Exploring the impact of metformin on testicular aging in Syrian hamsters

Abstract

Biochemical and molecular mechanisms associated with testicular aging are still poorly understood. Here, using the Syrian hamster as a natural model of aging, we observed a disturbed spermatogenesis with reduction of the testicular weight and the gonadosomatic index, altered histology including tubular wall fibrosis, increased collagen deposition, and diminished steroidogenesis in testes of aged animals. These changes took place in parallel with an increase in the levels of inflammatory and oxidative stress markers and a reduction in the cell proliferative, survival, DNA repair, and autophagic capacities. Metformin, beyond its current clinical applications, has been proposed as an anti-aging drug. In vitro incubations of testicular fragments from old hamsters with metformin revealed beneficial effects of this drug on the testicular inflammatory-oxidative state together with stimulation of autophagy and cell ability to fix DNA damage. However, in vivo daily oral administration of metformin to aged hamsters for 2 months in a dose equivalent to that usually received by patients with type 2 diabetes mellitus, reduced body and testicular weights, gonadosomatic index and blood glucose, concomitantly with increased levels of indicators of testicular inflammation, oxidation and fibrosis, decreased signs of autophagy, steroidogenesis and DNA repair capacity, and impaired spermatogenesis. Overall, while in vitro studies suggested beneficial effects of metformin in the aging testis evidenced through anti-inflammatory, anti-oxidant, and pro-autophagic actions, in vivo experiments in aged hamsters supplemented with metformin exhibited completely opposite effects. Therefore, the future of metformin as a testicular anti-aging agent should be further investigated, thoroughly reconsidered and, should the need arise, disregarded.

Supplementary information

The online version contains supplementary material available at 10.1007/s11357-025-01871-8.

Introduction

Literature indicates that aging affects the testes decreasing their volume and weight, reducing the diameter of the seminiferous tubules, thickening the basement membrane, triggering tubular fibrosis and sclerosis, impairing Leydig cell function and steroidogenesis, decreasing the number of Sertoli cells and germ cells, arresting spermatocytes, releasing spermatids prematurely, and increasing the number of germ cells showing cytoplasmic vacuolization and multinucleation [1–3].

Local inflammation, oxidative stress and apoptosis have been seen to be strongly associated with testicular aging [1–3]. These age-dependent morphological, histological, hormonal and functional changes that take place in the testes have been proposed to impact on male fertility reducing sperm production and quality [1–3].

Previously, our group described the development of local inflammatory processes in the testis of aged Syrian hamsters including upregulation of cyclooxygenase 2 (COX2) expression in parallel with a significant rise in the production of PGD2, an increment in the number of interstitial macrophages and induction of the nucleotide-binding oligomerization domain, leucine rich repeat and pyrin domain containing 3 (NLRP3) inflammasome [4]. The NLRP3 inflammasome is a multiprotein complex activated by cellular infection or stress that mediates caspase-1 activation leading to secretion of certain proinflammatory cytokines such as interleukin 1β (IL1β) and interleukin 18 (IL18), and an inflammatory form of cell death known as pyroptosis [5]. The binding protein called as never in mitosis gene A (NIMA)-related kinase 7 (NEK7) forms high-molecular-weight complexes with NLRP3 that are essential for oligomerization and activation of the inflammasome [6].

An imbalance between oxidant and antioxidant compounds in favor of oxidants is referred to as oxidative stress. Although aerobic organisms have non-enzymatic (β-carotene, glutathione and vitamins A, C and E) and enzymatic (superoxide dismutase, catalase, glutathione peroxidase, thioredoxin, peroxiredoxin, glutathione transferase) antioxidant systems that are usually effective in blocking the harmful effects of reactive oxygen species (ROS), in certain pathophysiological situations these antioxidant systems fail to restore cellular homeostasis [7]. While low levels of ROS might serve as effector molecules [8], high levels of free radicals act as trigger factors of inflammatory processes as a result of activation of the NLRP3 inflammasome. Under high levels of free radicals, lipids containing carbon–carbon double bond(s), mainly polyunsaturated fatty acids, are attacked by abstraction of a hydrogen or addition of an oxygen radical generating lipid peroxyl radicals and hydroperoxides [9]. This process leads to the eventual destruction of membrane lipids, with the generation of different breakdown products such as alkanes, alcohols, ethers, aldehydes and ketones [9].

We previously reported that inflammatory events in testes of old Syrian hamsters are accompanied by an enhanced lipid peroxidation and elevated expression levels of the anti-oxidant enzyme catalase [4]. A similar induction of inflammatory-oxidant processes was initially reported by our group in testes of a mouse model with accelerated aging (growth hormone‐transgenic mice) [10].

Not only inflammation and oxidative stress, but also autophagy has been extensively linked to aging. While an enhanced autophagy seems to partially delay aging-related frailties and the natural aging process, an impaired or defective autophagy has been associated with accelerated aging [11]. Autophagy is an intracellular process involved in the degradation and recycling of long-lived proteins and functionally redundant or damaged intracellular organelles. Cell components to be degraded are engulfed in a double-membrane structure called the phagophore. Then, the edges of the phagophore elongate and close forming the autophagosome which finally fuses with the lysosomal membrane for cargo degradation [12]. P62 (also known as sequestosome 1 -SQSTM1-) is an autophagy receptor that clusters ubiquitinated proteins previous to autophagosome engulfing [13]. The fusion of the autophagosome and lysosome requires several essential molecules including the lysosome-associated membrane protein 2 (LAMP2) [14]. LAMP2 maintains lysosomal stability and is critical for lysosomal function and promotion of the autophagic flux [14].

