Accepted Manuscript Title: Correlation study between molecular structure of sesquiterpene lactones and the selective adsorption performance of molecularly imprinted polymers Author: Xiaoying Yin Qingshan Liu Xingxia Ma Xudong Zhou Mingbo Zhao id="aut0030" orcid="0000-0001-8018-322X"> Pengfei Tu PII: DOI: Reference:
S0021-9673(14)00809-7 http://dx.doi.org/doi:10.1016/j.chroma.2014.05.043 CHROMA 355435
To appear in:
Journal of Chromatography A
Received date: Revised date: Accepted date:
14-2-2014 4-5-2014 18-5-2014
Please cite this article as: X. Yin, Q. Liu, X. Ma, X. Zhou, M. Zhao, P. Tu, Correlation study between molecular structure of sesquiterpene lactones and the selective adsorption performance of molecularly imprinted polymers, Journal of Chromatography A (2014), http://dx.doi.org/10.1016/j.chroma.2014.05.043 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
*
2
the selective adsorption performance of molecularly imprinted polymers
Correlation study between molecular structure of sesquiterpene lactones and
3 Xiaoying Yina,1, Qingshan Liu b,1, Xingxia Maa, Xudong Zhouc, Mingbo Zhaoc, Pengfei Tuc*
5
a College of Pharmacy, Jiangxi University of Traditional Chinese Medicine, Nanchang 330004,
6
China
7
b National Research Center for Chinese Minority Medicine, Minzu University of China, Beijing
8
100081, China
9
c School of Pharmaceutical Sciences, Peking University Health Science Center, Beijing 100191,
cr
China.
us
10
ip t
4
Ac ce p
te
d
M
an
11
#
Qingshan Liu and Xiaoying Yin are co-first authors.
*Corresponding author. Tel.: +86 10 82802750. E-mail address: [email protected] 1
Page 1 of 29
11
Abstract We preliminarily report that the structure of template molecules and target components
13
correlates with the selective adsorption performance of molecularly imprinted polymer (MIPs) in
14
sesquiterpene lactones. Template molecules involved three categories of sesquiterpene lactones
15
with distinct ring systems: 5-mem lactone ring atractylenolide , 7-mem lactone ring
16
dehydrocostus lactone, and 10-mem lactone ring costunolide lactone, of which the conformations
17
were verified by variable-temperature 1H NMR spectroscopy. Reciprocal MIPs were prepared by
18
precipitation polymerization and employed as selective sorbents in the columns of solid phase
19
extraction (SPE). These columns were further used for enriching the mixed adsorption solution of
20
sesquiterpene lactone ingredients and reference components. Finally, the extract of Radix
21
Aucklandiae, a Chinese medicine herb, was used to verify the efficiency of this method. Our
22
results demonstrate that the steric conformational stability of molecules is associated with the
23
selective adsorption of their corresponding MIPs. We’ve further observed that the maximum
24
adsorption capacity occurs when the target molecule conformation is consistent with that of the
25
template molecule. The addition of more hydrophilic groups correlates with weaker adsorption of
26
MIPs. Our findings provide important information to help guide the selection of appropriate
27
template molecules for synthesis of MIPs with specific adsorption.
28
Keywords: Molecularly imprinted polymer; Sesquiterpene lactones; Selective adsorption
30 31 32 33 34
cr
us
an
M
d
te
Ac ce p
29
ip t
12
performance; Molecular structure; Variable-temperature 1H NMR spectroscopy
1. Introduction
Molecularly imprinted polymers (MIPs) are polymers that have been processed using a
molecular imprinting technique which leaves cavities in the polymer matrix that have affinity to a chosen "template" molecule. Due to their high selectivity, stability and efficiency, MIPs have been
35
applied to many fields, such as environmental analysis, immunoassays biosensors, catalysts and
36
chiral separation of drugs [1-5]. In recent years, molecular imprinting technique (MIT) has been
37
applied to the directional separation of active ingredients in traditional Chinese medicines [6-11].
38
MIT has been mostly focused on investigating the preparation, characterization, adsorption
39
properties and performance evaluation of MIPs [12-17], but further research regarding MIPs is 2
Page 2 of 29
rare. Several key questions about the application of MIT to Chinese medicine preparation remain
41
unanswered: 1) Which structures in active ingredients are suitable as a template molecules to
42
prepare MIPs? 2) What is the difference between the selective adsorption performances of MIPs
43
for template molecules and for analogous structural compounds? 3) What variables exist between
44
the structure of template molecules or target components and the adsorption performance of MIPs?
45
Elucidating the answer to one or several of these questions would make a significant contribution
46
to the field of active ingredient separation in traditional Chinese medicine.
In our recent research, more than 20 kinds of active ingredients in traditional Chinese
48
medicines were chosen as template molecules to prepare the corresponding MIPs by precipitation
49
polymerization. The experimental results showed that not all compounds were suitable for the
50
MIPs preparation, and that the selective adsorption performance of MIPs prepared by some
51
template molecules was poor. It was noted that the selective adsorption performance of MIPs
52
prepared by lipophilic template molecules was significantly better than MIPs prepared by
53
hydrophilic template molecules. This may be because the supramolecular interactions between the
54
template molecule and the functional monomer were not easily damaged. Similarly, the binding
55
sites with strong affinity to the target molecules on the molecular cavities of synthetic MIPs still
56
remained during the process of polymerization. A question as of yet unanswered is whether the
57
lipophilic compounds are suitable as template molecules.
58
63 64
with three similar sesquiterpene lactones ingredients and one reference compound. The stability of
65
template molecule conformation was tested by variable-temperature 1H NMR spectroscopy. The
66
relationship between MIPs structure and its selective adsorption performance was explored using
67
the Chinese medicine Radix Aucklandiae extract. Furthermore, the difference of the adsorption
68
performance for sesquiterpene lactones between MIP-SPE columns and C18-SPE columns was
69
detected by the adsorption test solutions. This study offers important references for effectively
59 60 61 62
te
d
M
an
us
47
Ac ce p
cr
ip t
40
In this study some lipophilic compounds were used as template molecules, and the
correlation between template molecule structures, target molecule structures and the selective adsorption performance of the corresponding MIPs was explored. Three types of sesquiterpene lactones with different structures and activities were selected as template molecules to prepare the corresponding MIPs by precipitation polymerization and then used to prepare MIP-SPE columns. These columns were used for enrichment and separation of the adsorption test solutions mixed
3
Page 3 of 29
70
designing and synthesizing MIPs separation materials with acceptable selective adsorption
71
properties.
72 73
2.
74
2.1 Materials and apparatus
ip t
Experimental
All reagents were of analytical grade or better. Atractylenolide , costunolide lactone and
76
dehydrocostus lactone were purchased from Zelang Pharmaceutical Co., Ltd. (Nanjing, China,
77
more than 99% purity). 1-vinylimidazole (1-Viny) and 4-vinylbenzoic acid were obtained from
78
Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Ethylene glycol dimethacrylate
79
(EDMA) and trimethylolpropane trimethacrylate (TRIM) were purchased from TCI Development
80
Co., Ltd. (Shanghai, China). 2’, 2-Azo-bis-iso-butyronitrile (AIBN) was Shanghai No. 4 Reagent
81
& H.V. Chemical Co., Ltd. (Shanghai, China) product. Acetonitrile and methanol of HPLC grade
82
were purchased from Merck (Darmstadt, Germany). Water was doubly distilled.
M
an
us
cr
75
Isofraxidin, istanbulin A and lasianthuslactone A were separated by our laboratory from the
84
herbal medicine Chloranthus henryi Hemsl, and their purities were 98.2%, 98.6% and 98.0%,
85
respectively.
The UV spectrometer used was UV-1750 (Shimadzu, Japan). The HPLC system was an
87
Agilent-1200 (Agilent, USA). 1H NMR spectra were recorded on a Varian VNMRS-500
88
93 94
The mixture was sparged with oxygen-free nitrogen for 15 min and sealed under vacuum. The
95
polymerization was carried out in a water bath at 60℃ for 24 hours. After the reaction, the
96
obtained polymers were collected by centrifugation at 10,000 rpm for 10 min. Methanol-acetic
97
acid (9:1, v v-1) was used to remove the andrographolide using the Soxhlet extraction until there
98
was no template molecule to be detected by UV (at 220 nm) in the eluate. The polymers were then
99
eluted by methanol to remove the remaining acetic acid and then dried under vacuum for 2 hours.
89 90 91 92
te
86
Ac ce p
d
83
spectrometer. The incubation oscillator was a vortex genie 2 (Scientific Industries, USA).
2.2 Synthesis of sesquiterpene lactones MIPs A specific amount of atractylenolide was dissolved in acetonitrile and then mixed in a 250
mL round bottom flask with the functional monomer (1-Viny) and the cross-linker (EDMA) according to the molar ratio of 1:5:8. Next, 23 mg of AIBN was added into the mixed solution.
4
Page 4 of 29
100
The costunolide lactone MIPs and the dehydrocostus lactone MIPs were prepared by the
101
same protocols as the atractylenolide MIPs, with specific conditions listed in Table 1. The
102
flowchart of MIP preparation can be seen in Figure 1.
104
2.3 Preparation of molecularly imprinted solid phase extraction column
ip t
103
Next, 100 mg of atractylenolide Ⅲ, costunolide lactone and dehydrocostus lactone were
106
packed into a solid phase extraction cartridge. At first, the polymers were pre-equilibrated with
107
methanol and then dried. After the preparation of MIP - SPE columns, the steps of loading,
108
washing and elution vary slightly according to the different experimental conditions (see section
109
2.4, 2.6 and 2.7). The process of MIP-SPE preparation is illustrated in Figure 2.
us
cr
105
an
110
2.4 Determination of maximum loading capacity and enrichment factor of SPE columns
112
2.4.1 Determination of maximum loading capacity and enrichment factor of SPE columns The standard solutions of atractylenolide Ⅲ, costunolide lactone and dehydrocostus lactone
114
were respectively prepared by methanol as a solvent with a concentration of 30 g mL-1. A 5 mL
115
atractylenolide Ⅲ standard solution was added gradually into the atractylenolide Ⅲ MIP-SPE
116
column, costunolide lactone MIP-SPE column, dehydrocostus lactone MIP-SPE column and C18
117
-SPE column respectively. The flow rate was set at 0.2 mL min-1 and the filtrate was collected.
118 119 120 121 122 123
te
d
113
Ac ce p
M
111
The SPE columns were eluted by 5 mL methanol and the eluate was collected and then the eluant was filtered by 0.22 μm filter for HPLC analysis. Standard solutions of costunolide lactone and dehydrocostus lactone were dealt with by costunolide lactone MIP-SPE columns and dehydrocostus lactone MIP-SPE columns respectively, and the remaining steps follow the same protocol as above.
