Modified cytoskeletal and extracellular matrix composition in the canine testis after long‐acting gonadotropin‐releasing hormone agonist deslorelin treatment

Abstract

Background

Long‐acting gonadotropin‐releasing hormone (GnRH) agonists are commonly used for fertility control in male dogs. Their effect on testicular histomorphology has not been clarified.

Objectives

To elucidate the mechanisms underlying androgen withdrawal in response to long‐acting GnRH agonist deslorelin 4.7 mg treatment on testicular histomorphology and key extracellular matrix (ECM) and cytoskeletal components.

Materials and methods

Testes of deslorelin‐treated dogs (n = 5) were evaluated at maximum downregulation of germinative and steroidogenic function, and compared with untreated (control) adult (n = 7) and prepubertal immature (n = 8) dogs. Morphological analysis was performed with Crossman's triple staining and periodic‐acid‐Schiff (PAS). Localization of smooth muscle α‐actin (αSMA/ACTA2), desmin (DES), elastin (ELN), collagen type I (COL1A1), collagen type IV (COL4A1) was detected by immunohistochemistry.

Results

Deslorelin treatment resulted in complete absence of germ cells except for spermatogonia. Crossman's triple staining identified thickening of peritubular connective tissue characterized by small infoldings toward the lumen of atrophied seminiferous tubules, and the expansion of interstitial area. PAS staining revealed a thickened basement membrane and phagosome‐like particles within the seminiferous tubules. In control dogs, αSMA/ACTA2, DES, ELN, COL1A1, and COL4A1 were localized to blood vessels and the peritubular wall of seminiferous tubules. ELN, COL1A1, and COL4A1 were also found in stromal compartments. A similar localization pattern with signals within a thickened peritubular area was found in deslorelin‐treated dogs. Prepubertal dogs had similar distribution but weaker signal intensity in the peritubular wall of seminiferous tubules. Positive area percentages of αSMA/ACTA2, DES, ELN, and COL4A1 were the highest in deslorelin‐treated dogs, while COL1A1 was the lowest in control, intermediate in deslorelin‐treated, and highest in prepubertal dogs.

Conclusion

Deslorelin treatment caused infertility with accumulation of collagen (COL4A1, COL1A1) and deposition of αSMA/ACTA2, DES, and ELN within the thickened lamina propria surrounding seminiferous tubules. Peritubular myocytes maintained their smooth muscle phenotype despite androgen withdrawal. The immature testis had lower abundance of most cytoskeletal and ECM proteins.