On the other hand, excess deposition of extracellular matrix (ECM) leads to fibrosis, an event associated with inflammation. In fact, fibroinflammation is a hallmark in aging tissues [15]. Collagen is one of the main components of ECM. Particularly in the testis, our group previously reported thickening of the wall of the seminiferous tubules in parallel with a greater degree of disorganization in the structure of the collagen fibers in the tubular wall of 22-month-old hamsters compared to young adult animals [4].

Another common event in mammalian aging is deoxyribonucleic acid (DNA) double-strand breaks (DSBs). The histone variant H2AX is a key player in DNA repair. The early cellular response to DNA DSBs involves phosphorylation on Ser-139 of histone H2AX generating γH2AX which facilitates the assembly of checkpoint and DNA repair factors as well as the activation of transducer kinases (Chk1 and Chk2) associated with the p53 ⁄ p21 signal and, in humans, also to the p16 signal [16, 17].

Expression of the proliferating cell nuclear antigen (PCNA) is correlated with cell proliferation, particularly, with DNA synthesis in S phase of the cell cycle [18]. An age-dependent decline of PCNA protein expression has been associated with increased DNA damage and elevated levels of the cyclin-dependent kinase inhibitor p21 in several models of aging and senescence [19–21]. Binding of PCNA to a functional domain of p21 affects the interaction between PCNA and the DNA polymerase δ blocking the ability of PCNA to activate DNA synthesis and, consequently, deteriorating the DNA damage tolerance pathway commonly used by cells to overcome lesions encountered throughout DNA replication [22].

On the other hand, phosphatase and tensin homolog on chromosome 10 (PTEN) is an inhibitor of the PI3K/AKT pathway known as a tumor suppressor and a contributor to longevity that exerts physiological functions by inhibiting cell proliferation, inducing cell cycle arrest and promoting apoptosis [23, 24].

In the current work, we continued and expanded our earlier studies on testicular aging in the Syrian hamster. Beyond its action on the oxidative status and the activation of the NLRP3 pathway, the impact of aging on testicular histology, autophagic process, cell proliferation/survival and DNA repair ability, steroidogenic capacity, as well as indicators of sperm production and maturation were investigated.

Besides, metformin is a safe and effective anti-hyperglycemic compound postulated to attenuate the hallmarks of biological aging in different models of organisms, tissues, organs and cell lines [25]. Therefore, initial in vitro experiments followed by in vivo studies that provide valuable insights into the physiological relevance and, consequently, a clear translation value to humans, were carried out in an attempt to bring to light the potential effects of metformin on testicular function in aging.

Material and methods

Animals

Syrian (Golden) hamsters (Mesocricetus auratus) (Charles River descendants) were raised in our Laboratory Animal Care unit at the Instituto de Biología y Medicina Experimental (IBYME-CONICET), Ciudad Autónoma de Buenos Aires, Argentina. Animals were housed from birth in rooms at 23 ± 2 °C under a photoperiod of 14 h Light, 10 h darkness (Lights on 7:00–21:00 h) with free access to water and Purina formula chow. Hamsters were handled and humanely euthanized in the home cage by carbon dioxide (CO₂) inhalation or, when it was required, by decapitation according to protocols for the use of laboratory animals approved by the Institutional Animal Care and Use Committee (CICUAL, IBYME-CONICET, Ciudad Autónoma de Buenos Aires, Argentina, Protocol Number CE-024/2023), and following the Guide for Care and Use of Laboratory Animals published by the National Institutes of Health (NIH, USA).

Twelve animals were euthanized at the age of 5 months, while 16 hamsters were kept under the conditions previously described until the age of 22 months. At the experimental endpoint, body weight was determined on an analytical scale, and trunk blood was collected and processed by centrifugation to obtain the serum that was stored at −20 °C for determination of luteinizing hormone (LH), follicle stimulating hormone (FSH) and testosterone by radioimmunoassay (RIA). Right testes were weighed and processed for histological studies, while left testes were preserved at −80 °C for additional analyses (immunoblotting, reverse transcription (RT) quantitative real-time polymerase chain reaction (qPCR), Thiobarbituric Acid Reactive Substances (TBARS), and catalase activity assay).

Additional groups of 4–8 aged hamsters (22-month-old) were used for in vitro incubations of testicular fragments as described below.

Finally, 29 young hamsters (5-month-old) and 24 aged hamsters (20-month-old) were used to investigate the impact of in vivo treatment with metformin.

In vitro incubation of hamster testicular fragments

Testes from 22-month-old hamsters (n = 4–8) were cut into fragments of approximately equal size. Then, fragments were weighed, transferred to 12-well cell culture plates containing RPMI1640 medium (Sigma-Aldrich, Saint Louis, MO, USA), and incubated for 90 min at 37 °C under 95% O₂: 5% CO₂ in the presence or absence of 50 µM metformin (Cayman Chemical Company, Ann Harbor, MI, USA) as previously described [26]. Incubations in the presence of 10 µM MCC950 (Cayman Chemical Company) and 200 nM rapamycin (drugs known to inhibit inflammation and oxidative stress, and to stimulate autophagy) were used as positive controls. Metformin and MCC950 were dissolved in RPMI1640 medium. Rapamycin was dissolved in dimethyl sulfoxide (DMSO) (MP Biomedicals, Santa Ana, CA, USA). An appropriate volume of RPMI 1640 or DMSO was added to the corresponding control (Co) group.