2.4.2 The competitive adsorption experiments
124
MIP-SPE column was pre-treated with the solution of target components. One is template
125
molecule, the other is reference component. The reference component is also a competitive
126
compound. The composition and concentration of solutions that flow through each MIP-SPE
127
column are shown in table 3. Five mL of mixed solutions were added into the corresponding
128
MIP-SPE columns. The flow rate is 0.2 mL min-1 and the MIP-SPE columns are eluted with 5 mL
129
of methanol. Then, collect and filter the eluent respectively by 0.22µm filter membrane for HPLC 5
Page 5 of 29
130
analysis.
131 132
2.5 Preparation of adsorption test solution Next, 10 mg standard samples of atractylenolide Ⅲ, isofraxidin, istanbulin A and
134
lasianthuslactone A were weighed and placed in 10 mL flasks respectively and prepared by a stock
135
solution of methanol with a concentration of 1 mg mL-1. Then, 50 L of each of the four stock
136
solutions was mixed and diluted with methanol to 50 mL. The adsorption test solutions were
137
prepared and used for the following experiments.
cr
ip t
133
139
us
138 2.6 Selective adsorption experiments
A 1 mL adsorption test solution was added into the atractylenolide MIP-SPE column, the
141
costunolide lactone MIP-SPE column, the dehydrocostus lactone MIP-SPE column and the C18
142
column in turn. The flow rate was set at 0.2 mL·min-1. After the sample solution passing through
143
the SPE column, the SPE column was washed by 3 mL of 20% methanol. Then, 5 mL of methanol
144
was used to elute the SPE columns, and the eluent was collected and filtered by a 0.22 μm filter
145
for HPLC analysis.
147
2.7 Preparation of real samples
148
153
Ac ce p
te
146
d
M
an
140
154
and left overnight. After ultrasonic extraction for 30 min, the conical flask was cooled to room
155
temperature and weighed again. Finally, the methanol was used to supplement the weight loss and
156
then the solution was shook and filtered to prepare the test sample.
149 150 151 152
The extract of Chinese traditional medicine Radix Aucklandiae was prepared according to the
method of Chinese Pharmacopoeia, 2010 Edition [18]. The standard solution of costunolide lactone and dehydrocostus lactone were respectively prepared by methanol with a concentration of 0.1 mg·mL-1.
Next, 0.3 g of Radix Aucklandiae powder was weighed precisely and placed in a conical flask
and 50 mL methanol was add into the flask. Subsequently, the conical flask was sealed, weighed
157
1 mL extract mentioned above exactly was loaded into the dehydrocostus lactone and the
158
costunolide lactone MIP-SPE columns successively, and the flow rate was set at 0.2 mL·min-1;
159
When the sample solution reaches the SPE column, wash it with 3 mL of 20% methanol and elute 6
Page 6 of 29
160
it with 5 mL of methanol. Then, the eluent was collected and filtered respectively by 0.22µm filter
161
membrane for HPLC analysis.
163
2.8 Chromatographic conditions
164
2.8.1 Separation of the four compounds from the adsorption test solution
ip t
162
Eilte HypersiL ODS C18 (200 mm×4.6 mm, 5 µm, Dalian, China) column was used. Organic
166
phase acetonitrile (A) and aqueous phase water (B) were used as the mobile phases. The program
167
of gradient elution was as follows: 0 min，18% A; 10 min, 24% A; 19 min, 32%; A27 min, 37% A;
168
34 min, 47% A; and 45 min, 53% A. The flow rate was 1.0 mL·min-1, the injection volume was 20
169
μL. The column temperature was maintained at 30 ℃ and the detection wavelength was set at 220
170
nm.
171
2.8.2 Separation of the active compounds from the extract of Radix Aucklandiae
an
us
cr
165
Eilte HypersiL ODS C18 (200 mm×4.6 mm, 5 µm, Dalian, China) column was used; the
173
mobile phase is methanol: water = 65:35; The flow rate was 1.0 mL·min-1, and the injection
174
volume was 20 μL. The column temperature was maintained at 30 ℃ and the detection
175
wavelength was set at 225 nm..
177
2.9 Variable temperature 1H NMR spectroscopy experiment
178
Ac ce p
te
176
d
M
172
179 180 181 182 183
NMR spectra were recorded in CD3OD on a Varian VNMRS-500 spectrometer that operated
at 500 MHz for variable-temperature 1H NMR. Low temperature NMR experiments were carried out by cooling the probe with liquid nitrogen blow off. For the variable temperature studies the sample was placed in the probe and allowed to equilibrate to the required temperature for 20 min prior to shimming. The chemical shift was standardized using the residual protonated signal of the solvent or 0.03% TMS.
184 185
3. Results and discussion
186
3.1 Preparation and characterization of sesquiterpene lactones MIPs
187
The active sesquiterpene lactones, atractylenolide Ⅲ, costunolide lactone and dehydrocostus
188
lactone, were all selected as template molecules [19-21]. These compounds have a low polarity
189
and belong to the family of oleophyllic compounds. Their structures are shown in Figure 3a-c. 7
Page 7 of 29
Atractylenolide Ⅲ MIPs, costunolide lactone MIPs and dehydrocostus lactone MIPs were
191
prepared by precipitation polymerization under the experimental conditions described in Section
192
2.2. The functional monomers in precipitation polymerization were chosen based on the
193
intermolecular interactions between template molecules and functional monomers, which was
194
detected by UV spectroscopy. If there was strong intermolecular interaction, the MIPs would
195
demonstrate better adsorption performances [22-23]. Thus the functional monomer for
196
atractylenolide Ⅲ and dehydrocostus lactone was 1-vinylimidazole, and the functional monomer
197
for costunolide lactone was 4 - vinyl benzoic acid. EDMA and TRIM were chosen as the
198
cross-linkers (Detailed polymerization conditions shown in Table 1).
cr
ip t
190
Certain effecting factors including the template molecule ratio, functional monomers and
200
cross-linkers, amount and type of initiators, and the porogenic solvent ratio were systematically
201
studied on the morphology and adsorption performance of MIPs during the polymerization [24].
202
The polymerization conditions of three kinds of sesquiterpene lactones MIPs had been obtained
203
after optimization, as shown in Table 1. The morphologies of these sesquiterpene lactone MIPs
204
were characterized by scanning electron microscopy (SEM). Figure 4 shows an SEM image of
205
atractylenolide Ⅲ. The figure showed good spherical polymer particles (4-8 m diameter) size, so
206
ultimately the particle sizes were particularly suitable as a sorbent for SPE columns.
d
M
an
us
199
208
3.2 Performance evaluation of three sesquiterpene lactones MIP- SPE columns Three kinds of sesquiterpene lactone MIPs, atractylenolide , costunolide lactone and
215
Ac ce p
209
te
207
216
Where Cs was the concentration of the eluent (moI·L-1). Vs was the volume of the eluent (L).
217
WMIP was the weight of MIP (g). The enrichment factor was calculated as following [22]:
218
EF = Ccon C0
219
Ccon was the concentration of analyte in the concentrated phase (moI·L-1). C0 was the initial
210 211 212 213 214
dehydrocostus lactone, were used as sorbents to prepare MIP-SPE columns. To further evaluate adsorption performances of these sesquiterpene lactone MIP-SPE columns for sesquiterpene lactones components, two performance parameters of MIP-SPE columns were determined: maximum loading capacity and enrichment factor. The maximum loading capacity was calculated as following [25]:
Lmax = CsVs WMIP
(1)
(2)
8
Page 8 of 29
220
concentration of analyte within the sample (moI·L-1). Table 2 lists the maximum loading capacities and enrichment factors of three kinds of
222
MIP-SPE columns for their corresponding template molecules. Results show that the maximum
223
loading capacities of the three kinds of MIP-SPE column for their corresponding template
224
molecules were very similar about 4.0 µmoL·g-1. Likewise, the enrichment factor of each was up
225
to 80%. When the target molecule was replaced by similar compound with template molecules
226
in the structure, the enrichment factor decreased substantially to 60%. These results suggested that
227
three kinds of sesquiterpene lactones MIP-SPE columns all had good adsorption performance for
228
their corresponding template molecules.
The competitive adsorption experiment was aimed to investigate different MIP-SPE
230
columns’ performance on the selective adsorption for the target components, and distribution
231
coefficient KD and selectivity factor were used as parameters. Lasianthuslactone A was selected
232
as a competitive compound, due to its similar structure. The results demonstrate that KD and of
233
the template molecule were higher than those of the competitive compound MIP-SPE columns
234
listed in table 3. They also suggest that the binding sites and molecular cavities in the polymers
235
have good selective recognition and adsorption capacity to the target compound with the same
236
structure as template molecules. However, the competive compounds mainly relied on physical
237
adsorption due to its lack of imprinting process.
238
243 244
structure different from the template molecules as the enrichment object (target components). The
245
four compounds were isofraxidin, lasianthuslactone A, atractylenolide Ⅲ and istanbulin A
246
respectively. Among them, the first one compound was a coumarin compound and the last three
247
were sesquiterpene lactones compounds. Their structures are presented in Figure 3a, d, e and f.
248
HPLC chromatograms of the four compounds are shown in Figure 5.
239 240 241 242
249
te
d
M
an
229
Ac ce p
us
cr
ip t
221
3.3 Selective adsorption performance of sesquiterpene lactones MIP-SPE columns 3.3.1 The effect of structural differences in template molecules and target components on the adsorption performance of the MIP-SPE column In order to test the adsorption performance, we used a solution mixture composed of four
compounds, three of which had structures similar to the template molecules and one with a
In Figure 5, four main chromatographic peaks were isofraxidin (retention time 9.003 min), 9
Page 9 of 29
lasianthuslactoneA (retention time 19.528 min), atractylenolide (retention time 23.739 min) and
251
- (-) istanbulin A (retention time 37.715 min) in turn. The above adsorption test solution was
252
respectively enriched by three kinds of sesquiterpene lactones MIP-SPE columns according to
253
Section 2.6. Then the eluate was determined under the conditions of chromatography described in
254
Section 2.8 (The results are shown in Table 4).
As shown in Table 4, after the adsorption test solution flowed through the atractylenolide
256
MIP column, the most abundant component in the eluent was atractylenolide and its
257
concentration was up to 0.932 µg·mL-1. The concentrations of lasianthuslactone A and istanbulin
258
A (which have similar structures to atractylenolide ) were 0.066 µg·mL-1 and 0.104 µg·mL-1,
259
respectively. However, the concentration of isofraxidin with an obviously different structure from
260
atractylenolide was only 0.010 µg·mL-1. Next, after the adsorption test solution went through
261
the
262
lasianthuslactone A, atractylenolide and istanbulin were relatively higher and had maximum
263
concentrations of 0.416 µg·mL-1, 0.724 µg·mL-1 and 0.806 µg·mL-1, respectively. The
264
concentration of isofraxidin was only 0.048 µg·mL-1 when run through the dehydrocostus lactone
265
MIPs-SPE column, which was much less than the other three sesquiterpene lactone compounds.