1. INTRODUCTION

Ovario(hyster)ectomy (spaying) and orchidectomy (castration/neuter) have been traditionally used to control fertility in domestic female and male cats and dogs. However, not all owners prefer surgical sterilization for their pets as a permanent solution. Anesthesia and surgery may also be contraindicated in animals with certain medical conditions. Therefore, safe, reliable, and reversible pharmacological neutering solutions are becoming increasingly popular alternatives to surgical castration in owned male dogs. For this purpose, long‐acting slow‐release gonadotropin‐releasing hormone (GnRH) agonist deslorelin implants (Suprelorin 4.7 mg or 9.4 mg, Virbac, France) have been approved for fertility control for male dogs and cats in several countries. Deslorelin shows a high binding affinity to GnRH receptors and initially induces secretion of follicle‐stimulating hormone (FSH) and luteinizing hormone (LH) from the anterior pituitary gland, which leads to increased release of testosterone from the testes (“flare‐up effect”). ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ After this initial “flare‐up”, canine pituitary gonadotrophs become desensitized to the effects of GnRH demonstrated by a lack of LH response to a GnRH challenge. ⁶ The exact mechanism of pituitary desensitization has not been elucidated in the dog, but is likely to occur via inhibition of the expression of the β‐subunits of gonadotropins, similarly to deslorelin‐implanted rats which show reduced pituitary FSH β‐subunit expression, ⁹ and perhaps also through altered GnRH receptor expression. ¹⁰ Pituitary desensitization is followed by reduced testicular LH sensitivity, which is mediated by decreased LH receptor expression. ¹ , ⁶ In response to decreased circulating gonadotropin levels, steroid hormone synthesis will be disrupted, and spermatogenesis will cease. Aspermia and complete sterility occur within 2–12 weeks due to atrophy of the seminiferous tubules and spermatogenic arrest. ¹ , ² , ³ , ⁴ , ⁵ , ¹¹ , ¹² At the molecular level, a decrease in steroidogenic acute regulatory (STAR) protein and some of the steroidogenic enzymes in Leydig cells was observed, ¹³ explaining the low testosterone concentrations. The significant alterations in Leydig and Sertoli cell function are further substantiated by increased expression of anti‐Müllerian hormone (AMH) and its type 2 receptor (AMH2), and decreased expression of insulin‐like 3 peptide (INSL3) and its receptor (relaxin receptor type 2, RXFP2) at the time of maximum downregulation. ¹ Furthermore, treatment with the deslorelin implant also affects the antioxidant enzyme system of the testis, including superoxide dismutase 1 (SOD1) and glutathione peroxidase 1 (GPx1) expression. ¹⁴ These molecular changes are further supported by histological and ultrastructural changes, showing atrophic seminiferous tubules, as well as smaller or differentially positioned nuclei, atrophied nucleoli, lipid droplet and glycogen accumulation, and an increase in phagosome numbers in Leydig or Sertoli cells. ¹³ , ¹⁵ , ¹⁶ , ¹⁷ However, details on the histological changes within the seminiferous tubules and interstitium, especially with regards to the extracellular matrix (ECM) as they accompany the deslorelin‐induced endocrine and functional alterations of the gonad, are lacking. The ECM, which is part of the seminiferous tubules’ basal membrane and lamina propria, plays an important role in supporting Sertoli and germ cell function in the seminiferous epithelium, including the blood–testis barrier, as well as testicular development, and spermatogenesis. Collagens and laminins of the ECM, functioning in concert with proteases, protease inhibitors, cytokines (e.g., TNFα) and focal adhesion components, were shown to participate in regulating blood–testis barrier dynamics. ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² The main components of the basement membrane are type IV collagen (COL4/Col IV), laminin, heparan sulfate proteoglycans, and entactin, and this matrix, that is, the basement membrane, is closely surrounded by type I collagen (COL1) fibrils (reviewed by ²³ ). Both peritubular myoid cells and Sertoli cells contribute to the secretion of ECM proteins; peritubular myoid cells produce fibronectin, COL1 and COL4/Col IV, and proteoglycans, while Sertoli cells produce laminin, COL4/Col IV and proteoglycans. ²⁴ , ²⁵ Coating cell culture plates with COL4/Col IV, laminin, heparan sulfate proteoglycan, and entactin‐containing ECM induced proliferation and differentiation of bovine spermatogonial stem cells. ²⁶ On the other hand, coating culture plates with fibronectin or COL4/Col IV decreased the secretion of testosterone in rat Leydig cells in vitro. ²⁷ ECM expression is also modulated by infertility, cryptorchidism and developmental changes of the testis. In infertile men with Sertoli cell‐only syndrome (germ cell aplasia), COL1 was overexpressed in the thickened basement membrane of the seminiferous tubules, while COL4/Col IV was down‐regulated. ²⁸ In boys and adult men, decreased COL4/Col IV was detected in the lamina propria surrounding the seminiferous tubules while completely disappearing from around the Leydig cells of the cryptorchid compared with the normal testis. ²⁹ In prepubertal mice, high expression of procollagen I, collagen type I alpha 1 chain (COL1A1), and collagen type I alpha 2 chain (COL1A2) was detected in the seminiferous tubules when compared with adult mice. ³⁰ Furthermore, proteomic and secretomic analyses of the aging common marmoset testis suggest impairments in ECM remodeling and protein secretion, along with reduced antioxidant capacity to manage reactive oxygen species. ³¹

Peritubular myoid cells surround the seminiferous tubules in close proximity of Sertoli cells, and express both smooth muscle α‐actin (αSMA/ACTA2) and desmin (DES) among other markers. ²⁵ , ³² The expression of αSMA/ACTA2 as well as elastin (ELN) secretion seem to be regulated through androgens and increase during puberty, ³² , ³³ , ³⁴ marking the adult phenotype of myoid cells. In line with this, no elastic fibers are seen in the peritubular wall of seminiferous tubules in men with Klinefelter syndrome or hypogonadotropic hypogonadism. ³⁵ Peritubular myoid cells are organized in lamellar monolayers (the number depends on the species) alternating with ECM layers, forming the peritubular compartment of the seminiferous tubules. ²⁵ , ³⁶ They have smooth muscle cell characteristics and are responsible for the contraction of the seminiferous tubules and transport of spermatozoa, while also giving support to the structural integrity of the germinal epithelium. ³⁷ , ³⁸ Peritubular myoid cells also support spermatogonial stem cells ³⁷ and regulate Sertoli cell function and spermatogenesis. ³⁹ Their decreased numbers in sub‐ or infertile humans indicate their involvement in regulating fertility. ³⁸ A recent study using single‐cell transcriptome analyses showed that in men with Sertoli cell‐only syndrome, which is the most severe form of male infertility, the top upregulated pathways in myoid, Leydig and stromal cells (pericytes or vascular smooth muscle cells) were related to ECM deposition and organization, ²⁸ pinpointing the crucial role of the modulation of testicular ECM in infertility.

