Journal of Pharmaceutical and Biomedical Analysis 62 (2012) 191195

Contents lists available at SciVerse ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis

journal homepage: www.elsevier.com/locate/jpba

Short communication

Identication of ve gelatins by ultra performance liquid

chromatography/time-of-ight mass spectrometry (UPLC/Q-TOF-MS) using
principal component analysis
Xian-Long Cheng a,b , Feng Wei a , Xin-Yue Xiao a, , Ying-Yong Zhao c , Yan Shi a , Wei Liu a ,
Ping Zhang a , Shuang-Cheng Ma a , Shou-Sheng Tian d , Rui-Chao Lin a,b,
a
Research and Inspection Center of Traditional Chinese Medicine and Ethnomedicine, National Institutes for Food and Drug Control,
State Food and Drug Administration, 2 Tiantan Xili, Beijing 100050, China
b
School of Chinese Pharmacy, Beijing University of Chinese Medicine, No. 6 Wangjing Zhong Huan Nan Lu, Chaoyang District, Beijing 100102, China
c
Biomedicine Key Laboratory of Shaanxi Province, Northwest University, No. 229 Taibai North Road, Xian, Shaanxi 710069, China
d
Shandong Dong E E Jiao Co., Ltd., Liaocheng 252201, China

a r t i c l e

i n f o

Article history:
Received 26 September 2011
Received in revised form
19 December 2011
Accepted 19 December 2011
Available online 29 December 2011
Keywords:
UPLC/Q-TOF-MS
Gelatin
Polypeptides
PCA
Identication

a b s t r a c t
An ultra-performance liquid chromatography-quadrupole-time of ight mass spectrometry (UPLC/QTOF-MS) method coupled with a principal component analysis (PCA) was developed and applied toward
identifying donkey-hide gelatin, bovine-hide gelatin, pig-hide gelatin, tortoise shell glue, and deerhorn
glue. The UPLC-MS data of the trypsin digested samples were subjected to principal component analysis
(PCA) in order to classify these ve gelatins. Additionally, marker peptides given by the loadings plot of
PCA were identied based on a comparison of recorded LCMS data with a previously reported database
of the corresponding gelatin variants. The results from this study indicate that the proposed method is
reliable, and it has been successfully applied to the identication of variants of gelatins commonly used
in Traditional Chinese Medicine (TCM).
2011 Elsevier B.V. All rights reserved.

1. Introduction
Gelatin is a mixture of high molecular weight polypeptides from
the hydrolysis of collagen, and it is widely used in food, cosmetic,
and pharmaceutical industries [13]. There are many sources of collagen for gelatin production, i.e., bovine, porcine, and sh skin. In
China, donkey-hide gelatin, bovine-hide gelatin, pig-hide gelatin,
glue of tortoise shell, and deerhorn glue are widely used as Traditional Chinese Medicines (TCM) [46].
In Islamic countries, porcine gelatin is forbidden, and in some
cases, bovine gelatin carrying prion proteins associated with bovine
spongiform encephalopathy should be identied. An immune
response will occur when gelatins from certain sources are used
incorrectly [7]. These gelatins are believed to produce different

Corresponding author. Tel.: +86 10 67095432; fax: +86 10 67023650.

Corresponding author at: Research and Inspection Center of Traditional Chinese
Medicine and Ethnomedicine, National Institutes for Food and Drug Control, State
Food and Drug Administration, 2 Tiantan Xili, Beijing 100050, China.
Tel.: +86 10 67095432; fax: +86 10 67023650.
E-mail addresses: xiaoxxy@sina.com (X.-Y. Xiao), linrch307@sina.com (R.-C. Lin).
0731-7085/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jpba.2011.12.024

clinical effects, and they command different prices on the open

market in China. Therefore, it is important to complete comprehensive studies on identifying and distinguishing these ve gelatins.
These ve gelatins have similar chemical properties because
they are homologous [8], and it is difcult to distinguish them
using conventional spectroscopic methods [9]. The immunochemical method has been used to identify collagen [10], but this method
may be inuenced by the extent of proline hydroxylation, which
plays an important role in determining the antigenicity of collagen [11]. The polymerase chain reaction (PCR) method is not used
in gelatin identication because DNA integrity is destroyed during
gelatin processing, although this method has been widely applied
in collagen identication [12,13]. To the best of our knowledge,
amino acids from the hydrolysis of collagen were analyzed using
HPLC coupled with PCA to classify bovine and porcine gelatin, but
amino acids do not reveal the properties of primary gelatins. Collagen marker peptides, combined with sequence searching, allow
differentiation of bovine and porcine gelatin [8,14,15].
Time of ight mass spectrometry (TOF) has been developed as
an efcient method for protein analysis [1619]. To the best of our
knowledge, there has been little use of UPLC QTOF MS and PCA
analysis for the identication of these ve gelatins. In this work,

