituitary level, but the complete pathways are still not
entirely understood. The pharmacologically relevant doses
predominantly depend on the dosage form, and hyper- as well
as hypoeective doses are conceivable for cheating athletes.1,3,13

ecently, a class of small peptides with known up-regulating

eect on endogenous growth hormone (hGH) production
and secretion has received considerable attention of doping
control authorities.15 Due to their biological eects, these
compounds are summarized as GH-releasing peptides
(GHRP), and their analysis in a doping control setting is not
accomplished by means of established screening assays for GH,
e.g., based on the dierential isoform test.6 Findings of such
GHRPs in nutritional supplements and the availability by
Internet/black market providers demonstrate the necessity to
develop test methods allowing for the ecient detection of
these substances and/or their degradation products.5,7
Due to their low molecular mass (<1000 Da), GHRPs are
bioavailable after intravenous, subcutaneous, intranasal, buccal,
and in some cases also oral (iv, sc, in, b, po) administration and
increase age-related the amount of hGH and insulin-like growth
factor-1 (IGF-1) in the treated individuals independently from
gender.811 Due to their variable bioavailability and application
routes, their metabolic fate is expected to be extensive.
Consequently, the knowledge of metabolic pathways is of
utmost importance for eective doping control procedures as
2012 American Chemical Society

Received: July 20, 2012

Accepted: October 26, 2012
Published: October 26, 2012
10252

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Article

Table 1. Growth Hormone Releasing Peptides, Metabolite for GHRP-2, and the Used ISTDs with Their Amino Acid Sequence,
Elemental Composition, Monoisotopic Masses, and Dominant Charge Statea

name

amino acid sequence

monoisotopic mass [Da]

elemental composition

dominant charge state (ESI)

GHRP-2
GHRP-1
GHRP-6
GHRP-5
GHRP-4
alexamorelin
hexarelin
ipamorelin
GHRP-2 metabolite
ISTD1
ISTD2

(D-Ala)-(D--Nal)-Ala-Trp-(D-Phe)-Lys-NH2
Ala-His-(D--Nal)-Ala-Trp-(D-Phe)-Lys-NH2
His-(D-Trp)-Ala-Trp-(D-Phe)-Lys-NH2
Tyr-(D-Trp)-Ala-Trp-(D-Phe)-NH2
(D-Trp)-Ala-Trp-(D-Phe)-NH2
Ala-His-(D-Mrp)-Ala-Trp-(D-Phe)-Lys-NH2
His-(D-Mrp)-Ala-Trp-(D-Phe)-Lys-NH2
Aib-His-(D--Nal)-(D-Phe)-Lys-NH2
(D-Ala)-(D--Nal)-Ala
(D-[2]H3-Ala)-(D--Nal)-Ala
(D-Trp)-[2]H4-Ala-Trp-(D-Phe)-NH2

817.427
954.486
872.444
770.354
607.292
957.497
886.460
711.385
357.168
360.187
611.315

C45H55N9O6
C51H62N12O7
C46H56N12O6
C43H46N8O6
C34H37N7O4
C50H63O7N13
C47H58N12O6
C38H49N9O5
C19H23N3O4
C19H20[2]H3N3O4
C34H33[2]H4N7O4

2+
2+
2+
1+
1+
2+
2+
1+/2+
1+
1+
1+

Nonstandard abbreviations: Nal = naphthylalanine, Mrp = 2-methyltryptophan, Aib = aminoisobutyric acid.