266
Lastly, the enrichment concentrations of lasianthuslactone A, atractylenolide and istanbulin A,
267
having been pretreated by the costunolide lactone MIP-SPE column, were very close and were all
268
273 274
3.3.2 Analyzing the steric conformation of three template molecules
269 270 271 272
lactone
MIP-SPE column,
and
the
enrichment concentration of
te
d
M
dehydrocostus
an
us
cr
255
Ac ce p
ip t
250
about 0.6 μg·mL-1. Curiously, the concentration of the reference compound isofraxidin was as high as 0.216 μg·mL-1 when measured through costunolide lactone MIP-SPE columns after the adsorption test solution. This was completely different from the other two MIP-SPE columns. The results showed that the effects of molecular structural difference between the template molecules and target components were very significant on the adsorption performance for MIP-SPE column.
275
To further investigate the influence for the space conformation of the template molecules on
276
the adsorption performance of MIPs, 1H NMR spectra of three template molecules were
277
determined at the series of temperature from -20 to 50. As shown in Figure 6, the peaks in 1H
278
NMR spectrum of atractylenolide in CD3OD did not have any change when the temperature
279
changed from -20 to 50. The result suggests that the space conformation of atractylenolide is 10
Page 10 of 29
stable and exist only in one conformation. Similar result has been observed for dehydrocostus
281
lactone. However, the 1H NMR spectrum of costunolide lactone in CD3OD has showed a great
282
change. As shown in Figure 7, the peaks in the down field olefinic region and high field methyl
283
region had great differences when the temperature changed. The sharp peak for vinyl protons at H
284
4.79 (3H, J=11.5 Hz, 5H) at -20 changed as small broad peak at 50 in the downfield olefinic
285
region. Meanwhile, the peak height for methyl protons H 1.45 (d, s, 14H) changed greatly in the
286
high field as well. The above result suggests that there be a variety of stable conformation isomers
287
in the solution, and the inner ring and outer ring olefinic double bonds in the 10-mem lactone ring
288
costunolide lactone convert into each other between the cis and trans isomers. Together, the data
289
here hints that if there are morerings and olefinic double bonds either in the inside ring or in the
290
outer ring, the molecular steric conformation would be more unstable and multiple conformational
291
isomers would more possibly exist.
an
us
cr
ip t
280
M
292
3.3.3 The adsorption performance of MIP-SPE columns correlates with the structure of template
294
molecules and target components
When comparing the structures of the above three template molecules and four kinds of
296
target components, we found two rules: 1) If the steric conformations were stable, there would be
297
stronger selection and also specific adsorption perormances for corresponding MIPs. For the three
298
303 304
costunolide lactones is a ten-membered ring. Since seven and ten-membered ring sizes are
305
relatively large, the steric conformation of their molecules are more prone to torsion and can
306
pucker, which may lead to a smaller molecular ring size. On the other hand, there were a variety of
307
stable conformation isomers during the polymerization for the template molecule costunolide
308
lactone. Therefore, the selective adsorption performances using these molecules as template
309
molecules would decrease, and nonspecific adsorption performance increases accordingly. With
299 300 301 302
te
295
Ac ce p
d
293
kinds of sesquiterpene lactone MIPs, atractylenolide , a five-membered ring lactone with a stable steric conformation and strong lipophilicity, its corresponding MIP had strong adsorption performance and good selectivity. The enrichment factor of its MIP-SPE column for atractylenolide reached as high as 93.2%, but the enrichment factor of the column was very low for the six-membered ring lactone isofraxidin, which has major structural differences when compared with atractylenolide . The dehydrocostus lactones is a seven-membered ring and
11
Page 11 of 29
the increase of ring members, the nonspecific adsorption performance of MIPs can also increase
311
accordingly. The nonspecific adsorption of dehydrocostus lactone and costunolide lactones
312
MIP-SPE columns for isofraxidin reached 4.8% and 21.6% respectively. 2) When the target
313
molecule was consistent with the template molecule, its corresponding MIP would have the
314
greatest adsorption performance for MIPs. Moreover, when more hydrophilic functional groups
315
were introduced into target molecule, the weaker selective absorption performance was observed
316
for the MIPs. From Table 4, we can see that the adsorption performance of three kinds of
317
sesquiterpene lactone MIPs for istanbulin A were all higher than that adsorption for
318
lasianthuslactone A. This is because the number of hydroxyl introduced into the structure of
319
lasianthuslactone A is larger than that of hydroxyl introduced into the structure of istanbulin A.
320
an
us
cr
ip t
310
3.3.4 The contrast of three kinds of sesquiterpene lactone MIP-SPE columns and C18-SPE
322
columns in selective adsorption performance
As shown in Table 2, the enrichment factor of the C18-SPE column for atractylenolide was
324
only 60.1% which was significantly lower than that of the atractylenolide MIP-SPE column
325
(82.1%) and slightly lower than that of the costunolide lactone and dehydrocostus lactone
326
MIP-SPE columns. The concentration of four components including isofraxidin, lasianthuslactone
327
A, atractylenolide and istanbulin A in the eluate was increased to 88.2%, 60.4%, 42.4% and
328 329 330 331 332 333
te
d
323
Ac ce p
M
321
89.4% respectively, after the adsorption test solutions flowed through the C18-SPE column, as showed in Table 4. The results showed the C18-SPE column, as a classic column, had good adsorption performance with various ingredients although it is nonspecific adsorption. When compared with the C18-SPE column, the sesquiterpene lactone MIP-SPE column had better selectivity and stronger adsorption performances for various sesquiterpene lactone target components.
334
In sum, the adsorption performance of C18-SPE columns for various components was
335
superior to MIP-SPE columns, but the selective adsorption performance of MIP-SPE column for
336
the certain target components was much higher than that of the C18-SPE column. Therefore, the
337
C18-SPE column belongs to universal type of enrichment column, and MIP-SPE column belongs
338
to the specificity type of enrichment column. They should now be specifically selected based on
339
the experiment. 12
Page 12 of 29
340 341
3.4 Confirmation of real samples In order to further verify the results of the experiment, the extract of Chinese medicine Radix
343
Aucklandiae was tested as a real sample. The extract Radix Aucklandiae was passed through
344
costunolide lactone and dehydrocostus lactone MIP-SPE columns successively, and then the
345
eluent was detected and analyzed.
Figure 8a shows an HPLC chromatogram of Chinese medicine Radix Aucklandiae extract,
347
where costunolide and dehydrocostuslactone are the two major components. Figure 6b clearly
348
describes the dehydrocostus lactone with a high peak and an enrichment rate of 81.0% in the
349
elutent after the extract passing through dehydrocostus lactone MIP-SPE column. However, the
350
enrichment rate of this column for costunolide lactone was only 14.6%. Similarly, the costunolide
351
lactone in the eluent was also high and its enrichment rate reached 70.2% after the extract passed
352
through the costunolide lactone MIP-SPE column. Likewise, the enrichment rate of this column
353
for dehydrocostus lactone reached 30.0% in Figure 6c. These experimental results suggested that
354
costunolide lactone and dehydrocostus lactone MIPs have a large enrichment rate for target
355
components, when consistent with template molecules. As the ring system of dehydrocostus
356
lactone gets smaller than the ring system of costunolide lactone, the steric conformation of
357
dehydrocostus lactone becomes more stable than that of costunolide lactone. The enrichment
358
363 364
MIPs prepared by a ten-membered ring lactone. Thus, enrichment rates for costunolide lactone
365
MIPs for the dehydrocostus lactone were larger than those of the dehydrocostus lactone MIP for
366
the costunolide lactone. Thus, the enrichment rate of costunolide lactone MIPs for dehydrocostus
367
lactone was relatively high (30%). Of course, these adsorptions included two parts, selective
368
adsorption and nonspecific performance. The data was consistent with the results discussed above.
369
In summary, experimental results of real samples were entirely consistent with related
359 360 361 362
te
d
M
an
us
cr
346
Ac ce p
ip t
342
factor of the dehydrocostus lactone MIP-SPE column for the template molecule was higher than that of the costunolide lactone MIP-SPE column. Besides, the steric conformation of the costunolide lactone molecule with a ten-member ring was more prone to qaqatorsion and puckering. Also, the nonspecific adsorption of its MIP-SPE column for target molecules was relatively high. On the other hand, the molecular cavities of the dehydrocostus lactone MIP, prepared by seven-membered ring lactones, were smaller than those of the costunolide lactone
13
Page 13 of 29
370
regularities between molecular structures and the selective adsorption performances of MIPs in
371
our work.
372 4. Conclusions In this study, sesquiterpene lactone compounds with different structures were chosen as
375
template molecules to prepare MIPs, and these MIPs were used as a sorbent to prepare MIP-SPE
376
columns. The relationship between the selective adsorption performance of MIP-SPE columns and
377
the structures of template molecules and target molecules was explored preliminarily. Our results
378
showed that if there were more stable steric conformations of template molecules, then there
379
would be better selectivity and specific adsorption performance of their corresponding MIPs. Also,
380
certain MIPs had different adsorption performances for the template molecule analogues. When
381
the target molecule was consistent with the template molecule, MIPs reached a maximum
382
adsorption capacity. Furthermore, when more hydrophilic groups were introduced into the target
383
molecule, it resulted in weaker selective absorption performance level of various MIPs. If the
384
structure of target components and template molecules had major differences, the selective
385
adsorption performance of MIPs for it would be poor. When the steric conformation of template
386
molecules developed into multiple conformational isomers or easily puckered and created torsion,
387
its corresponding MIPs were easily lead to a strong nonspecific adsorption.
388
393 394
Acknowledgments
389 390 391 392
395
te
d
M
an
us
cr
ip t
374
Ac ce p
373
The correlation study between selective adsorption performance of various MIP-SPE
columns and the structures of target components is beneficial to the design and synthesis of MIPs enrichment materials with good selective adsorption properties. It provides new technology and methods for efficient separation of the active ingredients from traditional Chinese medicine. In the future, this technology will hopefully have a broad application prospect.
The authors would like to thank Dr. Zongwei Wang at Massachusetts General Hospital,
396
Harvard Medical School and Dr. Mark levine from USA for their kind help in manuscript
397
preparation, as well as Mr. Qiaoliang at Peking University school of Pharmaceutical Sciences for
398
his assistance in 1H NMR experiment. The authors acknowledge financial supports from National
399
Natural Science Foundation, PR China (Grant Nos. 81260690 and 81173657), the Education 14
Page 14 of 29
400
Research Project of Science and Technology Foundation, Jiangxi Province, PR China (Grant No.
