Chinese Medicine, 2010, 1, 91-97
doi:10.4236/cm.2010.13017 Published Online December 2010 (http://www.SciRP.org/journal/cm)
Metabolic and Behavioral Patterns in a Pre-Menstrual
Syndrome Animal Model with Liver-qi Invasion and Their
Reversal by a Chinese Traditional Formula
Peng Sun, Sheng Wei, Huiyun Zhang, Mingqi Qiao*
Lab of TCM Classical Theory, MOE, Shandong University of Traditional
Chinese Medicine, Jinan, China
E -mail: [e-mail],
Received December 8, 2010; revised December 10, 2010; accepted December 14, 2010
Abstract
This work successfully used model rats with Pre-Menstrual Syndrome (PMS) liver-qi invasion in the early de-
velopment days to detect the Metabolic and Behavioral Patterns and their reversal by a Chinese traditional for-
mula. Our aim is to verify the reliability of PMS liver-qi invasion rat model and explore some micro- mecha-
nism of the syndrome of the liver failing to maintain the normal flow of qi. 30 rats with estrous cycles not in
accepting time were selected and divided randomly into three groups: the normal control group, PMS liver-qi
invasion model group and PMS liver-qi invasion medication-administered group. Emotional stimulation and
multiple factors combination were used to prepare the PMS liver-qi invasion model. Baixiangdan Capsules (a
Chinese traditional formula) were administered to rats to interfere with the PMS liver-qi invasion mode.
Open-field test was used to explore behavioral aspects of the model. Urine samples, from the three groups,
were collected and analyzed with UPLC-Q-TOF method to detect changes in metabolites related to liver func-
tions. In the open-field experiment, the crossing scores, rearing scores and open-field experiment total scores of
rats in the PMS liver-qi invasion model group increased remarkably (P < 0.05) compared with the scores of the
normal control group, the tendency was retrieved remarkably after medications (P < 0.05). Metabolic finger-
prints between the PMS liver-qi invasion model group and the normal control group had also distinguished
changes through principal component analysis, and an evident restoration trend occurred after Baixiangdan
Capsules administration. Taken together, behavioral and metabolic patterns can differentiate the PMS liver-qi
invasion rat models from the normal rats. Our results identified potential biological markers that might reflect
metabolic pathologies associated with PMS liver-qi invasion.
Keywords: Premenstrual Syndrome, Liver-Qi Invasion, Open-Field, Metabonomics, Baixiangdan Capsules
1. Introduction
Pre-menstrual Syndrome (PMS) is the syndrome which
regularly happens before emmenia, and affects women’s
daily life and work [1]. Animal models of PMS are valu-
able tools to investigate behavioral, physiological, and
molecular aspects associated with PMS [2,3]. From the
Chinese Traditional Medical point of view, PMS is a re-
sult from the failure of the liver to maintain the normal
flow of qi. Its pathogenesis refers to three systems the
mind, nerves and internal hormones. The involvement of
liver qi is supported by studies on animal models of PMS.
For example, increase in anxiety behavior, that is similar
to liver-qi invasion induced anxiety, was reported in an
animal model of PMS [3]. Although genetic and pro-
teomics can reveal the pathophysiology in a great extent,
they also have the defects in the interaction of protein
signal pathway and the difficult to locate target sites, and
can’t dynamically, real-timely reflect the overall infor-
mation. The present study provides new insights into the
pathogenesis of PMS from traditional Chinese medical
view. We investigated behavioral characteristics and the
metabolic markers in urine in premenstrual syndrome
liver-qi invasion rat model. Furthermore, we investigated
the potential effects of a traditional Chinese formula
(Baixiangdan Capsules) on reversing PMS associated
*Sun Peng and Wei Sheng contribute equally to the work.

92 P. SUN ET AL.
behavioral and metabolic changes.
2. Methods and Materials
2.1. Sample Animals
Thirty healthy Wistar female rats were chosen with
body weights from 180 g to 220 g. They were provided
by Experimental Animal Centre of Shandong Univer-
sity of Traditional Chinese Medicine where they ini-
tially came from Chinese Rodents Experimental Animal
Seed Bank Shanghai Center with production license:
SCXK (Shandong) 20050015. The rats were bred in
perversion of day and night with light on at 9 am and
off at 9 pm every day. Rats had free access to food and
water except during the experiment time. The rats were
handled on daily basis.
