Journal of Chromatography A 1673 (2022) 463192

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Determination of polypeptide antibiotics in animal tissues using liquid
chromatography tandem mass spectrometry based on in-line
molecularly imprinted solid-phase extraction
Xuqin Song a,b, Esther Turiel c, Jian Yang a, Antonio Martín-Esteban c,∗, Limin He b,∗
a

Laboratory of Animal Genetics, Breeding and Reproduction in the Plateau Mountainous Region (Ministry of Education), College of Animal Science, Guizhou
University, Guiyang, Guizhou 550025, China
National Reference Laboratory of Veterinary Drug Residues (SCAU), College of Veterinary Medicine, South China Agricultural University, Guangzhou 510642,
China
c
Departamento de Medio Ambiente y Agronomía, INIA, CSIC, Carretera de A Cora km 7.5, Madrid 28040, Spain
b

a r t i c l e

i n f o

Article history:
Received 20 April 2022
Revised 31 May 2022
Accepted 31 May 2022
Available online 1 June 2022
Keywords:

Polypeptide antibiotics
Molecular imprinting
Imprinted stationary phase
Sample preparation
Animal tissue

a b s t r a c t
Effective purification and enrichment of polypeptide antibiotics in animal tissues is always a challenge,
due to the co-extraction of other endogenous peptides which usually interfere their final determination.
In this study, a molecularly imprinted column was prepared by packing polymyxin E-imprinted particles
into a 100 mm × 4.6 mm i.d. HPLC column. The as-prepared imprinted columns were able to tolerate 100% aqueous phase and exhibited good stability and high column efficiency. Polypeptides antibiotics with similar molecular size or spatial structure to polymyxin E were well retained by the imprinted
column, suggesting class selectivity. After optimization of mobile phase conditions of imprinted column,
polypeptide antibiotics in animal tissue extracts were enriched and cleaned up by in-line molecularly imprinted solid-phase extraction, allowing the screening of target analytes in complex samples at low concentration levels by UV detection. Eluate fraction from the imprinted column was collected, and further
dried and re-dissolved with methanol-0.5% formic acid aqueous solution (80:20, v/v) for final LC-MS/MS
analysis. Analysis was accomplished using multiple reaction monitoring (MRM) in positive electrospray
ionization mode and analytes quantified using the matrix-matched external calibration curves. The results showed high correlation coefficients for target analytes in the linear range of 2 ∼ 200 μg kg−1 . At
four different concentration levels (limit of quantification, 50, 100 and 200 μg kg−1 ), recoveries of four
polypeptide antibiotics in swine, cattle and chicken muscles ranged from 66.7 to 94.5% with relative standard deviations lower than 16.0%. The limits of detection (LOD) were 2.0 ∼ 4.0 μg/kg, depending upon
the analyte and sample. Compared with a conventional pretreatment method, the imprinted column was
able to remove more impurities and to significantly reduce matrix effects, allowing the accurate analysis
of polypeptide antibiotics.
© 2022 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license
( />
1. Introduction
Polypeptide antibiotics (PPTs) are a class of antibiotics isolated
from Bacillus, Streptomyces or Actinomyces, and consist of a cyclic
polypeptide structure formed by 4 ∼ 16 amino acids. The antibacterial mechanism of PPTs differs depending upon the antibiotic. For instance, bacitracin (BTC) and virginiamycin (VGM) disrupt
the bacterial cell wall, while polymyxins affect the bacterial mem-


∗

Corresponding authors.
E-mail addresses: (A. Martín-Esteban),
(L. He).

branes. Due to the favorable antibacterial effect, PPTs are widely
used in animal husbandry to treat many bacterial infections, such
as dysentery, mastitis, enteritis, etc [1]. PPTs like polymyxins, BCT
and VGM are often added at subtherapeutic level to animal feed as
growth promoters for animals. Although PPTs are beneficial to animal production, their long-term or illegal addition to feed could
cause drug residues in animal derived food and further threaten
human health through the food chain [2]. In addition, PPTs as a last
resort against multiple drug resistance infection have been threatened by drug-resistant bacteria. It has been reported that the longterm addition of VGM to chicken feed could increase drug resistance rate of Escherichia coli from 27% to 70% [3]. Moreover, after

/>0021-9673/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( />

X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

avoparcin (a glycopeptide antibiotic) use as antimicrobial growth
promoter, vancomycin (VCM) resistance was common detected in
intestinal enterococci, not only in exposed animals, but also in surrounding hospitals [4]. The recent emergence of MCR-1 resistance
genes in bacteria has also been associated with the excessive consumption of polymyxins in animal husbandry [5].
In order to curb the rise of antimicrobial resistance, many countries and governments have adopted measures. Glycopeptide antibiotics, that are extremely important to humans and could pose
a serious threat to public health safety if used, were firstly banned
for the treatment of animal diseases and growth promotion [6,7].
The application of other peptide antibiotics in animal production
is severely restricted as well. BTC and VGM were banned as animal feed additives by the European Communities in 1998 [8]. PPTs