Incubations were stopped by transferring the 12-well cell culture plates into an ice bath. Testicular fragments were preserved at −80 °C for subsequent immunoblotting, RT-qPCR, TBARS, and catalase activity assays. Media were centrifuged for 10 min at 1200 xg and the supernatants were stored at -20 °C until determination of testosterone by RIA.

In vivo metformin treatment in young and aged hamsters

Three-month-old hamsters (n = 17) and 20-month-old hamsters (n = 15) received metformin (kindly provided by Química Montpellier S.A., Laboratorios Roemmers and Fundación Alberto J. Roemmers) dissolved in drinking water, for 2 months, in a dose of 170 mg/kg body weight/day. This dose is equivalent to the daily dose usually prescribed to patients with type 2 diabetes mellitus (1500 mg orally administered in 3 daily doses of 500 mg). Dose translation from human to hamster studies was calculated using the formula based on body surface area [hamster equivalent dose (mg/kg) = human dose (mg/kg) x (human Km/hamster Km); Km = body weight (kg)/body surface area (m²)] as was previously described (see further details in [27]). Metformin administration was periodically adjusted taking into account changes in the body weight and volume of water intake. In addition, groups of 5-month-old (n = 12) and 22-month-old (n = 9) hamsters continually received only water, without any supplementation.

At the time of euthanasia, body and testicular weights were determined, and trunk blood was collected and processed by centrifugation to obtain the serum for quantification of LH, FSH and testosterone by RIA. Right testes were weighed and processed for histological studies while left testes were also weighed and used in subsequent biochemical and molecular analyses.

In a second group of animals receiving the same treatments, testes were dissected, flash frozen in liquid nitrogen and stored at − 80 °C for further determination of daily sperm production (DSP), while both epididymides were employed for evaluation of sperm count and motility.

Finally, a third set of animals was fasted for 6 h prior to euthanasia. A drop of trunk blood was collected and used for determination of blood glucose levels using a glucometer (One Touch Ultra, Johnson & Johnson, Argentina).

Immunoblotting

Hamster testes were homogenized in RIPA buffer supplemented with 1 mM sodium orthovanadate, 1 mM sodium fluoride and a commercial mixture of protease inhibitors (Roche Applied Science GmbH, Penzberg, Germany). Homogenates were cooled in ice for 30 min, and then centrifuged at 13,000 × g for 10 min. Supernatants were used to determine protein concentrations by the method of Bradford [28], and subsequently heated at 95 °C for 5 min under reducing conditions (10% β-mercaptoethanol), loaded onto 10% tricine-SDS–polyacrylamide gels, electrophoretically separated and blotted onto nitrocellulose membranes. Blots were performed as previously described [29] using the following mouse monoclonal antibodies from Santa Cruz Biotechnology (Santa Cruz, CA, USA): anti-NLRP3 (1:500, catalog # sc-134306), anti-caspase 1 (1:500, catalog # sc-398715), anti-IL1β (1:100, catalog # sc-32294), anti-catalase (1:500, catalog # sc-271803), anti-p62 (1:500, catalog # sc-28359), anti-LAMP2 antibody (1:500, catalog # sc-18822), anti-PCNA (1:300, catalog # sc-25280), and anti-PTEN (1:500, catalog # sc-7974). Additionally, these antibodies/antisera were used: rabbit monoclonal anti-NLRP3 antibody (1:200, catalog # ab263899, Abcam Inc., Waltham, MA, USA), mouse monoclonal anti-IL1β antibody (1:200, catalog # MAB501, R&D Systems, Minneapolis, MN, USA), mouse monoclonal anti-γH2AX antibody (1:500, catalog # 05–636, Upstate®, Merck KGaA, Darmstadt, Germany), mouse monoclonal anti-actin antibody (1:5000, catalog # CP01, Calbiochem, La Jolla, CA, USA), rabbit polyclonal anti-caspase 1 antiserum (1:200, catalog # NBP1-45433, Novus Biologicals, Bio-Techne, Minneapolis, MN, USA) and rabbit polyclonal anti-steroidogenic acute regulatory (StAR) protein antiserum (1:500) kindly provided by Dr. Stocco (Texas Tech University Health Sciences Center, Lubbock, Texas, USA) (see additional information in [30]).

Subsequently, blots were incubated with the following peroxidase-conjugated secondary antisera: goat anti-mouse IgG antiserum (1:3000; catalog # 1721011, Bio-Rad Laboratories, Hercules, California, USA) for detection of NLRP3, catalase, p62, caspase 1, PCNA, PTEN (Santa Cruz Biotechnology), γH2AX (Upstate®, Merck KGaA) and IL1β (R&D Systems); goat anti-rabbit IgG antiserum (1:3000; catalog # A0545, Sigma-Aldrich) for detection of StAR; goat anti-rabbit IgG antiserum (1:3000; catalog # 1706515, Bio-Rad) for detection of caspase 1 (Novus Biologicals) and NLRP3 (Abcam Inc.); and goat anti-mouse IgM antiserum (1:2000, catalog # sc-2064, Santa Cruz Biotechnology) for detection of actin.

