Estradiol plays a role in regulating the expression of lysyl oxidase family genes in mouse urogenital tissues and human Ishikawa cells

Abstract

The lysyl oxidase (LOX) family encodes the copper-dependent amine oxidases that play a key role in determining the tensile strength and structural integrity of connective tissues by catalyzing the crosslinking of elastin or collagen. Estrogen may upregulate the expression of LOX and lysyl oxidase-like 1 (LOXL1) in the vagina. The objective of this study was to determine the effect of estrogen on the expression of all LOX family genes in the urogenital tissues of accelerated ovarian aging mice and human Ishikawa cells. Mice and Ishikawa cells treated with estradiol (E2) showed increased expression of LOX family genes and transforming growth factor β1 (TGF-β1). Ishikawa cells treated with TGF-β1 also showed increased expression of LOX family genes. The Ishikawa cells were then treated with either E2 plus the TGF-β receptor (TGFBR) inhibitor SB431542 or E2 alone. The expression of LOX family genes induced by E2 was reduced in the Ishikawa cells treated with TGFBR inhibitor. Our results showed that E2 increased the expression of the LOX family genes, and suggest that this induction may be mediated by the TGF-β signal pathway. E2 may play a role in regulating the expression of LOX family genes.

1. Introduction

Pelvic organ prolapse (POP) affects many women, leading to a decreased quality of life (Handa et al., 2008). There are multiple causes of POP, with aging, obesity, vaginal childbirth, and menopause as risk factors and with most patients being older and postmenopausal (Olsen et al., 1997; Hunskaar et al., 2005). Lower serum concentrations of estrogen and lower estrogen receptor (ER) expression in patients with POP have been observed (Lang et al., 2003; Bai et al., 2005). Estrogen deficiency may be an important factor in the pathogenesis of POP.

Lysyl oxidase (LOX) is a copper-dependent amine oxidase. Four paralogs, lysyl oxidase-like 1, 2, 3, and 4 (LOXL1, LOXL2, LOXL3, and LOXL4), have been identified. Most previous studies have focused on the role of LOX and LOXL1, whose main functions are catalyzing the crosslinking of elastin or collagen in the extracellular matrix (Hornstra et al., 2003; Liu et al., 2006). LOXL1-deficient mice display POP, thinner vaginal wall, and lower urinary tract dysfunction through an inability of urogenital tissues to replenish elastic fibers after parturition (Liu et al., 2006). LOX and LOXL3 protein levels are significantly decreased in POP patients (Alarab et al., 2010; Zhao and Zhou, 2012). LOXL2 mRNA levels are significantly decreased in postmenopausal POP patients compared with asymptomatic postmenopausal controls (Drewes et al., 2005). There exists a close relationship between POP and the expression of LOX family members.

High expression of LOXL1 can be detected in the urogenital tract of adult female mice, but decreases during aging (Liu et al., 2006). The mRNA levels of the LOX family genes are lower in the urogenital tissues of naturally aging mice than in those of young control mice (Jiang et al., 2014). The level of estrogen is also significantly decreased during aging. Previous studies have revealed that the mRNA expression of LOX and LOXL1 genes in the vaginal wall of mice can be upregulated by estradiol (E2) treatment and LOX protein expression and activity were increased by E2 treatment in rat cardiac fibroblasts (Drewes et al., 2007; Voloshenyuk et al., 2012). E2 can upregulate mRNA expression of LOX and LOXL1 in the vaginal wall of humans (Drewes et al., 2005). E2 treatment also results in increases in LOX mRNA levels in the vagina of guinea pigs after an injury (Voloshenyuk et al., 2012; Balgobin et al., 2013).