such as polymyxin E (PME), VGM and BTC were allowed to be
added in feed at 4 ∼ 100 mg kg−1 as antimicrobial growth promoters in China [9]. However, the Ministry of Agriculture of China
announced prohibition of all antimicrobial growth promoters except for Chinese medicine since 2020 [10]. Furthermore, the maximum residues limits (MRL) of PPTs in the latest standards have
decreased from 50 ∼ 500 μg kg−1 to 50 ∼ 300 μg kg−1 in animal
food [11].
Considering the weak UV and fluorescence absorption of some
peptides, high performance liquid chromatography (HPLC) coupled
to evaporative light-scattering or mass spectrometry detectors, particularly liquid chromatography tandem mass spectrometry (LCMS/MS), are the techniques mainly used for PPTs analysis [12,13].
Due to the high polarity of PPTs, their extraction is usually performed using mixtures of acetonitrile (ACN) or methanol (MeOH)
and acidic water, resulting in the insufficient precipitation of proteins and fats in biological samples. Accordingly, the development
of a proper cleanup protocol is very important to reduce complex matrix interferences and improve the accuracy of the analysis.
Although various solid-phase extraction (SPE) cartridges such as
HLB, C18 and Strata-X could be currently used to enrich and clean
up PPTs from animal tissues, but these sorbents led to the coextraction of impurities and serious matrix effects [2,13,14]. It has
been reported that the absorption behavior of endogenous peptide
disruptors in animal tissues is similar to that of PPTs, and thus
they are co-extracted onto SPE cartridges [15], disrupting the accurate characterization and quantification of drug residue would be
affected even by highly sensitive LC-MS/MS. Nowadays, on-line/inline sample pretreatments are preferred over traditional SPE since
on-line/in-line SPE integrates sample loading, washing and elution, which greatly simplifies the pretreatment process, reducing
loss of analytes as well as eventual sample contamination [16–
19]. Besides, the volume of sample extract injected into the analytical instrument is very small and the consumption of organic
solvents is quite low, which is consistent with the principles of
green chemistry. As a highly stable and durable recognition material, molecularly imprinted polymers (MIP) have become an alternative to selectively extract trace drugs from complex matrices. The use of MIPs in on-line/in-line SPE procedures (so called
MISPE) allows to simplify SPE steps, improving selective recognition and reducing matrix effects and it has been successfully applied in the on-line analysis of several veterinary drugs such as
tetracyclines [20], sulfonamides [21] and quinoxaline [22] in animal tissues, milk and eggs and in the in-line analysis of phenylurea herbicides in vegetable samples [23], thiabendazole in fruits
[24] and fluoroquinolones in soils [25].
The aim of this study was to develop a selective analytical method by in-line MISPE for enrichment and purification
of selected polypeptide antibiotics in animal tissues (vancomycin
(VCM), teicoplain (TEC), polymyxin B (PMB) and bacitracin A
(BTCA)) prior their determination by LC-MS/MS. The MIP particles of polymyxin E were packed in an HPLC column and used as


stationary phase. After the optimization of mobile phase (loading,
washing and elution conditions), the imprinting column was able
to in-line enrich analytes from complex matrix and finally combined with LC-MS/MS to detect PPTs.
2. Material and methods
2.1. Reagents and chemicals
HPLC grade reagents including ACN, MeOH and formic acid
(FA) were provided by Fisher Scientific (Fairlawn, NJ, USA). Ultrapure water was obtained by a Millipore MilliQ equipment.
Trichloroacetic acid (TCA) was obtained from Guangzhou Chemical
Reagent Factory (Guangzhou, China). Oasis HLB SPE cartridge (60
mg, 3 mL) was purchased from Waters Co. (Milford, MA, USA).
2.2. Standards and stock solutions
The reference standards of VCM, BTCA and TEC were purchased from National Institutes for Food and Drug Control (Beijing,
China). Daptomycin, PME and PMB (containing two major components of PMB1 and PMB2) were obtained from Dr. Ehrenstrofer
GmbH (Augsburg, Germany). Enrofloxacin, sulfadimidine and virginiamycin was available from TRC (Toronto, Canada). The purity
of each standard was higher than 84.7%. Stock standard solution (1
mg mL−1 ) was prepared by weighing each reference standard into
a 10 mL brown volumetric flask and diluting with appropriate solvent as follows: daptomycin (DAP), enrofloxacin, sulfadimidine and
virginiamycin (VGM) dissolved with MeOH; VCM and TEC with water; PMB, PME and BTCA with 0.1% FA aqueous solution. Each stock
solution should be stored at -20 °C for a maximum of 3 months.
Mixed standard working solutions at 10 μg mL−1 concentration
were prepared by diluting stock solution with the mixed solution
of MeOH and 0.1% FA aqueous solution (50:50, v/v), which should
be stored at -20 °C during no more than a month.
2.3. Preparation of imprinted chromatographic column
The MIP particles of polymyxin E were obtained by precipitation polymerization according to our previous study [26]. After removing the template, the MIP particles were washed with water
and MeOH for three times and dried in an oven at 60 °C under vacuum for 24 h. An amount of 3 g MIP particles were dispersed into
45 mL of HPLC grade isopropanol. After sonication for 20 min, MIP
particles were packed into an empty column (100 mm × 4.6 mm
i.d.) under high pressure (80 0 0 psi) provided by a CP12 liquid

chromatographic column packing machine (Scientific Systems Inc.
(SSI), CA, USA). Finally, the MIP column was rinsed with MeOH at
a flow rate of 0.2 mL min−1 for 24 h.
2.4. Chromatographic performance of MIP column
The mobile phases including MeOH, ACN, water or FA in water were tested. The chromatographic performance of MIP column,
including background pressure and stability, were evaluated using
polymyxin E as the target analyte. The retention factor (k) and theoretical plates number (N) of both MIP and non-imprinted (NIP)
columns were calculated as following equations:

k=

tR − t0
t0

where, tR (min) is the retention time of polymyxin E; t0 (min) is
the void time of solvent peak.

N = 5.54
2

tR
w 1/2

2

= 16

tR
w


2


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

where, w1/2 (min) is the peak width at half peak height; w (min)
is the peak width at peak base.

200 ng/mL) were prepared to plot calibration curves. Mean recoveries of polypeptide antibiotics from pork, beef and chicken at the
spiked concentrations of limit of quantification (LOQ), 50, 100 and
200 μg kg−1 were calculated.