192

X.-L. Cheng et al. / Journal of Pharmaceutical and Biomedical Analysis 62 (2012) 191195

we analyzed tryptic peptides of gelatins using UPLC/Q-TOF-MS and

used principal component analysis (PCA) to classify donkey-hide
gelatin, bovine-hide gelatin, pig-hide gelatin, tortoise shell glue,
and deer-horn glue.

was used to obtain the precursor ion (MS) and their fragmentation
data (MSE ). All the acquisition and analysis of data were controlled
using Waters MassLynx v 4.1 software.
2.5. Multivariate data analyses

2. Experimental
2.1. Materials and reagents
Leucine enkephalin and formic acid were purchased from
SigmaAldrich (St. Louis, MO, USA). Acetonitrile (HPLC grade) was
purchased from J.T. Baker (Baker Mallinckrodt, Phillipsburg, NJ,
USA). Methanol (HPLC grade) was purchased from Fisher Scientic
(Pittsburgh, PA, USA). Ultra high purity water was prepared using
a Milli-Q water purication system (Millipore Corporation, Bedford, MA). Trypsin (sequencing grade) was obtained from Promega
(Madison, WI, USA). Syringe lters (0.22 m) were purchased from
Millipore (Billerica, MA, USA). All other chemicals used were of analytical grade. A total of 25 gelatin samples were collected by the
National Institute for Food and Drug Control. All samples were prepared in the laboratory by decoction at 110 C for 2 h in water and
concentration to a solid at 100 C. The three unknown samples were
provided by the manufacturer and were prepared under severe
conditions by decoction at 120 C for 4 h in an aqueous solution
of 0.6% NaCO3 and concentration to a solid at 100 C.
2.2. Sample preparation
First, 100 mg of the gelatin standard was dissolved in 50 mL of
a 1% NH4 HCO3 solution (pH 8.0). The solution was ltered through
a 0.22 m syringe lter. Then, 100 L of the gelatin solution was
drawn, and 10 L of trypsin solution (1 mg/ml in 1% NH4 HCO3 , pH
8.0) was added. The mixture was incubated at 37 C for 12 h.
2.3. Chromatographic separation
The UPLC analysis was performed using a Waters AcquityTM
Ultra Performance LC system (Waters MS Technologies, Manchester, UK). Chromatographic separation was carried out on an
ACQUITY UPLC BEH C18 column (100 mm 2.1 mm, 1.7 m) at a
column temperature of 45 C. The mobile phase consisted of solvent
A (0.1% formic acid in water, v/v) and solvent B (acetonitrile). The
optimized UPLC elution conditions were as follows: 0.025.0 min,
5%20% B; 25.040.0 min, 2050% B; 40.041.0 min, 5099% B;
40.045.0 min, 99% B; 45.045.1 min, 99.05.0% B; 45.155.0 min,
5.0%. The ow rate was set at 300 L/min. The column and autosampler were maintained at 40 C and 10 C, respectively. The injection
volume was 5 L.
2.4. Mass spectrometry
Mass spectrometry was performed on a XevoTM QTof (Waters
MS Technologies, Manchester, UK). The scan range was from 50 to
2000 m/z. For positive electrospray modes, the capillary and cone
voltage were set at 3.0 kV and 30 V, respectively. The desolvation
gas was set to 600 L/h at a temperature of 450 C, and the cone gas
was set to 50 L/h, and the source temperature was set to 150 C.
The MCP detector voltage was set at 2200 V. Data were collected
in continuum mode, in which the data acquisition rate was set
to 0.1 s, with a 0.1 s interscan delay. All analyses were acquired
using the lock spray feature to ensure accuracy and reproducibility. Leucineenkephalin at a concentration of 2 g/mL and a ow
rate of 3 L/min was used for the lock mass. The lock spray frequency was set at 10 s, and the data were averaged over 10 scans.
An alternation low-energy (collision cell energy of 4 V) and elevated
energy (collision cell energy ramped from 20 to 30 V) acquisition

