Dystroglycan patterns on the prostate of non-obese diabetic mice submitted to glycaemic control

Abstract

Dystroglycan (DG) is an adhesion protein which plays a crucial role in the maintenance of tissue integrity. Diabetes has been pointed out as a disease which causes harmful effects on prostate function. Therefore, the main objective of this work was to verify DG distribution and structure features in diabetic mice with and without glycaemic control and to relate these parameters to prostate pathogenesis. Thirty mice (Nod and BALB/c) were divided into three groups after 20 days of diabetic state: the control group received a 5 ml/kg dose of physiological saline daily for 20 days; the diabetic group had the same treatment as the control group; the diabetic-insulin group received 4–5 IU doses of Neutral Protamine Hagedorn (NPH) insulin daily for 20 days. After 20 days of treatment, all animals were killed and samples from the ventral prostate were processed for immunological and light microscopy analyses. The results showed diminished β- and α-DG receptors in the diabetic group. However, there was a recovery of both β-and α-DG receptor immunolocalization after insulin administration. Epithelial and stromal morphological changes were verified in the diabetic group, which also presented recovery after insulin treatment. Thus, it could be concluded that diabetes disturbed prostate structure integrity and altered the occurrence of α and β-DG receptors, indicating decreased cell-matrix extracellular and cell-basal membrane attachment. However, insulin treatment could partially restore glandular homeostasis. The decrease in epithelial–stromal interaction certainly predisposes this gland in diabetic mice to be a prostate disease target.

Diabetes mellitus is a disease which causes protein, carbohydrate and lipid metabolism alterations leading to hyperglycaemia because of insulin deficiency (Mokdad 2001; Conget 2002). Moreover, diabetic patients showed various alterations in their organic system such as vascular, urogenital, neuronal and digestive disorders among others (Ciardullo et al. 2004).

Different experimental studies in the male reproductive system, which exhibited type 1 diabetes mellitus because of either chemical induction or spontaneous development presented retrograde ejaculation (Ellemberg 1980), urinary bladder dysfunction (Buck et al. 1976; Stefan 1996), sexual impotence and decreased spermatozoid number in the seminal fluid (Frenkel et al. 1978; Mcculloch et al. 1984). Cagnon et al. (2000) and Ribeiro et al. (2006) verified atrophied secretory cells, stromal hypertrophy, inflammatory processes, prostatic intraepithelial neoplasia and dilated secretory organelles in the ventral lobe of the prostate in non-obese diabetic mice. Also, Carvalho et al. (2003) observed morphological changes in the coagulating gland of diabetic mice such as inflammatory cells in the stroma and hypertrophied extracellular matrix elements. Other studies revealed diminished testosterone serum levels and a low androgen receptor rate in the diabetic animals resulting from imbalanced hormone metabolism (Oksanen 1975;Tesone et al. 1976). Furthermore, Wang et al. (2000) stated that the diabetic rats, which underwent glycaemic control using insulin, did not recover their prostate weight. Also, Suthagar et al. (2008) verified that the diabetic state caused diminished testosterone and oestrogen receptor levels and insulin replacement was a recovery factor for these steroid hormone levels. In clinical observation, diabetes was pointed out as being a feasible aetiological factor in prostate cancer development (Ilic et al. 1996). According to Hammarsten and Hogstedt (2002), the increase in benign prostatic hyperplasia, as well as its quick progression in diabetic men, could be considered a risk factor towards the development of clinical cancer. On the other hand, there are clinical studies which demonstrated that diabetic patients had a reduction in the risk for prostate cancer (Kasper et al. 2008). In addition, another study in men presenting with prostate cancer and who received androgen suppression therapy showed that there was a high risk of these men developing insulin resistance and showing hyperglycaemia (Basaria et al. 2006).

The prostate is one of the most important accessory sex glands with its secretion being fundamental for the male reproductive process (Setchell & Brooks 1988; Marker et al. 2003; Untergasser et al. 2005).

The rodent prostate is divided into three pairs of lobes: ventral, dorsal and lateral according to their anatomic positions in relation to the urethra (Jesik et al. 1982; Sugimura et al. 1986; Aumüller & Seitz 1990). These lobes show different characteristics in relation to morphology, secretion and hormone dependence and the ventral lobe is primarily regulated by androgens (Costello & Franklin 1994; Colombel & Buttyan 1995).

