Analytical Biochemistry 376 (2008) 229–234

Contents lists available at ScienceDirect

Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio

Impurity profiling quality control testing of synthetic peptides using liquid

chromatography-photodiode array-fluorescence and liquid chromatography-
electrospray ionization-mass spectrometry: The obestatin case
Bart De Spiegeleer a,*, Valentijn Vergote a,b, Adel Pezeshki a,c, Kathelijne Peremans b, Christian Burvenich b,c,1
a
Drug Quality and Registration group, Faculty of Pharmaceutical Sciences, Ghent University, Harelbekestraat 72, B-9000 Ghent, Belgium
b
Department of Medical Imaging, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, B-9820 Merelbeke, Belgium
c
Department of Physiology and Biometrics, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, B-9820 Merelbeke, Belgium

a r t i c l e i n f o a b s t r a c t

Article history: Following several conflicting publications, the inability to reproduce the original findings on in vitro
Received 14 December 2007 obestatin binding and activation of GPR39 receptors was recently reported by its discoverers, and several
Available online 23 February 2008 hypotheses to rationalize these findings were proposed. Based on one of these postulations (i.e., presence
of impurities), peptide identity and impurity profiles were thoroughly evaluated on obestatin peptides
Keywords: obtained from five different manufacturers, as used by the different research groups. We found that
Obestatin GPR39 binding one of the products examined was in reality a totally different peptide and that the quality of two-thirds
Peptide quality
of the other peptides was insufficient for in vitro and in vivo experiments (i.e., peptide purity less than 95%
Impurity profile
Synthesis impurities
and/or individual impurities exceeding 1%). These observations question the divergent conclusions
HPLC-PDA reported in the literature about the activity of obestatin. Therefore, we strongly recommend appropriate
Fluorescence quality control testing before using any peptides for biomedical research purposes.
ESI-ion trap MS/MS Ó 2008 Elsevier Inc. All rights reserved.

Obestatin is a recently discovered 23-amino-acid peptide aris-
ing from the same 117-amino-acid preprohormone as ghrelin, in
which it is flanked by potential convertase cleavage sites. This gas-
tric-mucosa-derived peptide has been recognized as an anorexi-
genic (i.e., antagonist of ghrelin), thus explaining the failure to
find impaired growth or appetite in mutant mice with a deletion
of the ghrelin gene [1]. Due to the rising prevalence of obesity,
obestatin currently receives much attention in view of the devel-
opment of drugs for the treatment of eating disorders [2].
The orphan G-protein-coupled receptor GPR39 was originally
proposed as the target receptor of obestatin [1]. However, the
molecular mechanism remains controversial because this has not
been confirmed by other studies [3–5]. Moreover, following these
conflicting publications, the discoverers of obestatin, Zhang et al.,
reported their inability to reproduce the original findings on
in vitro obestatin binding and activation of GPR39 receptors and
proposed several hypotheses to rationalize these results [6]. It
was noted that the initial findings may be attributed to impurities
in obestatin obtained from one of their suppliers. Because radioio-
dine can be incorporated into human and mouse obestatin on two
* Corresponding author. Fax: +32 9 264 8193.
E-mail address: Bart.DeSpiegeleer@UGent.be (B. De Spiegeleer).
1
All authors are members of Phimadran (Physiological Imaging and Drug Analysis)
group.
different amino acid residues in their respective sequences [7,8], it
was also hypothesized that purified obestatin lost bioactivity after
iodination, whereas the contaminated peptide retained some
bioactivity.
Misinterpretation of biological results due to by-products from
peptides synthesis is not uncommon [9,10]. Because the binding
affinities of peptide receptors can be high (i.e., low Kd), even low
concentrations of impurities may have a strong impact on the out-
come of biological assays. Hence, a routine determination of impu-
rity profiles of (synthetic) peptides for research purposes is
recommended because it may be helpful to explain eventual
artifacts.
Even for synthetic peptides for medicinal use, no detailed regu-
latory guidelines or standards currently exist, although some pro-
posals were recently made related to the thresholds, e.g., a
qualification threshold of 1.0% [11,12]. In principle, these peptides
are subject to general guidelines, falling between biotherapeutics
and small molecule drugs [13,14]. For research purposes, where
these compendia and regulatory guidelines are not applicable sen-
su strictu, peptide purity acceptance criteria ranging from not less
than 50 to 95%, are applied, whereas for ligand binding assays,
in vitro bioassays, and in vivo studies, a total impurity level of less
than 5% is generally recommended [10,15]. Most impurities in syn-
thetic peptides are considered to arise from deletion, additional
residue incorporation, truncation, and incomplete deprotection