2.5. Sample preparation based on in-line separation using HPLC
Muscle samples including beef, pork and chicken were obtained
from local supermarkets (Guangzhou, China). After homogenization, 2 g samples were accurately weighed into a 50 mL polypropylene centrifuge tube and spiked with appropriate standard working
solutions for incubation at room temperature for 30 min. The extraction was performed with 5 mL of ACN-10% TCA in water (1:1,
v/v) through sonication for 5 min and shaking for 20 min. After centrifugation, the supernatant was collected and the residue
was re-extracted following the same above extraction procedure.
All the supernatants were combined and the final volume was
adjusted to 10 mL. A volume of 2 mL of the extract was collected and evaporated to near dryness. The residues were dissolved
in 0.5 mL of MeOH-0.5% FA aqueous solution (80:20, v/v) before
in-line MISPE by HPLC-UV. Once the chromatographic run time
reached 8 min, the target fraction of analytes was started and collected into a 5 mL test tube until a run time of 10.5 min. Then,
the eluate fraction of analytes at their retention time was collected
and evaporated to dryness at 45 °C. Finally, the residues were redissolved with 0.2 mL of the above reconstitution solution and analyzed by LC-MS/MS.

3. Results and discussion
3.1. Optimization of mobile phase

Chromatographic conditions have a significant impact on the
retention performance and impurity removal capability of the imprinted column during in-line MISPE. Generally, in-line MISPE consists of three steps [23–25]: first, during the loading step, an organic solvent (eg. ACN) is prioritized as the mobile phase, since
the co-extracted fat-soluble impurities are eluted rapidly by the organic mobile phase, while target analytes are retained well on the
MIP column with specific recognition. Thereafter, a mixture of water and organic solvent is commonly used as the washing solution
to reduce the interference of complex matrix but without disrupting the specific interactions between analytes and MIP. Finally, a
suitable eluent is selected to elute the target analytes still retained
by the MIP. In order to make sure that polypeptides could be selectively bound to MIP column whereas impurities co-extracted were
removed as much as possible, the in-line MISPE conditions (loading, washing and elution) should be optimized through adjusting
the composition of mobile phase and the elution program. In this
study, several mobile phase compositions, including MeOH, ACN,
MeOH-water/acid water, ACN-water/acid water were tested. The
results showed that a very high baseline background and an important peak tailing of polymyxin E were observed when MeOH
was used in the mobile phase, which could be explained by the
strong UV absorption of MeOH at low UV wavelengths. Accordingly, MeOH was discarded for further experiments and the effect of the presence of ACN in the mobile phase on the retention/separation of polymyxin E was studied. It was observed that
polymyxin E was completely retained onto the MIP column, making necessary the addition of formic acid to the component C of
mobile phase (see Section 2.6) in order to disrupt the hydrogen
bonding interactions occurring between target analyte and binding
sites, thus allowing the elution of polymyxin E [23]. As shown in
Fig. 1A, the retention time and peak shape of polymyxin E were
not significantly affected by the increase of FA concentration, but,
high FA concentration caused baseline instability due to FA strong
absorption at low UV wavelengths. Thus, a 0.02% FA was chosen as
optimum to be present in the component C of the mobile phase to
elute polymyxin E. Besides, it was observed that polymyxin E was
eluted faster with the increase of ACN concentration in component
C of the mobile phase (Fig. 1B), which suggests that MIP column
exhibited also a reversed-phase retention mechanism alongside hydrogen bonding interactions as mentioned above.

2.6. HPLC-UV and LC-MS/MS conditions
The chromatographic performance of MIP column and corresponding in-line MISPE procedures were performed by HPLC-UV.

The mobile phase includes ACN (A), 50% ACN in water (B) and 50%
ACN in water containing 0.02% FA (C). The in-line MISPE onto the
imprinted column was carried out with the following gradient elution program: 0∼6 min for loading, 100% A; 6.1∼7 min for washing, 100% B; 7.1∼11 min for eluting, 100% C; 11.5∼16 min for reconditioning for the next run. The flow rate was 0.75 mL min−1 and
the injection volume was 100 μL. Target analytes were monitored
at 205 nm.
Due to the trace amounts of polypeptides residues in animal tissues, final sample analysis was performed by LC-MS/MS. A
Shimadzu HPLC system (Shimadzu, Kyoto, Japan) and an Applied
Biosystems Sciex Triple Quad 5500 triple-quadrupole mass spectrometer were used to detect the analytes. Chromatographic and
mass conditions such as mass parameters, mobile phase and elution program were the same as our previous report [12]. Briefly,
a Phenomenex Kinetex Biphenyl column (50 mm × 2.1 mm i.d.,
2.6 μm, Phenomenex, Torrance, CA) was used to separate the analytes. The mobile phase consisted of 0.1% FA in ACN solution (A)
and 0.1% FA in water solution (B) with the following gradient elution: 0 min, 6% A; 2 min, 6% A; 5 min, 40% A; 14 min, 70% A;
14.1 min, 6% A; 18 min, 6% A. The mass conditions were acquired
in multiple reaction monitoring (MRM) mode. The tune parameters
were carried out as follows: ionspray voltage, 5500 V; nebulizing
gas pressure, 55 psi; auxiliary gas,50 psi; curtain gas, 40 psi; ion
source temperature, 600 °C; entrancepotential, 10 V and collision
cell exit potential, 12 V. At least two product ions of each compound were monitored under the ESI+ mode and the mass parameters are given in Table S1.

