Use of Granulocyte Colony-Stimulating Factor for the Treatment of Thin Endometrium in Experimental Rats

Granulocyte colony-stimulating factor (G-CSF) induces stem cells to mobilize to the injury site, which have beneficial effect on tissue repair. The aim of this study was to investigate the effect of G-CSF on the thin endometrium in rat models. In the present study, rats with thin endometrium were divided into 4 groups (experimental group I: administrated with G-CSF (40 mg/kg/d) 4–6 hours post-modeling; control group I: administrated with saline 4–6 hours post-modeling; experimental group II: administrated with G-CSF (40 mg/kg/d) 12 days post-modeling; control group II: administrated with saline 12 days post-modeling. The agentia was given once daily and last for 5 days. Endometrial morphology was analyzed by Hematoxylin-Eosin staining, and the regeneration of endometrial cells was evaluated by immunohistochemistry and western-blot with cytokeratin and vimentin. We found that endometrial thickness and morphology presented a significant difference between experimental groups and control groups. No matter when we start with G-CSF, there was a significantly thicker endometrium and stronger expression of cytokeratin/vimintin in the experimental groups compared with the control groups (P,0.01). There were significant thicker endometrial lining and stronger expression of cytokeratin/vimintin in experimental group I than that of experimental group II (P,0.05), but there was no difference in the endometrial lining and the expression of cytokeratin/vimintin between the two control groups (P.0.05). In conclusion, G-CSF can promote the regeneration of endometrial cells in animal research, especially when the G-CSF was administrated earlier. Citation: Zhao J, Tian T, Zhang Q, Wang Y, Li Y (2013) Use of Granulocyte Colony-Stimulating Factor for the Treatment of Thin Endometrium in Experimental Rats. PLoS ONE 8(12): e82375. doi:10.1371/journal.pone.0082375 Editor: Meijia Zhang, China Agricultural University, China Received August 19, 2013; Accepted October 30, 2013; Published December 23, 2013 Copyright: 2013 Zhao et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: no funding sources Competing Interests: The authors have declared that no competing interests exist. * E-mail: zhaojing19840427@sina.com . These authors contributed equally to this work.


Introduction
Well known, both high-quality embryo and receptive endometrium are essential to successful pregnancy. The minimum endometrial thickness that may maintain a successful implantation is called ''threshold thickness''. The endometrium that thinner than ''threshold thickness'' is called ''thin endometrium'' [1], [2]. Thin endometrium is an iatrogenic response to the aggression of some medical treatment in some cases, such as repeated dilatation and curettages, but most of the time the causative agent remains unknown. Approximately 0.6%-0.8% of patients cannot reach the minimum thickness [3]. Although the researches concerning treatment for thin endometrium are quite a few, these efforts have yield disappointing results with controversial conclusions. It was necessary to explore the innovative intervention for the cure of thin endometrium.
So, the aim of the present study was to investigate the effect of G-CSF on thin endometrium in experimental animals. We established thin endometrium rat models by perfusing 95% ethanol in uterus as described [21]. Accordingly, G-CSF (40 mg/ kg/d) was administrated subcutaneously after the development of models. 12 days later, we evaluated the effect of G-CSF on endometrium and investigated the possible mechanism.

Experimental Animals
A total of 40 female adult Sprague-Dawley (SD) rats weighing 220-280 g were maintained in temperature controlled (24uC) quarters with free access to food and water. A 14-hour light and 10-hour dark cycle was maintained. The study was reviewed and approved by the Institutional Review Board and the Ethics Committee of Xiangya Hospital, Changsha, China. All the animal procedures in the present study were approved by the Institutional Animal Care and Use Committee.