401
GJJ13618), Jiangxi science and technology support project (Grant No. 20111BBG70004-2).
402 References
404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438
[1] Y. P. Duan, C.M. Dai, Y.L. Zhang, C. Ling. Anal. Chim. Acta. 758(2013)93. [2] G. G. Paniagua, H.P. Fernández, A.J. Durand. Anal. Bioanal. Chem. 394(2009)963. [3] T. Alizadeh, A. Akbari. Biosens. Bioelectron. 43(2013)321. [4] R.J. Ansell, J.K. Kuah, D. Wang, C.E. Jackson, K.D. Bartle, A.A. Clifford. J. Chromatogr. A 1264(2012)117. [5] G. Wulff, J. Liu. Acc Chem Res. 45(2012)239. [6] K.T. Stella, M. Panagiotis, N.A. Andreana, P. P. Vassilios. J. Chromatogr. A 1315(2013)15. [7] L. Zhu, L. Chen, X. Xu. Anal Chem. 75( 2003)6381. [8] F.F. Chen, R. Wang, Y. P. Shi. Talanta 89(2012)505. [9] X. Chen, Z. Zhang, X. Yang, J. Li, Y. Liu, H. Chen, W. Rao, S. Yao. Talanta 99(2012)959. [10] J. Luo, L. Zhang, D. Chen, P. Wang, J. Zhao, Y. Peng, S. Du, Z. Zhang. Analyst 137(2012)2891. [11] F.F. Chen, G.Y. Wang, Y.P. Shi. J. Sep. Sci. 34(2011)2602. [12] F. Tan, D. Sun, J. Gao, Q. Zhao, X. Wang, F. Teng, X. Quan, J. Chen. J. Hazard. Mater. 244-245(2013)750. [13] L. Ban, L. Zhao, B.L. Deng, Y.P. Huang, Z.S. Liu. Anal. Bioanal. Chem. 405(2013) 2245. [14] H. Shaikh, N. Memon, H. Khan, M.I. Bhanger, S.M. Nizamani. J. Chromatogr. A 1247(2012) 125. [15] M. Singh, A. Kumar, N. Tarannum. Anal. Bioanal. Chem. 405(2013)4245. [16] Z. Lin, F. Yang, X. He, X. Zhao, Y. Zhang. J. Chromatogr. A1216(2009) 8612. [17] W. Zhang, X.W.He, Y. Chen, W.Y. Li, Y.K. Zhang. Biosens. Bioelectron 31(2012)84. [18] Chinese Pharmacopoeia Commission, Pharmacopeia of the People’s Republic of China, Beijing, 2010 ed., 2010. [19] K.T. Wang, L.G. Chen, C.H. Wu, C.C. Chang, C.C. Wang. J. Pharm. Pharmacol. 62(2010) 381. [20] S.H. Choi, E. Im, H.K. Kang, J.H. Lee, H.S. Kwak, Y.T. Bae, H.J. Park, N.D. Kim. Cancer Lett. 227(2005)153. [21] H. Matsuda, T. Kagerura, I. Toguchida, H. Ueda, T. Morikawa, M. Yoshikawa. Life Sci.66 (2000)2151. [22] X.Y. Yin, Q.S. Liu, Y.F. Jiang, Y.M. Luo. Spectrochim. Acta A 79(2011)191. [23] X.Y. Yin, Y.M. Luo, J.J Fu,Y.Q. Zhong, Q.S. Liu. J. Sep. Sci. 35(2012)384. [24] X.Y. Yin, X.H Xu, Y.F. Jiang, Y.M. Luo, L.Y. Luo. Advanced Materials Research 287-290(2011)334. [25] J.P. Lai, X.W. He, Y. Jiang, F.Chen. Anal. Bioanal. Chem. 375(2003)264.
Ac ce p
te
d
M
an
us
cr
ip t
403
15
Page 15 of 29
438 Figure legends
440
1. Figure 1 The flowchart of atractylenolide Ⅲ MIP preparation.
441
2. Figure 2 The flowchart of MIP-SPE procedure.
442
3. Figure 3 Structures of sesquiterpene lactones template molecules and the adsorption test
443
compounds.
444
4. Figure 4 SEM of atractylenolide Ⅲ imprinted polymers with 3000 magnification.
445
5. Figure 5 HPLC chromatogram of the adsorption test solutions.
446
6. Figure 6 1H NMR spectrum of atractylenolide Ⅲ.
447
7. Figure 7 1H NMR spectrum of costunolide lactone.
448
8. Figure 8 HPLC chromatogram of the extract of Chinese medicine Radix Aucklandiae (a). HPLC
449
chromatogram of the eluent for the extract of Chinese medicine Radix Aucklandiae after passing
450
through dehydrocostus lactone MIP-SPE column (b). HPLC chromatogram of the eluent for the
451
extract of Chinese medicine Radix Aucklandiae after passing through costunolide lactone
452
MIP-SPE column (c).
an
us
cr
ip t
439
Table legends
455
1. Table1. The polymerization condition of sesquiterpene lactones MIPs by optimization.
456
2. Table 2. The results of the adsorption performance parameter for sesquiterpene lactones
457
MIP-SPE columns to target molecule.
458
3. Table 3. The results of the selective adsorption coefficient for sesquiterpene lactones MIP-SPE
459
columns to target molecule.
460
4. Table 4. The content results of the elutent after the adsorption test solution passing through
461
te
d
M
454
Ac ce p
453
sesquiterpene lactones MIP-SPE columns and C18-SPE column.
16
Page 16 of 29
*Highlights (for review)
Highlights 1. The regularity between structure and selective adsorption performance of MIP was explored. 2. The steric conformation of template molecules is related to the MIP adsorption performance. 3. The conformation isomers were verified by variable temperature 1H NMR spectroscopy.
Ac
ce pt
ed
M
an
us
cr
ip t
4. The hydrophilic functional groups can affect the selective absorption performance of MIP.
1
Page 17 of 29
an
us
cr
ip t
Figure
Ac ce pt e
d
M
Figure 1 The flowchart of atractylenolide Ⅲ MIP preparation
Page 18 of 29
ip t cr us an
Ac ce pt e
d
M
Figure 2 The flowchart of MIP-SPE procedure
Page 19 of 29
H
CH3 H
OH O O
O
H
H O
CH3
O b:Dehydrocostus lactone
O O
OCH3
e:LasianthuslactoneA
f:- -£©IstanbulinA
us
d:Isofraxidin
OH
O
（
O
cr
HO
CH2
c:Costunolide O H OH O
OH O
H3CO
H
ip t
a:AtractylenolideIII
O
Figure 3 Structures of sesquiterpene lactones template molecules and the adsorption test
Ac ce pt e
d
M
an
compounds.
Page 20 of 29
ip t cr us
Ac ce pt e
d
M
an
Figure 4 SEM of atractylenolide Ⅲ imprinted polymers with 3000 magnification.
Page 21 of 29
25
0 0
10
20
30
40
us
Retaintion(min)
ip t
-(-)istanbulin A isofraxdin atraylenoideIII lasianthuslactone A
cr
mAu
50
Ac ce pt e
d
M
an
Figure 5 HPLC chromatogram of the adsorption test solutions.
Page 22 of 29
ip t cr us an M
Ac ce pt e
d
Figure 6 1H NMR spectrum of atractylenolide Ⅲ.
Page 23 of 29
ip t cr us an
Ac ce pt e
d
M
Figure 7 1H NMR spectrum of costunolide lactone.
Page 24 of 29
ip t cr us
an
Figure 8 HPLC chromatogram of the extract of Chinese medicine Radix Aucklandiae (a). HPLC chromatogram of the eluent for the extract of Chinese medicine Radix Aucklandiae after passing through dehydrocostus lactone MIP-SPE column (b). HPLC chromatogram of the eluent for the
Ac ce pt e
d
column (c).
M
extract of Chinese medicine Radix Aucklandiae after passing through costunolide lactone MIP-SPE
Page 25 of 29
Tables
ip t
Table1. The polymerization condition of sesquiterpene lactones MIPs by optimization. Initiator Template functional Molar ratio of T, F and Cross-linker molecules monomer C （mg） atractylenolide 1-Viny EDMA 1:5:8 23 Ⅲ dehydrocostus 1-Viny EDMA 1:4:5 10 lactone costunolide 4 - vinyl TRIM 1:4:4 10 lactone benzoic acid
Ac
ce pt
ed
M
an
us
cr
T, F and C stand for template molecule, functional monomer and Cross-linker.
Page 26 of 29
Table 2. The results of the adsorption performance parameter for sesquiterpene lactones MIP-SPE columns to target molecule. Maximum loading capacity (µmoL g-1)
Enrichment factor
atractylenolide Ⅲ
4.1
82.1%
atractylenolide Ⅲ
dehydrocostus lactone
3.2
64.0%
atractylenolide Ⅲ
4.0
81.0%
dehydrocostus lactone
3.1
62.2%
3.9
80.1%
3.0
60.1%
ip t
cr
costunolide lactone
atractylenolide Ⅲ
Ac
ce pt
ed
M
C18
atractylenolide Ⅲ
us
costunolide lactone
Target molecule
（%）
an
MIPs-SPE column type
Page 27 of 29
Table 3. The results of the selective adsorption coefficient for sesquiterpene lactones MIP-SPE columns to target molecule.
dehydrocostus lactone costunolide lactone
Target molecule
atractylenolide Ⅲ
atractylenolide Ⅲ
31.6
+ lasianthuslactone A
lasianthuslactone A dehydrocostus lactone lasianthuslactone A costunolide lactone lasianthuslactone A
2.23
dehydrocostus lactone + lasianthuslactone A costunolide lactone + lasianthuslactone A
KD(mL/g)
28.8 13.1 28.0 18.5
14.2
2.20
ip t
atractylenolide Ⅲ
Mixture solution
1.51
cr
MIPs-SPE column type
* KD=Cp/Cs, Cp was the amount of target molecule bound to the MIP (mg·g-1); Cs was the
Ac
ce pt
ed
M
an
mixture solution all was 0.3 mgmL-1, V=5 mL, m=100mg.
us
concentration of target molecule in the sample solution (mgmL-1) . = KDi / KDj, KDi,KDj was distribution coefficient of template molecular and competitive compound. Cs of target molecule in the
Page 28 of 29
Table 4. The content results of the elutent after the adsorption test solution passing through sesquiterpene lactones MIP-SPE columns and C18-SPE column. Concentration （µg·mL-1） istanbulin A
1
1
1
1
0.010
0.066
0.932
0.104
0.048
0.416
0.724
0.216
0.628
0.652
0.882
0.604
0.424
0.806 0.658 0.894
Ac
ce pt
ed
M
an
MIPs-SPE column By dehydrocostus lactone MIPs-SPE column By costunolide lactone MIPs-SPE column By C18-SPE column
atractylenolide Ⅲ
ip t
By atractylenolide Ⅲ
lasianthuslactone A
cr
Initial solution
isofraxidin
us
The eluent
Page 29 of 29
S0021-9673(14)00809-7 http://dx.doi.org/doi:10.1016/j.chroma.2014.05.043 CHROMA 355435
To appear in:
Journal of Chromatography A
Received date: Revised date: Accepted date:
14-2-2014 4-5-2014 18-5-2014
Please cite this article as: X. Yin, Q. Liu, X. Ma, X. Zhou, M. Zhao, P. Tu, Correlation study between molecular structure of sesquiterpene lactones and the selective adsorption performance of molecularly imprinted polymers, Journal of Chromatography A (2014), http://dx.doi.org/10.1016/j.chroma.2014.05.043 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
*
2
the selective adsorption performance of molecularly imprinted polymers
Correlation study between molecular structure of sesquiterpene lactones and
3 Xiaoying Yina,1, Qingshan Liu b,1, Xingxia Maa, Xudong Zhouc, Mingbo Zhaoc, Pengfei Tuc*
5
a College of Pharmacy, Jiangxi University of Traditional Chinese Medicine, Nanchang 330004,
6
China
7
b National Research Center for Chinese Minority Medicine, Minzu University of China, Beijing
8
100081, China
9
c School of Pharmaceutical Sciences, Peking University Health Science Center, Beijing 100191,
cr
China.
us
10
ip t
4
Ac ce p
te
d
M
an
11
#
Qingshan Liu and Xiaoying Yin are co-first authors.