0003-2697/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2008.02.014
230 Impurity profiling quality control testing of synthetic peptides / B. De Spiegeleer et al. / Anal. Biochem. 376 (2008) 229–234

[10,16–22]. Often, no acceptance criteria on individual impurities obestatin peptides are given in Table 1. The flow rate was set at
are disputably applied during the early research stages. 1.0 mL/min. Under these conditions, the obestatin peptides elute
The aim of the study described here was to investigate the qual- at approximately 40 min.
ity of commercially available obestatin synthetic peptides, as used Size-exclusion chromatography for confirmation of high-molec-
by the different obestatin research groups reporting conflicting ular-weight oligomeric impurities in obestatin samples was per-
conclusions, by liquid chromatography with UV, native fluores- formed using a BioSep-SEC-S2000 (300 mm  7.8 mm I.D., 300 Å,
cence, and electrospray ion trap mass spectrometry detection. 5 lm particle size) column (Phenomenex, Torrance, CA, USA) with
a mixture of 17 volumes of aqueous formate buffer (50 mM, pH
Materials and methods 2.5) and 3 volumes of acetonitrile as mobile phase. The flow rate
was set at 0.75 mL/min. Under these conditions, the obestatin pep-
Materials tides and their dimeric impurities elute at approximately 12.8 and
11.5 min, respectively.
Human obestatin was obtained from five different suppliers:
California Peptide Research (Napa, CA, USA), GL Biochem (Shang-
Results
hai, China), Global Peptide Services (Fort Collins, CO, USA), Penta
Biotech (Union City, CA, USA), and Phoenix Pharmaceuticals (Bur-
All peptides were supplied by their respective manufacturers as
lingame, CA, USA) [coded CPR, GLB, GPS, PB, and PP, respectively].
lyophilized powder in clear glass or polypropylene containers to-
Mouse obestatin was obtained from four different suppliers,
gether with a certificate of analysis. Each material was properly la-
[coded CPR, GLB, GPS, and PP]. Bovine, canine, ovine, and porcine
beled (i.e., product description, name of manufacturer, and batch
obestatin were obtained from PP only.
number) and stored in its original packaging material at 35 °C un-
HPLC- gradient-grade acetonitrile was purchased from Fisher
til analysis. All suppliers stated the peptide purity on their respec-
Scientific (Leicestershire, UK). LC-MS-grade formic acid and triflu-
tive certificates of analysis (see Table 2). However, the exact
oroacetic acid were purchased from Fluka (Buchs, Switzerland).
peptide content or net amount of peptide (both based upon amino
Water was purified using an Arium 611 purification system (Sarto-
acid analysis) was given only for obestatins from CPS and PP. The
rius, Göttingen, Germany).
other three suppliers could provide only an estimate for peptide
content.
Peptide lyophilization
Prior to analysis, the peptide samples were dissolved in acetoni-
Peptide identification
trile/water 5:95 (v/v) containing 0.1% (w/v) trifluoroacetic acid;
40-nmol aliquots were dispensed into low-volume polypropylene
All peptides were positively identified based on their respective
HPLC vials and lyophilized using a Lyovac GT4 precooled shelve
mass spectra (i.e., molecular mass and typical y- and b-type frag-
freeze-dryer (Leybold, Cologne, Germany).
ment ions in tandem mass spectra after collision-induced dissoci-
ation), except for mouse obestatin obtained from GLB.
Liquid chromatography
RP-HPLC analysis of this latter peptide revealed a main peptide
The HPLC-PDA/FL2 apparatus consisted of a Waters Alliance 2695
closely eluting at the normal retention time of mouse obestatin.
