Introduction

The level of oxidative damage in an organism is tightly related to development of several age-related diseases, such as cancer, neurodegenerative diseases and diabetes [1]. The resistance to oxidative damage appears to be sex related, with females being more resistant to oxidative damage [2]. Females show lower incidence of some age-related pathologies linked with oxidative stress and this sex difference disappears after menopause, which led to conclusion that this protection is attributed to sex hormones. 17-β estradiol (E2) plays a key role in development and maintenance of normal sexual and reproductive function, and exerts a vast range of biological effects in the cardiovascular, immune and central nervous systems in both females and males [3]. E2 is a powerful endogenous antioxidant that is able to reduce lipid peroxidation in liver and blood [4]. Its deficiency leads to increased vascular reactive oxygen species (ROS) production resulting in endothelial dysfunction, whilst E2 replacement therapy can prevent these pathological changes [5]. It was shown that at physiological concentrations E2 does not act as a chemical antioxidant per se, due to its phenolic structure, but through mechanisms that involve oestrogen receptors and tyrosine kinase signalling through MAP kinase and NFκB [6]. Consequently, several antioxidant enzymes, including manganese superoxide dismutase (MnSOD) and HO-1, were shown to be up-regulated in MCF-7 cells [6]. Increased expression of HO-1 and several other antioxidant enzymes induced in myocardial cells by E2 treatment was explained by increased nuclear translocation of transcription factor Nrf2 [7]. HO is the rate-limiting enzyme for haeme degradation in mammals. It is a stress-response enzyme which is highly induced by variety of agents causing oxidative stress, hypoxia, hyperoxia, proinflammatory cytokines [8], and as such regarded as sensitive and reliable indicator of cellular oxidative stress. In comparison to the constitutively expressed HO-2, it is presumed that inducible isoform HO-1 makes greater contribution to the maintenance of oxidant/antioxidant homoeostasis during changes in cellular environments. In response to oxidative stress, HO-1 induction provides cell protection by promoting the catabolism of pro-oxidant metalloporphyrins to bile pigments (biliverdin and bilirubin) which are considered to have free radical scavenging properties [9]. As such, HO-1 acts as a potent antioxidant. Transcriptional control of the expression of HO-1 is mediated through the antioxidant response element (ARE), a transcriptional regulatory element located in the upstream regulatory region of many phase II enzymes that play important protective role against oxidative toxicity [10]. ARE is activated via translocation of Nrf2 transcription factor into nucleus upon oxidative stress and other stimuli. Keap1 is a substrate adaptor protein for the Cul3–Rbx1 E3 ubiquitin ligase complex that represses Nrf2 by targeting it for proteasomal degradation [11]. Most recently, it was demonstrated that cytosolic protease dipeptidyl peptidase III (DPP III) binds Keap1 to displace Nrf2, thus inhibiting Nrf2 ubiquitination and driving Nrf2-dependent transcription [12]. DPP III is a monozinc exopeptidase that hydrolyses dipeptides from the N-terminal of its substrates consisting of three or more amino acids [13]. In mammalian tissues, it is broadly distributed and thought to contribute in the final steps of normal intracellular protein catabolism [14]. There are strong indications of its role in the endogenous pain-modulation system [15] as well as in the endogenous defense against oxidative stress [16]. Pathophysiological roles of human DPP III are indicated in cataractogenesis [17] and malignant growth [18]. Recent studies showed that DPP III is a member of a six-gene signature that accurately predicts human breast cancer patient survival [19]. All these findings characterize human DPP III as a valuable drug target.

Although the effect of E2 on HO-1 expression and its actions as an antioxidant have been well documented [6, 7], currently there are no data about the in vivo influence of E2 on sex-related HO-1 expression. Moreover, it is not known whether E2 influences DPP III. The liver was chosen, because of its high susceptibility to oxidative stress due to high metabolic activity. In addition, Keap1–Nrf2 pathway is indicated to counteract liver diseases. Therefore, the aim of this study was to examine the effect of E2 on the expression of DPP III, the protease involved in Nrf2–Keap1 signaling pathway—the main pathway responsible for cell defense against oxidative stress [11], and to associate it with the expression of known antioxidant enzyme HO-1 in the liver of adult CBA/H mice of both sexes.