Animal Model
95% ethanol has dehydration and protein denaturation efficacy, and aspiration and sclerotherapy with 95% ethanol is an effective treatment of ovarian endometromas; we developed thin endometrium rat model according to our previous study [21], in which thin endometrium rat model were established successfully without obvious adverse effects. Rats were anesthetized with 10% Chloral Hydrate (0.4 g/kg, i.p.) as described by authors [22], [23]. Exposed the uterus, clipped each uterine horn with vascular clip, and injected 95% alcohol (0.5 mL) into the uterine horn with a 1 ml syringe with a 16-gauge needle. After modeling, rats were placed in temperature and humidity controlled incubation chambers until they awoke.
Application of G-CSF. Forty female rats were randomly divided into four groups with different treatments: (1) G-CSF (40 mg/kg/d) was administrated subcutaneously daily lasting for 5 days starting 4-6 hours after modeling (experimental group I, n = 10), (2) Saline (40 mg/kg/d) was administrated subcutaneously daily lasting for 5 days starting 4-6 hours after modeling (control group II, n = 10), (3) G-CSF (40 mg/kg/d) was administrated subcutaneously daily lasting for 5 days starting 12 days after modeling (experimental group II, n = 10), (4) Saline (40 mg/kg/d) was administrated subcutaneously daily lasting for 5 days starting 12 days after modeling (control group II, n = 10). They were then transferred to the cages, and the health condition of the animals was monitored every 4-6 hours after surgery.

Specimen Collection
All rats were anaesthetized by intraperitoneal injection of overdose 10% Chloral Hydrate (1.0 g/kg body weight) at the forth estrus phase after injection of G-CSF/Saline. The phases of estrous cycle were determined by observing the vaginal smear. The uterus were removed and placed into 4% paraformaldehyde (PFA) or liquid nitrogen, and waited for further research.
Hematoxylin-Eosin staining. The sections (10-30 mm) on slides were immersed in xylene (10 min, twice), and rehydrated in a decreasing ethanol series diluted in distilled water (100%, 100%, 95%, 95%, 75%, 0%, 1 min each). The sections were rinsed in deionized water, stained in hematoxylin for 45 sec, rinsed in deionized water, and finally stained in eosin for 1 sec. After the color reaction, sections were dehydrated through an ethanol series into xylene and mounted using Permount mounting medium (Fisher Scientific, PA). The thickness and the morphology of endometrium were observed and measured under semi-automatic inverted biological microscope (with imaging system, DMI4000B, Leica, Germany).

Immunohistochemistry
The uterine horns were embedded in paraffin and about 6 micrometer serial sections were placed on SuperfrostPlus microscope slides. Sections were deparaffinized in xylene, rehydrated and rinsed in water. Endogenous peroxidase activity was blocked by incubating sections in 0.3% H 2 O 2 in methanol for 40 min at room temperature. Slides were blocked for 1 h in PBS supplemented with 10% normal goat serum. Localization of cytokeratin and vimentin proteins was performed by incubating section of rat uteri with either rabbit polyclonal antibodies against cytokeratin (3.3 mg/mL) or vimentin (5.0 mg/mL) over night at 4uC. Negative controls included substitution of the primary antibodies with the same concentration of normal rabbit IgG. Sections were incubated with 1:3000 HRP-conjugated goat antirabbit IgG in 10% goat serum for 1 h at room temperature. Sections were then briefly counterstained (10 s) with hematoxylin solution (Gill no. 3, Sigma), and examined using a Nikon microscope. For each protein studied, the immunohistochemical staining was repeated twice at each time point with sections obtained from for different rats. The regeneration of endometrial cells was analyzed via immunohistochemistry with cytokeratin, vimentin and was compared among groups.

Western Blotting
The tissues were homogenized in solubilization buffer. The homogenate was centrifuged at 10,000 g for 10 minutes at 4uC. The supernatant was removed. The protein concentration was determined using a detergent-compatible protein assay with a bovine serum albumin standard. For detection of cytokeratin and vimentin, 20 mg of protein from each sample was loaded onto an 8% SDS-polyacrylamide gel (PAGE) and then transferred to a polyvinylidene fluoride (PVDF) membrane. The blots were blocked with 5% milk in tris-buffer saline (TBS) buffer and then incubated with the primary antibody overnight at 4uC with anticytokeratin antibodies (1:500) and anti-vimentin antibodies (1:500). The membrane was washed with TBS and incubated with anti-rabbit IgG (1:3,000). The immunoreactivity was detected using enhanced chemiluminescence (ECL, Amersham). The loading and blotting of the amount of protein was verified by reprobing the membrane with anti-b-actin antiserum (Sigma, St. Louis, MO) and with Coomassie Blue staining.