Despite the well‐known role of the ECM in testicular physiology, little is known about how pharmacological suppression of the hypothalamic–pituitary–gonadal axis, particularly via GnRH agonist therapy, affects these components. Most studies that evaluated the effects of GnRH suppression have primarily focused on histomorphological alterations, such as changes in testicular volume, germinal epithelium height, or Sertoli and Leydig cell numbers in different mammalian species. ¹³ , ¹⁵ , ¹⁶ , ¹⁷ , ⁴⁰ , ⁴¹ , ⁴²

However, it is unclear how the long‐acting GnRH agonist deslorelin implant affects the testicular architecture in dogs, more specifically, the cytoskeletal components and ECM, which are critically important for spermatogenesis and normal endocrine function of the gonad. Therefore, investigating ECM responses in deslorelin‐treated male dogs offers a valuable opportunity to explore mechanisms underlying fertility suppression and its potential reversibility. Here, we hypothesized that treatment with a long‐acting GnRH agonist deslorelin implant leads to specific alterations in the expression and organization of key ECM and cytoskeletal (αSMA/ACTA2, DES) components of the testis of adult male dogs. Elucidating these changes will provide novel insights into testicular remodeling during pharmacological downregulation, and improve our knowledge of the structural factors in relation to infertility and reproductive recovery potential. The goal of this study was to elucidate the histomorphological changes and the expression and cellular localization of cytoskeletal (αSMA/ACTA2, DES) and ECM components (ELN, COL1A1, and collagen type IV alpha 1 chain (COL4A1)) in the testis of healthy adult dogs treated with the 4.7 mg deslorelin implant (Suprelorin), and compare them to untreated adult and prepubertal immature male dogs.

2. MATERIALS AND METHODS

2.1. Sample collection

Testes samples were collected following routine castration and were divided into the following groups: adult untreated control male dogs (n = 7, 1.25–4.25 years old, 12.4–44 kg, mixed breeds), prepubertal untreated male dogs (n = 8, 2–2.5 months old, 4.1–6.3 kg, mixed breed), and deslorelin‐treated dogs (n = 5, 2.5–3.75 years old, 10.9–13.9 kg, Beagles). Deslorelin‐treated dogs were implanted subcutaneously (Day 0) with a 4.7 mg deslorelin implant (Suprelorin, Virbac, France), and castrations were performed 16 weeks after implant insertion, at the time of maximum testicular downregulation. Prepubertal dogs were compared to deslorelin‐treated adult dogs to assess testicular alterations associated with immaturity versus acquired/induced infertility by manipulation of the hypothalamo–pituitary–testis axis in fully mature dogs, respectively. All animals were found healthy on clinical examination without the presence of reproductive diseases and disorders affecting the internal and external reproductive organs, and had no history of such conditions and diseases. All experimental procedures were performed in accordance with national and institutional guidelines and animal welfare legislation. Approval of procedures was granted by the Institutional Animal Care and Use Committee (permitted by the Food Chain Safety and Animal Health Directorate of the Government Office for Pest County, Hungary, permit number PEI/001/4557–4/2014).

For histology and immunohistochemistry (IHC) analyses, surrounding tissues were carefully removed from the testes immediately after surgery, and samples were fixed in 10% neutral phosphate‐buffered formalin for 24 h at 4°C. Afterward, phosphate‐buffered saline was used to wash tissue samples daily for 1 week. Graded ethanol series were used to dehydrate the tissue, after which they were embedded in paraffin.

2.2. Histological analysis

Morphological analysis of testis samples was performed with Crossman's triple staining ⁴³ and periodic‐acid‐Schiff (PAS) methods as described previously. ⁴³ , ⁴⁴ Briefly, samples were cut with a rotary microtome at 5–6 µm thickness, mounted on Epredia SuperFrost Plus Adhesion slides (Epredia, Microm International GmbH, Germany), and deparaffinized in xylene and rehydrated. The slides were then stained either with Crossman's triple staining method ⁴³ or PAS method ⁴⁴ for a detailed structural analysis.

For each tissue section, 10 images from the same histological section were systematically taken from nonoverlapping, randomly distributed fields across the entire section to ensure spatial representation using an Olympus BX51 microscope equipped with an Olympus DP72 camera (Olympus Europa SE & Co. Hamburg, Germany). Images were analyzed using Image J software ⁴⁵ to measure the diameter of the tubuli seminiferi contorti and to calculate the proportion (%) of the area occupied by the tubuli seminiferi contorti and intertubular area.