3.2. Chromatographic performance and selectivity of MIP column
The chromatographic parameters affecting MIP column performance, including background pressure, stability and retention factor, were examined. Considering the commonly used solvents in
HPLC analysis, the background pressure of MIP column in MeOH,
ACN and water was evaluated at flow rates ranging from 0.2 to
1.0 mL min−1 . As presented in Fig. S1-A, the background pressures in different solvents are in the order: water>MeOH>ACN.
The highest is the water with a background pressure of 125 bar, indicating that the MIP column could tolerate 100% aqueous phase.
The reproducibility of retention times, thus the performance stability of MIP column, was assessed by injecting the standard solution of polymyxin E for eight times in a row and the chromatograms obtained are shown in Figure S1-B. The retention time

2.7. Method validation
Eluate fraction of analytes at their retention time (8 to
10.5 min) was collected and further analyzed by LC-MS/MS. Under

optimum conditions, method parameters for validation including
linearity, accuracy, precision and sensitivity were assessed. Matrixmatched standard solutions (0.5, 1, 2, 5, 10, 20, 50, 80, 100 and
3


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

Fig. 1. Effect of different percentage of FA concentration (A) and ACN (B) in component C of mobile phase on the retention of polymyxin E. ACN: acetonitrile; FA: formic
acid.

Fig. 2. HPLC-UV chromatograms of polymyxin E on MIP column and NIP column. Chromatographic conditions: see Experimental section.

was 9.90 min with a relative standard deviation (RSD) of 0.04%,
whereas the theoretical plate number was 3298 with the RSD of
9.84%, demonstrating the excellent stability of the MIP column,
which would allow to be routinely used in HPLC analysis.
Under the optimal HPLC conditions, the selectivity was investigated by comparing the retention of polymyxin E on MIP and
non-imprinted (NIP) columns. The results showed (Fig. 2) that
polymyxin E was well retained and eluted from the MIP column
with sharp peak, while the retention time of polymyxin E on NIP
was 8.8 min with serious peak tailing. The theoretical plate number obtained from NIP column was 136, much lower than that
from MIP column. The stronger retention of polymyxin E on MIP
than on NIP, as well as the different peak shape demonstrate the
presence of selective binding sites on MIP column. It is important
to point out that, strictly speaking, the chromatographic parameters measured (k and N) are accurate only for isocratic elution
conditions and for chromatographic Gaussian peaks and thus the
reported values can only be used for comparison purposes of MIP
and NIP columns under the experimental conditions indicated in

the present paper.
The cross-reactivity was estimated by analyzing five different
polypeptide antibiotics (TEC, VCM, VGM, PMB and DAP) and two
antimicrobials (enrofloxacin and sulfadimidine) with a large consumption in animal production. Each compound was analyzed at
its optimal UV wavelength. As illustrated in Fig. 3, polypeptide

antibiotics, except for VGM, were well retained and eluted from
the MIP column with a rather symmetrical peak shape and negligible tailing, while enrofloxacin and sulfadimidine were not retained with retention times lower than 2 min. This result confirmed the existence of imprinted cavities, which can well match
the polymyxin E shape, size and functional groups. Since TEC, VCM,
DAP and PMB have large molecular weight and complex cyclic
polypeptide structure similar to polymyxin E, all of them were well
retained by the imprinted column. On the contrary, although VGM
belongs to polypeptides, its structure is a large lactone ring with
a molecular weight lower than 600 Da, which significantly differs
from that of polymyxin E (template). Enrofloxacin and sulfonamide
whose molecular structures are quite different from template could
not be retained on the MIP column. Therefore, these results revealed that specific binding sites on the MIP contribute to the retention of analytes, allowing the determination of several polypeptides simultaneously.
3.3. Preparation of animal tissue extracts
Proteins in animal tissues could strongly interact with polypeptide antibiotics, and thus low pH of the extraction solvent (TCA,
sulfuric acid and hydrochloric acid) is utilized to extract the
analytes. Several studies have confirmed that the mixture of
ACN/MeOH and 10% TCA aqueous solution is able to quantitatively
4


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

Fig. 3. HPLC-UV chromatograms of polypeptide antibiotics, enrofloxacin, and sulfadimidine on MIP column. Chromatographic conditions: see Experimental section.


recover polypeptides from biological samples [27,28]. Accordingly,
MeOH-10% TCA aqueous solution (1:1, v/v) and different proportions of ACN in 10% TCA aqueous solutions for the extraction of
polypeptide antibiotics from animal tissue samples were tested. As
shown in Fig. S2, MeOH-10% TCA aqueous solution was not able
to completely disrupt the interactions of target analytes with sample matrix, leading to recoveries lower than 80 % for TEC. However,
ACN-10% TCA aqueous solution (1:1, v/v) allowed to quantitatively
extract all the analytes under study reaching recoveries higher than
95%, and thus was selected for further experiments. Furthermore,
since it was necessary to dry and reconstitute sample extracts, the
effect of different ratios of MeOH:0.5% formic acid in water as reconstitution solution was evaluated. It was observed that the recoveries of analytes increased with the presence of MeOH. Finally,
an 8:2 (v/v) ratio was the optimum, providing recoveries higher
than 91 % for all the target analytes.
3.4. In-line MISPE of polypeptide antibiotics from animal tissue
extracts

Fig. 4. HPLC-UV chromatograms obtained after the injection of spiked at 200 μg
kg−1 in pork sample and non-spiked pork sample directly onto the imprinted column. Chromatographic conditions: see Experimental section.