Recommendations of oral daily amounts of GHRP-2 range

from 150 g to 10 mg, whereby the intravenously administered
amounts range from 50 to 200 g/kg body weight.
To date, comprehensive screening methods for GHRPs are
not established, mainly due to the lack of knowledge
concerning adequate target analytes, particularly for urine
analysis. First analytical studies dealing with these peptides were
published, and for GHRP-2 one major metabolite was
characterized.3,4 However, most other methods aim for the
active and intact peptides in blood or urine due to the lack of
knowledge concerning their metabolic degradation and
elimination. These detection assays are based on either liquid
chromatography (LC) coupled to mass spectrometry (MS) or
bioassays with receptor binding units.2,14
The analytical approach to identify metabolites of peptides
diers signicantly from those dealing with low molecular mass
nonpeptidic drugs as the metabolism follows deviating
mechanisms.1518 The liver is the main organ being responsible
for the biotransformation and metabolism of nonpeptide
(small-molecule) drugs but not regarding peptides.19 Here
proteolytic processes by proteases that are present in nearly all
tissues in the organism are known for their degradation
potential rather than metabolism by cytochrome-based
enzymes.18 Accordingly, the degradation of peptides takes
place in dierent body compartments, and here blood and
kidneys were found to be the major sites of metabolism. Due to
the known cleavage potential of endopeptidases and aminoexoor carboxyexopeptidases for circulating peptides, at least one
modied amino acid is incorporated in each analogue to
prolong the half-life of the drug.18 Also, the amidation of the Cterminus provides protective features against degradation in
addition to the importance for the bioactivity in the secretory
pathway.
The present study illustrates a comprehensive approach to
investigate the largely unknown metabolism of GHRPs by
combining dierent preanalytical techniques (in vivo/in vitro
experiments) with sophisticated instrumental tools (highresolution mass spectrometry with Orbitrap and quadrupole/
time-of-ight mass analyzers) after liquid chromatographic
separation. Hereby 28 dierent metabolites (for each GHRP at
least three) are identied with highest degree of condence by
their accurate mass as well as at least two diagnostic accurate
mass product ions.

EXPERIMENTAL SECTION
Chemicals and Reagents. Acetonitrile (gradient grade),
formic acid (LCMS grade), methanol, glacial acetic acid
(p.a.), and ammonium carbonate (p.a.) were purchased from
Sigma-Aldrich (Deisendorf, Germany). The reference substances of GHRP-2 and hexarelin, were obtained from
Purepetides (San Diego, CA). References for GHRP-1,
GHRP-4, labeled GHRP-4 ((D-Trp)d4AlaTrp(D-Phe)-NH2,
used as ISTD2), GHRP-5, GHRP-6, alexamorelin, and
ipamorelin were purchased from BMFZ (Dusseldorf, Germany). The metabolite of GHRP-2 and its labeled analogue
(ISTD1) were synthesized in house as described elsewhere.4
The model peptides TrpAlaTrpPheNH2 (analogue to GHRP4), TrpAlaTrpPheLys, and TrpAlaTrpPheLys-NH2 were
purchased from Centic Biotec (Weimar, Germany). The purity
and identity of all compounds were conrmed by means of
liquid chromatography and high-resolution/high-accuracy
tandem mass spectrometry before administration.4 Recombinant amidase (Enzyme Commission (EC) no. 3.5.1.4, expressed in E. coli) was obtained from Sigma-Aldrich
(Deisendorf, Germany), and pure water in Milli-Q quality
was used for all dilution steps.
Detailed information about the administration experiments
(in vivo), the sample collection procedure, and the in vitro
experiments are provided in the Supporting Information.
Sample Preparation. The basic conditions of the sample
preparation procedure were adapted from a former study and
included an ecient purication by means of weak cationexchange solid-phase extraction.4 Details are also provided in
the Supporting Information.
Liquid Chromatography. In the present study, an Agilent
(Waldbronn, Germany) 1260 or a Thermo (Bremen,
Germany) Accela 1200 liquid chromatograph was used, both
equipped with a Hypersil Gold C18 analytical column (Thermo,
Bremen, Germany), 2.1 mm 50 mm, 1.8 m particle size.
The gradient started at 95% solvent A (0.1% formic acid) and
linearly changed to 65% solvent B (acetonitrile) within 8 min.
In the next 2 min the gradient raised to 90% B, followed by a 5
min re-equilibration time at starting conditions. The ow rate
was set to 200 L/min, and the overall run time was 15 min.
Liquid Chromatographic Fractionation. The previously
described conditions for sample preparation and liquid
chromatography were also used for fractionation purposes.
Here the euent of the analytical column was split via a teesplitter to 199.5 L/min into the fraction collector (Triversa,
Advion, Ithaca, NY) and 500 nL/min was directed to a chip10253

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Figure 1. Chromatographic separation of

chromatography (UHPLC).