In general, the prostate lobes show a simple epithelium surrounded by stroma (Aumüller & Ader 1979). The prostate stroma is made up of a complex net of stromal cells and extracellular matrix associated with growth factors, regulatory molecules and enzyme restructuring, which provide biological signals and have mechanical influence on the epithelial cells (Tuxhorn et al. 2001; Cunha & Matrisian 2002). The fibroblasts and smooth muscle cells are important cellular types in the prostate stroma synthesizing structural and regulatory compounds of the extracellular matrix. The extracellular matrix is a net of fibrillar proteins, adhesion glycoproteins and proteoglycans (Lin & Bissel 1993;Kreis & Vale 1999; Tuxhorn et al. 2001). Thus, the association of stromal cells with extracellular matrix provides a microenvironment, which regulates the growth and functional differentiation of the adjacent cells, where each of these elements plays a crucial role in the structural maintenance and tissue function (Tuxhorn et al. 2001). The stroma–epithelium interaction has an important role in prostatic structural maintenance and function (Ekman 2000). The basal membrane is an interaction link, offering mechanical and physiological support which is specially made up of collagen type IV and laminin (Knox et al. 1994). The imbalance in the stroma-epithelium interaction leads to different prostate diseases including prostate cancer (Wong et al. 2000; Cunha et al. 2001, 2003; Cornell et al. 2003).

Currently, different studies suggest that dystroglycan (DG), which is an adhesion protein, plays a role in epithelial and neuronal cell development, formation of basal membrane and maintenance of tissue integrity (Henry & Campbell 1998). DG, despite having first been discovered in the skeletal muscle as a compound of the dystrophin–glycoprotein complex, is found in many other non-muscle tissues such as smooth muscle, epithelia and peripheral nerves (Ibraghimov-Beskrovnaya et al. 1992; Henry et al. 2001). Dystroglycan is linked to extracellular matrix proteins such as laminin, perlecan and agrin proteoglycans, in addition to cytosolic proteins (Losasso et al. 2000; Sugita et al. 2001). DG is formed by two protein subunits, β and α, interacting to form a non-covalent complex, which is identified by the same gene (Brennan et al. 2004). α-DG links extracellular molecules, whereas the transmembrane β-DG anchors α-DG to the cell membrane (Sgambato et al. 2003).

Dystroglycan expression changes have been verified in the occurrence of cancer in different organs including the prostate (Losasso et al. 2000; Henry et al. 2001; Sgambato et al. 2003; Brennan et al. 2004). According to Henry et al. (2001), decreased DG expression was observed in high grade prostatic cancer, leading to abnormal prostate cell interaction with the extracellular matrix causing metastasis. Thus, these studies indicated that changes in the DG expression are a determining factor for prostate pathogenesis.

Thus, based on the negative influence of diabetes mellitus on the prostate function as well as the fundamental DG participation in the cell–extracellular matrix interactions in the connection between epithelial cells and basal lamina and in the cytoskeleton organization, the main objective of this work was to characterize DG distribution and structure features in diabetic mice with and without glycaemic control and to verify the possibility of the association of these parameters on prostate pathogenesis.

Animals and tissue preparation

A total of 20 mice (Nod) and 10 control mice (BALB/c/Uni), all 18 weeks old, from the Bioterism Center/Unicamp were used. The blood glucose level was measured by capillary glycaemia, utilizing the Optium Advanced Diabetes Management System (MediSense, Abingdon, UK). Thus, the animals showing blood glucose level >300 mg/dl were considered diabetic (Shirai et al. 1998). Twenty days into a diabetic state, the mice were divided into three groups: the control group received a 5 ml/kg dose of 0.9% physiological saline subcutaneously daily for 20 days (Fresenius Kabi, São Paulo, Brazil); the diabetic group received the same treatment as the control group; the diabetic-insulin group received 4–5 IU doses of NPH insulin subcutaneously daily for 20 days (Biobrás, Montes Claros, Minas Gerais, Brazil) (Anderson 1983). Insulin administrations were interrupted 24 h prior to the mice being killed. After 20 days of experimental treatment, the animals were anaesthetized with a 0.25 ml/100 g body weight dose of Francotar/Virbaxyl (1:1; Vibra® Roseira, SP, Brazil), and samples of the intermediate and distal regions from the ventral lobe of the prostate were collected under a DF Vasconcellos Steromicroscopy which allows the withdrawal of prostatic samples from these specific regions to be processed for morphometric, structural and immunological analyses. Hormonal dosages were also carried out on the blood samples.