Sample extracts were injected in the chromatographic system
for the in-line MISPE of polypeptide antibiotics under optimum
conditions. Polymyxin E was not included in this study since it was
used as template for MIP preparation. It is well-known that, even
after exhaustive washing of MIPs, template leaking might occur
which would compromise the accurate determination of polymysin
E at trace level in real samples. Fig. 4 shows the LC-UV chromatogram at 205 nm obtained in the analysis of non-spiked and
spiked pork sample at the concentration of 200 μg kg−1 . As can

be observed, the mixture of polypeptide antibiotics was unambiguously detected in the spiked sample thanks to the high selectivity
provided by the MIP. Target analytes were recognized and eluted

free of co-extractives from the MIP column, with retention times
from 8 to 10.5 min, whereas the matrix interferences were rapidly
eluted, allowing the detection of polypeptide antibiotics at very
low concentration level in the pork sample extract without any
5


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

Table 1
Recovery and precision of polypeptide antibiotics in animal muscles (n = 5).

Compound
Vancomycin

Teicoplain
A2-1

Teicoplain
A2-2&2-3

Polymyxin
B1

Polymyxin
B2

Bacitracin A


a

Spiked
(μg kg−1 )

Average recovery (RSD) a , %
Beef
Pork
Intra-batch
Inter-batch
Intra-batch

Inter-batch

Chicken
Intra-batch

Inter-batch

LOQ
50
100
200
LOQ
50
100
200
LOQ
50

100
200
LOQ
50
100
200
LOQ
50
100
200
LOQ
50
100
200

71.3(9.7)
78.1(11.3)
89.0(3.2)
83.9(3.3)
69.1(9.9)
75.4(9.6)
81.0(14.3)
82.4(6.4)
69.5(10.2)
68.2(11.1)
89.4(13.2)
87.2(6.4)
71.0(9.1)
79.1(4.9)
85.2(1.7)

79.6(2.4)
88.1(7.5)
80.0(6.6)
89.0(3.0)
83.4(2.6)
71.1(4.5)
83.5(5.2)
89.6(2.7)
84.6(2.8)

69.7(10.8)
84.4(8.9)
88.9(12.2)
86.9(6.2)
66.7(13.4)
76.0(14.3)
86.7(10.3)
84.2(7.5)
73.5(12.6)
75.7(16.0)
86.7(11.8)
82.3(9.1)
81.9(6.6)
85.3(7.9)
83.9(12.3)
86.3(5.4)
79.7(9.9)
87.2(7.4)
89.9(11.1)
87.5(3.5)

74.5(11.6)
88.1(4.9)
88.1(9.5)
90.8(3.0)

68.1(8.5)
83.0(6.6)
80.6(5.4)
89.2(2.3)
83.8(7.8)
82.9(6.7)
77.4(7.6)
77.4(7.1)
77.0(5.8)
69.4(6.7)
83.9(4.4)
81.7(8.8)
86.5(4.0)
81.3(4.2)
81.6(10.8)
80.8(10)
86.5(10.5)
84.3(4.6)
84.6(8.4)
82.6(12.4)
84.1(3.4)
83.7(6.8)
78.0(8.9)
83.9(6.1)


68.2(6.2)
81.3(7.2)
83.6(6.0)
82.1(10.6)
81.5(7.7)
85.1(7.5)
77.0(8.2)
85.0(10.6)
79.7(6.3)
71.9(8.2)
78.2(8.2)
81.2(12.3)
80.8(10.7)
83.6(4.7)
85.6(6.9)
81.3(8.0)
79.4(15.0)
83.8(3.5)
87.0(5.7)
82.8(9.6)
83.9(6.0)
84.6(5.4)
83.5(7.5)
85.7(8.2)

75.5(11.6)
79.9(9.1)
88.1(5.6)
82.2(6.4)
72.1(13.8)

81.8(12.1)
85.7(9.1)
85.5(7.1)
71.6(9.5)
76.7(13.5)
90.8(11.2)
86.5(7.2)
79.5(11.9)
78.6(5.6)
84.0(3.2)
80.1(4.4)
85.9(11.1)
80.5(6.0)
88.3(4.2)
83.4(3.9)
74.1(7.3)
82.8(3.8)
88.5(2.9)
83.9(2.8)

73.5(11.5)
78.8(5.9)
85.7(11.2)
90.5(3.3)
66.6(11.1)
76.1(11.5)
94.5(2.8)
85.9(7.5)
77.4(13.0)
75.2(11.8)

83.1(7.3)
84.3(3.7)
86.6(4.5)
90.0(3.0)
85.1(10.5)
88.4(4.7)
84.3(4.6)
85.4(2.4)
86.8(9.2)
85.2(3.0)
69.2(13.9)
89.6(3.0)
87.2(1.5)
89.2(2.1)

RSD, relative standard deviation.

3.6. Matrix effect

sample clean-up. However, due to the slight differences observed
in the retention times (see Fig. 3 ), it was not possible to resolve
target analytes when the mixture was injected into the MIP column. Thus, from the obtained results, it can be concluded that the
in-line MISPE procedure is suitable for the screening of polypeptide
antibiotics in crude animal tissue extracts at the concentration levels required without any other sample clean-up. In this sense, only
those positive samples would be subjected to further analysis by
fraction collection and subsequent LC-MS/MS analysis as described
below.