Article

D/L-amino

acid containing GHRP-4 and its natural analogue by ultrahigh-performance liquid

with a collision energy of 35 eV, and product ions were

subsequently recorded in time-of-ight (TOF) mode covering
the scan range from m/z 100 to 1100. The TOF was calibrated
frequently (after 10 injections) via the DuoTurbo-V ion source
by a calibrant delivery system containing the manufacturers
calibrants for positive ionization. The nitrogen for the ion
source as well as collision gas supply was delivered by the
nitrogen generator (CMC, Eschborn, Germany).

based nanoelectrospray source coupled to an Exactive mass

spectrometer (Thermo, Bremen, Germany) enabling an online
fractionation control by means of nanospray high-resolution
mass spectrometry. The described basic metabolite screening
parameters in positive mode (see the Mass Spectrometry
section) were used for the mass spectrometer. The fraction
collector started 3 min after sample injection and collected
fractions of 50 L for 10 min into a 384 polypropylene well
plate (200 L, low protein binding coating, Eppendorf,
Hamburg, Germany). Approximately 20 fractions were
collected per sample. Fractionation was performed on selected
samples containing a substantial amount of a target metabolite.
The plates were stored frozen until further analysis.
Mass Spectrometry. Two dierent high-resolution mass
spectrometers were used for identication and characterization
of all metabolites: a Q Exactive benchtop Orbitrap mass
spectrometer (Thermo, Bremen, Germany) and an AB Sciex
Triple TOF 5600 (AB Sciex, Darmstadt, Germany). By
sensitive scan-to-scan polarity switching full scan analysis with
a resolution of 70 000 (full width at half-maximum, fwhm) with
electrospray ionization (HESI-II) and a scan range from m/z
300 to 1000, the samples were screened for potential
metabolites by means of the benchtop Orbitrap device. The
instrument was previously calibrated according to the
manufacturers recommendations (using caeine, the tetrapeptide MRFA, and Ultramark) to ensure mass accuracies below 5
ppm. The nitrogen used for the ion source as well as collision
gas was produced by a nitrogen generator (CMC, Eschborn,
Germany). Diagnostic ion traces (10 ppm of the theoretical
mass) of the singly or doubly charged precursor ions were
extracted for identication. After identication of precursor ions
of potential metabolites, high-resolution product ion scan
experiments were performed with the Triple TOF 5600. Here,
precursor ions were isolated in Q1, collisionally activated in Q2

RESULTS AND DISCUSSION

Metabolism. All target peptides bear an amidated Cterminus and comprise at least one articial amino acid (see
Table 1). Consequently, their metabolism is strongly inuenced
and characteristic metabolites are formed due to incomplete
degradation. Whether and which of these metabolites are still
active was not investigated in this project. All identied
metabolic products comprise at least one non-natural amino
acid, and accordingly, the dierentiation from endogenous or
natural counterparts is unequivocally given. This is exemplarily
shown for the articial peptide GHRP-4 ((D-Trp)AlaTrp(DPhe)-NH2 and its potential natural analogue TrpAlaTrpPheNH2 containing L-amino acids only. Figure 1 illustrates
the retention time shift for the two peptides, whereby a mass
spectrometric dierentiation is not as straightforward as shown
in the corresponding product ion mass spectra. Using identical
dissociation conditions, the generated product ions dier only
in relative abundance but support the identication of the
chromatographically separated analytes.
Mass Spectrometry. High-resolution/high-accuracy mass
spectrometry is a key technology for peptide and protein
analysis, especially in metabolism studies with unknown target
analytes. Due to the possibility to data-mine chromatograms for
the exact masses (<5 ppm) of potential metabolites in dierent
10254

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Article

Figure 2. Extracted ion chromatograms of product ion scans aiming at GHRP-6 metabolites in a rat urine sample collected 9 h after intravenous
administration of GHRP-6 (G5-IV-9-R1-9h). Abundant signals are observed for ISTDs 1 and 2, GHRP-6, and the metabolites of GHRP-6, M1M4.
The small inset gives an example for the corresponding product ion spectra.