*Corresponding author. Tel.: +86 10 82802750. E-mail address: [email protected] 1
Page 1 of 29
11
Abstract We preliminarily report that the structure of template molecules and target components
13
correlates with the selective adsorption performance of molecularly imprinted polymer (MIPs) in
14
sesquiterpene lactones. Template molecules involved three categories of sesquiterpene lactones
15
with distinct ring systems: 5-mem lactone ring atractylenolide , 7-mem lactone ring
16
dehydrocostus lactone, and 10-mem lactone ring costunolide lactone, of which the conformations
17
were verified by variable-temperature 1H NMR spectroscopy. Reciprocal MIPs were prepared by
18
precipitation polymerization and employed as selective sorbents in the columns of solid phase
19
extraction (SPE). These columns were further used for enriching the mixed adsorption solution of
20
sesquiterpene lactone ingredients and reference components. Finally, the extract of Radix
21
Aucklandiae, a Chinese medicine herb, was used to verify the efficiency of this method. Our
22
results demonstrate that the steric conformational stability of molecules is associated with the
23
selective adsorption of their corresponding MIPs. We’ve further observed that the maximum
24
adsorption capacity occurs when the target molecule conformation is consistent with that of the
25
template molecule. The addition of more hydrophilic groups correlates with weaker adsorption of
26
MIPs. Our findings provide important information to help guide the selection of appropriate
27
template molecules for synthesis of MIPs with specific adsorption.
28
Keywords: Molecularly imprinted polymer; Sesquiterpene lactones; Selective adsorption
30 31 32 33 34
cr
us
an
M
d
te
Ac ce p
29
ip t
12
performance; Molecular structure; Variable-temperature 1H NMR spectroscopy
1. Introduction
Molecularly imprinted polymers (MIPs) are polymers that have been processed using a
molecular imprinting technique which leaves cavities in the polymer matrix that have affinity to a chosen "template" molecule. Due to their high selectivity, stability and efficiency, MIPs have been
35
applied to many fields, such as environmental analysis, immunoassays biosensors, catalysts and
36
chiral separation of drugs [1-5]. In recent years, molecular imprinting technique (MIT) has been
37
applied to the directional separation of active ingredients in traditional Chinese medicines [6-11].
38
MIT has been mostly focused on investigating the preparation, characterization, adsorption
39
properties and performance evaluation of MIPs [12-17], but further research regarding MIPs is 2
Page 2 of 29
rare. Several key questions about the application of MIT to Chinese medicine preparation remain
41
unanswered: 1) Which structures in active ingredients are suitable as a template molecules to
42
prepare MIPs? 2) What is the difference between the selective adsorption performances of MIPs
43
for template molecules and for analogous structural compounds? 3) What variables exist between
44
the structure of template molecules or target components and the adsorption performance of MIPs?
45
Elucidating the answer to one or several of these questions would make a significant contribution
46
to the field of active ingredient separation in traditional Chinese medicine.
In our recent research, more than 20 kinds of active ingredients in traditional Chinese
48
medicines were chosen as template molecules to prepare the corresponding MIPs by precipitation
49
polymerization. The experimental results showed that not all compounds were suitable for the
50
MIPs preparation, and that the selective adsorption performance of MIPs prepared by some
51
template molecules was poor. It was noted that the selective adsorption performance of MIPs
52
prepared by lipophilic template molecules was significantly better than MIPs prepared by
53
hydrophilic template molecules. This may be because the supramolecular interactions between the
54
template molecule and the functional monomer were not easily damaged. Similarly, the binding
55
sites with strong affinity to the target molecules on the molecular cavities of synthetic MIPs still
56
remained during the process of polymerization. A question as of yet unanswered is whether the
57
lipophilic compounds are suitable as template molecules.
58
63 64
with three similar sesquiterpene lactones ingredients and one reference compound. The stability of
65
template molecule conformation was tested by variable-temperature 1H NMR spectroscopy. The
66
relationship between MIPs structure and its selective adsorption performance was explored using
67
the Chinese medicine Radix Aucklandiae extract. Furthermore, the difference of the adsorption
68
performance for sesquiterpene lactones between MIP-SPE columns and C18-SPE columns was
69
detected by the adsorption test solutions. This study offers important references for effectively
59 60 61 62
te
d
M
an
us
47
Ac ce p
cr
ip t
40
In this study some lipophilic compounds were used as template molecules, and the
correlation between template molecule structures, target molecule structures and the selective adsorption performance of the corresponding MIPs was explored. Three types of sesquiterpene lactones with different structures and activities were selected as template molecules to prepare the corresponding MIPs by precipitation polymerization and then used to prepare MIP-SPE columns. These columns were used for enrichment and separation of the adsorption test solutions mixed
3
Page 3 of 29
70
designing and synthesizing MIPs separation materials with acceptable selective adsorption
71
properties.
72 73
2.
74
2.1 Materials and apparatus
ip t
Experimental
All reagents were of analytical grade or better. Atractylenolide , costunolide lactone and
76
dehydrocostus lactone were purchased from Zelang Pharmaceutical Co., Ltd. (Nanjing, China,
77
more than 99% purity). 1-vinylimidazole (1-Viny) and 4-vinylbenzoic acid were obtained from
78
Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Ethylene glycol dimethacrylate
79
(EDMA) and trimethylolpropane trimethacrylate (TRIM) were purchased from TCI Development
80
Co., Ltd. (Shanghai, China). 2’, 2-Azo-bis-iso-butyronitrile (AIBN) was Shanghai No. 4 Reagent
81
& H.V. Chemical Co., Ltd. (Shanghai, China) product. Acetonitrile and methanol of HPLC grade
82
were purchased from Merck (Darmstadt, Germany). Water was doubly distilled.
M
an
us
cr
75
Isofraxidin, istanbulin A and lasianthuslactone A were separated by our laboratory from the
84
herbal medicine Chloranthus henryi Hemsl, and their purities were 98.2%, 98.6% and 98.0%,
85
respectively.
The UV spectrometer used was UV-1750 (Shimadzu, Japan). The HPLC system was an
87
Agilent-1200 (Agilent, USA). 1H NMR spectra were recorded on a Varian VNMRS-500
88
93 94
The mixture was sparged with oxygen-free nitrogen for 15 min and sealed under vacuum. The
95
polymerization was carried out in a water bath at 60℃ for 24 hours. After the reaction, the
96
obtained polymers were collected by centrifugation at 10,000 rpm for 10 min. Methanol-acetic
97
acid (9:1, v v-1) was used to remove the andrographolide using the Soxhlet extraction until there
98
was no template molecule to be detected by UV (at 220 nm) in the eluate. The polymers were then
99
eluted by methanol to remove the remaining acetic acid and then dried under vacuum for 2 hours.
89 90 91 92
te
86
Ac ce p
d
83
spectrometer. The incubation oscillator was a vortex genie 2 (Scientific Industries, USA).
2.2 Synthesis of sesquiterpene lactones MIPs A specific amount of atractylenolide was dissolved in acetonitrile and then mixed in a 250
mL round bottom flask with the functional monomer (1-Viny) and the cross-linker (EDMA) according to the molar ratio of 1:5:8. Next, 23 mg of AIBN was added into the mixed solution.
4
Page 4 of 29
100
The costunolide lactone MIPs and the dehydrocostus lactone MIPs were prepared by the
101
same protocols as the atractylenolide MIPs, with specific conditions listed in Table 1. The
102
flowchart of MIP preparation can be seen in Figure 1.
104
2.3 Preparation of molecularly imprinted solid phase extraction column
ip t
103
Next, 100 mg of atractylenolide Ⅲ, costunolide lactone and dehydrocostus lactone were
106
packed into a solid phase extraction cartridge. At first, the polymers were pre-equilibrated with
107
methanol and then dried. After the preparation of MIP - SPE columns, the steps of loading,
108
washing and elution vary slightly according to the different experimental conditions (see section
109
2.4, 2.6 and 2.7). The process of MIP-SPE preparation is illustrated in Figure 2.
us
cr
105
an
110
2.4 Determination of maximum loading capacity and enrichment factor of SPE columns
112
2.4.1 Determination of maximum loading capacity and enrichment factor of SPE columns The standard solutions of atractylenolide Ⅲ, costunolide lactone and dehydrocostus lactone
114
were respectively prepared by methanol as a solvent with a concentration of 30 g mL-1. A 5 mL
115
atractylenolide Ⅲ standard solution was added gradually into the atractylenolide Ⅲ MIP-SPE
116
column, costunolide lactone MIP-SPE column, dehydrocostus lactone MIP-SPE column and C18
117
-SPE column respectively. The flow rate was set at 0.2 mL min-1 and the filtrate was collected.
118 119 120 121 122 123
te
d
113
Ac ce p
M
111
The SPE columns were eluted by 5 mL methanol and the eluate was collected and then the eluant was filtered by 0.22 μm filter for HPLC analysis. Standard solutions of costunolide lactone and dehydrocostus lactone were dealt with by costunolide lactone MIP-SPE columns and dehydrocostus lactone MIP-SPE columns respectively, and the remaining steps follow the same protocol as above.