Matrix effect (ME) caused by the co-extracted impurities could
suppress or enhance the mass spectrum response of analytes under ESI mode. The matrix effect was calculated by the following

equation:
ME(% )
Slope of matrix matched standard curve − Slope of standard curve
Slope of standard curve
× 100%
=

where the slope of standard curve is obtained by the linear fitting equation of pure standard solutions (without matrix). The positive and negative values of ME represent the signal enhancement
and signal suppression, respectively. In addition, the values of ME
within the range of ±0∼20%, ±20∼50% and >±50% mean the soft,
medium and strong matrix effect, respectively. As demonstrated
in Fig. 6, except that BTCA in chicken (-23.1%) and pork matrix (25.3%) showing medium matrix effect, other target analytes exhibited slight signal suppression as the matrix effects were soft (less
than ±0∼20%). Many studies confirmed that strong signal suppression for polypeptides was observed in biological sample matrices
such as chicken (-41∼-52%) [2], pork (-24∼-86%) [13] and fish (<85%) [14], even though the SPE strategy was used for sample purification in the cited above studies. Considering severe matrix effects for polypeptides, the developed approach of in-line molecularly imprinted solid-phase extraction provided less matrix interference, which could become a better alternative for sample purification.

3.5. Analysis of polypeptide antibiotics from in-line MISPE extracts by
LC-MS/MS and method validation
As indicated in Section 2.5, the eluate fraction of analytes at
their retention time (8 to 10.5 min) was collected and further analyzed by LC-MS/MS. Under optimum conditions, method parameters for validation including linearity, accuracy, precision and sensitivity were assessed. Matrix-matched standard solutions (0.5, 1,
2, 5, 10, 20, 50, 80, 10 0 and 20 0 ng/mL) were prepared to plot
calibration curves. Good linearities at the spiked concentrations
of 5∼200 ng mL−1 for VCM and TEC, and 2∼200 ng mL−1 for
BTCA and PMB were observed with correlation coefficients (r) better than 0.99. Obtained recoveries and their corresponding relative
standard deviation (RSD) are summarized in Table 1. Mean recoveries of polypeptide antibiotics from pork, beef and chicken at the
spiked concentrations of limit of quantification (LOQ), 50, 100 and
200 μg kg−1 were within 66.7%∼94.5% interval with RSDs lower
than 14.3%, which are within the acceptable limits of residue analysis. The typical MRM chromatograms acquired under ESI mode
from spiked pork at 100 μg kg−1 are depicted in Fig. 5. The limit
of detection (LOD) and LOQ ranged from 2 to 4 μg kg−1 and 5 to
12.5 μg kg−1 , respectively, which are lower than the MRLs recommended by the Ministry of Agriculture of China. All results have

demonstrated that proposed method is accurate, reliable and sensitive for the monitoring of polypeptides residues in animal tissues
(Table 2).

3.7. Comparison with literature method
The developed method was compared with the literature
method [29], involving in the extraction with ACN-10% TCA aqueous solution and the purification with Oasis HLB SPE cartridge. As
exhibited in Fig. S3, there were serious impurity interferences after extraction using a conventional SPE cartridge. A relatively clean
chromatogram was observed when only 1 mL of the extract was
6


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

Fig. 5. Typical MRM chromatograms obtained from spiked (100 μg kg−1 ) (A) and non-spiked pork sample extracts (B) from the elute fraction obtained by in-line MISPE.
Chromatographic conditions: see Experimental section.

Table 2
The calibration curve, linearity, limit of detection, limit of quantitation of the developed method for polypeptide antibiotics in animal muscles.
Compound

Muscle

Calibration curve

Vancomycin

chicken
beef

pork
chicken
beef
pork
chicken
beef
pork
chicken
beef
pork
chicken
beef
pork
chicken
beef
pork

y
y
y
y
y
y
y
y
y
y
y
y
y

y
y
y
y
y

Teicoplain
A2-1
Teicoplain
A2-2&2-3
Polymyxin
B1
Polymyxin
B2
Bacitracin A

a
b

=
=
=
=
=
=
=
=
=
=
=

=
=
=
=
=
=
=

2.64
2.23
2.85
1.18
2.59
9.14
1.06
1.68
7.72
5.88
5.21
4.99
2.25
1.94
1.82
1.09
1.20
5.25

×
×
×

×
×
×
×
×
×
×
×
×
×
×
×
×
×
×

103 x + 4.08 × 102
103 x + 2.66 × 102
103 x + 1.23 × 102
103 x + 8.69 × 102
103 x + 3.48 × 102
102 x + 1.46 × 103
103 x + 5.95 × 102
103 x + 1.13 × 103
102 x + 4.02 × 102
104 x + 1.45 × 105
104 x + 1.09 × 105
104 x + 7.33 × 104
104 x + 1.44 × 105
104 x + 3.33 × 104

104 x + 1.01 × 105
104 x + 1.10 × 103
104 x + 1.50 × 103
103 x + 5.30 × 103

Correlationcoefficient (r2 )

Linear range(ng mL−1 )

LODa (μg kg−1 )

LOQb (μg kg

0.9912
0.9920
0.9943
0.9939
0.9912
0.9918
0.9925
0.9978
0.9976
0.9916
0.9905
0.9915
0.9929
0.9907
0.9930
0.9957
0.9973

0.9945

5∼200
5∼200
5∼200
5∼200
5∼200
5∼200
5∼200
5∼200
5∼200
2∼200
2∼200
2∼200
2∼200
2∼200
2∼200
2∼200
2∼200
2∼200

4
4
4
4
4
4
4
4
4

2
2
2
2
2
2
2
2
2

12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
5
5
5
5
5
5
5
5
5

LOD, limit of detection

LOQ, limit of quantitation.

7

-1

)


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

rapid separation provided by the present developed method open
new bright and exciting research avenues in the analysis of antibiotics in complex matrices.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
CRediT authorship contribution statement
Xuqin Song: Methodology, Validation, Writing – original draft,
Funding acquisition. Esther Turiel: Conceptualization, Supervision, Writing – review & editing. Jian Yang: Methodology.
Antonio Martín-Esteban: Conceptualization, Methodology, Supervision, Writing – review & editing, Resources, Funding acquisition.
Limin He: Conceptualization, Supervision, Writing – review & editing, Resources, Funding acquisition.