Serum testosterone and glucose levels

At the end of the 20-day treatment, blood samples from all animals in each group were collected. The blood samples were collected through a cardiac puncture, 24 h after administrating the last dose of insulin. The serum concentrations of testosterone were determined by radioimmunoassay using Coat-a-Count total testosterone kit (Diagnostic Products Corporation, Los Angeles, CA, USA) and expressed in ng/dl. The serum concentration of glucose was determined by chemiluminescense and expressed in mg/dl.

Light microscopy: immunolabelled α-DG and β-DG and morphometric analyses

Samples of the ventral prostate were collected from 10 animals in each group for histological and immunological analyses and then fixed by immersion in Bouin's solution, embedded in paraplast (Paraplast Plus, São Paulo, Brazil), cut into 5–6 μm thick sections. These samples were submitted to the following staining procedures: haematoxylin-eosin (Behmer et al. 1976), Picrossirius red (Behmer et al. 1976) and immunostaining for α- and β-DGs. The slides were photographed with a Nikon Eclipse E-400 photomicroscope (Nikon Corporation, Tokyo, Japan). Epithelial, luminal and stromal areas in the ventral lobe of the prostate were measured (25 fields per animal) for structural analyses in the sections stained for light microscopy in 10 animals per group, using 200× magnification. The photomicroscope Nikon Eclipse E-400 and the NIS-Elements: Advanced Research (USA) computerized image analysis system were used.

For immunological proceeding, the sections were deparaffinized in xylene, hydrated through graded alcohol concentrations and rinsed under tap water. Antigens were retrieved by boiling the sections in 10 mM citrate buffer, pH 6.0, three times for 5 min in a microwave oven. After that, the sections were incubated in 0.3% H₂O₂ for 15 min to block endogenous peroxidase. Non-specific binding was blocked by incubating the sections in a blocking solution for 1 h at room temperature. Primary rabbit H-300 (sc-28534) (Santa Cruz Biotechnology, California, USA) for the α-DG and mouse (NCL-b-DG) (Novocastra Laboratories, Newcastle, UK) for the β-DG antibodies were diluted in 1% BSA (1:50) and applied to the sections overnight at 4 °C. The Envision HRP Kit (Dako Cytomation Inc., Carpenteria, CA, USA) was used to visualize the bound antibody according to the manufacturer's instructions. The sections were washed for 15 min. with Tris Buffered Saline-Tween 20 (TBS-T) and secondary labelled polymer from the Envision HRP Kit (Dako) was applied for 40 min at room temperature. After washing in TBS-T, peroxidase activity was detected using a diaminobenzidine chromogen kit from Envision HRP Kit (Dako) for 10 min. Sections were lightly counterstained with methyl green dehydrated in an increasing ethanol series and xylene, mounted in entellan (Merck, Darmstadt, Germany).

The immunolocalization of the α-DG and β-DG were measured in 10 animals in each experimental group. Five microscopic fields per animal were measured with 40× objective lens and corresponded to a total area of 90,304.7 μm². α-DG and β-DG were quantified based on the area of positive immunostaining expressed as a percentage of the total area examined. In addition, staining intensity was graded as strong, moderate or weak according to concentration and distribution of the receptor in the sectioned tissues (Markopoulos et al. 2000).

Statistical analysis

The serum glucose and testosterone levels, the percentage of α-DG and β-DG positive stained areas, epithelial, luminal and stromal areas were compared between groups and analysed statistically by means of analysis of variance and Tukey multiple range test, with the level of significance set at 1% (Zar 1999).

Results

Serum testosterone and glucose levels

The glucose average level of the animals from the diabetic group was over 300 mg/dl. In contrast, the levels observed in the animals from the control and diabetic-insulin groups were lower than the value verified in the diabetic group (Table 1).

Table 1.