separations module, a Waters 2996 photodiode array detector, and a
However, the UV, fluorescence, and mass spectra obtained on the
Waters 2475 multiwavelength fluorescence detector with Empower
main peak indicated that the peptide did not correspond to the
2 software for data acquisition (all Waters, Milford, MA, USA). For
mouse obestatin amino acid sequence: absence of tyrosine residue
PDA detection, the UV spectrum was recorded at 190–400 nm. For
(i.e., no absorbance maximum at approximately 275 nm in UV
fluorescence detection, the excitation wavelength was set at
spectrum and no native fluorescence observed for kEX = 230 nm
230 nm with emission spectrum recorded at 240–400 nm.
with kEM = 296 nm) and a monoisotopic molecular mass of approx-
The LC-UV/MS apparatus consisted of a Spectra System SN4000
imately 1032.5 evidenced by peaks at m/z 517.3 and 1033.5 in
interface, a Spectra System SCM1000 degasser, a Spectra System
mass spectrum (i.e., [M+2H]2+ and [M+H]1+, respectively).
P1000XR pump, a Spectra System AS3000 autosampler, and a Finn-
Using a Procise Model 494 HT Edman sequencing system (Ap-
igan LCQ Classic ion trap mass spectrometer in positive ion mode
plied Biosystems, Foster City, CA, USA; by Alta Bioscience, Birmin-
(all Thermo, San José, CA, USA) equipped with a SPD-10A UV-VIS
ghan, UK), the N-terminal amino acid sequence of this incorrect
detector (Shimadzu, Kyoto, Japan) and Xcalibur 2.0 software (Ther-
peptide was determined to be SIIKFEKL. Using tandem mass spec-
mo) for data acquisition. ESI was conducted using a needle voltage
trometry, the sequence of this product was confirmed as: SIIK-
of 4.5 kV. Nitrogen was used as the sheath and auxiliary gas with
FEKLG-NH2, i.e., a nonapeptide totally unrelated to obestatin.
the heated capillary set at 250 °C. ESI mass spectra were deconvo-
luted manually and checked afterward using ProMass Deconvolu-
Impurity profile
tion 2.5 software (Thermo). Tandem mass spectrometry (MS/MS)
data on impurities were examined using SEQUEST seach in Bio-
Several peptide-related impurities were encountered in the
Works 3.3.1 software (Thermo) based upon FASTA database files
products tested, most of them eluting between 19 and 51 min.
developed for the individual obestatin peptides (which include
For each peptide, the impurity concentrations were determined
deletions, additional residue incorporations, and truncations).
by area normalization using UV detection at 195 nm. Also, in UV,
LC determination of impurities in obestatin samples was per-
the peptide counter-ions (e.g., trifluoroacetate) were clearly visible
formed using an Everest C18 238EV54 (250 mm  4.6 mm I.D.,
as early eluting peaks (at approximately 5 min). Some typical chro-
300 Å, 5 lm particle size) column (Grace Vydac, Hesperia, CA,
matograms are shown in Fig. 1. The impurity results obtained on
USA) in an oven set at 30 °C, with a mobile phase consisting of
each peptide (using a reporting threshold of 0.1%) are summarized
(A) 0.1% w/v formic acid in water and (B) 0.1% w/v formic acid in
in Table 2.
acetonitrile. The gradient programs employed for the different
Fluorescence detection (kEM = 296 nm) could be used only for
obestatin peptides containing a tyrosine amino acid residue: i.e.,
2 canine, human and mouse obestatin only. In these cases, peptide
Abbreviations used: PDA, photodiode array; FL, flourescence; ESI, electrospray
ionization. purity results comparable to those for UV detection were obtained,
 Impurity profiling quality control testing of synthetic peptides / B. De Spiegeleer et al. / Anal. Biochem. 376 (2008) 229–234 231