Bands were detected with an enhanced chemiluminescence kit (Amersham Pharmacia Biotech, Amersham, United Kindom) and visualized using a chemiluminescence imaging equipment (G-Box, Syngene™, Cambridge, UK) or, alternatively, by exposing an x-ray film (HR-U, FUJIFILM Corporation, Steinhausen, Switzerland) to the membrane in a dark room.

Some additional blots were incubated with a horse anti-mouse IgG biotin-conjugated antiserum (1:2000, catalog # BA-2000, Vector Laboratories, Burlingame, CA, USA) for detection of NLRP3, IL1β, p62 and LAMP2 (Santa Cruz Biotechnology); and a goat anti-rabbit IgG biotin-conjugated antiserum (1:2000, catalog # BA-1000, Vector Laboratories) for detection of caspase 1 (Novus Biologicals). Afterwards, an avidin biotin–peroxidase system (Vector Laboratories) was used for band visualization.

In all cases, bands were quantified by densitometry using the FIJI/ImageJ program (NIH, Bethesda, MA, USA). Actin was used as loading control for normalization of protein levels.

Following the principles of the 3Rs (Replacement, Reduction, and Refinement), our Institute’s Animal Care and Use Committee strongly recommended us to minimize the number of animals used in a manner consistent with the scientific aims of this study. Therefore, blots were cut into two or three sections. Each section was used to detect one or more target proteins within its corresponding molecular mass range, as was previously reported [31].

RT-qPCR analyses

Total RNA was prepared from hamster testes using TRIzol reagent (Life Technologies, Carlsbad, CA, USA) following the manufacturer’s instructions. RT reactions were performed using 500–1000 ng total RNA and dN6 random primers. qPCR assays were performed as previously described [29] using oligonucleotide primers for Nek7 (5′-CACAGCTGCACATTCTTTAGT and 5′-ACAGCCAAGAGACCAGATGT), StAR (5′-AACGGGGATGAGGTGCTGAG and 5′-CCACTCTCCCATTGCCTC), and 17β-hydroxysteroid dehydrogenase (17β-HSD) (5′-GCCGTCGCAGAAGAGATT and 5′-AGTGGATGAGACTCTGGA). Glyceraldehyde 3-phosphate dehydrogenase (Gapdh) was used as the housekeeping gene (5′-TGCACCACCAACTGCTTAGC and 5′-GGCATGGACTGTGGTCATGAG). Reactions were conducted under the following conditions: one cycle of 10 min at 95 °C and then 40 cycles of 30 s at 95 °C, 30 s at 55 °C and 1 min at 60 °C, using SYBR Green PCR Master Mix (Roche Applied Science GmbH) and the CFX96-Touch Real-Time PCR Detection System (Bio-Rad). As was described [32], the fold change of mRNA gene expression was determined for each sample by calculating (E target)^ΔCt(target)/(E housekeeping)^{ΔCt(housekeeping)}, where E is the efficiency of the primer set, Ct the cycle threshold, and ΔCt = Ct(normalization cDNA)-Ct(experimental cDNA). The amplification efficiency of each primer set was calculated from the slope of the standard amplification curve of log microlitres of cDNA per reaction vs. Ct value over at least four orders of magnitude (E = 10^−(1/slope)).

TBARs assay

Lipid peroxidation was determined by the colorimetric TBARs assay which mainly detects malondialdehyde (MDA) following the protocol previously described [29]. Hamster testes were homogenized in RIPA buffer. Results were expressed as nmol MDA per g protein.

Catalase activity assay

Catalase activity was determined according to the method described earlier [29]. Hamster testes were homogenized in 50 mM phosphate buffer containing 30 mM H₂O₂ at room temperature and spectrophotometrically measured at 240 nm for 3 min every 30 s. Catalase activity was expressed as μmol of H₂O₂ consumed/min/μg protein.

LH and FSH assays

Serum levels of LH and FSH were measured with a validated heterologous RIA [33] using materials and protocols supplied by the National Hormone and Pituitary Program, National Institute of Diabetes and Digestive and Kidney Diseases (Bethesda, MD, USA) and Dr A. F. Parlow (National Hormone and Pituitary Program, Torrance, CA, USA) as previously described [34].

The sensitivities of the assays were 0.15 ng/ml for LH and 1.17 ng/ml for FSH. All serum samples were measured, in duplicate, in the same assay to avoid inter-assay variance. The intra-assay coefficients of variation were 7.2% for LH and 8.0% for FSH.

Testosterone assay

Testosterone levels in sera and in incubation media were determined by RIA using a testosterone-7α-butyrate-BSA antibody (Immunotech Diagnostic, Montreal, Canada) that presents a 35% cross-reaction with dihydrotestosterone (DHT) as previously described [33]. The minimal detectable assay concentration was 0.215 pmol/ml. Intra-assay and inter-assay coefficients of variation were less than 12% and 15%, respectively. A previous extraction procedure using diethyl ether was required for quantification of serum testosterone levels [33].

Biometric measurements and histological assays

For histological examinations, hamster testes were fixed in Bouin or 10% paraformaldehyde (PFA) solutions, dehydrated, embedded in paraffin wax and sections of 5 μm were obtained from three different levels.

The number of seminiferous tubules per unit area (1 mm²), the percentage of seminiferous tubules with premature detachment of spermatocytes and spermatids, the percentage of seminiferous tubules with epithelial alterations, the percentage of seminiferous tubules with vacuolization within the cytoplasm of Sertoli cells, the percentage of seminiferous tubules at stage VII of the epithelial cycle as well as the tubular diameter were determined in testicular sections that had been deparaffinized, hydrated, stained with hematoxylin–eosin, and examined by bright-field microscopy. Seminiferous tubules at stage VII of the epithelial cycle were identified by the presence of mature elongated spermatids and residual bodies in the lumen as described [35, 36].