Most of these studies focused on LOX and LOXL1 in the vagina. The relationship between E2 and the other three LOX family genes remains unclear, and few studies of the relationship between LOX families and E2 in the uterus or bladder are reported. In our study, we used accelerated ovarian aging mice and Ishikawa cells derived from well-differentiated human endometrial adenocarcinoma as the experimental materials to investigate these questions. The accelerated ovarian aging mice express lower levels of E2 and LOX family genes. Furthermore, these mice can easily be developed, and this saves time compared with allowing mice to age naturally for use in investigating the role of estrogen (Jiang et al., 2013). We examined the expression of LOX family genes in the urogenital tissues of accelerated ovarian aging mice and Ishikawa cells following treatment with E2. Previous studies have shown that TGF-β family members may regulate LOX family genes (Sethi et al., 2011) and that E2 could significantly increase the expression of TGF-β family members in the uterus and vagina of mice (Takahashi et al., 1994). Thus, a preliminary investigation of the role of the TGF-β signal pathway in the relationship between E2 and the LOX family genes was included.

2. Materials and methods

2.1. Animals and treatments

Mice with a CD-1 background were used in these studies. The mouse model of accelerated ovarian aging was generated according to the study previously described (Jiang et al., 2013). Twelve female mice at two months of age were injected with busulfan (Bu) in dimethyl sulfoxide (DMSO; 12 mg/kg, subcutaneously) and cyclophosphamide (Cy) in 0.9% (9 g/L) sterile sodium chloride solution (120 mg/kg, intraperitoneally). The Bu/Cy-treated mice showed a significant reduction of primordial and primary follicles by 30 d after the treatment. The treated mice showed lower levels of estrogen and decreased expression of LOX family genes and elastin (Jiang et al., 2013). We randomly divided the Bu/Cy-treated mice into two groups: an E2-treated group and a control group (n=6 each group). The E2-treated group was injected with E2 in soybean oil (0.5 mg/(kg∙d), subcutaneously) for 12 d. The control group was injected with soybean oil. Twelve days after the onset of treatment, the mice in both groups were weighted, anesthetized with sodium pentobarbital (5 μg/g), and sacrificed. Urogenital tissues (vaginas, uteri, and bladders) were obtained for morphological and real time polymerase chain reaction (PCR) analyses.

2.2 Cells and treatments

Ishikawa cells were human endometrial adenocarcinoma cells cultured in minimum essential medium (MEM; Gibco, USA) supplemented with 10% (0.1 g/ml) heat-inactivated fetal bovine serum (JBI, Daegu, Korea) and 1% (0.01 g/ml) penicillin streptomycin (Gibco) at 37 °C in a 5% CO₂ atmosphere. Before treatment, the cells were cultured in MEM without phenol red (Gibco), supplemented with 10% charcoal stripped fetal bovine serum (JBI) and 1% penicillin streptomycin, at 37 °C in a 5% CO₂ atmosphere for 24 h. The cells were then treated with TGF-β1 (Peprotech; 0 and 10 ng/ml), E2 (Sigma; 0 and 10 nmol/L), or E2 (0 and 10 nmol/L) plus TGFBR inhibitor SB431542 (Abcam; 0 and 10 μmol/L).

2.3 Real-time PCR

Tissues and cells were analyzed for LOX family gene, TGF-β1, and housekeeping gene mRNA using real-time PCR as previously described by Jiang et al. (2014). Primers are listed in Tables 1 and 2 (Kenyon et al., 2003; Liu et al., 2004; 2006; Guo et al., 2014).

Table 1.