Table 1
Peptide information and gradient programs

Peptide Sequence Monoisotopic molecular Time (min) Typical retention

mass [expected m/z values] time (min)
0 6 66 96
Bovine obestatin FNAPFNIGIKLAGAQSLQHGQTL-NH2 2423.30746 [1212.7 and 808.8] 95% A + 5% B 86% A + 14% B 76% A + 24% B 5% A + 95% B 40.2
Canine obestatin FNAPFDVGIKLSGPQYHQHGQAL-NH2 2522.28197 [1262.1 and 841.8] 89% A + 11% B 79% A + 21% B 39.4
Human obestatin FNAPFDVGIKLSGVQYQQHSQAL-NH2 2545.30785 [1273.7 and 849.4] 89% A + 11% B 79% A + 21% B 36. 3
Mouse obestatin FNAPFDVGIKLSGAQYQQHGRAL-NH2 2515.30853 [1258.7 and 839.4] 86% A + 14% B 76% A + 24% B 40.5
Ovine obestatin FNAPFNIGIKLSGAQSLQHGQTL-NH2 2439.30237 [1220.7 and 814.1] 86% A + 14% B 76% A + 24% B 39.9
Porcine obestatin FNAPCDVGIKLSGAQSDQHGQPL-NH2 2380.15948 [1191.1 and 794.4] 91% A + 9% B 81% A + 19% B 44.4

Table 2
Peptide purity specifications and results

Peptide Manufacture Peptide purity (%) Number of impurities at P 1.0% Largest impurity (%) [RRT]
Specification Result
COA COA QC QC
Bovine obestatin PP P 97 98.70 88.32 3 3.43 [0.94]
Canine obestatin PP P 95 100.00 79.21 9 3.31 [1.08]
Human obestatin CPR – 98.80 99.66 0 0.19 [0.91]
GLB P 95 95.37 92.74 3 2.65 [0.88]
GPS P 95 95.86 93.79 3 1.67 [0.91]
PB – 99.69 99.49 0 0.17 [0.97]
PP P 96 96.75 93.18 4 1.99 [0.97]
Mouse obestatin CPR – 99.96 99.88 0 0.12 [0.97]
GLB P 95 98.94 N/A N/A N/A
GPS P 95 98.40 96.41 0 0.71 [0.90]
PP P 95 100.00 80.82 3 7.92 [1.09]
Ovine obestatin PP P 95 100.00 81.14 5 6.47 [0.92]
Porcine obestatin PP P 95 100.00 97.74 0 0.48 [0.87]

COA, values as given in the certificates of analysis issued by the respective manufacturers using their in-house HPLC methods; N/A, not applicable because the GLB peptide did
not correspond to mouse obestatin; QC, quality control results obtained using UV detection at 195 nm [reporting threshold = 0.1%]; RRT, relative retention time vs. obestatin
peak.

Using mass spectrometry, the molecular masses of the impuri-

Fig. 1. Typical chromatograms (UV at 195 nm) obtained on human obestatin pep-
tides using the chromatographic conditions described in Section 2.2.
although late-eluting peaks (i.e., after 66 min) can be detected
more easily with fluorescence due to the strongly increasing UV
absorbance as the percentage amount of acetonitrile in the mobile
phase rises.
ties were determined: almost all of these masses were between
706.1 and +137.1 (i.e., addition of one histidine residue) com-
pared to their respective target peptide. Although most of the tan-
dem mass spectra obtained on the impurities are considered
suitable for, e.g., de novo sequencing, it is desirable to know details
on the synthetic process followed in each case to unambiguously
identify these by-products. For each obestatin peptide, the most
abundant impurity above the 1.0% qualification threshold was
identified based upon molecular mass, database search on tandem
mass spectrometry (MS/MS) data, and/or Edman N-terminal
sequencing: see Table 3.
In some peptides, impurities with higher molecular masses
were found to elute after 66 min (i.e., after shallow gradient):
e.g., canine obestatin impurities at relative retention times of
1.79 and 1.80 corresponding to molecular masses of 5061.2 and
5028.1 (i.e., dimeric obestatins), respectively (see Fig. 2). However,
in other obestatin materials, similar dimeric peaks were either ab-
sent or below the 0.1% reporting limit using UV detection, e.g.,
mouse obestatin from CPR and GPS and human obestatin from
PB and GPS, for which these higher-molecular-weight peptides
could be observed only by using fluorescence and MS detection.
The differences between the molecular masses of the dimeric
impurities and twice the mass of the respective monomers re-
vealed that there are several types of dimers (e.g., for canine obest-
atin: +17 and 17). However, no clear sequence information could
be obtained.
For some obestatin samples, the degradation products produced
by acid-catalyzed cleavage after the aspartic acid amino acid resi-
due at position 6 (i.e., FNAPxD and heptadecapeptide) were also
observed, although below the 0.1% reporting limit.
232 Impurity profiling quality control testing of synthetic peptides / B. De Spiegeleer et al. / Anal. Biochem. 376 (2008) 229–234