2.2. Drugs and Reagents
Baixiangdan Capsule is composed of White peony root,
Cyperus rotundus, moutan. Producted Shandong Univer-
sity of Tranditional Medicine, Batch number: 080201,
Acetonitrile (HPLC grade, Merck Company), Formic
acid (HPLC grade, U. S. Tedia company), MilliQ ul-
trapure water, Other reagents were all analytical grade.
2.3. Bolting Experiment Rats
Open-field experiment was used to bolt rats of which
scores were close to each other. The method of open-
field experiment is described below.
2.4. Method of Confirming the Estrous cycle of the
Rats
The estrum of rats was confirmed by observing the va-
gina cell shape. The estrous cycle of a rat generally con-
tains four stages: pre-estrous, estrous, meto-estrous and
anestrous. The specific procedure was as follows: a small
quantity of isotonic Na chloride was drawn by a dropper,
and was dropped repeatedly on a rat’s colpo for 2-3
times. The suction fluid was overlaid on glass slide to be
naturally air dried. Then it was held by absolute alcohol
for 2 to 5 min, followed by being thoroughly accreted by
Giemsa stain for 10 to30 minutes. The cell appearance
was observed every time (Table 1). Vaginal smear was
conducted 5-9 times every day for 28 days to confirm the
estrous cycle of every experimental rat [4].
2.5. Ethnology Evaluation Method of Estrous Cycle
Ethnology observation was used to evaluate the estrous
Table 1. Cell morphologic change features of rats’ estrous
cycle.
Stage Ovary Changes
Cell Morphologic Change
Features
Pre-estrous
egg chamber
accelerated
growth
All is caryon cellula epithe-
lialis with a small quantity
of keratinocyte
Estrous egg chamber
ovulating
All is acaryotic keratino-
cyte, or several cellula epi-
thelialis
Metao-estrous corepus luteum
forming
blood corpuscle, keratino-
cyte, caryon cellula epithe-
lialis
Anestrous corpus luteum
regression
generous blood corpuscles,
several cellula epithelialis
and cellula mucipara
cycle. The female rats were put into cages to be observed
for their estrous behaviors, such as jumping, crawling,
joggling ears, arching back, and the male rat attacking
female rat’s back (Kow, 1976, Sodersten and Eneroth,
1981). Male rats were not allowed to crawl across female
rats’ back. Rats showed active estrous behaviors at
pre-estrous and estrous (the accepting time), but estrous
behaviors decreased or even disappeared at the stages of
anestrous and metao-estrous (not accepting time). The
interval was two days. Rats which displayed active es-
trous behaviors were internalized [5].
Thirty rats whose estrous cycles are not in accepting
time were selected and weighed. They were randomly
divided into three groups: normal control group, PMS
liver-qi invasion model group (the invasion model group),
PMS liver-qi invasion medication-administeredn group
(the invasion medication group). Rats which were not in
the accepting time were stimulated continuously for two
weeks.
2.6. PMS Liver-qi Invasion model Method
The model rats were put into cages which could accom-
modate interval time and intensity of noise and electric-
ity stimulations freely. Rats were exposed to the noise
and electricity stimu-lations once per 5 min at daytime
and once per 10 min at night. Current flow of 0.5 mA
and voltage 2700-3300 V were applied. The nervure was
0.3 s wide. Photographic recording was taken at night
and photos were taken once every day [6].
2.7. Medication
Medication was orally administered once per day. The
administration dosage for rats was 1 ml per 100 g, for
five days [7]. Control rats received equal volume of ster-
ile water orally.

P. SUN ET AL.
93
2.8. Urine Samples
Rats fed 4 ml sterile water at each day at 9:00 am, and
then placed in metabolic cages, using the cylinder col-
lected 2 ml of urine. Urine samples were centrifuged, 20
min with 10000 rpm, and the supernatant was taken for
further analysis. Following metabolic analysis, rats were
tested in the open field.
2.9. Open-field Experiment Method
The open-field box was by 100 cm 100 cm 50 cm. The
paries and undersurfaces were black. The undersurface
was partitioned into 25 areas with same size by white
lines which were called periphery grilles porch. The wall
and the rest were called central grilles. The rat’s tail was
gripped at 1/3 of the root by the runner, and was put into
the mesh grille. The rat’s behavior changes were record
by photographic recording system for three minutes. (1)
Crossing score: the score is the number of times that
animals cross undersurface grilles. Four claws must all
enter the grille. (2) Rearing score: the score is the num-
ber of times that animals make a perpendicular act. Four
claws must leave empty or hold on to the wall. The
open-field experiment score is the total of the crossing
score and the rearing score. Simultaneously record the
modified time, modified number of times, the number of
times staying on the mesh grille, and dejecta number.
