Endostatin Gene Therapy for Endometriosis in Rats

Abstract

Objective:

Endostatin gene therapy for endometriosis was studied in an experimental autotransplantation model in rats.

Methods:

Endometriotic lesions were transfected by intralesional injection of the plasmid lipofectamine-endostatinpBud (group 1), lipofectamine-pBud (empty vector; group 2) or phosphatebuffered saline (group 3). Endostatin mRNA and protein levels in lesions were evaluated by quantitative real-time reverse transcription-polymerase chain reaction and Western blot analysis. Endostatin and vascular endothelial growth factor (VEGF) protein levels in serum, and microvessel density (MVD) and matrix metalloproteinase (MMP)-2 protein levels in endometriotic lesions, were also determined.

Results:

Lipofectamine- endostatin-pBud injection increased endostatin mRNA and protein levels in lesions. Lesions were significantly smaller, and serum VEGF levels significantly lower, in group 1 versus controls. Serum VEGF was significantly and negatively correlated with serum endostatin. In group 1, MMP-2 levels and MVD were significantly lower versus controls. MMP-2 level was negatively correlated with endostatin.

Conclusions:

Gene therapy with endostatin appears to be an effective treatment for endometriosis. Restoration of endostatin gene expression by gene transfer in vivo might be a potential gene therapy approach for human endometriosis.

Introduction

Endometriosis is a benign, oestrogen-dependent, gynaecological disorder characterized by the presence of endometrial tissue (glands and stroma) outside the uterine cavity, mainly on the pelvic peritoneum and ovaries, and in the rectovaginal septum. ¹ It is a chronic recurrent disease affecting a large number of women: Approximately 10% of the general female population and 70% of young women with chronic pelvic pain have the disorder. ² Endometriotic lesions can be removed surgically to ease pain, but the disease may recur and patients may not accept a second operation. ³ In addition, some women prefer a medical rather than a surgical approach to treatment. Current therapeutic approaches include various hormone treatments that are designed to decrease circulating oestrogen to postmenopausal levels. Such treatments are of limited benefit, however, and are associated with severe adverse effects including hot flushes, weight gain and mood disorders. ⁴ Thus, the treatment itself reduces the woman's health-related quality of life. In the long term, lower oestrogen levels induce osteoporosis and increase the risk of heart disease, therefore hormone therapies can only be used for ≤ 6 months. ⁵ For such reasons, new therapeutic strategies are under development.

Recent studies have indicated that angiogenesis is a prerequisite for endometriosis development.^6–8 According to the transplantation theory, endometrial fragments lodged in the peritoneal cavity require the establishment of a new blood supply for the survival of implants and development of disease. According to this hypothesis, early endometriotic lesions are characterized by dense vascularization. Several angiogenic factors (such as vascular endothelial growth factor [VEGF], interleukin-8 and placental growth factor) have been shown to contribute to the establishment and progression of endometriosis, ⁹ and it has been suggested that antiangiogenesis therapy is a potential treatment.

Endostatin is an antiangiogenic factor. It is a 22 kDa polypeptide, derived from a carboxyl-terminal fragment of type XVIII collagen. ¹⁰ Endostatin derived from Escherichia coli inhibits human umbilical vascular endothelial cell proliferation, migration and tube formation, and suppresses in vivo tumour growth after inoculation of cancer cells into mice. ¹¹ Endostatin has also been reported to interfere with fibroblast growth factor-2-induced signal transduction. ⁹ On the basis of these findings, endostatin appears to have the potential to become one of the most valuable antiangiogenic agents in the treatment of cancer. ¹² The effect of transferring the endostatin gene into endometriotic lesions has not yet been investigated. Thus, the present study examined the therapeutic effect of endostatin gene therapy in a rat model of endometriosis.

Materials and methods

Animals

Fifty female Lewis rats (14 weeks old, mean ± SD weight, 240.6 ± 15.14 g) were purchased from the Institute of Laboratory Animals of the Chinese Academy of Medical Sciences, Beijing, China, and were housed under standard conditions of controlled light (12-h light/12-h dark cycle, lights on at 08.00 h) and temperature (mean ± SD 22 ± 2 °C). All experiments were conducted in accordance with the guidelines of the International Council for Laboratory Animal Science (ICLAS; http://www.iclas.org/). Protocols were approved by the Institutional Animal Ethics Committee of The Second Xiangya Hospital of Central South University, Changsha, Hunan, China.