Serum glucose (mg/dl) and testosterone (ng/ml) levels of mice in the different experimental groups

	Groups
Variants	Control	Diabetic	Diabetic-insulin
Glucose	180.1 ± 6.1a	839.1 ± 40.0b	365.4 ± 9.5c
Testosterone	33.3 ± 2.1a	6.6 ± 0.6b	11.1 ± 1.5b

Two averages, followed by the same small letter are not different from each other (P > 0.01) Tukey's test.

The average serum testosterone levels were higher in the control group than those found in the other experimental groups. Moreover, the diabetic and diabetic-insulin groups did not show any significant differences in relation to each other (Table 1).

Light microscopy: immunolabelled α-DG and β-DG and morphometric analyses

The control group

The ventral lobe of the prostate showed folded acini mucosa (Figure 1a,b). The secretory epithelium presented tall columnar cells with basal nuclei (Figure 1a,b). The prostatic stroma showed thin collagen fibres underlying the secretory epithelium and intermingled with smooth muscle cells (Figure 1a,b). The epithelial area was greater than that of the stromal, representing 39.6% and 18.0% respectively (Graph 1).

Figure 1.

Photomicrographs of the secretory epithelium of the ventral lobe of the prostate from the (a,b) control group; (c,d) diabetic group; (e,f) diabetic-insulin group. (a) Folded prostatic acini. Secretory epithelium (EP) with columnar and basal cells. Lumen (L). Stroma (St) collagen fibres underlying the epithelium. Hematoxylin–eosin. (b) Collagen fibres (col) underlying the epithelium (EP). Lumen (L). Stroma (St) with blood vessel (bv). Picrossirius red. (c) Acini showing poorly folded mucosa and intra-luminal secretion (L). Atrophied secretory cells (EP). Hypertrophied stroma (St). Hematoxylin–eosin. (d) Increased collagen fibres (col), distributed in all the stromal area (St). Epithelium (EP). Lumen (L). Picrossirius red. (e) Secretory epithelial cells (EP) with basal cells intermingled with cuboidal cell. Lumen (L). Collagen fibres (col) in the stroma (St). Inflammatory cells (arrow) in the fibrillar elements. Hematoxylin–eosin. (f) Increased collagen fibres (col), distributed in the stroma area (St). Glandular epithelium (EP) and intraluminal secretion (L). Picrossirius red.

Graph 1.

Mean and standard deviation of the areas (%) of the epithelium, lumen and stroma of the ventral lobe of the prostate in the experimental groups.

The secretory epithelium and smooth muscle cells showing intense α-DG immunolocalization representing 37.6% of the total area (Figure 2a and Graph 2). The β-DG immunoreactivity was observed in 43.9% of the total measured area found in the periacinal prostatic stroma (Figure 2b and Graph 3).

Figure 2.

(a) Ventral lobe of the prostate. α-DG immunolocalization from the control group: strong α-DG immunoreactivity (arrows) in the secretory epithelium (EP) and smooth muscle cells. Stroma (St). Lumen (L). (b) Ventral lobe of the prostate. β-DG immunolocalization from the control group: strong β-DG immunoreactivity (arrows) in the periacinal prostatic stroma. Epithelium (EP). Stroma (St). Lumen (L). (c) Ventral lobe of the prostate. α-DG immunolocalization from the diabetic group: weak α-DG immunoreactivity (arrows) in the secretory epithelium (EP) and smooth muscle cells. Stroma (St). Lumen (L). (d) Ventral lobe of the prostate. β-DG immunolocalization from the diabetic group: weak β-DG immunoreactivity (arrows) in the periacinal prostatic stroma. Secretory epithelium (EP). Stroma (St). Lumen (L). (e) Ventral lobe of the prostate. α-DG immunolocalization from the diabetic-insulin group: Moderate α-DG immunoreactivity (arrows) in the secretory epithelium (EP) and smooth muscle cells. Stroma (St). Lumen (L). (f) Ventral lobe of the prostate. β-DG immunolocalization from the diabetic-insulin group: Moderate β-DG immunoreactivity (arrows) in the periacinal prostatic stroma. Secretory epithelium (EP). Stroma (St). Lumen (L).

Graph 2.

Percentage of immunolabelled α-dystroglycan.

Graph 3.

Percentage of immunolabelled β-dystroglycan.