Each rat was used once after being made into the model,
3 min every time. The differences among groups were
compared. The photographic recordings were observed
by three staffs, and the concordance was confirmed
(Kappa exceeded 0.95) [8].
2.10. Analysis Conditions
Chromatographic conditions: Flow 0.4 ml/min, Sample
room temperature: 4, Column temperature: 60℃℃.
MS conditions:
Electrospray ion source, Positive ion mode V, 50-
1000 Da, scan time: 0.1s, inter time: 0.02 s, Lock mass:
0.01 ml/min LE ([M + H] + = 556.2771) (200 pg/ml),
Capillary voltage: 3 kv, Cone voltages: 60v, Ion source
temperature: 100℃, Desolvation temperature: 300℃,
Desolvation nitrogen flow: 700 L/hr, Cone gas flow: 50
L/hr.
2.11. Data Processing
Behavior of experimental data analysis using SPSS 13.0
statistical software, Multiple samples of the comparative
analysis of variance with one factor, UPLC-TOF-MS
spectra data use the Micromass MarkerLynx software
to identify peak and to peak matching, and use principal
component analysis to carry out pattern recognition to
the control group, model group and administration group.
3. Experimental Results
3.1. Comparison of Open-field Experiment Scores
In open-field experiments, compared with the normal
control group, the crossing scores, rearing scores and
open-field experiment total scores of rats in the inva-
sion model group remarkably increased (P < 0.05). The
tendency was retrieved remarkably after medication (P
< 0.05).
3.2. The Spectra Results of the urine Metabolic
Products
There was a remarkable difference between control
group and model group, but the contrasts were subtle
between control group and adminstration group from
Figure 1 showed as below.
3.3. Results of Principal Component Analysis
The scores chart of the urine metabolic product content
through principal component analysis showed as below
(Figure 2). There was so close position that we can
hardly differentiate the box marks (represent the con-
trol group) from the triangle marks (represent the ad-
ministration group), on the contrary, the diamond
marks (represent the model group) distributed far away
from other marks.
The box marks represent the control group, the dia-
mond marks represent the model group and the triangle
marks represent the administration group.
3.4. Analysis of Differences in Metabolic
Products
14 potential biological markers had been detected
through comparing the metabolic fingerprint difference
between model group and control group.
4. Discussion
Quantization evaluation of macroscopic appearance of
animal samples is very valuable to evaluate liver-related
metabolic aspects of the syndrome. Qualitative method
and half quantization bound with four-grade score
method were adopted to evaluate sample’s behaviors and
expressions in early works. The specific method was to
use four-grade score method (“+”expresses masc. result.)

P. SUN ET AL.
94
to process photographic recording survey results by half
quantization, Then, the results were compared with rats
in the normal control group. This method just made a
corresponding evaluation of model macroscopy appear-
ance by animal clinical symptoms. The modality can’t
realize the statistical analysis of experimental results.
This modality is sufficient as an aiding evaluation
method, but it will be too subjective and lacks objective
evidence as a major evaluation method for animal mac-
roscopy appearances. In the open-field experiment,
compared with the normal control group, the crossing
scores, rearing scores and open-field experiment total
scores of rats in the invasion model group increased re-
markably. Under stimulations, activities and excitement
PMS0103
Time
2.004.006.008.0010.00 12.00 14.00 16.0018.00 20.00 22.00 24.00
%
0
100
blank-4-8-1 1: TOF MS ES+
BPI
2.08e4
19.35
4.61
4.60
0.92
0.91
0.67
0.65
0.59
4.20
1.03
3.33
1.80 2.81 3.71
5.49
5.21
19.22
5.50
6.34
5.54
5.98
5.96 18.11
6.52
16.43
8.35
8.01
6.93 10.52
9.24
9.39
14.25
12.58
17.34 18.39
19.10
20.36
19.72
20.82 23.31
23.11
Control Group
PMS0103
Time
2.004.006.008.0010.00 12.0014.00 16.0018.00 20.00 22.00 24.00
%
0
100
model-4-8-3 1: TOF MS ES+
BPI
2.04e4
5.78
5.77
4.58
0.91
0.90
0.67
4.56
4.19
1.03 3.31
1.82
1.63 2.74 3.65
5.53
5.51
4.61
5.50
5.26
6.35
6.34
10.55
10.53
6.36
6.49
6.51
10.51
8.226.96
10.56
12.62
12.61
11.44
19.31 23.30
20.83
20.36
Model Group