Modelling Of Endometriosis and Assessment of Lesions

Autologous transplantation of endometrial tissue ¹³ was performed to establish a model of endometriosis. Four weeks later, the peritoneum and visceral organs were examined using a dissecting microscope and lesions exhibiting the distinctive morphology of endometriosis were removed. The examination was performed in sterile conditions, in rats under anaesthesia. Modelling was successful in 45 of the 50 Lewis rats included in the study. Endometriotic lesions were fixed in 10% neutral-buffered formalin, then embedded in paraffin wax. Paraffin wax-embedded tissues (3 μm) were then deparaffinized with xylene and rehydrated in a graded ethanol series for haematoxylin and eosin staining.

Two experienced pathologists (T.T.Z., J.G.) examined all slides in a blinded manner. Endometriosis was diagnosed if ectopic foci were present and contained both viable endometrial glands and stroma. The volume of endometriotic lesions was calculated according to the following formula: ¹³ V = L × W² × 0.5, where L denotes lesion length and W denotes lesion width.

Endostatin Expression Vectors and Vector Administration

The endostatin expression vector ES-pBudCE4.1 was kindly provided by Professor Zhixiong Lin, Fujian Medical University, Fuzhou, Fujian, China, and consisted of a cDNA encoding mouse endostatin that had been subcloned into the commercially available pBudCE4.1 vector (Invitrogen, Carlsbad, CA, USA). Expression of the vector is driven by a cytomegalovirus immediate-early promoter, and a secretion signal from the immunoglobulin κ chain is fused to the N-terminus of the endostatin gene. DNA sequence analysis confirmed that the cDNA sequences were inserted in the proper reading frame, and that no mutations had been incorporated.

One week after the modelling was confirmed to be successful, the 45 animals were randomly divided (according to a computer-generated randomization schedule) into three groups (15 per group): group 1 received an intralesional injection of 2 μl lipofectamine–ES-pBud; group 2 received an intralesional injection of 2 μl lipofectamine-pBud as a blank vector control; group 3, serving as an untreated control group, received an equivalent volume of 0.01 mM phosphate-buffered saline (PBS; pH 7.4). This experiment was performed in sterile conditions, in rats under anaesthesia.

Quantitative Real-Time RT–PCR Analysis of Endostatin Gene Expression

Two weeks after injection, endostatin gene expression in the endometriotic lesions of the three groups was detected by quantitative realtime reverse transcription–polymerase chain reaction (RT–PCR). Total RNA was extracted from 50 mg of tissue using the RNeasy^® Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Single-stranded cDNAs were generated using cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA) following the manufacturer's instructions.

Quantitative real-time PCR was performed using the SYBR^® Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA), according to the manufacturer's instructions, in a 50-μl reaction mixture containing 18 μl diethylpyrocarbonate water, 2 μl primers (10 μm), 25 μl SYBR^® Green Master Mix and 5 μl cDNA. The primer sequences, synthesized by SBS Genetech (Beijing, China), were: β-actin, forward 5′-GGA GAT TAC TGC CCT GGC TCC TA-3′, reverse 5′-GAC TCA TCG TAC TCC TGC TTG CTG-3′; endostatin, forward 5′-GCC CGC ATC TTT TCT TTC G-3′, reverse 5′-TGT AAC CCC AGT AGC TTC AGT CC-3′. The thermal cycling conditions were as follows: 10 s at 95 °C, 5 s at 95 °C and 20 s at 55 °C (45 cycles) and 60 °C for 30 s. Duplicate PCR amplifications were performed for each sample.

Using β-actin plasmid DNA constructed by the authors as a positive control, and based on 10-fold concentration gradients, fluorogenic quantitative PCR was performed over a concentration gradient between 2 and 20 000 μg/l. The level of target gene expression was extrapolated from the standard curve and normalized to β-actin. Data were analysed using Rotor Gene™ version 5.0 application software (Corbett Research Pty Ltd, Mortlake, Australia).