The diabetic group

The prostate acini presented less folded mucosa than observed in the control group (Figure 1c,d). Atrophied secretory epithelium with cuboidal cells was verified (Figure 1c,d). Also, hypertrophied stroma was found, representing an area of 36.6% in contrast to 19.3% of the epithelial area (Figure 1c,d and Graph 1).

The secretory epithelium and smooth muscle cells showed weak α-DG immunolocalization, representing 7.8% (Figure 2c and Graph 2). The β-DG immunoreactivity was observed in 13.6% of the total measured area found in the periacinal prostatic stroma (Figure 2d and Graph 3).

The diabetic-insulin group

The prostatic acini demonstrated less folded mucosa than observed in the control group; however, this folding was intensified in relation to that verified in the diabetic group (Figure 1e,f). Atrophied secretory epithelium with cuboidal cells and basal nuclei was seen (Figure 1e,f). Hypertrophied stroma was verified in relation to the control group; however, there was a lower occurrence of collagen fibres than verified in the diabetic group (Figure 1e,f). Some inflammatory cell foci were found (Figure 1e). The glandular epithelial area was approximately 1.5 times smaller than the stromal area (Graph 1). The secretory epithelium and smooth muscle cells showed a moderate α-DG immunolocalization, containing 19.9% of the immunoreactivity (Figure 2e and Graph 2). Also, moderate β-DG immunoreactivity was observed, occupying 29.5% of the total measured area in the periacinal prostatic stroma (Figure 2f and Graph 3).

Discussion

This work showed high serum glucose levels in diabetic mice. Nevertheless, these values were close to the normal parameter in the diabetic animals submitted to insulin treatment. Early studies have shown that hyperglycaemia in diabetic rodents is a determining characteristic in both diabetic rodents by chemical induction and spontaneously developed (Makino et al. 1980; Tesone et al. 1980; Ho 1991; Saito et al. 1996; Ader et al. 1998; Ribeiro et al. 2006; Caldeira & Cagnon 2007; Ohta et al. 2007). Also, another study verified that insulin administration in diabetic rats by chemical induction led to unchanged blood glucose levels with values similar to those of healthy rats (Ohta et al. 2007). Thus, in this study, the glycaemic levels showed that there was an effective diabetic state in the analysed mice as well as confirmed insulin action on glycaemic control.

Another result of this work is that the prostate exhibited significant molecular and structural changes in diabetic animals. The diabetic mice presented atrophied epithelium, hypertrophied stroma, low α- and β-DG receptor localization and low serum testosterone level. In contrast, these structural alterations, α- and β-DG receptor immunolocalization and testosterone levels were recovered in the diabetic animals which received insulin treatment.

Clinical and experimental studies demonstrated that diabetes led to reduced gonadotrophic hormones by means of hypothalamic–hypophyseal–gonodal axis imbalance which causes physiopathological changes in male reproductive organs, including the prostate (Tesone et al. 1980; Ho 1991; Saito et al. 1996; Ballester et al. 2004). Other studies observed low serum testosterone levels with diminished androgen receptor in the prostate of diabetic animals as a result of hormone metabolic imbalance, provoked by a negative feed-back of the hypothalamic–hypophyseal–gonadal axis (Oksanen 1975; Tesone et al. 1976; Daubrese et al. 1978). Nevertheless, Jackson and Hutson (1984) emphasized that the changes in the accessory sex glands of diabetic rodents occurred as a result of two aspects; decreased testosterone levels and lack of insulin which could alter the cellular mechanism, damaging the normal androgen action.

In addition, Cagnon et al. (2000) and Ribeiro et al. (2006) showed atrophied epithelium, hypertrophied stroma, inflammatory cells and occurrence of prostatic intraepithelial neoplasia in the prostate ventral lobe of Nod mice. In another experimental study, Carvalho et al. (2003) detected coagulating gland alterations in diabetic mice such as inflammatory cells and atrophied acini. In contrast, clinical studies by means of epidemiological evidence suggested that diabetes mellitus is associated with the decrease in prostate cancer risk. However, the same authors believe that these findings are based on a small number of observations and require further investigation (Kasper et al. 2008; Pierce et al. 2008). According to Kasper et al. (2008), the IGF-1R levels appear to be decreased in diabetic patients when compared with non-diabetic ones. These authors thought that the possible reduction of prostate cancer in diabetic patient could be related to the hypothesis of varying hormonal profiles. Nevertheless, other epidemiological data showed that men who have been diabetic for five years or more have more chances of developing prostate cancer than healthy men (Will et al. 1999).