Table 3
Characterization and identification of largest impurities above the 1.0% qualification threshold

Peptide Manufacturer m/z values (MS) Molecular mass Proposed sequencea

Bovine obestatin PP 771.1 and 1156.0 2310.1 FNAPFIGIKLAGAQSLQHGQTL-NH2 [Deletion of N6]
Canine obestatin PP 632.2 and 842.4 and 1262.8 2524.1 FNAPFDVGIKLSGPQYHQHGQAL [Desamidated peptide]
Human obestatin GLB 773.3 and 1159.9 2317.2 FAPFVGIKLSGVQYQQHSQAL-NH2 [Des-N2,D6 peptide]
GPS 811.4 and 1216.4 2431.5 FNAPFVGIKLSGVQYQQHSQAL-NH2 [Deletion of D6]
PP 678.8 and 1017.7 2033.4 FNAPFLSGVQYQQHSQAL-NH2 [Deletion of DVGIK]
Mouse obestatin PP 630.4 and 840.1 and 1259.4 2517.2 FNAPFDVGIKLSGAQYQQHGRAL [Desamidated peptide]
Ovine obestatin PP 776.3 and 1163.9 2325.6 FNAPFIGIKLSGAQSLQHGQTL-NH2 [Deletion of N6]
a
Based upon calculated molecular mass and database search on MS2 data and/or Edman sequencing.

1265.7 - [M+4H]4+ 1258.2 - [M+4H]4+

100 100

95 95

90 90

85 85

80 80

75 75

70 70

65 65
Relative Abundance

Relative Abundance

60 60

55 55
844.2 - [M+6H]6+
839.2 - [M+6H]6+
50 50

45 45

1012.4 - [M+5H]5+ [M+7H]7+ - 719.5

40 40

35 35
[M+7H]7+ - 723.5 1006.7 - [M+5H]5+

30 30

25 25

20 20

15 15 1677.9 - [M+3H]3+

[M+8H]8+ - 633.2 1688.0 - [M+3H]3+

10 10

5 5

0 0
500 1000 1500 2000 500 1000 1500 2000
m/z m/z

Fig. 2. Mass spectra of dimeric peptides found in canine obestatin from PP using RP-HPLC: (left) 5059-Da impurity at 70.8 min and (right) 5029-Da impurity at 71.9 min.

Discussion gradient programs on a C18 column with a mobile phase consisting

of water and acetonitrile with 0.1% trifluoroacetic acid: standard
As a response to comments on their work, the discoverers of gradient slopes of 0.5 to 2% acetonitrile per minute were applied.
obestatin, Zhang et al., recently admitted not being able to repro- To maximize the resolution of the HPLC method, a very shallow
duce their original finding on the in vitro binding and suggest that gradient was chosen for thorough quality control testing of impu-
the initial results may be caused by impurities in the synthesized rities, 0.17% acetonitrile per minute, while still retaining a total
peptide used for these experiments [6]. Following this hypothesis, chromatographic analysis time of 60 min. This 60 min gradient
a total of 13 obestatin peptides from five different manufacturers, program for the elution of obestatin and its most common impuri-
which are known to have supplied materials to obestatin research ties is adapted for each type of obestatin with the target peptide
groups, were screened for the presence of related peptide impuri- eluting in the middle. This shallow gradient is preceded by a short
ties using reversed-phase HPLC with UV, fluorescence, and ESI-MS/ gradient of 5 min starting from 5% acetonitrile to elute the peptide
MS detection. counter-ion. Afterward, a steep gradient to 95% acetonitrile over
The data given in the certificates of analysis issued by the man- 30 minutes was applied to check for apolar peptide impurities
ufacturers suggest that all 13 obestatin peptides under examina- and column cleaning.
tion are suitable for most in vitro and in vivo experiments. The QC analyses on the 13 obestatin peptides revealed that 1 of
peptide purities were determined by all manufacturers using linear these materials (mouse obestatin from GLB) is in fact a totally
 Impurity profiling quality control testing of synthetic peptides / B. De Spiegeleer et al. / Anal. Biochem. 376 (2008) 229–234 233