Primers for quantification of gene expression in mouse^*

Gene	GenBank accession No.	Sequence
		Forward primer (location)	Reverse primer (location)
mLOX	NM_010728.2	5'-TGCCAGTGGATTGATATTACAGATGT-3'	5'-AGCGAATGTCACAGCGTACAA-3'
mLOXL1	NM_010729.3	5'-AAGGCACAGCGGACTTTCTC-3'	5'-GAACTCGTCCATGCTGTGGTAA-3'
mLOXL2	NM_033325.2	5'-CAACCCCAAAGCCTATAAAACCT-3'	5'-GCCCGTGCAGTTCATAGAAAA-3'
mLOXL3	NM_013586.4	5'-GGGTGGACTCATAGTGCCAAATA-3'	5'-TCCCCCTGCAGCTCAGATT-3'
mLOXL4	NM_053083.3	5'-TGGTGACCTGTCGGCAACT-3'	5'-TCCCCCTGCAGCTCAGATT-3'
mTGFβ1	NM_011577.1	5'-TGAGTGGCTGTCTTTTGACG-3'	5'-TCTCTGTGGAGCTGAAGCAA-3'
mGAPDH	NM_008084.2	5'-TGAAGGTCGGTGTGAACGGATTTGGC-3'	5'-CATGTAGGCCATGAGGTCCACCAC-3'

^*Data refer to those reported previously (Kenyon et al., 2003; Liu et al., 2004; 2006; Guo et al., 2014)

Table 2.

Primers for quantification of gene expression in human cells^*

Gene	GenBank accession No.	Sequence
		Forward primer (location)	Reverse primer (location)
hLOX	NM_001178102.1	5'-TGCCAGTGGATTGATATTACAGATGT-3'	5'-AGCGAATGTCACAGCGTACAA-3'
hLOXL1	NM_005576.2	5'-TGAGGCCACCGACTACGATG-3'	5'-GAACTCGTCCATGCTGTGGTAA-3'
hLOXL2	NM_002318.2	5'-GATGACGACTTCTCCATCCACG-3'	5'-GTCGCCTCGTTGCCAGTACAG-3'
hLOXL3	NM_032603.2	5'-CGATGATGACTTCACGCTGC-3'	5'-CTCAAGTTGTCCAGCCAGATGC-3'
hLOXL4	NM_032211.6	5'-GCCTGTCGACAGCCACTACTACAG-3'	5'-ACACAGCTGACCACAGCGTGCATG-3'
hTGFβ1	NM_000660.5	5'-CGCGTGCTAATGGTGGAA-3'	5'-CGCTTCTCGGAGCTCTGATG-3'
hGAPDH	NM_002046.3	5'-CACATCGCTCAGACACCATGG-3'	5'-AATGAAGGGGTCATTGATGGCAAC-3'

^*Data refer to those reported previously (Guo et al., 2014)

2.4 Data analyses

All data are presented as the mean±standard error of the mean (SEM), and data analyses were performed using the SPSS statistical program, version 17.0. Student’s t test was used for single factor experiments involving two groups. A one-way analysis of variance (ANOVA) with Bonferroni testing was used for single factor experiments involving more than two groups. A significance level was set to P<0.05 for all statistical analyses.

3. Results

3.1 Increased expression levels of LOX family genes and TGF-β1 in urogenital tissues of accelerated ovarian aging mice treated with E2

Mice in the E2 and control groups were weighed and sacrificed, and their uteri, vaginas, and bladders were harvested for morphological analysis. After E2 treatment for 12 d, a small increase of 9% in the body weight of the E2 group was observed compared with the soybean oil vehicle control group (P<0.05; Fig. 1a). The uterine and vaginal walls of the E2 group were thicker than those of the control group (Fig. 1b). The uterine wet weight of the E2 group increased by 60% (P<0.01; Fig. 1c). Moreover, a 30% increase was observed in the ratio of uterus to body weight of E2 mice relative to that of the control group (P<0.05; Fig. 1d).

Fig. 1.