2.4.2 The competitive adsorption experiments
124
MIP-SPE column was pre-treated with the solution of target components. One is template
125
molecule, the other is reference component. The reference component is also a competitive
126
compound. The composition and concentration of solutions that flow through each MIP-SPE
127
column are shown in table 3. Five mL of mixed solutions were added into the corresponding
128
MIP-SPE columns. The flow rate is 0.2 mL min-1 and the MIP-SPE columns are eluted with 5 mL
129
of methanol. Then, collect and filter the eluent respectively by 0.22µm filter membrane for HPLC 5
Page 5 of 29
130
analysis.
131 132
2.5 Preparation of adsorption test solution Next, 10 mg standard samples of atractylenolide Ⅲ, isofraxidin, istanbulin A and
134
lasianthuslactone A were weighed and placed in 10 mL flasks respectively and prepared by a stock
135
solution of methanol with a concentration of 1 mg mL-1. Then, 50 L of each of the four stock
136
solutions was mixed and diluted with methanol to 50 mL. The adsorption test solutions were
137
prepared and used for the following experiments.
cr
ip t
133
139
us
138 2.6 Selective adsorption experiments
A 1 mL adsorption test solution was added into the atractylenolide MIP-SPE column, the
141
costunolide lactone MIP-SPE column, the dehydrocostus lactone MIP-SPE column and the C18
142
column in turn. The flow rate was set at 0.2 mL·min-1. After the sample solution passing through
143
the SPE column, the SPE column was washed by 3 mL of 20% methanol. Then, 5 mL of methanol
144
was used to elute the SPE columns, and the eluent was collected and filtered by a 0.22 μm filter
145
for HPLC analysis.
147
2.7 Preparation of real samples
148
153
Ac ce p
te
146
d
M
an
140
154
and left overnight. After ultrasonic extraction for 30 min, the conical flask was cooled to room
155
temperature and weighed again. Finally, the methanol was used to supplement the weight loss and
156
then the solution was shook and filtered to prepare the test sample.
149 150 151 152
The extract of Chinese traditional medicine Radix Aucklandiae was prepared according to the
method of Chinese Pharmacopoeia, 2010 Edition [18]. The standard solution of costunolide lactone and dehydrocostus lactone were respectively prepared by methanol with a concentration of 0.1 mg·mL-1.
Next, 0.3 g of Radix Aucklandiae powder was weighed precisely and placed in a conical flask
and 50 mL methanol was add into the flask. Subsequently, the conical flask was sealed, weighed
157
1 mL extract mentioned above exactly was loaded into the dehydrocostus lactone and the
158
costunolide lactone MIP-SPE columns successively, and the flow rate was set at 0.2 mL·min-1;
159
When the sample solution reaches the SPE column, wash it with 3 mL of 20% methanol and elute 6
Page 6 of 29
160
it with 5 mL of methanol. Then, the eluent was collected and filtered respectively by 0.22µm filter
161
membrane for HPLC analysis.
163
2.8 Chromatographic conditions
164
2.8.1 Separation of the four compounds from the adsorption test solution
ip t
162
Eilte HypersiL ODS C18 (200 mm×4.6 mm, 5 µm, Dalian, China) column was used. Organic
166
phase acetonitrile (A) and aqueous phase water (B) were used as the mobile phases. The program
167
of gradient elution was as follows: 0 min，18% A; 10 min, 24% A; 19 min, 32%; A27 min, 37% A;
168
34 min, 47% A; and 45 min, 53% A. The flow rate was 1.0 mL·min-1, the injection volume was 20
169
μL. The column temperature was maintained at 30 ℃ and the detection wavelength was set at 220
170
nm.
171
2.8.2 Separation of the active compounds from the extract of Radix Aucklandiae
an
us
cr
165
Eilte HypersiL ODS C18 (200 mm×4.6 mm, 5 µm, Dalian, China) column was used; the
173
mobile phase is methanol: water = 65:35; The flow rate was 1.0 mL·min-1, and the injection
174
volume was 20 μL. The column temperature was maintained at 30 ℃ and the detection
175
wavelength was set at 225 nm..
177
2.9 Variable temperature 1H NMR spectroscopy experiment
178
Ac ce p
te
176
d
M
172
179 180 181 182 183
NMR spectra were recorded in CD3OD on a Varian VNMRS-500 spectrometer that operated
at 500 MHz for variable-temperature 1H NMR. Low temperature NMR experiments were carried out by cooling the probe with liquid nitrogen blow off. For the variable temperature studies the sample was placed in the probe and allowed to equilibrate to the required temperature for 20 min prior to shimming. The chemical shift was standardized using the residual protonated signal of the solvent or 0.03% TMS.
184 185
3. Results and discussion
186
3.1 Preparation and characterization of sesquiterpene lactones MIPs
187
The active sesquiterpene lactones, atractylenolide Ⅲ, costunolide lactone and dehydrocostus
188
lactone, were all selected as template molecules [19-21]. These compounds have a low polarity
189
and belong to the family of oleophyllic compounds. Their structures are shown in Figure 3a-c. 7
Page 7 of 29
Atractylenolide Ⅲ MIPs, costunolide lactone MIPs and dehydrocostus lactone MIPs were
191
prepared by precipitation polymerization under the experimental conditions described in Section
192
2.2. The functional monomers in precipitation polymerization were chosen based on the
193
intermolecular interactions between template molecules and functional monomers, which was
194
detected by UV spectroscopy. If there was strong intermolecular interaction, the MIPs would
195
demonstrate better adsorption performances [22-23]. Thus the functional monomer for
196
atractylenolide Ⅲ and dehydrocostus lactone was 1-vinylimidazole, and the functional monomer
197
for costunolide lactone was 4 - vinyl benzoic acid. EDMA and TRIM were chosen as the
198
cross-linkers (Detailed polymerization conditions shown in Table 1).
cr
ip t
190
Certain effecting factors including the template molecule ratio, functional monomers and
200
cross-linkers, amount and type of initiators, and the porogenic solvent ratio were systematically
201
studied on the morphology and adsorption performance of MIPs during the polymerization [24].
202
The polymerization conditions of three kinds of sesquiterpene lactones MIPs had been obtained
203
after optimization, as shown in Table 1. The morphologies of these sesquiterpene lactone MIPs
204
were characterized by scanning electron microscopy (SEM). Figure 4 shows an SEM image of
205
atractylenolide Ⅲ. The figure showed good spherical polymer particles (4-8 m diameter) size, so
206
ultimately the particle sizes were particularly suitable as a sorbent for SPE columns.
d
M
an
us
199
208
3.2 Performance evaluation of three sesquiterpene lactones MIP- SPE columns Three kinds of sesquiterpene lactone MIPs, atractylenolide , costunolide lactone and
215
Ac ce p
209
te
207
216
Where Cs was the concentration of the eluent (moI·L-1). Vs was the volume of the eluent (L).
217
WMIP was the weight of MIP (g). The enrichment factor was calculated as following [22]:
218
EF = Ccon C0
219
Ccon was the concentration of analyte in the concentrated phase (moI·L-1). C0 was the initial
210 211 212 213 214
dehydrocostus lactone, were used as sorbents to prepare MIP-SPE columns. To further evaluate adsorption performances of these sesquiterpene lactone MIP-SPE columns for sesquiterpene lactones components, two performance parameters of MIP-SPE columns were determined: maximum loading capacity and enrichment factor. The maximum loading capacity was calculated as following [25]:
Lmax = CsVs WMIP
(1)
(2)
8
Page 8 of 29
220
concentration of analyte within the sample (moI·L-1). Table 2 lists the maximum loading capacities and enrichment factors of three kinds of
222
MIP-SPE columns for their corresponding template molecules. Results show that the maximum
223
loading capacities of the three kinds of MIP-SPE column for their corresponding template
224
molecules were very similar about 4.0 µmoL·g-1. Likewise, the enrichment factor of each was up
225
to 80%. When the target molecule was replaced by similar compound with template molecules
226
in the structure, the enrichment factor decreased substantially to 60%. These results suggested that
227
three kinds of sesquiterpene lactones MIP-SPE columns all had good adsorption performance for
228
their corresponding template molecules.
The competitive adsorption experiment was aimed to investigate different MIP-SPE
230
columns’ performance on the selective adsorption for the target components, and distribution
231
coefficient KD and selectivity factor were used as parameters. Lasianthuslactone A was selected
232
as a competitive compound, due to its similar structure. The results demonstrate that KD and of
233
the template molecule were higher than those of the competitive compound MIP-SPE columns
234
listed in table 3. They also suggest that the binding sites and molecular cavities in the polymers
235
have good selective recognition and adsorption capacity to the target compound with the same
236
structure as template molecules. However, the competive compounds mainly relied on physical
237
adsorption due to its lack of imprinting process.
238
243 244
structure different from the template molecules as the enrichment object (target components). The
245
four compounds were isofraxidin, lasianthuslactone A, atractylenolide Ⅲ and istanbulin A
246
respectively. Among them, the first one compound was a coumarin compound and the last three
247
were sesquiterpene lactones compounds. Their structures are presented in Figure 3a, d, e and f.
248
HPLC chromatograms of the four compounds are shown in Figure 5.
239 240 241 242
249
te
d
M
an
229
Ac ce p
us
cr
ip t
221
3.3 Selective adsorption performance of sesquiterpene lactones MIP-SPE columns 3.3.1 The effect of structural differences in template molecules and target components on the adsorption performance of the MIP-SPE column In order to test the adsorption performance, we used a solution mixture composed of four
compounds, three of which had structures similar to the template molecules and one with a
In Figure 5, four main chromatographic peaks were isofraxidin (retention time 9.003 min), 9
Page 9 of 29
lasianthuslactoneA (retention time 19.528 min), atractylenolide (retention time 23.739 min) and
251
- (-) istanbulin A (retention time 37.715 min) in turn. The above adsorption test solution was
252
respectively enriched by three kinds of sesquiterpene lactones MIP-SPE columns according to
253
Section 2.6. Then the eluate was determined under the conditions of chromatography described in
254
Section 2.8 (The results are shown in Table 4).
As shown in Table 4, after the adsorption test solution flowed through the atractylenolide
256
MIP column, the most abundant component in the eluent was atractylenolide and its
257
concentration was up to 0.932 µg·mL-1. The concentrations of lasianthuslactone A and istanbulin
258
A (which have similar structures to atractylenolide ) were 0.066 µg·mL-1 and 0.104 µg·mL-1,
259
respectively. However, the concentration of isofraxidin with an obviously different structure from
260
atractylenolide was only 0.010 µg·mL-1. Next, after the adsorption test solution went through
261
the
262
lasianthuslactone A, atractylenolide and istanbulin were relatively higher and had maximum
263
concentrations of 0.416 µg·mL-1, 0.724 µg·mL-1 and 0.806 µg·mL-1, respectively. The
264
concentration of isofraxidin was only 0.048 µg·mL-1 when run through the dehydrocostus lactone
265
MIPs-SPE column, which was much less than the other three sesquiterpene lactone compounds.