2.3. Immunohistochemistry

Localization of αSMA/ACTA2, DES, ELN, COL1A1, and COL4A1 was detected in testis samples following our previously published IHC method. ¹⁴ Briefly, paraffin‐embedded samples were cut at 3–4 µm thickness and mounted on Epredia SuperFrost Plus Adhesion slides (Epredia, Microm International GmbH, Germany). Deparaffinization was achieved by xylene, and then samples were rehydrated using a graded ethanol series. 10 mM citrate buffer (pH 6.0) at 100°C for 3×5 min was used for epitope/antigen retrieval. Endogenous peroxidase activity was blocked by immersing slides in 0.3% hydrogen peroxide in methanol. Primary antibody‐associated nonspecific bindings were prevented by incubating slides with 10% goat serum for 20 min. Then, slides were incubated with the primary antibodies (Table 1) overnight at 4°C. Non‐immunized IgGs from the same species and at the same protein concentration as the primary antibody‐incubated slides were used as isotype controls. Slides incubated without the primary antibodies served as negative control (not shown). After washing with PBS (0.8 mM Na₂HPO₄, 1.47 mM KH₂PO₄, 2.68 mM KCl, 137 mM NaCl; pH 7.2–7.4), the secondary antibodies (Table 1) were applied onto the slides. Ready‐to‐use streptavidin peroxidase (Thermo Fisher Scientific, TS‐125‐HR) was used to enhance signal intensity for 20 min at room temperature. A liquid DAB+ substrate system (Agilent Dako, Santa Clara, CA, USA) was used for the visualization of immunohistochemical reactions and hematoxylin for the counterstaining of the slides. After that, slides were dehydrated in graded ethanol series followed by xylene, and were mounted with Entellan (Entellan new, Merck, 1079610500, Darmstadt, Germany).

TABLE 1.

List of primary and secondary antibodies used for indirect immunohistochemistry.

Antibody	Company	Reference number	Species /type	Dilution IHC
Actin, Smooth Muscle Ab‐1 (Clone 1A4)	Epredia, Microm International GmbH, Germany	MS‐113‐P1	Mouse monoclonal	1:200
Desmin (Muscle Cell Marker) Ab‐1 (Clone D33)	Epredia, Microm International GmbH, Germany	MS‐376‐S0	Mouse monoclonal	1:200
Elastin (BA‐4)	Santa Cruz Biotechnology Inc., CA, USA	sc‐58756	Mouse monoclonal	1:200
Anti‐Collagen I (COL 1)	Abcam, Cambridge, MA, USA	ab6308	Mouse monoclonal	1:200
Anti‐Collagen IV	Abcam, Cambridge, MA, USA	ab6586	Rabbit polyclonal	1:200
Biotinylated goat anti‐mouse IgG (H+L)	Vector Laboratories Inc., Burlingame, CA, USA	BA‐9200	Goat anti‐mouse IgG	1:100
Biotinylated goat anti‐rabbit IgG (H+L)	Vector Laboratories Inc., Burlingame, CA, USA	BA‐1000	Goat anti‐rabbit IgG	1:100

For each tissue section, 10 images from the same histological section were systematically taken from nonoverlapping, randomly distributed fields across the entire section to ensure spatial representation using an Olympus BX51 microscope equipped with an Olympus DP72 camera (Olympus Europa SE & Co. Hamburg, Germany). Images were analyzed using Image J software ⁴⁵ to determine the proportion (%) of the area occupied by positive staining of αSMA/ACTA2, DES, ELN, COL1A1, and COL4A1 out of the total tissue area.

2.4. Statistical analyses

Since the data obtained in this study showed the normal distribution based on the Shapiro–Wilk test, parametric one‐way analysis of variance (ANOVA) was used for global comparison across the experimental groups to evaluate the diameter of the tubuli seminiferi contorti, the proportion (%) of the area occupied by the tubuli seminiferi contorti and the intertubular area. The proportion (%) of the area occupied by ELN, αSMA/ACTA2, DES, COL1A1, and COL4A1 immunopositivity was also evaluated by ANOVA. SPSS version 24 (IBM, Armonk, NY, USA) was used for all statistical analyses. In the case of statistically significant differences (P < 0.05), the Tukey–Kramer multiple comparisons post‐hoc test was applied. The data are presented as the mean ± standard deviation (SD). The level of significance was set at P < 0.05.

3. RESULTS

3.1. Histomorphological evaluation of the effect of deslorelin treatment on the testis

In testicular sections of adult untreated control animals, the tunica albuginea exhibited typical morphology, that is, thick and consisting of dense, well‐organized connective tissue with a high concentration of collagen fibers as identified by Crossman's triple staining (Figure 1A). In deslorelin‐treated animals, the morphological structure of the tunica albuginea was similar to the control animals, but its thickness appeared to be increased (Figure 1D). In prepubertal dogs, the tunica albuginea exhibited similar morphology compared to both groups but seemed to have a thinner structure (Figure 1G).

FIGURE 1.

Crossman's triple staining in the testis at lower (A, D, G) and higher (B, C, E, F, H, I) magnifications. (A–C) Adult untreated control dogs; (D–F) deslorelin‐treated dogs; (G–I) prepubertal dogs. Seminiferous tubules (open arrows), peritubular interstitial tissue of seminiferous tubules (solid arrows), intertubular connective tissue (asterisks), Leydig cells (solid arrowheads).