Materials and methods

Animals and experimental design

The experiments were performed in accordance with the current laws of the Republic of Croatia and with the guidelines of European Community Council Directive of November 24, 1986 (86/609/EEC). Male and female CBA/H mice aged 4 months from breeding colony of the Ruđer Bošković Institute (Zagreb, Croatia) were used for all experiments. The animals were maintained under the following laboratory conditions: three to a cage; light on from 06:00 to 18:00; 22 ± 2 °C room temperature; access to food pellets, and tap water ad libitum. The experimental groups were as follows: male control (cm), male control treated with E2 (cme), female sham (cf), ovariectomized (ovx), ovariectomized and treated with E2 (ovxe) mice. The number of animals was three per each experimental group.

Ovariectomy procedure and E2 administration

Ovariectomy and sham surgery were performed under ketamine/xilazyne anaesthesia. Since low levels of E2 are normally detected in ovariectomized females due to other endogenous E2 sources [20], plasma E2 levels were not used as indicator of efficiency of ovariectomy. Instead, the success of ovariectomy was checked by analysing vaginal smear during five consecutive days after the surgery (data not shown). In our preliminary experiments, we established that anestrus phase seen on vaginal smear is reflected by uterus atrophy as evaluated post-mortem. Body weight gain was also used as a marker of successful ovariectomy, since ovariectomy-induced body weight gain is established phenomena. For E2 administration, a pellet containing E2 (50 μg, Innovative Research of America, Sarasota, FL) was placed into the interscapular subcutaneous space releasing a constant dose of 830 ng daily. The procedure was the same for both females and males. After 37 days, animals were subjected to experimental protocols.

RNA isolation and quantitative real-time PCR analysis

Total RNA was extracted from individual mouse livers in each group using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Reverse transcription and real-time PCR analysis were done as described previously [37], to quantify relative mRNA expression of DPP III and HO-1. Using the 2−ΔΔCt method, data are presented as the fold-change in DPP III and HO-1 gene expression normalized to endogenous reference gene (β-actin) and relative to the untreated control. Assays used in this study are listed in Table 1. All reactions were carried out in triplicate.

Table 1 Probes used for quantitative real-time PCR analysis
Full size table

SDS-PAGE and Western blotting

Liver was homogenized with RIPA buffer supplemented with proteinase inhibitors (10 % w/v) using an ice-jacketed Potter–Elvehjem homogenizer (1,300×g). After sonification (3 × 30 s), whole liver homogenates were centrifuged at 16,000 g for 20 min in refrigerated centrifuge. Supernatant was collected and total cellular proteins (100 μg per lane) were resolved by denaturing SDS-PAGE, transferred onto a PVDF membrane (Bio-Rad, Hercules, CA). Membranes were blocked in 5 % nonfat dry milk in TN buffer (50 mM TRIS, 150 mM NaCl, pH = 7.4) overnight, incubated with primary polyclonal rabbit antibody against DPP III (antiserum diluted 1:200 and incubated 3 h at room temperature) or with primary polyclonal rabbit anti-mouse HO-1 antibody (Abcam, Cambridge, UK, diluted 1:200 and incubated 18 h at +4 °C), followed by incubation with donkey anti-rabbit IgG, horseradish peroxidase-conjugated, secondary antibody (Amersham Biosciences Inc., USA) for 3 h at room temperature. Equality of loading was confirmed using AmidoBlack (Sigma Aldrich, St.Louis, USA), which was also used for normalization of the bands [21]. The chemiluminescence signals were detected and analysed with the Alliance 4.7 Imaging System (UVITEC, Cambridge, UK). The blots were repeated at least three times and representative blots are presented. A high-titre anti-DPP III polyclonal antibody was custom prepared by Abcore, USA, using purified recombinant human DPP III as antigen [22], injected into New Zealand white rabbits, 3- to 9-month old. Rabbits were bled from the auricular artery, blood was allowed to coagulate, clot was removed and serum was collected, clarified by centrifugation and stored at −20 °C.