The raw data were analyzed by the MarkerLynx applications

manager version 4.1 (Waters, UK). The parameters used were retention time (RT) in the range of 140 min, mass in the range of
502000 Da, mass tolerance 0.05 Da, noise elimination level was
set at 6.00, intensity threshold 100 counts, mass window 0.05 amu,
and retention time window of 0.2 min. No specic mass or adduct
was excluded. A list of the intensities of the peaks detected was
generated using RT and m/z data pairs. Ions in different samples
were considered to be the same ion when they demonstrated the
same RT (tolerance of 0.2 min) and m/z value (tolerance of 0.05 Da).
The ion intensities for each detected peak were normalized against
the sum of the peak intensities within that sample using MarkerLynx. The resulting three-dimensional data comprising peak number
(RT m/z pair), sample name (observations), and ion intensity were
analyzed by unsupervised principal component analysis (PCA). All
variables were pareto-scaled prior to analysis.
3. Results and discussion
3.1. Optimization of chromatographic separation and mass
spectrometry
An earlier study reported proling of tryptic peptides in bovine
and porcine gelatin with a 120 min analysis method [14]. In our
study, a 55 min UPLC QTOF MS tryptic peptides proling assay was
reported for the proling of gelatins using step-gradients, which
signicantly shortens the time of sample analysis without compromising chromatographic peak resolution. The signicant reduction
in analysis time was a result of a combination of the enhanced
selectivity of TOF/MS detection and the enhanced chromatographic
resolution due to the use of a sub-2 m particle column. This
method could reduce the severe ion suppression, although the complexity of the tryptic gelatin peptides makes separation difcult.
The peptides are amphiprotic substances and can be ionized in ESI+
and ESI mode, but sparkover will often occur at a high capillary
voltage in the ESI mode. Therefore, ESI+ was selected as the ionization mode for the TOF-MS experiments. The desolvation gas ow
was set at 600 L/h, and desolvation temperature was set at 450 C to
remove the abundant solvent that results from a higher proportion
of water in the mobile phase. A cone voltage of 30 V was selected
for use after optimizing the cone voltage in the range of 1545 V
to obtain a higher sensitivity. The data acquisition rate of 200 ms
was used to ensure that the chromatographic peaks are dened by
no less than 10 data points [20]. The comparison of the ve base
peak ion (BPI) chromatograms obtained from the ve tryptic samples is shown in Fig. 1. The UPLC-QTOF-MS system proved to be
robust for the retention time shift and the mass accuracy based on
the extracted ion chromatographic peaks of eight ions (402.2120,
469.2430, 473.2303, 802.4034, 672.3723, 765.8712, 714.3438, and
737.9980) were <0.02 min and 10 ppm, respectively.
3.2. Multivariate data analyses
It appears that the marker peptides of each gelatin were concealed within a larger number of tryptic peptides that were
indistinguishable at higher concentrations because of the homology of collagen in TIC chromatography. It is difcult to distinguish
each gelatin by visual observation of their chromatograms; therefore, further sample proling of the gelatins requires the use of
multivariate statistical tools. In our study, the rst step of the

X.-L. Cheng et al. / Journal of Pharmaceutical and Biomedical Analysis 62 (2012) 191195

95
%

(A)

193

-5
10.00

15.00

20.00

25.00

30.00

35.00

40.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

95
%

(B)

5.00

-5

95
%

(C)

-5

95
%

(D)

-5

95
%

(E)

-5

Time

Fig. 1. Positive ion base peak intensity chromatograms obtained from the analysis of tryptic digest of (A) donkey-hide gelatin, (B) bovine-hide gelatin, (C) pig-hide gelatin,
(D) glue of tortoise shell and (E) deerhorn glue.

multivariate statistical analysis of the UPLC/MS dataset was to convert the 3D LC/MS data into a 2D matrix expressed as an Exact
Mass Retention Time (EMRT) pair using Markerlynx, which is an
Application Manager for the MassLynxTM Software. The data set
was visualized using unsupervised PCA to check for outliers and
classication trends among the gelatins. Preliminary PCA was

Scores Comp[1] vs. Comp[2] colored by Sample Group

(A)
100

Loadings Comp[1] vs. Comp[2]

(B)
0.08

50

t[2]

performed on all observations using 8556 pareto-scaled variables.

The nal PCA score plot demonstrated that the ve different types
of gelatins cluster together, and all were found to lie inside the
Hotelling T2 (0.95) ellipse, as is shown in Fig. 2A. In the scores plot
obtained by PCA, donkey-hide gelatin, bovine-hide gelatin, pig-hide
gelatin, deer-horn glue are close to each other, and the tortoise shell

unknown

0.04

p[2] 0.00

unknown

-50

-0.04

-100

-0.08

-100

-50

t[1]
Fig. 2. PCA score plot (A) and loading plot (B) of (

50

100

-0.04

) donkey-hide gelatin, () bovine-hide gelatin, ( ) pig-hide gelatin, (

0.00

0.04

p[1]
) deerhorn glue and (

) glue of tortoise shell.