The thickness of the tubular wall and testicular collagen deposition were determined in specimens stained with 0.1% Sirius Red (Direct Red 80, Sigma-Aldrich) prepared in saturated picric acid solution (picrosirius red) as previously described [4, 37]. For evaluation of the features of the tubular wall, bright-field images and polarized light images of picrosirius red-stained sections were captured using an IX83-DSU microscope (Olympus Corporation, Tokyo, Japan) at a magnification of 400 × and a numerical aperture (NA) of 0.95. Stained tissues were also examined by bright-field microscopy using an XSZ-107BNT microscope (Arcano, China) at a magnification of 1600 ×. Images acquired were subsequently examined with the FIJI/ImageJ image analysis software platform for quantification of tubular wall thickness [4, 37].

In addition, scans comprised of a series of individual images (100 × magnification, NA: 0.40) were taken by bright-field microscopy across the section and then automatically stitched together to view entire testicular sections using Olympus cellSens Dimension software. Subsequently, color threshold images of picrosirius red-stained testicular sections were processed using the FIJI/ImageJ software as recently described [38]. Following the protocol published in https://imagej.net/ij/docs/examples/stained-sections/index.html, collagen deposition was quantified as the percentage of the total area stained with picrosirius red in the entire testicular section, in the parenchyma and in the capsule.

Daily sperm production assay

The number of spermatids produced in the testis per day was assessed as previously described [39, 40] with slight modifications. Testicular parenchyma was homogenized in 25 ml of 0.05% Triton X-100 buffer in 0.9% NaCl using an ULTRA-TURRAX (IKA-Labortechnik, Staufen, Germany) at 13,000 rpm for 4 min. After 30 min on ice, 100 µl aliquots of homogenization resistant (elongated) spermatids (stage XIV-XVI) were incubated with 800 µl of 0.9% NaCl and 100 µl of 0.4% Trypan blue (Sigma-Aldrich) for 5 min. Spermatids were counted using a Neubauer chamber under bright-field microscopy at 400 × magnification. DSP per testis was calculated by dividing the total number of elongated spermatids in the testis by 6.03 (which is the number of days that developing spermatids remain in stage XIV-XVI during spermatogenesis in hamsters) [41]. Finally, efficiency of sperm production was calculated by dividing DSP by testicular parenchyma weight (expressed per g) as described [42].

Sperm collection and evaluation of sperm parameters

Cauda epididymides were dissected in a droplet of 1 ml of modified Tyrode’s (m-TALP) medium [43] and sperm were allowed to swim-out for 10 min. Total motility was evaluated in fresh sperm suspensions placed on pre-warmed slides and subjectively analyzed under a bright-field microscope (400 ×). Sperm count was determined using a Neubauer chamber and a bright-field microscope (400 ×).

Statistical analyses

Initially, data were subjected to the Kolmogorov–Smirnov test for evaluation of normal distribution [44].

For two comparisons, Student’s t-test (with Welch’s correction in cases of unequal population variances) was performed when data were normally distributed, while the non-parametric Mann–Whitney test was performed when data were not normally distributed.

For multiple comparisons, one-way analysis of variance (ANOVA) followed by Student-Newman–Keuls test was performed when data were normally distributed. When data were not normally distributed, the non-parametric Dunn test was performed.

Differences were considered statistically significant when p < 0.05. Data are expressed as mean ± standard error of the mean (S.E.M.).

Results

Markers of inflammation, oxidative stress, autophagy, steroidogenesis, and proliferative/survival/DNA repair events in testes of young and aged adult hamsters

Protein expression levels of inflammatory markers linked to activation of the NLRP3 inflammasome (NLRP3, cleaved caspase 1, and IL1β) as well as mRNA expression levels of the NLRP3-binding protein Nek7 were significantly higher in testes of aged hamsters than in gonads of young animals (Fig. 1a).

Fig. 1.

Comparison of inflammatory-oxidative status, autophagy capacity, steroidogenic activity and proliferative/survival/DNA repair events in testes of young adult and aged adult Syrian hamsters. Protein expression levels of the inflammasome NLRP3, pro- and cleaved-caspase 1, and IL1β (panel a), catalase (panel b), p62 and LAMP2 (panel c), StAR (panel d), as well as PCNA, PTEN, and γH2AX (panel e) were determined by immunoblotting. Bar plot graphs represent the mean ± S.E.M. and show the quantification by densitometry of all bands obtained [n (young hamsters/aged hamsters) = 7/7 for NLRP3, caspase 1 and IL1β; n = 9/16 for catalase; n = 11/15 for p62, LAMP2, StAR, PCNA, PTEN and γH2AX]. Results are expressed as fold change relative to the group of young hamster testes and normalized to actin. Blots used in the quantitation analysis of protein bands density are shown in Fig. S1. mRNA expression levels of Nek7 (panel a), as well as StAR and 17β-HSD (panel d) were determined by RT-qPCR using Gapdh as the housekeeping gene. Bar plot graphs represent the mean ± S.E.M. of the mRNA levels obtained [n (young hamsters/aged hamsters) = 6/10]. Lipid peroxidation was determined using the TBARS assay (panel b). Bar plot graph represents the mean ± S.E.M. of MDA detected per g protein [n (young hamsters/aged hamsters) = 6/6]. Catalase activity was evaluated by a colorimetric assay (panel b). Bar plot graph represents the mean ± S.E.M. of enzyme activity expressed as μmol/min (international unit, IU) per µg protein [n (young hamsters/aged hamsters) = 12/10]. Student’s t-test (with Welch’s correction in cases of unequal population variances) was performed when data were normally distributed and the non-parametric Mann-Whitney test was used when data were not normally distributed. *p < 0.05