Immunohistochemistry

After perfusion, tissue was fixed by immersion in 4 % paraformaldehyde in 0.1 M PBS (pH 7.4), then tissue blocks were embedded in paraffin, sectioned (12 µm) and deparaffinized through a graded series of xylol and alcohols. Sections were processed with immunohistochemistry as previously described [23], with minor modifications. In brief, after incubation in 0.3 % H2O2 in PBS and blocking in 5 % normal goat serum to prevent nonspecific background staining, rabbit anti-Haeme Oxygenase 1 (HO-1) monoclonal antibody (Abcam, UK) or anti-DPP III polyclonal antibody (custom made by Abcore, USA) was diluted 1:200 and incubated overnight at 4 °C. Biotinylated anti-rabbit antibody from Vectastain ABC kit (Vector Laboratories Inc) was used according to the manufacturer’s protocol. For visualization of peroxidase activity, 3,3′-diaminobenzidine tetrahydrochloride (DAB) with metal enhancer (CoCl2, Sigma-Aldrich) was used to produce intense gray staining. Sections were rinsed, air dried, dehydrated and coverslipped with Histamount (National Diagnostic). Negative controls were included in all immunohistochemical experiments by replacing the primary antibody with blocking solution, or 5 % normal goat serum. No immunolabeling was detected in control sections. Tissue sections were photographed by slide scanner NanoZoomer 2.0 (Hamamatsu) and image assembled in CorelDRAW(R)—Version 11.0.

Statistical analysis

Statistical analyses of data were performed using R v2.15.3 (CRAN, http://cran.r-project.org) and RStudio for Windows, v 0.97 (http://www.rstudio.com/). For the analysis of real-time PCR data, all groups were tested for normality of distribution using Shapiro–Wilk test. The differences between multiple groups were compared with Kruskall–Wallis non-parametric ANOVA, followed by Wilcoxon signed-rank test for testing differences between two related groups. For all tests significance level was set at p < 0.05.

Results

DPP III gene and protein expression in the liver of CBA/H mice

To establish whether sex-related differences in the basal expression of hepatic DPP III exist, we measured DPP III gene and protein expression in intact male and female mice. Both gene and protein expression levels of DPP III were found to be unchanged between sexes, as revealed by real-time PCR and Western blot analysis (Fig. 1a, b). Next, we wanted to determine whether DPP III gene and protein expression were influenced by E2 in males, who basically have low levels of this hormone. Real-time PCR analysis showed no significant effect of E2 on DPP III gene expression in cme, compared to cm group (Fig. 2a). However, there was a marked increase in DPP III protein level in cme, compared to cm group (p = 0.001) (Fig. 2b). We then investigated the level of DPP III gene expression in the liver of cf, ovx and ovxe animal group. The transcript level remained unchanged across all experimental groups (Fig. 2c). However, Western Blot analysis showed that ovariectomy decreased DPP III protein level (p = 0.048 for ovariectomized vs. control females), whilst E2 administration abolished the effect of ovariectomy and increased DPP III expression even higher than control level (p = 0.007 for ovx vs. ovxe) (Fig. 2b).

Fig. 1
figure 1

DPP III expression in liver of control CBA/H mice. Real-time PCR analysis of DPP III gene expression in liver of control male (cm) and female (cf) CBA/H mice. Fold-change in gene expression was calculated with 2−ΔΔCT method, using β-actin as endogenous control (a). Western Blot analysis of DPP III protein expression in liver of control male and female CBA/H mice. Bands are normalized using AmidoBlack (b). All data are mean ± SD from three individual mice per group

Full size image
Fig. 2
figure 2

Effect of E2 on DPP III expression in liver of CBA/H mice. Real-time PCR analysis of DPP III gene expression in liver of control (cm) and E2-treated male (cme) CBA/H mice (a), and sham (cf), ovariectomized (ovx) and ovariectomized females treated with E2 (ovxe) (c). Fold-change in gene expression was calculated with 2−ΔΔCT method, using β-actin as endogenous control. Western Blot analysis of DPP III protein level in liver of control and E2-treated male CBA/H mice (b: p = 0.001 control vs. E2-treated group) and sham, ovariectomized, and ovariectomized females treated with E2 (d: p = 0.048 control vs. ovariectomized; p = 0.007 ovariectomized vs. ovariectomized treated with E2). Bands are normalized using AmidoBlack. All data are mean ± SD from three individual mice per group

Full size image

HO-1 gene and protein expression in the liver of CBA/H mice

The expression level of protective enzyme HO-1 in intact mice of both sexes was determined, since there are no such data for CBA/H strain of mice. There was no change in the hepatic expression of HO-1 between control animals of both sexes, as revealed by real-time PCR and Western Blot analysis (Fig. 3a, b). Similarly to DPP III, we wanted to determine whether HO-1 gene and protein expression were influenced by E2 in males. No change of expression in HO-1 mRNA and protein level between cm and cme group was found (Fig. 3c, d). However, real-time PCR data analysis showed significant decrease in HO-1 gene expression fold-change (fc = −2.85, p < 0.001) in liver of ovx group compared to both cf and ovxe group (Fig. 4a). Western Blot has shown that ovariectomy decreased HO-1 protein level (p = 0.042 for ovx vs. cf), whilst E2 abolished the effect of ovariectomy and increased HO-1 expression compared with cf (p = 0.002 ovx vs. ovxe) (Fig. 4b).