Fig. 6. Matrix effects of chicken, beef and pork matrices on the response of target
analytes (n = 3).

directly introduced into the LC-MS/MS, but the existence of many
impurities, especially from TCA, would not only reduce the efficiency of ion formation but also would affect the long-term service
life of analytical instrument. In sharp contrast, there were few impurity interferences after the in-line extraction using MIP column

with the matrix effect less than -20%, suggesting very week matrix
suppression effects. Consequently, the proposed analytical method
based on in-line MISPE might be a convenient alternative for the
determination of polypeptides in animal foodstuffs.

Acknowledgments
The authors are grateful to the assistance from Myriam
Díaz-Álvarez. This research was funded by the National Science
Foundation of China (grant number 31572562), the Key Program of Guangzhou Science and Technology Plan (grant number
201804020019) and the Special Funds of the National Natural Science Foundation of Guizhou University ([2020]25).

3.8. Applications to real samples

References

To demonstrate the feasibility and practicability of the developed method, 60 muscle samples (20 chicken, 20 beef and 20
pork) collected from different provinces were analyzed. All muscle
samples were free from VGM and TEC, perhaps because these two
compounds are expensive and more importantly, they have been
banned in food producing animal for a long time. The PMB and
BTCA were detected in 1 chicken and 4 pork samples, indicating
that they are still used in animal production to prevent bacterial
infection, or even use as growth promoters. However, the concentrations of PMB and BTCA detected in muscle samples were below
the corresponding MRLs. The highest concentration of BTCA detected was 43.6 μg kg−1 in pork sample and the lowest was only
3.9 μg kg−1 of PMB in chicken. Since the PMB and BTCA have low
oral absorption, the residues in animal tissues are low when drugs
are rationally used and withdrawal time are enough.

[1] I.L. Tsai, H.Y. Sun, G.Y. Chen, S.W. Lin, C.H. Kuo, Simultaneous quantification
of antimicrobial agents for multidrug-resistant bacterial infections in human

plasma by ultra-high-pressure liquid chromatography–tandem mass spectrometry, Talanta 116 (2013) 593–603.
[2] J. Boison, S. Lee, J. Matus, A multi-residue method for the determination of
seven polypeptide drug residues in chicken muscle tissues by LC-MS/MS, Anal.
Bioanal. Chem. 407 (2015) 4065–4078.
[3] F. Aarestrup, A. Seyfarth, H.D. Emborg, K. Pedersen, R. Hendriksen, F. Bager, Effect of abolishment of the use of antimicrobial agents for growth promotion on
occurrence of antimicrobial resistance in fecal enterococci from food animals
in Denmark, Antimicrob. Agents Chemother. 45 (2001) 2054–2059.
[4] A.E. van den Bogaard, E.E. Stobberingh, Epidemiology of resistance to antibiotics: links between animals and humans, Int. J. Antimicrob. Agents 14 (20 0 0)
327–335.
[5] Y.Y. Liu, Y. Wang, T.R. Walsh, L.X. Yi, R. Zhang, J. Spencer, Y. Doi, G. Tian,
B. Dong, X. Huang, L.F. Yu, D. Gu, H. Ren, X. Chen, L. Lv, D. He, H. Zhou, Z. Liang,
J.H. Liu, J. Shen, Emergence of plasmid-mediated colistin resistance mechanism
MCR-1 in animals and human beings in China: a microbiological and molecular biological study, Lancet Infect. Dis. 16 (2016) 161–168.
[6] Announcement No. 560. Veterinary Drug Local Standard Revocation Catalogue. 0806/t20 080606_1057240.htm
Ministry of Agriculture and Rural Affairs of the People’s Republic of China.
March 2008.
[7] US Food and Drug Administration, Code of Federal Regulations. Title 21,
Section 530.41: Drugs prohibited for extralabel use in animals, March
(2021) E/part- 530/
subpart- E/section- 530.41.
[8] Commission of the European communities, council regulation 2821/98, Off. J.
Eur. Commun. Legis. L351 (2012) 4.
[9] Announcement No. 168. Specifications for the Use of Feed Drug Additives.
July 2001. />Ministry of Agriculture and Rural Affairs of the People’s Republic of China.
[10] Announcement No. 194. January 2020. />201907/202001/t20200103_6334292.htm. Ministry of Agriculture and Rural Affairs of the People’s Republic of China
[11] Ministry of Agriculture and Rural Affairs of China, the National Health Commission of China, and the State Administration of Market Regulation, Document GB-31650-2019. National Food Safety Standard-Maximum Residue Limits
for Veterinary Drugs in Foods, September (2019) />zwdt/201910/t20191008_6329518.htm.
[12] X. Song, Q. Huang, Y. Zhang, M. Zhang, J. Xie, L. He, Rapid multiresidue analysis
of authorized/banned cyclopolypeptide antibiotics in feed by liquid chromatography–tandem mass spectrometry based on dispersive solid-phase extraction,
J. Pharm. Biomed. Anal. 170 (2019) 234–242.

[13] A. Kaufmann, M. Widmer, Quantitative analysis of polypeptide antibiotic
residues in a variety of food matrices by liquid chromatography coupled to
tandem mass spectrometry, Anal. Chim. Acta 797 (2013) 81–88.