Western Blot Analysis of Endostatin Protein Levels

Two weeks after injection, endostatin protein levels in the endometriotic lesions were detected by Western blot analysis using 50 mg of tissue from each experimental animal. Lesions were rinsed in ice-cold 0.01 mM PBS and homogenized in buffer containing 50 mmol/l Tris–HCl (pH 8.5), 150 mmol/l NaCl, 0.2 g/l NaN₃, 0.1 g/l sodium dodecyl sulphate [SDS], 100 g/ml phenylmethyl -sulphonyl fluoride, 1 μg/ml aprotinin, 10 g/l NP-40 and 5 g/l sodium deoxycholate. The products were centrifuged at 14 000 g for 15 min to remove cellular debris. Total protein concentration was determined by the Bradford method. A total of 100 μg protein per sample was separated by SDS–10% polyacrylamide gel electrophoresis then transferred to an Immun-Blot^® PVDF Membrane (Bio-Rad, Hercules, CA, USA). Membranes were blocked for 1 h at room temperature in Tris-buffered saline with 0.5% Tween (TBST; pH 7.6) containing 5% (w/v) nonfat dried milk, then probed overnight at 4 °C with a rabbit polyclonal antiendostatin antibody (1:200 dilution, Santa Cruz Biotechnology, Santa Cruz, CA, USA). Membranes were then washed three times in 0.01 mM PBS and incubated with horseradish peroxidase-conjugated goat antirabbit immunoglobulin G (1:2000 dilution; Santa Cruz Biotechnology) for 40 min at room temperature. Membranes were then washed again three times for 10 min in distilled water and immunoreactivity was detected by enhanced chemiluminescence (ECL™; Amersham Biosciences, Piscataway, NJ, USA), according to the manufacturer's instructions.

Enzyme-Linked Immunosorbent Assay

Two weeks after injection, serum levels of endostatin and VEGF in the three groups were measured with enzyme-linked immunosorbent assay kits (ELISA) specific for the respective rat proteins (Wuhan Boster Biological Engineering, Hubei, China) according to the manufacturer's instructions.

Immunohistochemical Analysis

Two weeks after injection, immunohisto -chemistry was used to determine lesion microvessel density (MVD) and the level of endostatin and matrix metalloproteinase (MMP)-2 protein. Endometriotic lesions from the three groups were fixed in 10% neutral-buffered formalin, then embedded in paraffin wax. The paraffin wax-embedded tissues were cut at 3 μm, deparaffinized with xylene, and rehydrated in a graded ethanol series for haematoxylin and eosin staining or peroxidase immunohistochemical staining.

After brief proteolytic digestion and peroxidase blocking, the slides were incubated overnight at 4 °C with the primary antibodies: rabbit antirat endostatin polyclonal antibody (100 g/ml at 1:100 dilution; Santa Cruz Biotechnology), or rabbit antirat MMP-2 polyclonal antibody (200 μg/ml at 1:100 dilution; Santa Cruz Biotechnology); 0.01 mM PBS was used in place of primary antibody as a negative control. Slides were washed three times with 0.01 mM PBS, then incubated with a biotinylated goat antirabbit antibody from a Histostain-Bulk-SP kit (Zymed Laboratories, San Francisco, CA, USA). After washing three times with 0.01 mM PBS, peroxidase-labelled polymer and diaminobenzidine as the chromogenic substrate (Dako EnVision™ System; Dako Diagnostics, Zug, Switzerland) were used to visualize proteins.

Microvessel density was evaluated semiquantitatively according to a method described by Weidner et al. ¹⁴ The 10 most intensely stained microvessel regions were selected per section. Each lesion was examined three times and the mean value was taken. Endostatin and MMP-2 staining was evaluated according to a previous method, with slight modification. ¹⁵ Distribution and intensity of staining within the glandular epithelium or the stroma were analysed. Histological scores (H) for endostatin and MMP-2 were calculated using the formula H = ∑Pi/2, where i is an intensity variable ranging between 0 (no staining) and 3 (high staining intensity) and P is the percentage of stained cells for each value of i; P values of 0, 1, 2, or 3 represented percentages of positively stained cells of < 5%, 5 – 25%, 26 – 50% and > 50%, respectively.