As expected, the septula testis, formed by extensions of the tunica albuginea that divide the testis into compartments, was observed as a thin layer of connective tissue surrounding the basement membrane of the seminiferous tubules in untreated adult dogs (Figure 1A). In contrast to the control dogs, the septula testis appeared more prominent in deslorelin‐treated animals, resulting in a relative increased distance between the seminiferous tubules (Figure 1D). In prepubertal dogs, the thickness of the septula testis also appeared greater compared to adult controls (Figure 1G).

As for the seminiferous tubules, fully differentiated germ cells were observed in all tubules in the testicular sections of control adult males (Figure 1B). Moreover, in the interstitial tissue, Leydig cell clusters were visible (Figure 1C). However, regressed tubules and the absence of spermatogenesis was observed in deslorelin‐treated dogs (Figure 1E). The majority of seminiferous tubules were composed almost entirely of Sertoli cells and type A and B spermatogonia (Figure 1E). Concomitantly, a reduction in Leydig cell volume and numbers, and accumulation of lipid droplets in the cytoplasm were observed (Figure 1F). Additionally, the peritubular interstitial tissue appeared thickened, likely due to the atrophy and retraction of the seminiferous tubules, and there were small infoldings of peritubular connective tissue visible along the circumference of the seminiferous tubules (Figure 1E). This clearly indicates the severe disruption of germ cell development in these animals. In prepubertal dogs, seminiferous tubules containing prespermatogonia (gonocytes) and pre‐Sertoli cells (Figure 1H) were separated by only a thin lamina propria from the interstitial area. In the interstitial tissue, immature Leydig cells had an irregular nucleus with relatively little heterochromatin (Figure 1I).

Next, PAS staining was used to evaluate the thickness of the basement membrane of the seminiferous tubules and the acrosomal cap of spermatids and spermatozoa. In the adult control animals, the acrosomal cap of spermatids and spermatozoa, the peritubular interstitial tissue, and the basement membrane were stained positively (Figure 2A,B). However, PAS‐stained sections from deslorelin‐treated dogs revealed an apparently thickened basement membrane along with a complete absence of the acrosomal cap, as spermatids and spermatocytes had disappeared (Figure 2C,D). Accompanying the loss of germinal cells, PAS‐positive phagosome‐like particles were seen within the cytoplasm of Sertoli cells (Figure 2C,D). In prepubertal dogs, only the peritubular interstitial tissue and the thin basement membrane stained positively with PAS (Figure 2E,F).

FIGURE 2.

Periodic acid Schiff (PAS) staining of the testis at lower (A, C, E) and higher (B, D, F) magnifications. (A, B) adult untreated control dogs; (C, D) deslorelin‐treated dogs; (E, F) prepubertal dogs. Seminiferous tubules (open arrows), basement membrane of seminiferous tubules (solid arrows). Acrosomal cap of developing germ cells (thin black arrows) is shown in (A,B). Phagosome‐like particles within the seminiferous tubules (open arrowheads) are depicted in (C,D).

3.2. The effect of deslorelin treatment on testicular histomorphometry

Testicular histomorphometry was assessed by the diameter of the seminiferous tubules, the area percentage of the seminiferous tubules, the intertubular area percentage, and tubuli seminiferi contorti/intertubular area ratio (Figure 3A–D). Deslorelin treatment resulted in a pronounced decrease in the diameter of the seminiferous tubules compared with control adult dogs (P  <  0.001) (Figure 3A). Accordingly, the area percentage of the seminiferous tubules was also reduced in deslorelin‐treated dogs compared to controls (P  <  0.05) (Figure 3B). This decrease in the diameter and area percentage of the seminiferous tubules was further evidenced by the significant increase in the intertubular area percentage of deslorelin‐treated dogs (P  <  0.05) (Figure 3C). The area percentage of the seminiferous tubules/intertubular area percentage ratio was also reduced in deslorelin‐treated dogs compared with untreated adults (P  <  0.05) (Figure 3D). Juvenile dogs had significantly smaller seminiferous tubules diameter than the adult dog groups, while the percentages of the intertubular area and the seminiferous tubules area, as well as their ratio, was similar to deslorelin‐treated dogs (Figure 3A–D).

FIGURE 3.

Histomorphometry of the testis showing (A) the diameter of seminiferous tubules, (B) the area percentage of the seminiferous tubules, (C) the intertubular area percentage, and (D) the ratio of seminiferous tubules/intertubular area. Data are presented as means ± SD. Different superscript letters (a,b,c) indicate significant (P < 0.05) differences between groups.