Histological differences between the genitourinary organs of E2-treated and control mice

(a) E2 group showed a 9% increase in the body weight compared with the control group (^* P<0.05); (b) The uterine and vaginal walls of the E2 group were thicker than those of the control group (U: uterine; V: vaginal; B: bladder); (c) E2 group showed a 60% increase in the wet weight of the uterus compared with the control group (^** P<0.01); (d) The ratio of the uterus weight to the body weight of the E2 group showed a 30% increase compared with the control group (^* P<0.05). Data are expressed as mean±SEM (n=6)

Uteri, vaginas, and bladders were obtained from the accelerated ovarian aging mice treated with E2 (0.5 mg/(kg∙d), subcutaneously) or with the soybean oil (control). The mRNA levels of the LOX family and the TGF-β1 genes were determined in these tissues. In the uteri, mRNA levels of LOX, LOXL1, LOXL2, LOXL3, LOXL4, and TGF-β1 were significantly higher in the E2-treated group than in the control group, with increases of 297%, 51%, 155%, 51%, 594%, and 176%, respectively (Fig. 2). Furthermore, the mRNA levels of those genes in the vaginas of the E2-treated group were upregulated by 242%, 498%, 157%, 287%, 12%, and 285%, respectively (Fig. 3). The expression levels in the bladders of the E2-treated group increased by 286%, 120%, 274%, 28%, 31%, and 177%, respectively (Fig. 4).

Fig. 2.

Effect of E2 treatment on LOX family and TGF-β1 mRNA levels in the uterus

The blank columns represent the control group (E2-untreated accelerated ovarian aging mice) and the grey columns represent the E2 group (0.5 mg/(kg∙d) E2-treated accelerated ovarian aging mice). Each bar represents the mean±SEM (n=6). ^* P<0.05, ^** P<0.01, E2-treated group vs. control group

Fig. 3.

Effect of E2 treatment on LOX family and TGF-β1 mRNA levels in the vagina

The blank columns represent the control group (E2-untreated accelerated ovarian aging mice) and the grey columns represent the E2 group (0.5 mg/(kg∙d) E2-treated accelerated ovarian aging mice). Each bar represents the mean±SEM (n=6). ^* P<0.05, ^** P<0.01, ^*** P<0.001, E2-treated group vs. control group

Fig. 4.

Effect of E2 treatment on LOX family and TGF-β1 mRNA levels in the bladder

The blank columns represent the control group (E2-untreated accelerated ovarian aging mice) and the grey columns represent the E2 group (0.5 mg/(kg∙d) E2-treated accelerated ovarian aging mice). Each bar represents the mean±SEM (n=6). ^* P<0.05, ^** P<0.01, ^*** P<0.001, E2-treated group vs. control group

3.2 Increased expression levels of LOX family genes and TGF-β1 in Ishikawa cells treated with E2

We also checked the expression of all five members of the LOX gene family in cultured cells. Ishikawa cells were used because our previous study showed that all the LOX family genes were expressed in this type of cell (data unpublished). We examined the levels of LOX and LOXL mRNA in the Ishikawa cells treated with E2. The mRNA levels of LOX, LOXL1, LOXL2, LOXL3, LOXL4, and TGF-β1 were significantly higher in the E2 (10 nmol/L)-treated cells than in the control cells, with increases of 832%, 218%, 436%, 629%, 569%, and 99%, respectively (Fig. 5).

Fig. 5.

Effect of E2 and E2 plus TGFBR inhibitor treatments on LOX family and TGF-β1 mRNA levels in Ishikawa cells

The blank columns represent the control cells (E2-untreated Ishikawa cells), the grey columns represent the 10 nmol/L E2-treated Ishikawa cells, and the black columns represent 10 nmol/L E2 plus 10 μmol/L SB431542-treated Ishikawa cells. Each bar represents the mean±SEM (n=6). ^* P<0.05, ^** P<0.01, ^*** P<0.001, E2-treated cells vs. control cells; ^# P<0.05, ^## P<0.01, E2+SB431542-treated cells vs. E2-treated cells

3.3 Increased expression levels of the LOX family genes in Ishikawa cells treated with TGF-β1

We examined the effect of TGF-β1 (10 ng/ml) on the expression of the LOX family genes in Ishikawa cells. TGF-β1 significantly increased the mRNA expression levels of LOX, LOXL1, LOXL2, LOXL3, and LOXL4 in the Ishikawa cells, with increases of 45%, 50%, 123%, 70%, and 109%, respectively, compared with controls (P<0.05; Fig. 6).