Statistical Analyses

All statistical analyses were carried out using the SPSS^® statistical package, version 13.0 (SPSS Inc., Chicago, IL, USA) for Windows^®. Continuous variables were expressed as mean ± SD. Analysis of variance with Fisher's post hoc test or Fisher's exact test was used to test for significant differences between groups. The nonparametric Kruskal–Wallis and Mann– Whitney U-tests were used to analyse non-normally distributed variables. Spearman's correlation coefficients were calculated between the levels of endostatin and MMP-2, and between those of endostatin and VEGF. Differences were considered statistically significant at P < 0.05.

Results

The 45 rats in whom the modelling was successful rats showed well-established endometriosis 4 weeks after endometrial fragments obtained from uterine horns had been implanted into the peritoneal cavity of the same rat. Nodules had commonly formed on the peritoneal layer of the anterior abdominal wall. In a few rats, endometriosis was also present on the surface of the intestine or in the pelvic peritoneum. On gross examination, the lesions varied among the different animals and included a single pink nodule, multiple nodules, and dense tanned lesions (Fig. 1A). This is similar to the gross variation seen in human endometriosis. ¹⁶ Microscopic examination showed the typical appearance of endometriosis, with endometrial glandular epithelium surrounded by an intact basement membrane and adjacent endometrial stroma (Fig. 1B).

FIGURE 1:

Morphology of experimental endometriosis 4 weeks after inoculation of endometrium into the peritoneal cavity of rats. (A) Gross morphology, indicating lesions as pink nodules (× 4 magnification); (B) formalin-fixed and paraffin wax-embedded endometriosis lesion 4 weeks after transplantation into the peritoneal cavity of a rat, demonstrating typical morphology of endometriosis with endometrial glandular epithelium surrounded by an intact basement membrane and adjacent endometrial stroma (× 400 magnification)

The expression of endostatin in endometriotic lesions after transfection was confirmed by quantitative real time RT–PCR and Western blot analysis. Table 1 shows that the endostatin mRNA and protein levels in group 1 were significantly higher than levels in groups 2 and 3 (both P < 0.01).

TABLE 1:

Transfection of the endostatin gene into endometriotic lesions in model rats: efffect on endostatin mRNAand protein levels in the lesions

		Endostatin/β-actin ratio
Group	n	mRNA	Protein
1	15	31 950.44 ± 10 507.82	1.08 ± 0.02
2	15	19 810.37 ± 14447.15 a	0.677 ± 0.01 a
3	15	17 803.22 ± 6650.74 a	0.696 ± 0.01 a

Data presented as mean ± SD.

a

P < 0.01 compared with group 1; normally distributed data were analysed by analysis of variance with Fisher's post hoc test or Fisher's exact test; non-normally distributed data were analysed using the nonparametric Kruskal–Wallis and Mann–Whitney U-tests.

Group 1 rats were transfected with the endostatin (ES) gene by intralesional injection of the lipofectamine–ES-pBud plasmid; group 2 received an intralesional injection of lipofectamine-pBud, as a blank vector control; group 3 (untreated control) received an equivalent volume of phosphate-buffered saline.

There was no significant difference in serum levels of endostatin and VEGF between the three groups before transfection (Table 2). After transfection, the endostatin serum level in group 1 was significantly higher than that in groups 2 and 3 (both P < 0.01), whereas the VEGF serum level in group 1 was significantly lower than that in groups 2 and 3 (P < 0.01). The serum level of VEGF was negatively correlated with the serum endostatin level (r = -0.81, P < 0.05).

TABLE 2:

Transfection of the endostatin gene into endometriotic lesions in model rats: effect on serum levels of endostatin and vascular endothelial growth factor (VEGF) in the lesions

Group	n	Transfection	Endostatin, pg/ml	VEGF pg/ml
1	15	Before	127.48 ± 17.69	58.97 ± 8.65
		After	173.03 ± 15.41	48.90 ± 10.18
2	15	Before	136.05 ± 16.70	56.70 ± 8.26
		After	135.28 ± 14.82 a	57.03 ± 7.81 a
3	15	Before	125.17 ± 19.48	59.29 ± 9.60
		After	126.64 ± 18.68 a	58.53 ± 8.57 a

Data presented as mean ± SD.

a

P < 0.01 versus group 1; there was no significant difference (P  0.05) between the three groups before transfection or between groups 2 and 3; normally distributed data were analysed by analysis of variance with Fisher's post hoc test or Fisher's exact test; non-normally distributed data were analysed using the nonparametric Kruskal–Wallis and Mann–Whitney U-tests.