Fig. 3
figure 3

HO-1 expression in liver of CBA/H mice. Real-time PCR analysis of HO-1 gene expression in liver of control male (cm) and female (cf) CBA/H mice (a), and in control (cm) and E2-treated male (cme) CBA/H mice (c). Fold-change in gene expression was calculated with 2−ΔΔCT method, using β-actin as endogenous control. Western Blot analysis of HO-1 protein expression in liver of control male (cm) and female (cf) CBA/H mice (b), and control male (cm) and E2-treated male (cme) CBA/H mice (d). Bands are normalized using AmidoBlack. All data are mean ± SD from three individual mice per group

Full size image
Fig. 4
figure 4

Effect of E2 on HO-1 expression in liver of CBA/H mice. Real-time PCR analysis of HO-1 gene expression in liver of sham (cf), ovariectomized (ovx) and ovariectomized treated with E2 (ovxe); p < 0.001 cf vs. ovx and ovx vs. ovxe. Fold-change in gene expression was calculated with 2−ΔΔCT method, using β-actin as endogenous control (a). Western Blot analysis of HO-1 protein expression in liver of sham (cf), ovariectomized (ovx) and ovariectomized females treated with E2 (ovxe); p = 0.042 control vs. ovariectomized; p = 0.002 ovariectomized vs. ovariectomized treated with E2 (b). Bands are normalized using AmidoBlack. All data are mean ± SD from three individual mice per group

Full size image

Immunohistochemical analysis of DPP III and HO-1 localization in mouse liver

Immunohistochemical analysis revealed that in the liver the expression of HO-1 was more pronounced in pericentral areas of hepatic lobules, and particularly in Kupffer cells, whilst DPP III protein was found to be expressed in hepatocytes and uniformly distributed throughout the hepatic tissue. Considering intracellular distribution, HO-1 was found more prominent in the cytoplasm and membranes of organelles, whilst strong DPP III immunoreactivity was detected, in addition to the cytoplasm, in nuclei and plasma membranes (Fig. 5).

Fig. 5
figure 5

Immunohistochemical localization of HO-1 and DPP III protein in mouse liver. a Negative control (liver tissue not incubated with primary antibody). b HO-1 immunoreactivity is more pronounced in pericentral areas, inside Kupffer cells (arrows), and in hepatocytes (H). c DPP III immunoreactivity is localized in hepatocytes (H) and uniformly distributed throughout the hepatic tissue

Full size image

Discussion

In this study the effect of E2 on DPP III and HO-1 gene and protein expression level in the liver of 4 month-old male and female CBA/H mice was investigated. The liver is particularly susceptible to oxidative stress, and E2, DPP III and HO-1 contribute to cellular antioxidant defense. Reactive oxygen species (ROS) are produced in the liver as byproducts of normal metabolism and detoxification [24]. Hepatic proteins, lipids and DNA are primarily affected by excessive and sustained ROS, which results in structural and functional abnormalities. The pathogenesis of the oxidative damage involves each hepatic cell type (i.e., hepatocytes and stellate cells, endothelial and Kupffer cells) [25]. It is generally accepted that oxidative stress plays a crucial role in promoting the progression of chronic liver diseases and hepatocellular carcinoma. The DPP III is involved as activator in Keap1–Nrf2–ARE signaling pathway. The protective roles of Nrf2 activation in the pathogenesis of liver diseases have been intensively investigated. In the animal model, Keap1–Nrf2–ARE pathway has been demonstrated to counteract non-alcoholic and alcoholic liver diseases, fibrosis and cancer, and to support liver regeneration [27]. Several Keap1–Nrf2-activating drugs and molecules derived from plants (e.g. sulforaphane) have entered clinical trials for several disease processes, in which oxidative stress and inflammation play a crucial role [28].