3.3. Localization of components of the cytoskeleton and extracellular matrix in the testes

The protein expression of αSMA/ACTA2, DES, ELN, COL1A1, and COL4A1 was detectable in all testes samples (Figures 4, 5, 6, 7, 8).

FIGURE 4.

Immunohistochemical localization of smooth muscle α‐actin (αSMA/ACTA2) in the testis of adult untreated control (A, B), deslorelin‐treated (C, D), and prepubertal dogs (E, F). Tunica media of blood vessels (solid arrowheads), peritubular myoid cells (solid arrows). There is no background staining in the isotype control (inserted in E).

FIGURE 5.

Immunohistochemical localization of desmin (DES) in the testis of adult untreated control (A, B), deslorelin‐treated (C, D), and prepubertal dogs (E, F). Tunica media of blood vessels (solid arrowheads), peritubular myoid cells (solid arrows). There is no background staining in the isotype control (inserted in E).

FIGURE 6.

Immunohistochemical localization of elastin (ELN) in the testis of adult untreated control (A, B), deslorelin‐treated (C, D), and prepubertal dogs (E, F). Peritubular connective tissue of seminiferous tubules (solid arrows), intertubular connective tissue (asterisks), blood vessels (solid arrowheads). There is no background staining in the isotype control (inserted in E).

FIGURE 7.

Immunohistochemical localization of collagen type I alpha 1 (COL1A1) in the testis of adult untreated control (A, B), deslorelin‐treated (C, D), and prepubertal dogs (E, F). Peritubular connective tissue of seminiferous tubules (solid arrows), intertubular connective tissue (asterisks), Sertoli cells (thin black arrows), blood vessels (solid arrowheads). There is no background staining in the isotype control (inserted in E).

FIGURE 8.

Immunohistochemical localization of collagen type IV alpha 1 chain (COL4A1) in the testis of adult untreated control (A, B), deslorelin‐treated (C, D), and prepubertal dogs (E, F). Peritubular connective tissue of seminiferous tubules (solid arrows), intertubular connective tissue (asterisks), Sertoli cells (thin black arrows), blood vessels (solid arrowheads). There is no background staining in the isotype control (inserted in E).

The testicular expression of αSMA/ACTA2 in adult control dogs was localized to the media of blood vessels and the entire peritubular cell area surrounding the seminiferous tubules (Figure 4A,B). In deslorelin‐treated dogs, the peritubular cell area appeared to have similar intensity but broader distribution of αSMA/ACTA2 immunostaining, with scattered small infoldings into the area of the seminiferous tubules along their circumference; the media of blood vessels also stained (Figure 4C,D). In the testis of prepubertal dogs, signals were detected in the media of blood vessels and the peritubular wall of seminiferous tubules (Figure 4E,F).

DES was localized in peritubular cells and media of blood vessels in adult control dogs (Figure 5A,B). In deslorelin‐treated dogs, signals appeared to occupy a broader area with similar localization pattern to control animals (Figure 5C,D). No or negligible immunostaining was seen in the testis of prepubertal dogs (Figure 5E,F).

In untreated adult control dogs, strong positive signals for ELN were observed in the peritubular wall of seminiferous tubules, blood vessels, and the interstitial area (Figure 6A,B). A similar localization pattern of ELN in the testes of deslorelin‐treated dogs was also observed; however, signals appeared stronger in the thickened peritubular wall (Figure 6C,D). Whereas ELN was clearly detectable in the interstitial area, no or only weak signals were observed in the peritubular cells of prepubertal dogs (Figure 6E,F).

In adult control dogs, testicular COL1A1 signals were localized to the peritubular wall of the seminiferous tubules, intertubular stroma and blood vessels (Figure 7A,B). Weak or no COL1A1 expression was observed in Sertoli cells (Figure 7A,B). Following deslorelin treatment, a similar testicular cellular localization pattern was observed, however, signals were more broadly distributed in the peritubular and intertubular stroma (Figure 7C,D). In the prepubertal dogs, peritubular cells seemed to stain sporadically and very weakly, while signals in the intertubular connective tissue layer appeared still strong (Figure 7E,F).

With regards to COL4A1, its expression in the testis of adult control dogs matched the localization pattern of COL1A1, with signals present in the peritubular area, intertubular stroma, Sertoli cells, and blood vessels (Figure 8A,B). Similar testicular signals were observed in deslorelin‐treated dogs; however, the signals were overall more distinct than in adult control dogs (Figure 8C,D). This was attributed to thickened peritubular wall and expanded interstitial areas, where COL4A1 was localized. (Figure 8C,D). Prepubertal dogs had similar signal distribution but seemingly weaker intensity for COL4A1 than the adult dog groups (Figure 8E,F).