P. SUN ET AL.
95
PMS0103
Time
2.004.006.008.0010.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00
%
0
100
drug-4-8-1 1: TOF MS ES +
BPI
1.45e4
21.58
4.44
4.42
4.41
0.64
0.54
0.68 3.24
1.12 1.30 2. 7 0
4.04
3.56
4.45
4.46
6.24
6.23
5.06
5.05
4.80
5.86
5.36
21.48
6.43
20.39
19.39
8.18 8.01
6.86
9.27 19.30
10.60
9.41 11.10 12.00
19.55
21.13 23.26
Medication Group
Figure 1. The urine test total ion chromatogram of rats.
Figure 2. Rat urine metabolic pattern analysis scores chart.
of rats in the invasion model group increased remarkably.
After intervention of medication, each group recovered
to normal (Table 3). Metabolic fingerprints between the
PMS liver-qi invasion model group and the normal
control group had also distinguished changes through
principal component analysis, which were restored
after Baixiangdan Capsules administration (Figure 1
and Figure 2). According to traditional Chinese medi-
cine’s theory of symptoms and signs, the model rats
displayed proximal clinic symptoms of patients with
PMS liver-qi invasion such as anxious emotion, and
hyperactivities. Therefore, above-mentioned conclu-

96 P. SUN ET AL.
Table 2. Rat urine UPLC chromatographic conditions.
Time(min) 0.2% Formic acid (%) Acetonitrile (%)
0 100 0
18 65 35
20 5 95
22 5 95
25 100 0
28 100 0
Table 3. C o mparison o f op en-f ield exp eriment s cor es (n = 10) .
Group Crossing score Rearing score Open-field score
Normal control
group 49.92 ± 29.11 11.58 ± 5.41 61.50 ± 30.93
Invasion model
group 64.00 ± 20.15☆ 15.92 ± 6.26☆ 79.92 ± 24.18☆
Medication
group 43.33 ± 11.11△ 9.75 ± 2.18△ 53.08 ± 12.56△
☆There was significant difference when compared with control group (P <
0.05); △There was significant difference when compared with model group
(P < 0.05).
Table 4. Information Sheet potential biological marker s.
No. Retention time Identification Trends
1 1.0 Methyl guanine ↑
2 1.1 2,3-dihydroxy-3-methyl valerate↑
3 4.6 D-galactosamine ↓
4 4.7 5-Amino acid ↓
5 5.7 2-aminoadipic acid ↑
6 5.8 D-Proline ↑
7 6.3 Melatonin ↑
8 6.4 3-hydroxy-pyruvate-5-carboxyl ↑
9 6.5 4-hydroxy-glutamic ↑
10 8.1 2,3-dihydro-pyridine acid ↓
11 8.3 Deoxy-adenosine ↓
12 12.6 5,7,4 '- genistein ↑
13 16.3 Sebacic acid ↓
14 19.3 Prostaglandin F2α ↓
sions of qualitative analysis, quantitative judgment and
medicine disproof were sufficient to confirm the most
important evidence of traditional Chinese medicine’s
theories of symptoms and signs to the metabolic and
behavioral patterns.
This study also found 14 potential biological markers
(Table 3) which were relevant with PMS liver-qi inva-
sion symptoms of rats and had not been reported before
although some markers thought to be related to the reap-
praisal regulation processing of emotion such as 2-ami-
noadipic acid, 4-hydroxy-glutamic. 2-aminoadipic acid
was the major metabolic product of lysine metabolism,
and the normal metabolism of lysine also depended on
the regulation of glutamate. Therefore, 2-aminoadipic
acid increasing was associated with glutamate increasing.