charge states, the evaluation of the processed samples is

performed in a straightforward approach. The unambiguous
identication of the metabolites is subsequently provided by
product ion scan experiments of the initially observed
candidates, which was accomplished by means of QTOF-MS/
MS in the present study. The identication criteria included
mass accuracy of precursor as well as product ions (<5 ppm)
and matching amino acid sequence tags according to Roepstor
and Fohlman.20 At least two product ions fullling all
requirements were considered necessary for an adequate
structure proposal.
Characterization of the Metabolites. For all target
compounds, a hydroxyl group transfer at the amidated Cterminus was identied in the in vivo samples due to amidase
activity. The identication of the active and intact drug
compound as well as the metabolites M1M4 by means of
chromatography and accurate mass product ion spectra is
depicted in Figure 2 for GHRP-6.
In general, the comparison of the administered peptide with
the resulting deamidated metabolites outlines a common
retention time shift (as well as a mass shift) yielding moderately
prolonged elution times under the chosen chromatographic
conditions. Due to exopeptidases, unchanged/natural amino
acids are eliminated from the N/C-terminus and various
truncated metabolites were identied. Hereby the cleavage of
the polar residue lysine yielded mainly singly charged precursor
ions (from formerly doubly charged species) with signicantly
prolonged retention times. In addition to the product ions
determined with accurate mass, these observations were taken
into consideration for conclusive identication. Altogether, 28
metabolites were characterized, and for ve out of the seven

drug candidates, also the intact peptide was found in urine

samples. The complete overview including amino acid
sequences, molecular masses, masses of diagnostic product
ions, and retention times of all target compounds is
summarized in Table 1, parts a and b, in the Supporting
Information. Quantitative result interpretations (providing
pharmacokinetic characteristics) were not conducted due to
the lack of certied reference material for the identied
metabolites. Generally, the estimated concentrations of the
intact peptide as well as their metabolites in urine after
administration (iv or po) to humans range in the low- to
subnanogram per milliliter level, which was shown in all
previous studies.24 Nevertheless, the present data show that,
with the exemption of alexamorelin and GHRP-1, all peptides
were excreted as intact drugs and their metabolites when
administered intravenously. Accordingly, a possible screening
procedure should include both species as analytes of interest. In
contrast to the other candidates, GHRP-1 and alexamorelin
exhibit two unmodied amino acid residues (Ala-His-) at the
N-terminus and are predestined for fast exoproteolytic
metabolism. The analysis of the placebo group samples showed
no interfering signals in the respective diagnostic ion traces for
the identied metabolites and supported their origin and
proposed structure.
In Vivo/in Vitro Experiments. The in vivo study urine
samples collected 9 h after iv application of the dierent
GHRPs contained the highest amounts of metabolites. In
contrast, specimens sampled after oral application yielded
undeterminable [less than the limit of detection (LOD)] or
trace amounts of the main metabolites. This is in accordance to
former studies where the oral bioavailability of various GHRPs
10255

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Article

Figure 3. Comparison of ionization response in positive and negative ionization mode (R/R+), calculated by the normalized largest (NL) peak
heights for the metabolites M1 at 5.9 min and M5 at 6.5 min of GHRP-1. Positive ions (monoisotopic mass) are extracted from the full scan
chromatograms as [M + 2H]2+ and negative ions as [M H]. The small inset gives an example for the corresponding high-resolution mass
spectrum of M5 in negative full scan mode yielding mass accuracies <3 ppm.