3.4. Positive area percentage of αSMA/ACTA2, DES, ELN, COL1A1, and COL4A1 in the testes

Protein expression, evaluated as the proportion (%) of the immunopositive testicular area, was compared between adult untreated control, deslorelin‐treated, and prepubertal dogs to assess deslorelin‐induced and age‐related changes (Figure 9A–E). Positive area percentage of αSMA/ACTA2 (Figure 9A), DES (Figure 9B), ELN (Figure 9C), and COL4A1 (Figure 9E) showed the highest protein abundance in deslorelin‐treated animals (P  <  0.001 for all) compared with adult control or prepubertal dogs. On the other hand, COL1A1 (Figure 9D) positive area percentage was the highest in the prepubertal group, while deslorelin‐treated dogs were intermediate, and the adult control group showed the lowest expression (P  <  0.001). αSMA/ACTA2 (Figure 9A) expression was higher in the adult control group compared with prepubertal dogs (P  <  0.001), while DES (Figure 9B), ELN (Figure 9C), and COL4A1 (Figure 9E) did not differ between adult control and prepubertal animals.

FIGURE 9.

The effect of in vivo deslorelin treatment on the proportion (%) of immunopositive testicular area for αSMA/ACTA2 (A), DES (B), ELN (C), COL1A1 (D), and COL4A1 (E). All numerical data are presented as means ± SD. Different superscript letters (a,b,c) indicate significant (P < 0.05) differences between groups.

A schematic illustration of the changes in testicular histomorphometry, and cytoskeletal protein and ECM composition in deslorelin‐treated and prepubertal dogs in comparison to the testis of healthy normal adult dogs are presented in Figure 10.

FIGURE 10.

Schematic representation of the changes in testicular histomorphometry, cytoskeletal proteins, and ECM composition, induced by the 4.7 mg deslorelin implant (Suprelorin 4.7 mg, Virbac, France) and compared to the testis of untreated adult (control) dogs. Age‐related differences are depicted by comparing the testis of prepubertal dogs with untreated adult (control) dogs. Arrows indicate the positive area percentage of particular factors and changes in the seminiferous tubules and intertubular area: high positive area percentage or increased diameter ↑; low area percentage or decreased diameter ↓; unchanged area or diameter →.

4. DISCUSSION

We found that androgen withdrawal through long‐acting GnRH agonist treatment induced infertility and modified the testicular architecture, ECM distribution and abundance in adult dogs. In contrast to previous studies which primarily focused on histomorphological alterations (e.g., changes in testicular volume and germinal epithelium height) after long‐acting GnRH agonist treatment via light microscopy using routine hematoxilin and eosin staining, ¹³ , ¹⁵ , ¹⁶ , ¹⁷ we applied special histological stains and IHC to highlight the specific structures and ECM and cytoskeletal components of the testis. In this context, Crossman's triple staining, ⁴³ one of the reference methods for visualizing connective tissue, particularly collagen and reticular fibers, allows for a better understanding of tubulo‐matrix ratio changes in the testis in the current study. Following this direction, COL1A1, and COL4A1 immunoreactivity further confirmed alterations in the tubulo‐matrix ratio in deslorelin‐treated dogs, that is, accumulation of collagen fibers within the interstitium and around the seminiferous tubules, which contrasted the relatively low abundance of collagen fibers within the testicular parenchyma and tunica albuginea of juvenile dogs. Thus, these findings pinpoint the marked differences in ECM reorganization between pharmacologically‐induced infertility in adult dogs and immaturity in prepubertal dogs.

An important and novel finding of our study was the thickening of the lamina propria and basement membrane around the seminiferous tubules in deslorelin‐treated dogs shown by Crossman's triple staining and PAS staining. This was also accompanied by increased deposition of ELN, αSMA/ACTA2, DES, COL1A1 and COL4A1 visualized by IHC. Furthermore, an interesting and previously unidentified feature was the presence of small infoldings of the lamina propria towards the seminiferous epithelium, which was clearly visible by Crossman's triple staining for collagen and αSMA/ACTA2 IHC. These infoldings likely result from architectural changes in the actin filaments’ arrangements within peritubular myoid cells and Sertoli cells similarly to cryptorchid testes, ²⁹ , ⁴⁶ , ⁴⁷ and from increased ECM synthesis or decreased glycoconjugates turnover causing thickening of the lamina propria. ⁴⁶ Clinically, the reversible nature of deslorelin‐induced gonadal inactivity in dogs is similar to the photoperiod‐induced testicular regression in seasonal breeders like the golden (Syrian) hamster, in which gonadal activity alternates with testicular regression during inactive states. In the Syrian hamster, thickening of the lamina basalis with abundant collagen fiber deposits and highly infolded, thickened, and irregularly contoured peritubular myoid cells characterize the inactive, regressed state of the testis. ⁴⁸

In deslorelin‐treated dogs, DES and αSMA/ACTA2 expression appeared strong within the thickened peritubular area, similarly to the control animals with undisturbed germinative function. This indicates that peritubular myoid cells maintain their mature smooth muscle cell phenotype despite baseline testosterone levels, in contrast to canine Leydig and Sertoli cells which show de‐differentiation‐like features under long‐acting GnRH agonist treatment. ¹ GnRH antagonists have a similar effect on peritubular myoid cells in terms of αSMA expression. ³² The stronger peritubular ELN expression in deslorelin‐treated compared to control adult dogs indicates that the continued expression of ELN in peritubular myoid cells, similarly to αSMA/ACTA2, is likely regulated by other endocrine and local factors besides androgens.