Our present research has shown that ovariectomy markedly lowered the level of both DPP III and HO-1 proteins in liver tissue, and that E2 administration abolished this effect and increased DPP III and HO-1 protein expression in both sexes. Using immunohistochemistry, we identified that HO-1 localizes in the pericentral areas of hepatic lobules, particularly in Kupffer cells, which is in accordance to previous studies [26]. In our knowledge, the present study is the first to have analysed the immunohistochemical localization of DPP III in the mouse liver. In contrast to HO-1 protein, we have found a uniform expression of DPP III within hepatic tissue, with a more intense signal in nuclei and membranes of hepatocytes. Interestingly, human DPP III protein is localized in the cytoplasm, plasma membranes and nuclei of cells (http://www.proteinatlas.org/).

Our present data clearly show for the first time that E2 influences the level of DPP III protein in mice of both sexes. It seems that E2 affects DPP III at a post-transcriptional level. This is in accordance with up to date knowledge on DPP III gene expression, since no oestrogen response element was recognized in its gene promoter, and transcription factors Ets-1/Elk-1 and C/EBP-β have been experimentally confirmed as relevant for human DPP III gene regulation [29, 30]. Human and mouse DPP III share 93 % of amino acid sequence identity. Mouse DPP III (protein) has not been characterized yet biochemically, whilst human and rat orthologs have been the subject of intensive investigation [13, 14, 29–31]. The increased level of DPP III protein observed in this study as a result of E2 action in vivo could be related to an imbalance between protein synthesis and degradation, i.e. it could be due to a decreased proteolytic degradation in E2-treated animals. It is known that E2 regulates several mammalian peptidases, such as neprilysin, cathepsin D and ubiquitin-specific peptidase 19 (USP 19) [32, 33]. Ogawa and colleagues have shown that E2 increases the expression of USP 19 in mouse myoblasts and satellite cells, both in vitro and in vivo [34]. As USP 19 does not associate with proteasome, it may release ubiquitin from specific proteins and may have regulatory function, since removal of ubiquitin chain from specific substrate proteins inhibits their degradation by proteasome. At this point, it is not known whether the DPP III is amongst substrates of USP 19 or some other deubiquitinating enzyme. However, several ubiquitination sites can be predicted in DPP III molecule, and experimental evidence on human DPP III ubiquitination has been reported [35].

Considering that males have lower serum E2 levels compared to intact females, and our finding that E2 administration increases the DPP III protein content, the observed lack of difference in protein DPP III expression in intact male and female mice is intriguing. This indicates that in vivo the regulation of DPP III protein is rather complex and that the existence of other factors (which are able to up-regulate DPP III) can be postulated. We can speculate that in male mice, androgens could up-regulate DPP III protein. However, up to date there are no data about the effect of testosterone or any other androgenic hormone on the regulation of DPP III.

Our results on the increased expression of HO-1 gene are in agreement with findings of Yu et al. [7] about E2 effects of transcription factor Nrf2 on myocardial cells. These authors have shown a concomitant upregulation of the transcription factor Nrf2 in nuclear extracts. Moreover, since HO-1 is one of the phase II enzymes with ARE in its promoter, and DPP III was shown to compete with Nrf2 for binding to the same site on cytosolic repressor Keap1, our finding of an oestrogen-induced increase in HO-1 mRNA and protein content is consistent with the notion that enhanced DPP III protein releases Nrf2 from its repressor and thereby augments the transcription of responsive genes, such as HO-1. E2 has well-established cytoprotective effect during oxidative stress and its depletion contributes to the pathogenesis of some age-related diseases [36]. We have previously demonstrated that the resistance to hyperoxia-induced oxidative stress appears to be female biased [37, 38] and that E2 deficiency caused an increase in oxidative damage in females exposed to hyperoxia, whilst E2 administration abolished this effect. Moreover, E2 treatment significantly decreased oxidative damage in males [39]. These results strongly suggested a beneficial effect of E2 in resistance to oxidative stress. Data from our present study clearly show that protein levels of two enzymes involved in endogenous response to oxidative stress, DPP III and HO-1, are increased upon administration of E2 in mice.

Although gender-specific differences in hepatic HO expression and activity have been shown in previous studies [40] here we demonstrate for the first time that E2 influences the protein level of DPP III in vivo. These results additionally confer new insights into the complexity of the protective action of this hormone, and in its relationship to Keap1–Nrf2–ARE pathway, which contains molecular targets potentially able to prevent and treat liver diseases.