The relation of 4-hydroxy-glutamic acid and glutamate
reflected by the following chemical equation:
=
+
+
4-hydroxy-2-ketoglutarate Ketoglutarate Glutamate 4-hydroxy-glutamic
Thus 4-hydroxy-glutamic content growing induced an
increase of the glutamate level inferred from above fig-
ure. As to the relation between glutamate level and PMS
liver-qi invasion symptoms we had done more works.
Our previous studies showed that glutamate level was
decreased in hypothalamus, while glutamate levels were
remarkably increased in cortex and hippocampus com-
paring PMS liver-qi invasion rats with the control rats [9].
Glutamate is the major excitatory brain neurotransmitter
involved in brain protein and glucose metabolism. It can
improve the function of the glutamate system neurons
[10]. Some researchers had reported that women with
premenstrual dysphoric disorder had a higher risk to suf-
fer from emotional disorders. May be glutamatergic sys-
tem involved in the pathogenesis of affective disorders
and it was proved that the drugs which reduce activity of
glutamatergic system or inhibit signal transduction of
glutamate receptor showed a approximate anti-manic
effects [11]. Maybe it can be inferred that glutamate
level plays a key role in the course of PMS liver-qi inva-
sion. Other metabolic products mentioned in Table 4
need further exploration about their relation with PMS
liver-qi invasion symptomes.
5. Acknowledgements
This study was supported by the National Natural Science
Foundation of China (No.30973688 and No.30930110),
the National Program of Key Basic Research Project
(973 Program) (No. 2011CB505102). The authors thank
Nashat Abumaria for his excellent manuscript prepara-
tion assistance.
6. References
[1] J. J. Li, “Clinical Gynecologic Endocrinology and Dys-
genesia,” Shandong Science and Technology Publishing
Company, Jinan, 2003.
[2] T. Schneider and P. Popik, “Increased Depress Ive-Like
Traits in All Animal Model of Premenstrual Irritabil-
ity,” Horm behav, Vol. 51, No. 1, January 2007, pp.
142-148.
[3] S. Wei, “Improvement of Premenstrual Syndrome Liver- Qi
Invasion, Dpression Animal Model and Study on Micro-

P. SUN ET AL.
97
mechanism of the Liver Failing to Maintain the Normal
Flow of Qi,” Journal of Shandong University of Traditional
Chinese Medicine, Vol. 31, No. 5, May 2007, pp. 1-5.
[4] X. F. Tu, “Effects of Excessive Swimming Exercise Oil the
Estrous Cycle of Mice,” Chinese Journal of Clinical Reha-
bilitation, Vol. 36, No. 8, Augest 2004, pp. 8316- 8317.
[5] H.-P. Ho, Marie Olsson and Lars Westberg, “The Sero-
tonin Reuptake Inhibitor Fluoxetine Reduces Sex Ster-
oid-Related Aggression in Female Rats: An Animal
Model of Premenstrual Irritability?” Neuropsychophar-
macology, Vol. 24, No. 5, May 2001, pp. 502-510.
[6] H. Y. Zhang, M. Q. Qiao and W. C. Zhu, “Variation on
Rats' Serum Hormones and Behavior of Premenstrual
Syndrome Liver-Qi Invasion,” Chinese Traditional Pat-
ent Medicine, Vol. 24, No. 2, February 2002, pp.
118-119.
[7] S. Y. Xu, R. L. Bian and X. Chen, “Pharmacological
Experiments Methodology,” People's Medical Publishing
House, Beijing, 2002.
[8] X. W. Liu, H. M. Zhang and H. D. Qu, “Observation and
Detection of Behavior on Rat of ‘Anger Hurt Qi’,” Jiang
Su Journal of Traditional Chinese Medicine, Vol. 26, No.
3, March 2005, pp. 53-55.
[9] L. Sun, “PMS Rat Model of Liver against the Permit and
Different Brain Regions of Serum Progesterone and
Amino Acid Detection Analysis,” Master dissertation,
Shandong University of Traditional Chinese Medicine,
Jinan, 2008.
[10] C. Konradi and S. Heckers, “Molecular Aspects of Glu-
tamate Dysregulation: Implications for Schizophrenia and
Its Treatment,” Pharmacology & Therapeutics, Vol. 97,
No. 2, February 2003, pp. 153-179.
[11] X. Zhou and X. Q. Wang, “Glutamate and Γ-Amino-
butyric Acid System in Mood Disorder,” Chinese Journal
of Neuroscience, Vol. 19, No. 2, February 2003, pp.
130-133.