Group 1 rats were transfected with the endostatin (ES) gene by intralesional injection of the lipofectamine–ES-pBud plasmid; group 2 received an intralesional injection of lipofectamine-pBud, as a blank vector control; group 3 (untreated control) received an equivalent volume of phosphate-buffered saline.

As shown in Table 3, MVD in the endometriotic lesions was significantly lower in group 1 than in groups 2 and 3 (both P < 0.01). The endostatin level in group 1 was significantly higher than that in groups 2 and 3 (both P < 0.01), whereas the MMP-2 level in group 1 was significantly lower than that in the other two groups (both P < 0.01). The level of MMP-2 was negatively correlated with that of endostatin (r = -0.70, P < 0.05).

TABLE 3:

Transfection of the endostatin gene into endometriotic lesions in model rats: effect on microvessel density (MVD) and immunostaining scores for endostatin and matrix metalloprotein-2 (MMP-2) in the lesions

			Immunostaining score
Group	n	MVD	Endostatin	MMP-2
1	15	7.52 ± 0.84	6.58 ± 0.27	2.08 ± 0.12
2	15	20.56 ± 2.34 a	2.53 ± 0.15 a	4.66 ± 0.21 a
3	15	21.37 ± 1.40 a	3.22 ± 0.20 a	4.69 ± 0.21 a

Data presented as mean ± SD.

a

P < 0.01 versus group 1; there was no significant difference (P  0.05) between groups 2 and 3; normally distributed data were analysed by analysis of variance with Fisher's post hoc test or Fisher's exact test; non-normally distributed data were analysed using the nonparametric Kruskal–Wallis and Mann–Whitney U-tests. Group 1 rats were transfected with the endostatin (ES) gene by intralesional injection of the lipofectamine– ES-pBud plasmid; group 2 received an intralesional injection of lipofectamine-pBud, as a blank vector control; group 3 (untreated control) received an equivalent volume of phosphate-buffered saline. MVD, microvessel density, expressed as number of microvessels per high-power field.

Endometriotic lesion volume after transfection was significantly greater in group 1 than in groups 2 and 3 (P < 0.01) (Table 4). Before transfection, lesion volume did not differ significantly among the three groups.

TABLE 4:

Transfection of the endostatin gene into endometriotic lesions in model rats: effect on lesion volume

Group	n	Transfection	Volume, mm3
1	15	Before	96.37 ± 10.74
		After	39.33 ± 7.38
2	15	Before	98.06 ± 11.54
		After	115.53 ± 13.64 a
3	15	Before	97.71 ± 9.73
		After	112.51 ± 11.75 a ,b

Data presented as mean ± SD.

a

P < 0.01 versus group 1; there was no significant difference (P  0.05) between the three groups before transfection or between groups 2 and 3; normally distributed data were analysed by analysis of variance with Fisher's post hoc test or Fisher's exact test; non-normally distributed data were analysed using the nonparametric Kruskal–Wallis and Mann–Whitney U-tests.

Group 1 rats were transfected with the endostatin (ES) gene by intralesional injection of the lipofectamine– ES-pBud plasmid; group 2 received an intralesional injection of lipofectamine-pBud, as a blank vector control; group 3 (untreated control) received an equivalent volume of phosphate-buffered saline.