4. Conclusions
In this study, a molecularly imprinted column was prepared using polymysin E-imprinted particles as stationary phase for the
in-line extraction of polypeptide antibiotics from animal foodstuffs. The imprinted column exhibited excellent stability and high
enough column efficiency. The polypeptide antibiotics with similar structure to polymyxin E were all well-retained on the MIP
column, further demonstrating that the MIP has class-specific
recognition abilities, allowing the selective enrichment of several
polypeptide antibiotics simultaneously from complex samples. After optimization, the imprinted-stationary phase was successfully
employed for the screening of polypeptide antibiotics from animal
muscles sample extracts by HPLC-UV without any further clean-up
at the concentration levels required. Finally, sample extracts were
collected and further analyzed by LC-MS/MS. Compared to conventional SPE, the developed method is able to efficiently remove matrix interferences, allowing to meet current requirements for analysis of trace amounts of sevral polypeptides in terms of precision
and accuracy. The efficient purification, automated operation and
8


X. Song, E. Turiel, J. Yang et al.

Journal of Chromatography A 1673 (2022) 463192

[14] Y. Xu, X. Tian, C. Ren, H. Huang, X. Zhang, X. Gong, H. Liu, Z. Yu, L. Zhang,
Analysis of colistin A and B in fishery products by ultra performance liquid
chromatography with positive electrospray ionization tandem mass spectrometry, J. Chromatogr. B 899 (2012) 14–20.
[15] E. Wan, C. Ho, D. Sin, Y.C. Wong, Detection of residual bacitracin A, colistin A,
and colistin B in milk and animal tissues by liquid chromatography tandem
mass spectrometry, Anal. Bioanal. Chem. 385 (2006) 181–188.
[16] S. Rodriguez-Mozaz, M.J. Lopez de Alda, D. Barceló, Advantages and limitations

of on-line solid phase extraction coupled to liquid chromatography–mass spectrometry technologies versus biosensors for monitoring of emerging contaminants in water, J. Chromatogr. A 1152 (2007) 97–115.
[17] A.L. Heffernan, K. Thompson, G. Eaglesham, S. Vijayasarathy, J.F. Mueller,
P.D. Sly, M.J. Gomez, Rapid, automated online SPE-LC-QTRAP-MS/MS method
for the simultaneous analysis of 14 phthalate metabolites and 5 bisphenol analogues in human urine, Talanta 151 (2016) 224–233.
[18] C. Moser, D. Zoderer, G. Luef, M. Rauchenzauner, L. Wildt, A. Griesmacher,
C. Seger, Simultaneous online SPE-LC-MS/MS quantification of six widely
used synthetic progestins in human plasma, Anal. Bioanal. Chem. 403 (2011)
961–972.
[19] C. Ramirez, C. Wang, P. Gardinali, Fully automated trace level determination
of parent and alkylated PAHs in environmental waters by online SPE-LC-APPI-MS/MS, Anal. Bioanal. Chem. (2013) 406.
[20] Y.K. Lv, J.Q. Zhang, Z.Y. Guo, W. Zhang, H.W. Sun, Determination of tetracyclines
residues in egg, milk, and milk powder by online coupling of a precolumn
packed with molecular imprinted hybrid composite materials to RP-HPLC-UV,
J. Liq. Chromatogr. Relat. Technol. 38 (2015) 1–7.
[21] J. He, S. Wang, G. Fang, H. Zhu, Y. Zhang, Molecularly imprinted polymer online solid-phase extraction coupled with high-performance liquid
chromatography−UV for the determination of three sulfonamides in pork and
chicken, J. Agric. Food Chem. 56 (2008) 2919–2925.

[22] Z.J. Duan, L.P. Fan, G.Z. Fang, J.H. Yi, S. Wang, Novel surface molecularly
imprinted sol–gel polymer applied to the online solid phase extraction of
methyl-3-quinoxaline-2-carboxylic acid and quinoxaline-2-carboxylic acid from
pork muscle, Anal. Bioanal. Chem. 401 (2011) 2291–2299.
[23] F.G. Tamayo, A. Martin-Esteban, Selective high performance liquid chromatography imprinted-stationary phases for the screening of phenylurea herbicides
in vegetable samples, J. Chromatogr. A 1098 (2005) 116–122.
[24] E. Turiel, J.L. Tadeo, P.A.G. Cormack, A. Martin-Esteban, HPLC imprinted-stationary phase prepared by precipitation polymerisation for the determination
of thiabendazole in fruit, Analyst 130 (2005) 1601–1607.
[25] E. Turiel, A. Martín-Esteban, J.L. Tadeo, Molecular imprinting-based separation
methods for selective analysis of fluoroquinolones in soils, J. Chromatogr. A
1172 (2007) 97–104.
[26] X. Song, E. Turiel, L. He, A. Martín-Esteban, Synthesis of molecularly imprinted

polymers for the selective extraction of polymyxins from environmental water
samples, Polymers 12 (2020) 131.
[27] Q. Fu, X. Li, K. Zheng, Y. Ke, Y. Wang, L. Wang, F. Yu, X. Xia, Determination of
colistin in animal tissues, egg, milk, and feed by ultra-high performance liquid
chromatography-tandem mass spectrometry, Food Chem. 248 (2017) 166–172.
[28] H. Kumar, D. Kumar, E. Nepovimova, D. Oulkar, A. Kumar, R. Mohammad, S. Budakoti, N. Upadhyay, R. Verma, K. Kuca, Determination of colistin B in chicken
muscle and egg using ultra-high-performance liquid chromatography-tandem
mass spectrometry, Int. J. Environ. Res. Public Health 18 (2021) 12.
[29] D. Zhang, J.-A. Park, D.-S. Kim, N.-H. Kim, S.-K. Kim, K.-S. Cho, D. Jeong,
J.-H. Shim, A.M. Abd El-Aty, H.-C. Shin, Simultaneous detection of bacitracin
and polymyxin B in livestock products using liquid chromatography with tandem mass spectrometry, J. Sep. Science 38 (2015) 2371–2380.

9