Statistical Analysis
Date was expressed as mean6standard error. Comparisons of parameters among the groups were made by the Student's t-test or nonparametric Mann-Whitney test using SPSS16.0 software (SPSS Institute, Inc. Cary, NC). A value of P,0.05 was considered statistically significant.

Histopathological Observations
All animals recovered satisfactorily from the procedure. Endometrial thickness, superficial epithelia of the endometrium, and the number of endometrial glands presented a significant difference in experimental group compared with that of control group. Histologic evaluation of the uterine in experimental group showed intact structure of the endometrial layer characterized by increased endometrial thickness, more endometrial glands and capillaries, and the uterine horn in control group was completely destroyed with extensive necrosis, even coagulative necrosis in both the endometrium layer and myometrium layer.
The endometrial thickness of experiment/control group I and experimental/control group II were as follows: 569.37637.01 mm, 234.96637.82 mm, 487.78619.48 mm, 231.21617.81 mm, respectively. Obviously, significantly thicker endometrial lining was observed in experimental group I and experimental group II compared with that of the corresponding control groups (P,0.01). The endometrial lining was significantly thicker in experimental group I than that of experimental group II (P,0.01), and there was no significant difference between control group I and control group II (P.0.05) (Table 1, Figure 1).

The Regeneration of Epithelial Cells
The regeneration of epithelial cells was evaluated by examining the expression of cytokeratin with Immunohistochemistry (IHC) and Western Blot (WB). Immunohistochemical staining showed that cytokeratin expression was mainly localized in the cytoplasm of the endometrial epithelium. The expression of cytokeratin in experimental group I and experimental group II were significantly stronger than that of corresponding control group (P,0.01). The expression of cytokeratin in the experimental group I was significantly stronger than that of the experimental group II(P,0.05), but there was no difference between control group I and control group II (P.0.05) (Figure 2).
The result of WB showed that the expression of cytokeratin in the experimental groups was significantly stronger compared with the control groups (P,0.05). However, there was no significant difference between group I and group II when the G-CSF was given or not given (Figure 3).

The Regeneration of Stromal Cells
The regeneration of stroma cells was evaluated by examining the expression of vimentin with Immunohistochemistry (IHC) and Western Blot (WB). Immunohistochemical staining showed that Vimentin staining mainly expressed in cytoplasm of endometrial stroma cells. The expression of vimentin in experimental group I and experimental group II were significantly stronger than that of corresponding control groups (P,0.01). The expression of vimentin in the experimental group I was significantly stronger than that of the experimental group II(P,0.05), but there was no difference between control group I and control group II(P.0.05) (Figure 4).
The result of WB showed that the expression of vimentin in the experimental groups was significantly stronger compared with the corresponding control group (P,0.05). However, there was no significant difference between group I and group II no matter the G-CSF was given or not ( Figure 3).