Discussion

The present study demonstrated for the first time that the inhibition of angiogenesis by the transfer of the endostatin gene is a potent treatment for endometriosis. Transient overexpression of the endostatin gene, produced by transfection into endometriotic lesions using the lipofectamine–pBudCE4.1 vector, successfully alleviated established endometriosis. Endostatin is a carboxyl-terminal fragment of type XVIII collagen, with strong antiangiogenic activity. ¹⁷ The present study demonstrated that transfection with the lipofectamine–ES-pBud construct had an apparent direct effect on serum VEGF in a rat model of endometriosis. In addition, MMP-2 protein levels, MVD and volume of endometriotic lesions were significantly decreased after endostatin gene transfer, indicating that endostatin inhibited the growth and angiogenesis of endometriotic lesions. These effects on endometriotic lesions strongly suggest that the mechanism involved in regression of endometrial tissue growth induced by endostatin gene transfer operates through an inhibitory effect on angiogenesis.

Endometriosis is an oestrogen-dependent condition that becomes quiescent after the menopause and may recur with oestrogen replacement therapy. Currently available therapies designed to induce the menopause chemically have severe adverse effects. ⁴ A drug that is effective while maintaining normal circulating oestrogen levels would, therefore, be an important advance in the treatment of endometriosis. The animal model of endometriosis used in the present study was produced by autotransplantation and does not result in immunological rejection; a success rate of 90% was achieved (45 model rats were obtained after 50 autotransplantations). Since autologous transplantation of endometrium offers a high survival rate for the implant, convenient observation and easy handling, the authors recommend it as a means of establishing a model of endometriosis in Lewis rats.

It is generally accepted that endometriosis is the result of implantation of shed endometrial tissues into the peritoneal surface after retrograde menstruation. A new blood supply is essential for the survival and development of the ectopic endometrium, and activation of angiogenesis might be a key factor in the pathogenesis of endometriosis. Recent studies have demonstrated that vascular distribution in the endometrium is irregular, and that the vessels are thick, dilated and/or reticular in patients with endometriosis.^18,19 Moreover, morphometric analysis of the endometrium showed that the mean and total surface areas (and the total number of capillaries) in patients with endometriosis were significantly higher than those in fertile women. ²⁰ In addition it has been demonstrated that, in patients with endometriosis, the mean MVD in endometriosis tissue was significantly higher than that in the adjacent endometrium, ²¹ showing that the ectopic endometrium has angiogenic properties. The volume and MVD of the endometrial lesions in the present study were significantly lower in rats treated with endostatin compared with controls, further showing that the endometrium has angiogenic potential and indicating that an adequate blood supply is essential for survival of the ectopic endometrium.

Vascular endothelial growth factor plays critical physiological and pathological roles in the regulation of vascular endothelial growth, angiogenesis and capillary permeability. ²² It is a basic heparin-binding homodimeric glycoprotein that binds to specific receptors of endothelial cells. VEGF is present in the corpora lutea of rats and primates, and in cultured human granulosa cells. ²³ It has also been reported that VEGF plays an important role in vascular development during follicular growth, luteal differentiation, oocyte maturation and fertilization. ²⁴ Thus, elevated VEGF levels in the follicles may be closely related to follicular visualization and maturation. In addition, members of the MMP family are believed to be the physiological mediators of matrix degradation. ²⁵ In a study where endometrial fragments from women with or without endometriosis were transplanted into mice, and the grafts were subsequently examined by immunohistochemistry, higher levels of VEGF and MMP-2 protein were found in grafts from women with endometriosis than in those without the disease, ²⁶ suggesting that VEGF and MMP-2 might play important roles in the formation and development of endometriosis. In a model where cultured human endometrial fragments were implanted into nude mice and two VEGF-A inhibitors (a truncated soluble inhibitory receptor and an affinity-purified antibody to human VEGF-A) were administered immediately after implantation, ²⁷ both angiostatic agents were effective in preventing blood-vessel growth and the development of the endometriotic explants. Like these previous studies, the present study showed that the serum level of VEGF and the immunostaining of MMP-2 in the endometrium of rats transfected with lipofectamine–ES-pBud were both significantly lower than in controls, indicating that endostatin is likely to interfere with the angiogenesis and growth of endometriotic tissue by antagonizing the actions of VEGF and MMP-2.

In conclusion, the present study demonstrated that gene therapy with the endostatin gene is a highly effective treatment for experimental endometriosis. Thus, restoration of endostatin gene expression by gene transfer in vivo might be a potential gene therapy strategy for human endometriosis.