Furthermore, there was a statistically significant increase in TBARS production (a nonspecific marker of lipid peroxidation and oxidative stress) and in the protein expression levels of the anti-oxidant enzyme catalase in testes of 22-month-old hamsters compared to gonads of 5-month-old animals (Fig. 1b). However, testicular catalase activity did not show significant changes between young and aged hamsters (Fig. 1b). Testes of old hamsters expressed higher protein levels of p62 (a marker of decreased autophagy) and lower protein levels of the lysosome marker LAMP2 (critical for an efficient lysosomal activity and autophagic flux) than gonads of young animals (Fig. 1c). The mRNA and protein expression levels of StAR as well as mRNA expression levels of the steroidogenic enzyme 17β-HSD decreased in aged hamsters when compared to those values detected in young hamsters (Fig. 1d).

Furthermore, a significant reduction in the protein expression levels of the proliferation marker PCNA and the DNA DSBs indicator γH2AX, was observed in testes of old animals compared to gonads of young hamsters (Fig. 1e). In contrast, PTEN levels (an inhibitor of proliferation and positive regulator of longevity) showed a clear increment in testes of aged hamsters (Fig. 1e).

Effects of metformin on the levels of markers of inflammation, oxidative stress, autophagy, steroidogenesis, and proliferative/survival/DNA repair events in in vitro incubations of testicular fragments from aged adult hamsters

MCC950 and rapamycin, inhibitors of the NLRP3 pathway known to also alleviate oxidative stress and to promote autophagy [45–49] were used as positive controls in in vitro incubations of testicular fragments from aged hamsters (Fig. 2). MCC950 (10 µM) and rapamycin (200 nM) diminished protein expression levels of inflammatory markers (NLRP3, IL1β) and also decreased lipid peroxidation and catalase expression (Fig. 2a and b). While expression of the indicator of reduced autophagy p62 was diminished by MCC950 and rapamycin, LAMP2 expression did not show statistically significant changes as a result of the presence of these drugs in the incubation media (Fig. 2c). MCC950 diminished catalase activity, stimulated the release of testosterone into the culture media and induced the expression of γH2AX (Fig. 2b, d and e). In contrast, these parameters were statistically unaffected by rapamycin (Fig. 2b, d and e). The expression levels of StAR, 17β-HSD, PCNA and PTEN were not modified by MCC950 or rapamycin (Fig. 2d and e).

Fig. 2.

Impact of metformin on the inflammatory-oxidative status, autophagy capacity, steroidogenic activity and proliferative/survival/DNA repair events in in vitro incubations of testicular fragments from aged adult Syrian hamsters. Testicular fragments from aged (22-month-old) hamsters were incubated for 90 min in the presence or absence of metformin (50 µM). The following drugs, MCC950 (10 µM) and rapamycin (200 nM), were used as positive controls. Protein expression levels of the inflammasome NLRP3 and IL1β (panel a), catalase (panel b), p62 and LAMP2 (panel c), StAR (panel d), as well as PCNA, PTEN and γH2AX (panel e) were determined by immunoblotting. Bar plot graphs represent the mean ± S.E.M. and show the quantification by densitometry of all bands obtained [n (Co/MCC950/rapamycin/metformin) = 6/6/6/6 for NLRP3, IL1β, and p62; n = 5/5/6/4 for catalase; n = 8/4/4/4 for LAMP2, StAR, PCNA, PTEN and γH2AX]. Results are expressed as fold change relative to the control group (Co, untreated conditions) and normalized to actin. Blots used in the quantitation analysis of protein bands density are shown in Fig. S2. mRNA expression levels of StAR and 17β-HSD (panel d) were determined by RT-qPCR using Gapdh as the housekeeping gene. Bar plot graphs represent the mean ± S.E.M. of the mRNA levels obtained [n (Co/MCC950/rapamycin/metformin) = 6/6/6/6]. Lipid peroxidation was determined using the TBARS assay (panel b). Bar plot graph represents the mean ± S.E.M. of MDA detected per g protein [n (Co/MCC950/rapamycin/metformin) = 6/6/6/6]. Catalase activity was evaluated by a colorimetric assay (panel b). Bar plot graph represents the mean ± S.E.M. of enzyme activity expressed as μmol/min (IU) per µg protein [n (Co/MCC950/rapamycin/metformin) = 8/4/4/4]. Testosterone levels secreted to the culture media in incubations of testicular fragments of aged hamsters were determined by RIA (panel d). Bar plot graphs represent the mean ± S.E.M of testosterone levels reported as pmol per g tissue [n (Co/MCC950/rapamycin/metformin) = 6/6/6/6]. ANOVA followed by Student-Newman–Keuls test was performed when data were normally distributed. When data were not normally distributed, the non-parametric Dunn test was performed. *p < 0.05

When the effect of metformin (50 µM) on the levels of markers of inflammation was evaluated, a significantly decreased was detected in the protein expression levels of NLRP3 and IL1β in testicular fragments of aged hamsters incubated for 90 min in the presence of this antidiabetic agent (Fig. 2a). Metformin also diminished lipid peroxidation as well as protein catalase expression and activity in in vitro incubations of testicular fragments from old hamsters (Fig. 2b). Furthermore, incubations of old hamster testes with metformin stimulated autophagy which was evidenced by a reduction in the protein expression levels of the marker of decreased autophagy p62 and by an increase in the protein expression levels of the marker of lysosomal activity and autophagic flux LAMP2 (Fig. 2c).