by accurate mass product ion scan experiments. Alternatively to

a deamidation process, the observed increment of 1 Da is
explainable by the transfer of a hydroxyl group to the -amino
function of the C-terminal lysine residue present in case of
GHRP-1, -2, -6, hexarelin, alexamorelin, and ipamorelin.3 In
order to verify or falsify the proposed deamidation process, the
ratio of the ionization response in positive and negative
ionization mode was evaluated. Here, the Q Exactive benchtop
Orbitrap mass spectrometer operated in scan-to-scan polarity
switch mode was used. Figure 3 shows exemplarily the
chromatograms for the analysis of two combined LC fractions
of the in vitro experiments of GHRP-1 containing the two
metabolites M1 and M5. If M5 is the deamidation product of
M1, this metabolite should be more acidic and yield a better
ionization response in negative mode due to a carboxylic acid at
the C-terminus. This phenomenon is indeed found as
demonstrated in Figure 3 by an approximately 10-fold
increased ionization eciency for M5 (R/R+ = 0.177 for
M5 and 0.020 for M1) corroborating the hypothesis of
deamidation rather than deamination (at the -position of Lys)
at the C-terminus accordingly. To our best knowledge, such a
simple experiment to distinguish between an amidated or
carboxylic C-terminus in a peptide was not described before.
Thus, the proof of principle was additionally tested for the two
model peptides (TrpAlaTrpPheLys and TrpAlaTrpPheLysNH2), yielding conrmatory results with a considerably
decreased ionization eciency in negative mode for the

was found to range far below 1%. After 3248 h post

administration nearly all samples did not generate evaluable
signals for the intact drugs and their metabolites, which
demonstrated the fast clearance of these polar compounds after
a single-dose administration. The in vitro metabolism with
human serum conrmed the occurrence of the main
metabolites for all target peptides although the conversion
extent was found to be lower than in vivo. Nevertheless, the
simple experiment supported the prospect to transfer results
from the animal to the human model. Noteworthy here is the
statement to use human serum rather than EDTA-stabilized
plasma due to known inhibitory eects of EDTA for proteolytic
processes.
Deamidation was observed for each GHRP (see Table 1,
parts a and b, in the Supporting Information); consequently, all
chromatograms yielded a metabolite eluting slightly after the
intact drug candidate with a mass increment of 1 Da (exchange
of NH2 with OH). Since the modication is located at the
C-terminus, all observed b-ions are identical to the unchanged
GHRP and are not appropriate for dierentiation of the
metabolite from the intact substrate. The y-ions showed the
described mass shift of 1 Da, but their utility is compromised by
the isotopic signal resulting from the unmodied GHRP even
at resolutions >20 000 fwhm, particularly when comparably
large amounts of the intact peptide are present. Thus, LC
fractionation was performed in order to separate the parent
drug from its metabolite enabling the identication of the latter
10256

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Article

Figure 4. Product ion chromatograms of the in vitro metabolism samples using recombinant amidase (EC 3.5.1.4) for the doubly charged precursors
[M + 2H]2+ of (a and b) GHRP-2 (TOF product of m/z 410), (c and d) hexarelin (TOF product of m/z 444), and (e and f) ipamorelin (TOF
product of m/z 357). After incubation for 36 h at 30 C considerable amounts of the peptides were deamidated yielding the respective metabolites.

amidated model peptide (data not shown). Despite the fact that
this approach does not allow us to estimate or calculate
absolute gas-phase acidity/basicity of peptidic ions, relative
ionization eciencies (R/R+) of dierent peptides with mass
spectrometry were obtained and support structural interpretation.21,22
Finally, one of the responsible enzymes for the metabolism
was investigated with in vitro experiments using recombinant
amidase. Figure 4 illustrates the eectiveness of this enzyme for
the deamidation of GHRP-2, hexarelin, and ipamorelin.
Especially for GHRP-2 and hexarelin the turnover was found
to be nearly complete (95%), and thus, this approach could
potentially serve as an interesting tool to generate reference
material for the main metabolites without chemical synthesis.
Additional degradation processes were not observed in these
experiments when monitoring the respective ion traces. For
ipamorelin the deamidation process under the described
conditions is less eective, which is obvious with regard to
the relative intensities of the diagnostic signals for ipamorelin
and its deamidated metabolite in Figure 4f.
GHRP-2 Metabolism. In the investigated group of
compounds, GHRP-2 represents an extraordinary candidate
as it is the only clinically approved pharmaceutical agent
(Kaken 100). Comprising three modied amino acid residues