The prepubertal dogs in our study (2–2.5 months old) had positive, albeit apparently weaker αSMA/ACTA2 and ELN immunolabeling in the peritubular area. This is comparable to prepubertal bulls and humans, ³³ , ⁴⁹ with ELN synthesis shown to start at puberty in the human testis under the stimulating effects of androgens. ³³

Despite several shared histological, immunohistochemical, and physiological attributes between the inactive gonad of GnRH agonist‐treated male dogs and retained testes, that is, atrophied seminiferous tubules in spermatogenic arrest, increased AMH protein expression and secretion by Sertoli cells, thickened peritubular area, ¹ , ²⁹ , ⁴⁶ , ⁵⁰ , ⁵¹ , ⁵² the fundamental difference is the maintenance of steroidogenic potential, albeit at a lower level, in canine cryptorchid compared to normally descended scrotal testes. ⁵³ Cryptorchid dogs also frequently have Sertoli cell‐only pathology and occasional to low DES immunoreactivity in Sertoli cells, ⁵¹ which is different from deslorelin‐treated dogs where Sertoli cells are devoid of DES.

Deslorelin‐treated dogs had similar intensity of immunoreactivity signals for COL1A1 and COL4A1 in the intertubular interstitium than control adult dogs. This is different from cryptorchidism, where decreased COL4A1 expression was found in the lamina propria around the seminiferous tubules in men ²⁹ despite the similar feature of increased lamina propria thickness. In addition to the strong COL1A1 immunoreactivity in the interstitial and peritubular areas, weak expression was also detected within Sertoli cells in both healthy adult and deslorelin‐treated dogs. This finding is in line with the study of Raychoudhury et al, ⁵⁴ which showed expression and production of COL1A1 in both Sertoli and peritubular myoid cells. Co‐culture of Sertoli and myoid cells resulted in a significant increase in relative collagen synthesis compared to monocultures of each cell type, suggesting that this may be attributed to paracrine interactions rather than being solely related to ECM components. ⁵⁴ However, further studies are needed to clarify the regulatory mechanisms and functional significance of COL1A1 expression in canine Sertoli cells.

One of the long‐known functions of Sertoli cells is their phagocytic activity. ⁵⁵ Electron microscopic studies showed an increased number of phagosomes within Sertoli cells of GnRH agonist‐treated dogs, ¹⁵ and a higher volume density of lysosomes in Sertoli cells of the Syrian hamster testis during the non‐breeding season. ⁵⁶ With a simple histochemical stain, we were also able to visualize PAS‐positive, polysaccharide containing phagosome‐like particles in the Sertoli cells of deslorelin‐treated dogs, which are likely the remnants of degenerated germ cells.

5. CONCLUSION

Cytoskeletal and ECM components show specific expression patterns within the canine testis according to developmental stage (adult versus. juvenile) and gonadal inactivity induced by a long‐acting GnRH agonist implant. Deposition of collagen fibers, as well as thickening and infolding of the lamina propria around seminiferous tubules accompanies deslorelin‐induced clinical infertility, which is not a feature of the normal untreated adult or immature, prepubertal testis. Despite the marked modulation of the testicular architecture in response to deslorelin administration, a similar expression pattern of cytoskeletal and ECM proteins was observed in both adult dog groups, likely representing the post‐pubertal phenotype of the testis. Androgen withdrawal and pharmacological suppression of fertility did not affect the expression of αSMA/ACTA2 and DES in peritubular myoid cells, which may indicate their functional reversibility potential after cessation of deslorelin's effects. This is further substantiated by the histomorphological similarities of the testis between deslorelin‐treated dogs, as we have shown here, and seasonally (photoperiodically) active mammals like the Syrian hamster.

AUTHOR CONTRIBUTIONS

Aykut Gram: Study design; knowledge transfer; critical discussion of data; drafting and revision of the manuscript. Murat Abay, Duygu Yaman Gram, Narin Liman, and Linda Müller: Knowledge transfer; critical discussion of data; and revision of the manuscript. Orsolya Balogh: Knowledge transfer; critical discussion of data; drafting and revision of the manuscript. All authors read and approved the final version of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the study reported.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding authors upon reasonable request.