Discussion
In the present study, we evaluated the therapeutic effect of G-CSF on thin endometrium in rats and compared the effect when the G-CSF was administrated at two different points in time. In clinical practice, acute or chronic thin endometrium usually means    The results showed that the expression of cytokeratin, vimentin in experimental groups were significantly stronger compared with the corresponding control groups, and there was no significant difference between group I and group II when the G-CSF was given or not given. Figure 3 is excluded from this article's CC-BY license. See the accompanying retraction notice for more information. doi:10.1371/journal.pone.0082375.g003 Therapeutical Effect of G-CSF on Thin Endometrium that the status have been last less than one menstrual cycle or more than 3 menstrual cycles, respectively. Considering that rats have an estrous cycle every 4-5 days, in our study we evaluated whether G-CSF has effect on thin endometrium after less one estrous (about 4-6 hours) post-modeling and three estrous cycles (about 12 days, but not exactly 12 days, should combine with vaginal smear) post-modeling. Histologic evaluation of the uterine showed intact structure of the endometrial layer characterized by increased endometrial thickness, more endometrial glands and capillaries after the administration of G-CSF. Our results showed that administration of G-CSF subcutaneously have beneficial effects on the regeneration of thin endometrium in rat model, especially when the administration of G-CSF was given earlier. On the contrary, administration of saline instead of G-CSF has no positive effect on the thin endometrium, which showed extensive necrosis, even coagulative necrosis in both the endometrium layer and myometrium layer.
The endometrial thickness was significantly thicker in experimental group I and experimental group II compared with that of the corresponding control groups, which was similar with the two studies done by Gleicher N, et al. One study [19] reported successful endometrial expansion to at least minimal thickness of 7 mm after uterine perfusion with G-CSF in four patients previously resistant to treatment with estrogen and vasodilators, and the other one found direct perfusion of the endometrial cavity with G-CSF can improve the thickness of endometrial lining in 21 IVF women with too thin endometrium ever after treatments [20]. Certainly, there were differences between our and their studies. One difference was that they found this interesting phenomenon in infertile women, whereas we used SD rats. Another difference was that the administration of G-CSG by intrauterine cavity perfusion in Gleicher's studies, while in our rodent model we administer G-CSF s.c.. Despite all this, we all found expanded endometrium, and all the results indicated that administration with G-CSF via uterine cavity perfusion or s.c. would be a promising treatment for the currently intractable problem of inadequate, thin endometrium in the near futher.
The expression of vimentin and cytokeratin evaluated by IHC and WB in experimental groups was significantly stronger than the control group, and the number of new capillary vessels was increased dramatically compared with control groups (date was not shown). All these results indicate that the administration of G-CSF could improve the regeneration of endometrial cells, including epithelial cell, stromal cells, and endothelial cells in rat model.
In the present study, we not only showed the effective of G-CSF on thin endometrium, but also explored the possible mechanisms underlying therapeutic effects of G-CSF.. Recently, adult stem cells have been identified in the endometrium. Regeneration of the endometrial cells may result from proliferation and differentiation of stem cells that were shown to be present in the endometrium [24]. If resident endometrial stem/progenitor cells were damaged or lost, particularly if the basalis layer was involved, the endometrium would be thin and atrophic with inactive glands and scant poorly vascularised stroma [25].
G-CSF was known as a powerful regulator of white blood cell proliferation and differentiation in mammals. Bone marrow stem cells and peripheral blood stem cells could be mobilized effectively by G-CSF. Some studies showed that administration of 40 mg/kg G-CSF for 5 days could mobilize bone marrow-derived cells (BMSCs) to peripheral blood the mostly effectively, and then these cells could differentiate into epithelial cells [26][27][28][29]. So the present study also choosed to apply G-CSF for 5 consecutive days. Treatment with the cytokine G-CSF, known to mobilize bone marrow hematopoietic stem cells (HSCs) [30], also caused dormant HSCs to proliferate. Based on previous research, we postulated that the G-CSF could promote bone marrow stem cell mobilization, proliferation, migration, and differentiation that may be involved in endometrial regeneration.
In addition, G-CSF may induce anti-apoptotic proteins and inhibit apoptotic process of endometrial cells. G-CSF may also reduce apoptosis of endothelial cells and increase vascularization in the thin endometrium, further protecting against ischemic injury [31].
Moreover, the process of modeling was similar to the local injury in clinical practice. Although experimental study using animal has difference with clinical practice, local injury to the endometrium may also evoke injury-induced inflammatory, and secrete growth factors, chemokines and chemotactic factors, which may promote homing of stem cells that mobilized by G-CSF and induce differentiation in microenvironment as homing induction [32].
Furthermore, in both animal experiments and clinical trials, G-CSF promoted follicle development in new-born rats and improved pregnancy rates in ovary poor responders [33], [34]. Therefore, it was suggested that G-CSF may also improve ovarian response and promote the regeneration of endometrium in turn.
However, the beneficial effect of G-CSF on thin endometrium was weakened by delayed administration of the G-CSF. Our result showed the thinker of endometrial lining and the weaker expression of vimentin and cytokeratin the experimental group II. It was proposed that administration of G-CSF 12 days postmodeling could not mobilize stem cells to regenerate endometrial cells because the inflammation induced by the modeling was diminished progressively. The expression of vimentin in experimental group I and experimental group II were significantly stronger than that of corresponding control groups. The expression of vimentin in the experimental group I was significantly stronger than that of the experimental group II, but there was no difference between control group I and control group II (PV, 6200). doi:10.1371/journal.pone.0082375.g004