Fig. 6.

Effect of TGF-β1 treatment on LOX family mRNA levels in Ishikawa cells

The blank columns represent the control cells (TGF-βluntreated Ishikawa cells) and the grey columns represent the 10 ng/ml TGF-β1-treated Ishikawa cells. Each bar represents the mean±SEM (n=6). ^* P<0.05, ^** P<0.01, ^*** P<0.001, TGF-β1-treated cells vs. control cells

3.4 Decreased expression levels of LOX family genes in Ishikawa cells treated with TGFBR inhibitor

To test the role of TGF-β1 in the relationship between E2 and LOX family genes, Ishikawa cells were treated either with E2 (10 nmol/L) plus TGFBR inhibitor SB431542 (10 μmol/L) or with only E2 (10 nmol/L). The mRNA expression of the LOX family genes in the E2 plus TGFBR inhibitor-treated cells was lower than that in the E2-treated cells, with decreases of 815%, 142%, 350%, 553%, and 426%, respectively (Fig. 5).

4 Discussion

Estrogen plays an important role in the development and growth of the urogenital system. The term “estrogen” includes a group of chemically similar hormones: estradiol (E2, the most abundant in women of reproductive age), estrone, and estriol (Shimizu, 2010). Estrogen, specifically E2, is secreted mainly by the ovaries in premenopausal women (Shimizu, 2010). Estrogen circulates in the bloodstream and affects many tissues, including the urogenital system (Yasui et al., 2009).

Estrogen has been demonstrated to affect the expression of LOX and LOXL1 genes in the vagina (Drewes et al., 2005; 2007; Balgobin et al., 2013), though little is known about its effects on other members of the LOX family. Thus, to investigate the roles of estrogen and LOX family genes in the vagina, we examined the expression of LOX family genes in accelerated ovarian aging mice undergoing treatment with E2. The expression of LOX and LOXL1 genes was significantly higher in the E2-treated group than in the control group, which is consistent with previous reports (Drewes et al., 2005; 2007; Balgobin et al., 2013). The expression of the other LOX family genes (LOXL2, LOXL3, and LOXL4) was also increased in the vaginas of the mice treated with E2. These results suggest that E2 may improve the expression of all LOX family genes in vagina of mice.

To investigate whether E2 affected the expression of LOX family genes in other urogenital tissues, we used accelerated ovarian aging mice to determine the expression of LOX family genes in the uterus and the bladder of mice treated with E2. The expression of LOX family genes was significantly increased in the uterus and the bladder of mice in the E2-treated group compared with controls. We used Ishikawa cells that express all five LOX family genes to confirm similar results in vitro. The mRNA levels of LOX, LOXL1, LOXL2, LOXL3, and LOXL4 in the 10 nmol/L E2-treated group increased to variable degrees. The results showed that in addition to affecting expression in the vagina, E2 can affect the expression of LOX family genes in other urogenital tissues (uterus and bladder).

Takahashi et al. (1994) showed a strong link between E2 and TGF-β family members. E2 was able to stimulate the mRNA and protein expression of TGF-β family members in the uterus and vagina of mice prior to initiation of DNA synthesis. TGF-β1 mRNA levels were dramatically increased in the vaginas of E2-treated guinea pigs (Balgobin et al., 2013). Topical application of E2 can upregulate the expression of tropoelastin by activating the TGF-β signaling pathway in aged human skin in vivo (Son et al., 2005). E2 induces the formation and secretion of TGF-β (Soares et al., 2003; Gantus et al., 2011; Yu et al., 2011) and is able to increase TGF-β mRNA level in mice (Lindberg et al., 2002). Consistent with the above reports, our results showed that E2 induced TGF-β1 mRNA expression in the urogenital tissues of accelerated ovarian aging mice and Ishikawa cells.