12.9
100 1277.2 1274.0
100 100
90
95 95
80
90 90
Relative Abundance

70
85 85
60
80 80
849.7
50
75 75
40
70 70
30
65 65

Relative Abundance

Relative Abundance
20
60 60
10
11.5 55 55

12.8
50 50
0.70
0.65 45 45
0.60
40 40
0.55
0.50 35 35

0.45
30 30
0.40
uAU

0.35 25 25
1021.8 1702.2
0.30 20 20
0.25
15 15
0.20
0.15 10 10 1697.9
851.8
0.10
5 5
0.05
11.5
0.00 0 0
10 11 12 13 14 500 1000 1500 2000 500 1000 1500 2000
Time (min) m/z m/z

Fig. 3. Size exclusion chromatography LC-MS analysis of GPS human obestatin: (left) chromatogram showing dimeric and monomeric obestatin at 11.5 and 12.9 min,
respectively; (middle) mass spectra of dimeric obestatin with calculated monoisotopic molecular mass of 5104.3 Da; and (right) mass spectra of monomeric obestatin with
calculated monoisotopic molecular mass of 2546.1 Da.

unrelated peptide, although this was not observed on HPLC reten-
tion with UV detection only, whereas only 4 peptides comply with
a peptide purity of not less than 95%. For these latter products,
maximum individual impurity levels ranging from 0.12 to 0.48%
were found, hence, complying with the generally approved accep-
tance criterion of not more than 1.0% for individual impurities.
Labeled as mouse obestatin, the peptide supplied by GLB (prob-
ably derived from ovalbumin) is clearly the result of a mix-up. This
finding clearly demonstrates the need for in-depth quality control
testing on peptides for biomedical research purposes.
When the impurity profiles obtained on human obestatin pep-
tides from different manufacturers are compared, the following
similarities are noted: impurities at relative retention times 0.89
and 0.94 for GLB vs PP (2317 & 2269 Da, respectively), impurity
at relative retention time 0.91 for GLB vs GPS (2431 Da), and impu-
rity at relative retention time 0.97 for GLB vs PB (2390 Da). For the
mouse obestatin peptides, the only similarity noted was the impu-
rity at a relative retention time of 0.98 for CPR vs PP (2129 Da).
When comparing the impurity profiles obtained on the different
obestatin peptides supplied by one manufacturer, again several
similarities are noted.
Interestingly, in some peptides, impurities with molecular
masses approximately twice that of the target peptide were found
eluting at a relative retention time of approximately 1.8, which
corresponds to about double the amount of acetonitrile in the mo-
bile phase vs the main peak. However, in most products tested, the
concentrations of these dimeric impurities (estimated using fluo-
rescence detection) do not rise above 1.0%, except for canine obest-
atin from PP where 1.7% dimers was found. The presence of these
high-molecular-weight impurities was unequivocally confirmed
by size-exclusion chromatography as shown for human obestatin
from GPS (Fig. 3).
Our findings indicate that there is a good possibility that one or
more of the impurities found may bind onto GPR39 receptors,
explaining the original findings by Zhang et al. on the in vitro bind-
ing [10]. The lack of consistent peptide quality may thus shed new
light on the discussion about the in vivo effects of obestatin [23].
Conclusions
Our results confirm that obestatin peptides supplied by relevant
manufacturers can be unsuitable for in vitro and in vivo testing as
even a mix-up with another peptide may occur. Overall, the quality
of more than half of the obestatin peptides supplied by the five
manufacturers was considered insufficient for in vitro and in vivo
experiments. These findings demonstrate the absolute necessity
of appropriate quality control testing and specification settings
on peptides for biomedical research purposes.
Acknowledgment
We thank Chris Vervaet (UGent) for the use of the Lyovac.
234 Impurity profiling quality control testing of synthetic peptides / B. De Spiegeleer et al. / Anal. Biochem. 376 (2008) 229–234