266
Lastly, the enrichment concentrations of lasianthuslactone A, atractylenolide and istanbulin A,
267
having been pretreated by the costunolide lactone MIP-SPE column, were very close and were all
268
273 274
3.3.2 Analyzing the steric conformation of three template molecules
269 270 271 272
lactone
MIP-SPE column,
and
the
enrichment concentration of
te
d
M
dehydrocostus
an
us
cr
255
Ac ce p
ip t
250
about 0.6 μg·mL-1. Curiously, the concentration of the reference compound isofraxidin was as high as 0.216 μg·mL-1 when measured through costunolide lactone MIP-SPE columns after the adsorption test solution. This was completely different from the other two MIP-SPE columns. The results showed that the effects of molecular structural difference between the template molecules and target components were very significant on the adsorption performance for MIP-SPE column.
275
To further investigate the influence for the space conformation of the template molecules on
276
the adsorption performance of MIPs, 1H NMR spectra of three template molecules were
277
determined at the series of temperature from -20 to 50. As shown in Figure 6, the peaks in 1H
278
NMR spectrum of atractylenolide in CD3OD did not have any change when the temperature
279
changed from -20 to 50. The result suggests that the space conformation of atractylenolide is 10
Page 10 of 29
stable and exist only in one conformation. Similar result has been observed for dehydrocostus
281
lactone. However, the 1H NMR spectrum of costunolide lactone in CD3OD has showed a great
282
change. As shown in Figure 7, the peaks in the down field olefinic region and high field methyl
283
region had great differences when the temperature changed. The sharp peak for vinyl protons at H
284
4.79 (3H, J=11.5 Hz, 5H) at -20 changed as small broad peak at 50 in the downfield olefinic
285
region. Meanwhile, the peak height for methyl protons H 1.45 (d, s, 14H) changed greatly in the
286
high field as well. The above result suggests that there be a variety of stable conformation isomers
287
in the solution, and the inner ring and outer ring olefinic double bonds in the 10-mem lactone ring
288
costunolide lactone convert into each other between the cis and trans isomers. Together, the data
289
here hints that if there are morerings and olefinic double bonds either in the inside ring or in the
290
outer ring, the molecular steric conformation would be more unstable and multiple conformational
291
isomers would more possibly exist.
an
us
cr
ip t
280
M
292
3.3.3 The adsorption performance of MIP-SPE columns correlates with the structure of template
294
molecules and target components
When comparing the structures of the above three template molecules and four kinds of
296
target components, we found two rules: 1) If the steric conformations were stable, there would be
297
stronger selection and also specific adsorption perormances for corresponding MIPs. For the three
298
303 304
costunolide lactones is a ten-membered ring. Since seven and ten-membered ring sizes are
305
relatively large, the steric conformation of their molecules are more prone to torsion and can
306
pucker, which may lead to a smaller molecular ring size. On the other hand, there were a variety of
307
stable conformation isomers during the polymerization for the template molecule costunolide
308
lactone. Therefore, the selective adsorption performances using these molecules as template
309
molecules would decrease, and nonspecific adsorption performance increases accordingly. With
299 300 301 302
te
295
Ac ce p
d
293
kinds of sesquiterpene lactone MIPs, atractylenolide , a five-membered ring lactone with a stable steric conformation and strong lipophilicity, its corresponding MIP had strong adsorption performance and good selectivity. The enrichment factor of its MIP-SPE column for atractylenolide reached as high as 93.2%, but the enrichment factor of the column was very low for the six-membered ring lactone isofraxidin, which has major structural differences when compared with atractylenolide . The dehydrocostus lactones is a seven-membered ring and
11
Page 11 of 29
the increase of ring members, the nonspecific adsorption performance of MIPs can also increase
311
accordingly. The nonspecific adsorption of dehydrocostus lactone and costunolide lactones
312
MIP-SPE columns for isofraxidin reached 4.8% and 21.6% respectively. 2) When the target
313
molecule was consistent with the template molecule, its corresponding MIP would have the
314
greatest adsorption performance for MIPs. Moreover, when more hydrophilic functional groups
315
were introduced into target molecule, the weaker selective absorption performance was observed
316
for the MIPs. From Table 4, we can see that the adsorption performance of three kinds of
317
sesquiterpene lactone MIPs for istanbulin A were all higher than that adsorption for
318
lasianthuslactone A. This is because the number of hydroxyl introduced into the structure of
319
lasianthuslactone A is larger than that of hydroxyl introduced into the structure of istanbulin A.
320
an
us
cr
ip t
310
3.3.4 The contrast of three kinds of sesquiterpene lactone MIP-SPE columns and C18-SPE
322
columns in selective adsorption performance
As shown in Table 2, the enrichment factor of the C18-SPE column for atractylenolide was
324
only 60.1% which was significantly lower than that of the atractylenolide MIP-SPE column
325
(82.1%) and slightly lower than that of the costunolide lactone and dehydrocostus lactone
326
MIP-SPE columns. The concentration of four components including isofraxidin, lasianthuslactone
327
A, atractylenolide and istanbulin A in the eluate was increased to 88.2%, 60.4%, 42.4% and
328 329 330 331 332 333
te
d
323
Ac ce p
M
321
89.4% respectively, after the adsorption test solutions flowed through the C18-SPE column, as showed in Table 4. The results showed the C18-SPE column, as a classic column, had good adsorption performance with various ingredients although it is nonspecific adsorption. When compared with the C18-SPE column, the sesquiterpene lactone MIP-SPE column had better selectivity and stronger adsorption performances for various sesquiterpene lactone target components.
334
In sum, the adsorption performance of C18-SPE columns for various components was
335
superior to MIP-SPE columns, but the selective adsorption performance of MIP-SPE column for
336
the certain target components was much higher than that of the C18-SPE column. Therefore, the
337
C18-SPE column belongs to universal type of enrichment column, and MIP-SPE column belongs
338
to the specificity type of enrichment column. They should now be specifically selected based on
339
the experiment. 12
Page 12 of 29
340 341
3.4 Confirmation of real samples In order to further verify the results of the experiment, the extract of Chinese medicine Radix
343
Aucklandiae was tested as a real sample. The extract Radix Aucklandiae was passed through
344
costunolide lactone and dehydrocostus lactone MIP-SPE columns successively, and then the
345
eluent was detected and analyzed.
Figure 8a shows an HPLC chromatogram of Chinese medicine Radix Aucklandiae extract,
347
where costunolide and dehydrocostuslactone are the two major components. Figure 6b clearly
348
describes the dehydrocostus lactone with a high peak and an enrichment rate of 81.0% in the
349
elutent after the extract passing through dehydrocostus lactone MIP-SPE column. However, the
350
enrichment rate of this column for costunolide lactone was only 14.6%. Similarly, the costunolide
351
lactone in the eluent was also high and its enrichment rate reached 70.2% after the extract passed
352
through the costunolide lactone MIP-SPE column. Likewise, the enrichment rate of this column
353
for dehydrocostus lactone reached 30.0% in Figure 6c. These experimental results suggested that
354
costunolide lactone and dehydrocostus lactone MIPs have a large enrichment rate for target
355
components, when consistent with template molecules. As the ring system of dehydrocostus
356
lactone gets smaller than the ring system of costunolide lactone, the steric conformation of
357
dehydrocostus lactone becomes more stable than that of costunolide lactone. The enrichment
358
363 364
MIPs prepared by a ten-membered ring lactone. Thus, enrichment rates for costunolide lactone
365
MIPs for the dehydrocostus lactone were larger than those of the dehydrocostus lactone MIP for
366
the costunolide lactone. Thus, the enrichment rate of costunolide lactone MIPs for dehydrocostus
367
lactone was relatively high (30%). Of course, these adsorptions included two parts, selective
368
adsorption and nonspecific performance. The data was consistent with the results discussed above.
369
In summary, experimental results of real samples were entirely consistent with related
359 360 361 362
te
d
M
an
us
cr
346
Ac ce p
ip t
342
factor of the dehydrocostus lactone MIP-SPE column for the template molecule was higher than that of the costunolide lactone MIP-SPE column. Besides, the steric conformation of the costunolide lactone molecule with a ten-member ring was more prone to qaqatorsion and puckering. Also, the nonspecific adsorption of its MIP-SPE column for target molecules was relatively high. On the other hand, the molecular cavities of the dehydrocostus lactone MIP, prepared by seven-membered ring lactones, were smaller than those of the costunolide lactone
13
Page 13 of 29
370
regularities between molecular structures and the selective adsorption performances of MIPs in
371
our work.
372 4. Conclusions In this study, sesquiterpene lactone compounds with different structures were chosen as
375
template molecules to prepare MIPs, and these MIPs were used as a sorbent to prepare MIP-SPE
376
columns. The relationship between the selective adsorption performance of MIP-SPE columns and
377
the structures of template molecules and target molecules was explored preliminarily. Our results
378
showed that if there were more stable steric conformations of template molecules, then there
379
would be better selectivity and specific adsorption performance of their corresponding MIPs. Also,
380
certain MIPs had different adsorption performances for the template molecule analogues. When
381
the target molecule was consistent with the template molecule, MIPs reached a maximum
382
adsorption capacity. Furthermore, when more hydrophilic groups were introduced into the target
383
molecule, it resulted in weaker selective absorption performance level of various MIPs. If the
384
structure of target components and template molecules had major differences, the selective
385
adsorption performance of MIPs for it would be poor. When the steric conformation of template
386
molecules developed into multiple conformational isomers or easily puckered and created torsion,
387
its corresponding MIPs were easily lead to a strong nonspecific adsorption.
388
393 394
Acknowledgments
389 390 391 392
395
te
d
M
an
us
cr
ip t
374
Ac ce p
373
The correlation study between selective adsorption performance of various MIP-SPE
columns and the structures of target components is beneficial to the design and synthesis of MIPs enrichment materials with good selective adsorption properties. It provides new technology and methods for efficient separation of the active ingredients from traditional Chinese medicine. In the future, this technology will hopefully have a broad application prospect.
The authors would like to thank Dr. Zongwei Wang at Massachusetts General Hospital,
396
Harvard Medical School and Dr. Mark levine from USA for their kind help in manuscript
397
preparation, as well as Mr. Qiaoliang at Peking University school of Pharmaceutical Sciences for
398
his assistance in 1H NMR experiment. The authors acknowledge financial supports from National
399
Natural Science Foundation, PR China (Grant Nos. 81260690 and 81173657), the Education 14
Page 14 of 29
400
Research Project of Science and Technology Foundation, Jiangxi Province, PR China (Grant No.
401
GJJ13618), Jiangxi science and technology support project (Grant No. 20111BBG70004-2).