Although a tendency to higher levels of testosterone secreted to the incubation media was seen when testicular fragments from old hamsters were incubated in the presence of metformin compared to the control group, this change was not statistically significant (Fig. 2d). In addition, metformin did not alter protein/mRNA expression levels of StAR or mRNA expression levels of 17β-HSD (Fig. 2d).

PCNA and PTEN expression levels were not affected by the addition of metformin in the incubation media (Fig. 2e). Instead, γH2AX levels were significantly higher when testicular fragments from aged hamsters were incubated in the presence of metformin (Fig. 2e).

Effects of metformin on the levels of markers of inflammation, oxidative stress, autophagy, steroidogenesis, and proliferative/survival/DNA repair events in testes of aged adult hamsters that were given a daily oral supplementation with this antidiabetic agent

No significant differences were detected in the amount of water consumed between elderly (20-month-old) hamsters receiving a daily oral dose of 170 mg/kg body weight/day metformin dissolved in drinking water (equivalent to that dose used for type 2 diabetes mellitus therapy in humans) for two months and aged animals of the control group receiving drinking water alone without any supplementation (data not shown).

Oral supplementation of old hamsters with metformin markedly increased the testicular expression levels of inflammatory markers (NLRP3, pro- and cleaved-caspase 1, IL1β and Nek7) (Fig. 3a) and oxidative state indicators (lipid peroxidation as well as catalase protein expression and activity) (Fig. 3b). When testicular expression of autophagy markers was evaluated, p62 showed higher levels while LAMP2 presented lower levels in metformin-treated hamsters than in animals receiving drinking water alone (Fig. 3c). On the other hand, oral metformin intake significantly decreased the mRNA and protein expression levels of StAR (Fig. 3d). The expression of 17β-HSD showed a statistically non-significant tendency to decrease in the metformin-treated group (Fig. 3d).

Fig. 3.

Impact of metformin daily oral supplementation on the inflammatory-oxidative status, autophagy capacity, steroidogenic activity and proliferative/survival/DNA repair events in testes of aged adult Syrian hamsters. Protein expression levels of the inflammasome NLRP3, pro- and cleaved-caspase 1, and IL1β (panel a), catalase (panel b), p62 and LAMP2 (panel c), StAR (panel d), as well as PCNA, PTEN and γH2AX (panel e) were determined by immunoblotting. Bar plot graphs represent the mean ± S.E.M. and show the quantification by densitometry of all bands obtained [n (Co/metformin) = 9/15 for NLRP3; n = 6/7 for caspase 1; n = 5/7 for IL1β; n = 9/15 for catalase, p62 and LAMP2; n = 9/14 for StAR, PCNA, PTEN and γH2AX]. Results are expressed as fold change relative to the control group (testes of aged hamsters that received drinking water alone) and normalized to actin. Blots used in the quantitation analysis of protein bands density are shown in Fig. S3. mRNA expression levels of Nek7 (panel a), StAR and 17β-HSD (panel d) were determined by RT-qPCR using Gapdh as the housekeeping gene. Bar plot graphs represent the mean ± S.E.M. of the mRNA levels obtained [n (Co/metformin) = 8/8]. Lipid peroxidation was determined using the TBARS assay (panel b). Bar plot graph represents the mean ± S.E.M. of MDA detected per g protein [n (Co/metformin) = 9/13]. Catalase activity was evaluated by a colorimetric assay (panel b). Bar plot graph represents the mean ± S.E.M. of enzyme activity expressed as μmol/min (IU) per µg protein [n (Co/metformin) = 9/13]. Student’s t-test (with Welch’s correction in cases of unequal population variances) was performed when data were normally distributed and the non-parametric Mann-Whitney test was used when data were not normally distributed. *p < 0.05

Testicular protein expression levels of PCNA and PTEN were unaffected by metformin supplementation, but γH2AX expression levels decreased in testes of aged hamsters that received metformin dissolved in drinking water (Fig. 3e).

Effects of aging and metformin daily oral supplementation on body weight, testicular weight, the gonadosomatic index and blood glucose in adult hamsters

No significant changes were seen in body weight between 5-month-old and 22-month-old hamsters. However, oral supplementation to young and aged hamsters for two months with a daily dose of 170 mg/kg/day metformin significantly diminished body weight when compared to 5-month-old and 22-month-old hamsters who were given drinking water alone without any supplementation (Table 1).

Table 1.