References [11] EDQM: Technical Guide for the Elaboration of Monographs on Synthetic
Peptides & Recombinant DNA Proteins, 2006.
[12] EDQM: Synthetic Peptides: Monograph specifications: EDQM Symposium:
[1] J.V. Zhang, P.G. Ren, O. Avsian-Kretchmer, C.W. Luo, R. Rauch, C. Klein, AJ.W.
New impurities control, Strasbourg, 22 September 2006.
Hsueh, Science 310 (2005) 996–999.
[13] ICH Harmonised Tripartite Guideline Q6B ‘Specifications: Test procedures and
[2] G.J. Lagaud, A. Young, A. Acena, M.F. Morton, T.D. Barrett, N.P. Shankley,
acceptance criteria for biotechnology/biological products’ dated 10 March 1999.
Biochem. Biophys. Res. Commun. 357 (2007) 264–269.
[14] ICH Harmonised Tripartite Guideline Q3A(R2) ‘Impurities in new drug
[3] E. Lauwers, B. Landuyt, L. Arckens, L. Schoofs, W. Luyten, Biochem. Biophys.
substances’ dated 25 October 2006.
Res. Commun. 351 (2006) 21–25.
[15] G. Schwach, N. Oudry, S. Delhomme, M. Lück, H. Lindner, R. Gurny, Eur. J.
[4] B. Holst, K.L. Egerod, E. Schild, S.P. Vickers, S. Cheetham, L.O. Gerlach, L.
Pharm. Biopharm. 56 (2003) 327–336.
Storjohann, C.E. Stidsen, R. Jones, A.G. Beck-Sickinger, T.W. Schwartz,
[16] V. Sanz-Nebot, F. Benavente, I. Toro, J. Barbosa, Anal. Chim. Acta 521 (2004)
Endocrinology 148 (2006) 13–20.
25–36.
[5] N. Chartrel, R. Alvear-Perez, J. Leprince, X. Iturrioz, A. Reaux-Le Goazigo, V.
[17] V. Sanz-Nebot, F. Benavente, I. Toro, J. Barbosa, Anal. Bioanal. Chem. 377 (2003)
Audinot, P. Chomarat, F. Coge, O. Nosjean, M. Rodriguez, J.P. Galizzi, J.A. Boutin,
306–315.
H. Vaudry, C. Llorens-Cortes, Science 315 (2007) 766.
[18] K. Hettiarachchi, S. Ridge, D.W. Thomas, L. Olson, C.R. Obi, D. Singh, J. Peptide
[6] J.V. Zhang, C. Klein, P.G. Ren, S. Kass, L. Verdonck, D. Moechars, A.J. Hsueh,
Res. 57 (2001) 151–161.
Science 315 (2007) 766.
[19] V. Sanz-Nebot, F. Benavente, A. Castillo, J. Barbosa, J. Chromatogr. A 889 (2000)
[7] V. Vergote, S. Bodé, K. Peremans, H. Vanbree, B. Baert, G. Slegers, C. Burvenich,
119–133.
B. De Spiegeleer, J. Chromatogr. B 850 (2007) 213–220.
[20] D. Yang, R. Wang, Y. Ma, Y. Di, X. Huo, Int. J. Pept. Res. Ther. 6 (1999) 391–
[8] V. Vergote, B. Baert, E. Vandermeulen, K. Peremans, H. van Bree, G. Slegers, C.
394.
Burvenich, B. De Spiegeleer, J. Pharm. Biomed. Anal. 46 (2008) 127–136.
[21] V. Sanz-Nebot, A. Garcés, J. Barbosa, J. Chromatogr. A 833 (1999) 267–
[9] J.W. de Beukelaar, J.W. Gratama, P.A. S Smitt, G.M. Verjans, J. Kraan, T.M.
275.
Luider, P.C. Burgers, Rapid Commun. Mass Spectrom. 21 (2007) 1282–
[22] P.A. D’Agostino, J.R. Hancock, L.R. Provost, P.D. Semchuk, R.S. Hodges, J.
1288.
Chromatogr. A 800 (1998) 89–100.
[10] J.W. Metzger, C. Kempter, K.H. Wiesmüller, G. Jung, Anal. Biochem. 219 (1994)
[23] A. Garg, J. Clin. Endocr. Metab. 92 (2007) 3396–3398.
261–277.