194

X.-L. Cheng et al. / Journal of Pharmaceutical and Biomedical Analysis 62 (2012) 191195

Fig. 3. Selected ion monitoring chromatograms of marker peptides in (A) donkey-hide gelatin, m/z 765.8556, doubly-charged ion of fragment GEAGPAGPAGPIGPVGAR. (B)
bovine-hide gelatin, m/z 641.3065, doubly-charged ion of fragment GEAGPSGPGPTGAR. (C) pig-hide gelatin, m/z 925.4326, doubly-charged ion of fragment GEPGPTGVQGPPGPAGEEGK. (D) glue of tortoise shell m/z 758.3530, sequence unknown. (E) deerhorn glue m/z 732.8282, sequence unknown.

glue is located much farther from these four. This is believed to be

because donkey, bovine, pig, and deer are all mammals, while the
tortoise is a reptile. The unknown samples were clustered in the
corresponding area, as expected.
3.3. Identication of the marker peptides
Results from this study demonstrate that it is possible to identify the markers that play an important role in the classication
of gelatins. The loading plot from the PCA based on UPLC/MS
data (8556 variables) is shown in Fig. 2B. A number of ions with
RT m/z pairs of 15.59 765.8665, 4.65 641.3065, 8.49 925.4326,
16.16 758.8589 and 8.53 732.8282 were chosen as marker peptides
for each gelatin. To dene the marker peptides, the sequence of
donkey collagen, bovine collagen, porcine collagen were obtained
from the swiss-prot database (http://www.expasy.org) and Biopharmalynx 1.2 was used to identify the sequence of the marker

peptides of each gelatin, which was accomplished using. The ion

at m/z 765.8556, a doubly charged ion of the fragment GEAGPAGPAGPIGPVGAR, is the marker of donkey-hide gelatin. The ion at m/z
641.3065, a doubly charged ion of the fragment GEAGPSGPGPTGAR,
is the marker of bovine-hide gelatin. The ion at m/z 925.4326, a doubly charged ion of the fragment GEPGPTGVQGPPGPAGEEGK, is the
marker of pig-hide gelatin. The ion at m/z 758.3530 is the marker
of the tortoise shell glue, the sequence for which is unknown.
Additionally, the ion at m/z 732.8282 is the deer-horn glue, the
sequence for which is also unknown. The selected ion monitoring chromatograms of the marker and corresponding MSE spectra
are shown in Fig. 3. However, the identication of most marker
peptides was not possible, despite searching available databases
using the accurate molecular weights and elemental compositions.
The markers of the tortoise shell glue and deerhorn glue were not
identied due to the lack of corresponding collagen sequences.
The gelatins from different sources can also be identied using the

X.-L. Cheng et al. / Journal of Pharmaceutical and Biomedical Analysis 62 (2012) 191195

marker peptide of each gelatin in selected ion monitoring chromatography.

4. Conclusion
A UPLC/Q-TOF-MS sample proling method coupled with PCA
has been successfully used to classify gelatins from different
sources. The UPLC/Q-TOF-MS detection coupled with PCA allowed
for details of the samples to be proled so that important markers
of differences can be measured, even at low concentration levels.
The markers of gelatins can be used to distinguish one gelatin from
a second or to identify gelatins in a mixture. As a result, this is the
rst time that differences among these gelatins have been observed
systematically.
Acknowledgements
This study was supported in part by grants from the National
Institute for Food and Drug Control, the State Food and Drug Administration (No. 2010A1), the National Natural Science Foundation
of China (Nos. 30873356, 81001622), the project The Platform of
Quality Standards and Information System Research of Traditional
Chinese Medicine (No. 2009ZX09308-004) from the Important
Program of Ministry of Science and Technology of the Peoples
Republic of China. The author wishes to thank Xu Bai from Waters
Technology (Shanghai) Co., Ltd. and Gui-Feng Zhang from State
Key Laboratory of Biochemical Engineering, Institute of Process
Engineering, and the Chinese Academy of Sciences for their expert
technical assistance.
References
[1] G.Y. Li, S. Fukunaga, K. Takenouchi, Comparative study of the physiological
properties of collagen, gelatin and collagen hydrolysate as cosmetic materials,
Int. J. Cosmet. Sci. 27 (2005) 101106.
[2] J. Poppe, in: A. Immeson (Ed.), Gelatin in Thickening and Gelling Agents for
Food, Blackie, London, 1992, pp. 98123.
[3] D. Shoshani, E. Markovitz, Y. Cohen, Skin test hypersensitivity study of a crosslinked, porcine collagen implant for aesthetic surgery, J. Periodontol. 78 (2007)
112121.