In another way, Soudamani et al. (2005) verified that type I diabetes mellitus compromised the differentiation development of the ventral prostate during sexual maturation and exogenous insulin minimized the harmful diabetic effects. Wang et al. (2000) verified that the prostate weight of diabetic rats who received insulin replacement did not recover. However, the simultaneous administration of insulin and testosterone was much more efficient in the restoration of morphological and functional prostate characteristics (Sufrin & Scott 1972; Tesone et al. 1980; Ho 1991). According to Suthagar et al. (2008), diabetes mellitus altered the biochemical parameters of the prostate as well as androgen and oestrogen expression receptors. In addition, the insulin replacement partially or completely minimized these changes, pointing to insulin as being essential for maintaining prostate functional integrity.

Regarding physiological and molecular aspects of the prostate, this gland is an androgen-mediated organ, which has been widely studied for its fundamental role in the male reproductive system as well as in the high occurrence of pathologies (Leav et al. 2001; Cunha et al. 2002). The cellular and molecular prostate complexity, especially in relation to paracrine events signalling the stroma–epithelium interaction, has pointed to different molecules, which are involved in the various prostate biological processes (Marker et al. 2003).

Dystroglycan is a non-integrin adhesion molecule expressed by a variety of tissues interacting with extracellular proteins including laminin, perlecan, agrin (Winder 2001). The DG biosynthesis is complex, which is the product of a single gene and the primary peptide is post-translationally cleaved, resulting in two protein subunits α- and β-DG (Winder 2001). DG was initially studied in skeletal muscles and its role has been limited to muscle physiopathology for a long time (Brennan et al. 2004). Nevertheless, currently, the adhesion protein role such as DG has shown multiple functions, which are involved in connecting cells to the basal lamina and in the exchange of information with the extracellular environment (Hood & Cheresh 2002; Lyons & Jones 2007). In addition, it is well known that the precise contact between epithelial cells and their underlying basement membrane is crucial to the maintenance of tissue architecture and function (Weir et al. 2006). Thus, various studies verified that different types of cancers such as breast, colon and prostate demonstrated heterogeneity of the β-DG expression and a low or absent α-DG one when compared to normal epithelial tissue (Sgambato et al. 2003). Moreover, reduced α-DG was associated with tumour progression (Henry et al. 2001). According to Sgambato et al. (2007), the DG overexpression inhibits the growth and tumourigenesis of these cells. Also, the same authors analysed the DG in patients with a diagnosis of prostate cancer who were submitted to radical prostectomy and anti-androgen treatment and who demonstrated a dose- and time-dependent decrease in the DG expression. Nevertheless, the molecular relationship between androgen and DG is not clear. Raz (2004) showed that reduced DG expression caused loss of function leading to aberrant cell–extracellular matrix interaction, resulting in an increase in invasion properties. This author suggested the DG as being a positive element towards tumourigenesis in the epithelial cells. Also, other results suggested that lack of DG could have early effects on the carcinogenesis events, much more than neoplasia transformation (Sgambato et al. 2003; Sgambato & Brancaccio 2005).

Regarding the animal model, it would be appropriate to state that the Nod mouse is a polygenic model of human insulin-dependent diabetes which presents an autoimmune genetic factor (Makino et al. 1980; Martin et al. 1997). In addition, different authors have shown that the autoimmune factor is not responsible for triggering glandular morphological changes because these animals develop moderate autoimmune alterations (Humprhreys-Beher et al. 1998; Yamano et al. 1999). According to Hu et al. (1992), the autoimmune effects could intensify the changes caused by the metabolic disorder provoked by type I diabetes.

Finally, it can be concluded that diabetes disturbed the prostate structure integrity and the altered occurrence of α- and β-DG receptors probably decreased the cell-matrix extracellular and cell-basal membrane attachment. However, insulin treatment was seen to restore glandular homeostasis partially. Also, it is possible to suggest that DG could be regulated by testosterone. However, the relationship between DG and testosterone is still not clear and new studies in this field will be necessary. Moreover, DG analysis is another important step towards knowledge of the prostate complexity and could certainly be a useful molecule as a marker in prostate diseases.