(and an amidated C-terminus), its bioavailability is described

after oral as well as iv application. Moreover, a potential
masking eect concerning doping control detection assays for
hGH employing the accepted GH assay were described.13 Due
to the established metabolism in humans, this peptide was not
included in the present in vivo studies. Former studies showed
the renal elimination of minor amounts of the parent drug as
well as one major metabolite after intravenous administration.3,4,13 Two additional metabolites were described, but their
abundance in urine was considered negligible.3 In the in vitro
experiments of the present study using human serum, another
two metabolites were identied as illustrated in Figure 4. In
accordance to the ndings for the other GHRPs, the C-terminal
deamidation was conrmed. The evidence of this structure
proposal (deamidation of the C-terminus instead of deamination of the -amino function of the lysine residue) was given
by high-resolution mass spectrometry (HRMS) product ion
scan experiments of the [M H] ion at m/z 817.40 in
negative ionization mode from the LC-fractionated in vitro
sample (data not shown). By LC fractionation it was possible to
separate the excess of nonmetabolized GHRP-2 from its
deamidated metabolite (M1) to yield sucient mass spectrometric information for proper identication.
10257

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Article

Figure 5. Extracted ion chromatograms of product ion scans aiming at ion chromatograms of GHRP-2 metabolites (M1 at m/z 410.2, M2 at m/z
358.2, and M3 at m/z 357.2) and product ion spectra of M1 and M3 (small insets) from an excretion study urine sample (5 h after 10 mg po).

Interestingly, an amidated metabolite (D-Ala)(-Nal)AlaNH2 (M3) was also found in minor extent in the incubated
serum samples. Its occurrence is not explainable by a linear
exopeptidic degradation process as described for the other
analogues before. This unusual nding strongly suggests that
the metabolism of GHRP-2 is due to endopeptidase activity in
addition to exopeptidases as selected endopeptidases are
described to induce amidation during the peptide bond
cleavage process.1518,23 Both new metabolites (together with
the former established M2) were also conrmed to be present
in human excretion study urine samples (see Figure 5) by
analysis with sensitive product ion scan experiments on the Q
Exactive mass spectrometer. Here it is obvious that the
amidated metabolite M3 is present only in trace amounts,
but in addition to the well-established M2, the deamidated
substrate (M1) is excreted in considerable extent (estimated
roughly by the peak areas in the extracted product ion
chromatograms). Intact GHRP-2 was found in this sample only
in trace concentrations.

the same procedure it is also possible to determine additional

prohibited peptides in urine (e.g., desmopressin), and thus, the
method provides the option to implement a screening assay for
a variety of dierent compounds instead of methods limited to
individual compounds only.24
Here it would be interesting to develop a sophisticated
purication procedure to detect the target analytes also in
blood. This is of particular interest as doping control tests for
hGH are conducted with serum. In case of atypically high
amounts of endogenous hGH, the same sample can/should be
analyzed for the presence of growth hormone releasing
peptides.

ASSOCIATED CONTENT

* Supporting Information
S

Additional information as noted in text. This material is

available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author

CONCLUSION
The investigation of metabolic processes for peptide hormones
represents a new challenge in doping controls and analytical
chemistry. The combination of animal in vivo and human in
vitro approaches together with hyphenated mass spectrometric
detection systems enables an ecient study design in order to
identify and characterize potential metabolites. Within this
study important qualitative data are provided for doping
control laboratories to implement a facile LCMS/MS-based
procedure for these prohibited compounds by means of
commonly and routinely employed instrumental equipment
(triple-quadrupole mass spectrometers or equivalent).
Although derived from animal in vivo studies, the model
appears valid in a qualitative manner since human in vitro
incubations corroborated the likely presence of the surrogate in
vivo generated compounds in human urine. Furthermore, with

*Phone: 0221-49827072. Fax: 0221-49827071. E-mail: a.

thomas@biochem.dshs-koeln.de.
Notes

The authors declare no competing nancial interest.

ACKNOWLEDGMENTS
The study was carried out with support of the Manfred Donike
Institute for Doping Analysis, Cologne, Germany, the Federal
Ministry of the Interior of the Federal Republic of Germany,
and the World Anti-Doping Agency (WADA, Grant No.
10B10MT).