TGF-β family members are primary regulators of many cellular processes including proliferation, growth, differentiation, and other functions in many cell types (Maurya et al., 2013). The functions of TGF-β family members are executed mainly through binding to type I and type II receptors, thus initiating a post-receptor signaling cascade (Derynck et al., 1996). The TGF-β signaling pathway has been shown to play an important role in many biological functions including the regulation of the extracellular matrix components and enzymes (Herpin et al., 2004). The TGF-β signaling pathway can regulate the amount and activity of the LOX family (LOX, LOXL1 to LOXL4) in human trabecular meshwork cells (Sethi et al., 2011). All five LOX family members can be upregulated by TGF-β1 in anterior cruciate ligament and medial collateral ligament fibroblasts after mechanical injury in humans (Xie et al., 2013). In our study, the E2 induction of LOX family gene expression was reduced by TGFBR inhibitors in the Ishikawa cells. Also, we found that the expression of all five LOX genes in Ishikawa cells was increased by treatment with TGF-β1. Thus, we suggest that the increased expression of LOX family genes by E2 treatment may occur through TGF-β signaling. In our results, the increase in mRNA expression of LOX family members induced by TGF-β1 was not as significant as the increase induced by E2 treatment. The reason is mainly that TGF-β family members exist in at least three isoforms, TGF-β1, TGF-β2, and TGF-β3. The regulation of LOX family gene expression by E2 was potentially via TGF-β1, TGF-β2, and TGF-β3, not only TGF-β1.

There has been no systematic study of the molecular mechanism of E2-induced TGF-β1 gene expression. However, Guo et al. (2006) showed that E2 binds to the ER and induces subsequent nuclear translocalization of the receptor dimmers. In the nucleus, E2 modulates the PI3K/Akt signal pathway via the action of ER at transcriptional level in Ishikawa cells (Guo et al., 2006). E2 also elicits non-transcriptional effects to activate the PI3K/Akt signal pathway in Ishikawa cells. In summary, E2 promptly activates the PI3K/Akt signal pathway in Ishikawa cells via ER-dependent and ER-independent mechanisms (Guo et al., 2006). The expression of TGF-β family members is reduced by blockade of the PI3K/Akt signal pathway in mouse osteosarcoma (Tsubaki et al., 2011). Activated Akt leads to upregulation of TGF-β1 in mesangial cells (Wu et al., 2009). In relation to E2-induction of TGF-β1 gene expression, it is likely that E2 stimulates the PI3K/Akt signal pathway, and activated Akt accumulation prompts the expression of TGF-β family members.

Studies have shown that deficient synthesis and degradation of elastic fibers are associated with POP (Liu et al., 2006). Decreased expression of elastin fibers in POP has been found (Karam et al., 2007; Goepel, 2008; Klutke et al., 2008; Zong et al., 2010). Functional elastin fibers are formed through a complex process, whereby one or more members of the LOX family of enzymes crosslink tropoelastin monomers to form polymers in the extracellular matrix (Liu et al., 2004). Human studies have revealed a significant decreased protein levels of LOX, LOXL1 and LOXL3 in the uterosacral and cardinal ligaments of POP patients (Kobak et al., 2005; Zhang et al., 2008; Zhou et al., 2013). A significantly decreased elastin content and suppression of mRNA expression from LOX, LOXL1, and LOXL2 are found in the uterosacral ligament tissue of women with prolapse compared with women in control groups (Klutke et al., 2008). In addition, LOXL1 knockout mice have a failure of elastic fiber homeostasis in the uterine tract postpartum and develop abnormalities, leading to POP (Liu et al., 2004; 2006). Thus, a lack of expression of LOX family genes may be a causal factor for POP in humans. In this study, we found that E2 could increase the expression of LOX family genes, and that this induction may be mediated by the TGF-β signal pathway. Thus, we suggest that E2 may potentially be used in interventions for POP in humans.