402 References
404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438
[1] Y. P. Duan, C.M. Dai, Y.L. Zhang, C. Ling. Anal. Chim. Acta. 758(2013)93. [2] G. G. Paniagua, H.P. Fernández, A.J. Durand. Anal. Bioanal. Chem. 394(2009)963. [3] T. Alizadeh, A. Akbari. Biosens. Bioelectron. 43(2013)321. [4] R.J. Ansell, J.K. Kuah, D. Wang, C.E. Jackson, K.D. Bartle, A.A. Clifford. J. Chromatogr. A 1264(2012)117. [5] G. Wulff, J. Liu. Acc Chem Res. 45(2012)239. [6] K.T. Stella, M. Panagiotis, N.A. Andreana, P. P. Vassilios. J. Chromatogr. A 1315(2013)15. [7] L. Zhu, L. Chen, X. Xu. Anal Chem. 75( 2003)6381. [8] F.F. Chen, R. Wang, Y. P. Shi. Talanta 89(2012)505. [9] X. Chen, Z. Zhang, X. Yang, J. Li, Y. Liu, H. Chen, W. Rao, S. Yao. Talanta 99(2012)959. [10] J. Luo, L. Zhang, D. Chen, P. Wang, J. Zhao, Y. Peng, S. Du, Z. Zhang. Analyst 137(2012)2891. [11] F.F. Chen, G.Y. Wang, Y.P. Shi. J. Sep. Sci. 34(2011)2602. [12] F. Tan, D. Sun, J. Gao, Q. Zhao, X. Wang, F. Teng, X. Quan, J. Chen. J. Hazard. Mater. 244-245(2013)750. [13] L. Ban, L. Zhao, B.L. Deng, Y.P. Huang, Z.S. Liu. Anal. Bioanal. Chem. 405(2013) 2245. [14] H. Shaikh, N. Memon, H. Khan, M.I. Bhanger, S.M. Nizamani. J. Chromatogr. A 1247(2012) 125. [15] M. Singh, A. Kumar, N. Tarannum. Anal. Bioanal. Chem. 405(2013)4245. [16] Z. Lin, F. Yang, X. He, X. Zhao, Y. Zhang. J. Chromatogr. A1216(2009) 8612. [17] W. Zhang, X.W.He, Y. Chen, W.Y. Li, Y.K. Zhang. Biosens. Bioelectron 31(2012)84. [18] Chinese Pharmacopoeia Commission, Pharmacopeia of the People’s Republic of China, Beijing, 2010 ed., 2010. [19] K.T. Wang, L.G. Chen, C.H. Wu, C.C. Chang, C.C. Wang. J. Pharm. Pharmacol. 62(2010) 381. [20] S.H. Choi, E. Im, H.K. Kang, J.H. Lee, H.S. Kwak, Y.T. Bae, H.J. Park, N.D. Kim. Cancer Lett. 227(2005)153. [21] H. Matsuda, T. Kagerura, I. Toguchida, H. Ueda, T. Morikawa, M. Yoshikawa. Life Sci.66 (2000)2151. [22] X.Y. Yin, Q.S. Liu, Y.F. Jiang, Y.M. Luo. Spectrochim. Acta A 79(2011)191. [23] X.Y. Yin, Y.M. Luo, J.J Fu,Y.Q. Zhong, Q.S. Liu. J. Sep. Sci. 35(2012)384. [24] X.Y. Yin, X.H Xu, Y.F. Jiang, Y.M. Luo, L.Y. Luo. Advanced Materials Research 287-290(2011)334. [25] J.P. Lai, X.W. He, Y. Jiang, F.Chen. Anal. Bioanal. Chem. 375(2003)264.
Ac ce p
te
d
M
an
us
cr
ip t
403
15
Page 15 of 29
438 Figure legends
440
1. Figure 1 The flowchart of atractylenolide Ⅲ MIP preparation.
441
2. Figure 2 The flowchart of MIP-SPE procedure.
442
3. Figure 3 Structures of sesquiterpene lactones template molecules and the adsorption test
443
compounds.
444
4. Figure 4 SEM of atractylenolide Ⅲ imprinted polymers with 3000 magnification.
445
5. Figure 5 HPLC chromatogram of the adsorption test solutions.
446
6. Figure 6 1H NMR spectrum of atractylenolide Ⅲ.
447
7. Figure 7 1H NMR spectrum of costunolide lactone.
448
8. Figure 8 HPLC chromatogram of the extract of Chinese medicine Radix Aucklandiae (a). HPLC
449
chromatogram of the eluent for the extract of Chinese medicine Radix Aucklandiae after passing
450
through dehydrocostus lactone MIP-SPE column (b). HPLC chromatogram of the eluent for the
451
extract of Chinese medicine Radix Aucklandiae after passing through costunolide lactone
452
MIP-SPE column (c).
an
us
cr
ip t
439
Table legends
455
1. Table1. The polymerization condition of sesquiterpene lactones MIPs by optimization.
456
2. Table 2. The results of the adsorption performance parameter for sesquiterpene lactones
457
MIP-SPE columns to target molecule.
458
3. Table 3. The results of the selective adsorption coefficient for sesquiterpene lactones MIP-SPE
459
columns to target molecule.
460
4. Table 4. The content results of the elutent after the adsorption test solution passing through
461
te
d
M
454
Ac ce p
453
sesquiterpene lactones MIP-SPE columns and C18-SPE column.
16
Page 16 of 29
*Highlights (for review)
Highlights 1. The regularity between structure and selective adsorption performance of MIP was explored. 2. The steric conformation of template molecules is related to the MIP adsorption performance. 3. The conformation isomers were verified by variable temperature 1H NMR spectroscopy.
Ac
ce pt
ed
M
an
us
cr
ip t
4. The hydrophilic functional groups can affect the selective absorption performance of MIP.
1
Page 17 of 29
an
us
cr
ip t
Figure
Ac ce pt e
d
M
Figure 1 The flowchart of atractylenolide Ⅲ MIP preparation
Page 18 of 29
ip t cr us an
Ac ce pt e
d
M
Figure 2 The flowchart of MIP-SPE procedure
Page 19 of 29
H
CH3 H
OH O O
O
H
H O
CH3
O b:Dehydrocostus lactone
O O
OCH3
e:LasianthuslactoneA
f:- -£©IstanbulinA
us
d:Isofraxidin
OH
O
（
O
cr
HO
CH2
c:Costunolide O H OH O
OH O
H3CO
H
ip t
a:AtractylenolideIII
O
Figure 3 Structures of sesquiterpene lactones template molecules and the adsorption test
Ac ce pt e
d
M
an
compounds.
Page 20 of 29
ip t cr us
Ac ce pt e
d
M
an
Figure 4 SEM of atractylenolide Ⅲ imprinted polymers with 3000 magnification.
Page 21 of 29
25
0 0
10
20
30
40
us
Retaintion(min)
ip t
-(-)istanbulin A isofraxdin atraylenoideIII lasianthuslactone A
cr
mAu
50
Ac ce pt e
d
M
an
Figure 5 HPLC chromatogram of the adsorption test solutions.
Page 22 of 29
ip t cr us an M
Ac ce pt e
d
Figure 6 1H NMR spectrum of atractylenolide Ⅲ.
Page 23 of 29
ip t cr us an
Ac ce pt e
d
M
Figure 7 1H NMR spectrum of costunolide lactone.
Page 24 of 29
ip t cr us
an
Figure 8 HPLC chromatogram of the extract of Chinese medicine Radix Aucklandiae (a). HPLC chromatogram of the eluent for the extract of Chinese medicine Radix Aucklandiae after passing through dehydrocostus lactone MIP-SPE column (b). HPLC chromatogram of the eluent for the
Ac ce pt e
d
column (c).
M
extract of Chinese medicine Radix Aucklandiae after passing through costunolide lactone MIP-SPE
Page 25 of 29
Tables
ip t
Table1. The polymerization condition of sesquiterpene lactones MIPs by optimization. Initiator Template functional Molar ratio of T, F and Cross-linker molecules monomer C （mg） atractylenolide 1-Viny EDMA 1:5:8 23 Ⅲ dehydrocostus 1-Viny EDMA 1:4:5 10 lactone costunolide 4 - vinyl TRIM 1:4:4 10 lactone benzoic acid
Ac
ce pt
ed
M
an
us
cr
T, F and C stand for template molecule, functional monomer and Cross-linker.
Page 26 of 29
Table 2. The results of the adsorption performance parameter for sesquiterpene lactones MIP-SPE columns to target molecule. Maximum loading capacity (µmoL g-1)
Enrichment factor
atractylenolide Ⅲ
4.1
82.1%
atractylenolide Ⅲ
dehydrocostus lactone
3.2
64.0%
atractylenolide Ⅲ
4.0
81.0%
dehydrocostus lactone
3.1
62.2%
3.9
80.1%
3.0
60.1%
ip t
cr
costunolide lactone
atractylenolide Ⅲ
Ac
ce pt
ed
M
C18
atractylenolide Ⅲ
us
costunolide lactone
Target molecule
（%）
an
MIPs-SPE column type
Page 27 of 29
Table 3. The results of the selective adsorption coefficient for sesquiterpene lactones MIP-SPE columns to target molecule.
dehydrocostus lactone costunolide lactone
Target molecule
atractylenolide Ⅲ
atractylenolide Ⅲ
31.6
+ lasianthuslactone A
lasianthuslactone A dehydrocostus lactone lasianthuslactone A costunolide lactone lasianthuslactone A
2.23
dehydrocostus lactone + lasianthuslactone A costunolide lactone + lasianthuslactone A
KD(mL/g)
28.8 13.1 28.0 18.5
14.2
2.20
ip t
atractylenolide Ⅲ
Mixture solution
1.51
cr
MIPs-SPE column type
* KD=Cp/Cs, Cp was the amount of target molecule bound to the MIP (mg·g-1); Cs was the
Ac
ce pt
ed
M
an
mixture solution all was 0.3 mgmL-1, V=5 mL, m=100mg.
us
concentration of target molecule in the sample solution (mgmL-1) . = KDi / KDj, KDi,KDj was distribution coefficient of template molecular and competitive compound. Cs of target molecule in the
Page 28 of 29
Table 4. The content results of the elutent after the adsorption test solution passing through sesquiterpene lactones MIP-SPE columns and C18-SPE column. Concentration （µg·mL-1） istanbulin A
1
1
1
1
0.010
0.066
0.932
0.104
0.048
0.416
0.724
0.216
0.628
0.652
0.882
0.604
0.424
0.806 0.658 0.894
Ac
ce pt
ed
M
an
MIPs-SPE column By dehydrocostus lactone MIPs-SPE column By costunolide lactone MIPs-SPE column By C18-SPE column
atractylenolide Ⅲ
ip t
By atractylenolide Ⅲ
lasianthuslactone A
cr
Initial solution
isofraxidin
us
The eluent
Page 29 of 29