REFERENCES

(1) Ghigo, E.; Arvat, E.; Muccioli, G.; Camanni, F. Eur. J. Endocrinol.
1997, 136, 445460.
10258

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259

Analytical Chemistry

Article

(2) Gil, J.; Cabrales, A.; Reyes, O.; Morera, V.; Betancourt, L.;
Sanchez, A.; Garcia, G.; Moya, G.; Padron, G.; Besada, V.; Gonzalez, L.
J. J. Pharm. Biomed. Anal. 2012, 60, 1925.
(3) Okano, M.; Sato, M.; Ikekita, A.; Kageyama, S. Rapid Commun.
Mass Spectrom. 2011, 24, 20462056.
(4) Thomas, A.; Hoppner, S.; Geyer, H.; Schanzer, W.; Petrou, M.;
Kwiatkowska, D.; Pokrywka, A.; Thevis, M. Anal. Bioanal. Chem. 2011,
401, 507516.
(5) Thomas, A.; Kohler, M.; Mester, J.; Geyer, H.; Schanzer, W.;
Petrou, M.; Thevis, M. Drug Test. Anal. 2010, 2, 144148.
(6) Bidlingmaier, M.; Suhr, J.; Ernst, A.; Wu, Z.; Keller, A.;
Strasburger, C. J.; Bergmann, A. Clin. Chem. 2009, 55, 445453.
(7) Kohler, M.; Thomas, A.; Geyer, H.; Petrou, M.; Schanzer, W.;
Thevis, M. Drug Test. Anal. 2010, 1112, 533537.
(8) Hashizume, T.; Tanabe, Y.; Ohtsuki, K.; Mori, A.; Matsumoto,
N.; Hara, S. Domest. Anim. Endocrinol. 2001, 20, 3746.
(9) Camanni, F.; Ghigo, E.; Arvat, E. Front. Neuroendocrinol. 1998,
19, 4772.
(10) Hartman, M. L.; Farello, G.; Pezzoli, S. S.; Thorner, M. O. J.
Clin. Endocrinol. Metab. 1992, 74, 13781384.
(11) Bowers, C. Y.; Alster, D. K.; Frentz, J. M. J. Clin. Endocrinol.
Metab. 1992, 74, 292298.
(12) Thevis, M.; Schanzer, W. Mini-Rev. Med. Chem. 2007, 7, 531
537.
(13) Okano, M.; Nishitani, Y.; Sato, M.; Ikekita, A.; Kageyama, S.
Drug Test. Anal. 2010, 2, 548556.
(14) Pinyot, A.; Nikolovski, Z.; Bosch, J.; Such-Sanmartin, G.;
Kageyama, S.; Segura, J.; Gutierrez-Gallego, R. Anal. Bioanal. Chem.
2012, 402, 11011108.
(15) Matsas, R.; Kenny, A. J.; Turner, A. J. Biochem. J. 1984, 223,
433440.
(16) Merkler, D. J. Enzyme Microb. Technol. 1994, 16, 450456.
(17) Tinoco, A. D.; Saghatelian, A. Biochemistry 2011, 50, 7447
7461.
(18) Werle, M.; Bernkop-Schnurch, A. Amino Acids 2006, 30, 351
367.
(19) Katsila, T.; Siskos, A. P.; Tamvakopoulos, C. Mass Spectrom. Rev.
2011, 31, 110133.
(20) Roepstorff, P.; Fohlman, J. Biomed. Mass Spectrom. 1984, 11,
601609.
(21) Touboul, D.; Jecklin, M. C.; Zenobi, R. J. Phys. Chem. B 2007,
111, 1162911631.
(22) Harrison, A. G. Mass Spectrom. Rev. 1997, 16, 201217.
(23) Kim, K. H.; Seong, B. L. Biotechnol. Bioprocess Eng. 2001, 6,
244251.
(24) Thomas, A.; Solymos, E.; Schanzer, W.; Baume, N.; Saugy, M.;
Dellanna, F.; Thevis, M. Anal. Chim. Acta 2011, 707, 107113.

10259

dx.doi.org/10.1021/ac302034w | Anal. Chem. 2012, 84, 1025210259