195

[4] The State Pharmacopoeia Commission of PR China, Pharmacopoeia of the Peoples Republic of China (English edition), vol. 1, Peoples Medical Publishing
House, Beijing, 2005, 32 p.
[5] The State Pharmacopoeia Commission of PR China, Pharmacopoeia of the Peoples Republic of China (English edition), vol. 1, Peoples Medical Publishing
House, Beijing, 2005, 31 p.
[6] The State Pharmacopoeia Commission of PR China, Pharmacopoeia of the Peoples Republic of China (English edition), vol. 1, Peoples Medical Publishing
House, Beijing, 2005, 33 p.
[7] M. Sakaguchi, T. Nakayama, S. Inouye, Food allergy to gelatin in children
with systemic immediate-type reactions, including anaphylaxis, to vaccines,
J. Allergy Clin. Immunol. 98 (1996) 10581061.
[8] M. Nemati, M.R. Oveisi, H. Abdollahi, O. Sabzevari, Differentiation of bovine and
porcine gelatins using principal component analysis, J. Pharm. Biomed. Anal.
34 (2004) 485492.
[9] J.Y. Hu, X.L. Cheng, X.Y. Xiao, Review on chemical composition and quality
evaluation of Gelatina Nigra, Chin. Pharm. Aff. 21 (2007) 193195.
[10] R.S. Sawhney, Immunological identication of types I and III collagen in bovine
lens epithelium and its anterior lens capsule, Cell Biol. Int. 29 (2005) 133137.
[11] A. Venien, D. Levieux, Differentiation of bovine from porcine gelatines using
polyclonal anti-peptide antibodies in indirect and competitive indirect ELISA,
J. Pharm. Biomed. Anal. 39 (2005) 418424.
[12] K. Tasanen, R. Palatsi, A. Oikarinen, Demonstration of increased levels of type
I collagen mRNA using quantitative polymerase chain reaction in brotic and
granulomatous skin diseases, Br. J. Dermatol. 139 (1998) 2326.
[13] S.G. Kauschke, A. Knorr, M. Heke, Two assays for measuring brosis: reverse
transcriptase-polymerase chain reaction of collagen alpha(1) (III) mRNA is an
early predictor of subsequent collagen deposition while a novel serum Nterminal procollagen (III) propeptide assay reects manifest brosis in carbon
tetrachloride-treated rats, Anal. Biochem. 275 (1999) 131140.
[14] G.F. Zhang, T. Liu, Q. Wang, L. Chen, J.D. Lei, J. Luo, G.H. Ma, Z.G. Su, Mass spectrometric detection of marker peptides in tryptic digests of gelatin: a new method
to differentiate between bovine and porcine gelatin, Food Hydrocolloids 23
(2009) 20012007.
[15] G.F. Zhang, A.M. Sun, W.J. Li, T. Liu, Z.G. Su, Mass spectrometric analysis of
enzymatic digestion of denatured collagen for identication of collagen type,
J. Chromatogr. A 1114 (2006) 274277.
[16] B. Fuchs, K. Arnold, J. Schiller, Mass spectrometry of biological molecules, in:
R.A. Meyers (Ed.), Encyclopedia of Analytical Chemistry, Wiley, Chichester,
2008, pp. 139.
[17] J.T. Wu, P.Q. Huang, M.X. Li, D.M. Lubman, Protein digest analysis by pressurized
capillary electrochromatography using an ion trap storage/reectron time-ofight mass detector, Anal. Chem. 69 (1997) 29082913.
[18] H.W. Xie, M. Gilar, J.C. Gebler, Characterization of protein impurities and sitespecic modications using peptide mapping with liquid chromatography
and data independent acquisition mass spectrometry, Anal. Chem. 81 (2009)
56995708.
[19] A. Nimptsch, S. Schibur, C. Ihling, A. Sinz, T. Rieme, D. l Huster, J. Schiller, Quantitative analysis of denatured collagen by collagenase digestion and subsequent
MALDI-TOF mass spectrometry, Cell Tissue Res. 343 (2011) 605617.
[20] Waters Corporation application note 72000091EN, 2004.