Body weight, testicular biometric measurements, and epididymal secretion determinations in young adult (5-month-old) and aged adult (22-month-old) Syrian hamsters that received drinking water alone, as well as in young and aged adult animals that were given metformin dissolved in drinking water (170 mg/kg/day) for two months prior to sacrifice

	Young adult hamsters	Young adult hamsters + Metformin	Aged adult hamsters	Aged adult hamsters + Metformin
Body weight (g)	149.16 ± 2.48 (12)	122.59 ± 5.82 (17) *	145.71 ± 5.81 (9)	129.08 ± 5.03 (14) *Ϯ
Testicular weight (g)	2.06 ± 0.04 (12)	1.41 ± 0.15 (17) *	1.34 ± 0.14 (9) *	0.89 ± 0.14 (14) *Ϯ
Gonadosomatic index (testicular weight as % body weight)	1.39 ± 0.03 (12)	1.12 ± 0.10 (17) *	0.91 ± 0.07 (9) *	0.66 ± 0.09 (14) *Ϯ
Fasted blood glucose (mg/dl)	129.25 ± 9.69 (4)	112.5 ± 8.5 (6)	149.33 ± 38.1 (3)	72.50 ± 5.69 (4) *Ϯ
FSH blood levels (ng/ml)	2.60 ± 0.73 (7)	4.08 ± 0.74 (7)	2.60 ± 0.52 (7)	3.10 ± 0.46 (7)
LH blood levels (ng/ml)	0.99 ± 0.29 (7)	0.33 ± 0.07 (7) *	0.27 ± 0.08 (7) *	0.13 ± 0.04 (7) *
Testosterone blood levels (pmol/ml)	16.10 ± 1.99 (7)	9.97 ± 1.24 (7) *	10.86 ± 1.75 (8) *	2.59 ± 1.08 (8) *Ϯ
Number of ST per unit area (1 mm2)	24.34 ± 0.68 (7)	22.84 ± 0.86 (6)	28.53 ± 2.11 (7) *	33.88 ± 3.01 (8) *
Diameter of ST (µm)	224.29 ± 4.36 (7)	229.8 ± 6.85 (6)	196.65 ± 11.1 (7) *	171.58 ± 12.0 (8) *
ST with epithelial alterations (%)	1.14 ± 0.67 (7)	0.50 ± 0.10 (6)	3.00 ± 1.41 (6) *	11.86 ± 7.71 (7) *
ST with premature detachment of spermatocytes and spermatids (%)	8.57 ± 1.39 (7)	16.5 ± 1.19 (6) *	18.67 ± 2.30 (6) *	40.00 ± 11.43 (7) *Ϯ
ST with epithelial vacuolization within the cytoplasm of Sertoli cells (%)	7.57 ± 2.05 (7)	4.50 ± 1.32 (6)	10.33 ± 4.67 (6)	25.71 ± 5.78 (7) *Ϯ
Number of adluminal vacuoles/ST	0.12 ± 0.03 (7)	0.12 ± 0.04 (6)	0.32 ± 0.22 (6)	0.71 ± 0.23 (7) *Ϯ
ST in stage VII of the epithelial cell cycle (%)	45.14 ± 2.56 (7)	51.75 ± 4.17 (6)	31.00 ± 7.05 (7)	26.00 ± 7.16 (7) *
Daily sperm production (DSP) per testis (× 106)	34.62 ± 2.21 (11)	33.30 ± 1.54 (6)	16.10 ± 5.86 (3) *	14.65 ± 1.34 (3) *
Sperm production efficiency (DSP per g of testicular parenchyma) (× 106)	17.32 ± 1.07 (11)	17.98 ± 1.15 (6)	12.48 ± 2.99 (3)	11.54 ± 0.82 (3) *
Epididymal sperm concentration (106/ml)	49.45 ± 9.10 (6)	38.75 ± 4.05 (6)	13.59 ± 3.78 (4) *	25.83 ± 11.38 (3)
Epididymal sperm motility (%)	74.68 ± 3.38 (6)	77.85 ± 3.01 (6)	56.13 ± 10.52 (4)	65.23 ± 9.01 (3)

ST: seminiferous tubules. The data represent the mean ± S.E.M. (n)

^*p < 0.05 vs. group of young adult (5-month-old) hamsters that received drinking water alone. Ϯ p < 0.05 vs. group of aged adult (22-month-old) hamsters that received drinking water alone

Testicular weight and the gonadosomatic index (testicular weight expressed as % of body weight) were lower in aged hamsters than in young adult animals (Table 1). Metformin treatment in both, young and elderly hamsters, further diminished both parameters (Table 1).

Blood glucose in 6-h-fasted hamsters was unaffected by age. However, blood glucose markedly diminished when aged hamsters were supplemented with metformin compared to the blood glucose levels detected in old hamsters receiving drinking water alone (Table 1). In contrast, metformin supplementation did not significantly affect blood glucose in young hamsters (Table 1).

Effects of aging and metformin daily oral supplementation on steroidogenesis and spermatogenesis in adult hamsters

In parallel to the changes detected in the expression levels of StAR and the steroidogenic enzyme 17β-HSD (Fig. 1d), serum LH and testosterone concentrations decreased in aged hamsters when compared to those values detected in young hamsters (Table 1). Following the change pattern previously described in the expression levels of StAR and 17β-HSD (Fig. 3d), serum LH levels showed a statistically non-significant tendency to decrease while testosterone blood diminished in the metformin-treated age group (Table 1). Similarly, circulating levels of LH and testosterone significantly decreased in young hamsters supplemented with metformin compared to those blood levels detected in young animals receiving drinking water alone (Table 1).

Similar to previously described findings in aged animals supplemented with metformin (Fig. 3b), impaired steroidogenesis in metformin-treated young hamsters was accompanied by elevated lipid peroxidation (nmol MDA/g protein; mean ± S.E.M., n; young adult hamsters: 210.04 ± 32.48, n = 12; young adult hamsters that were given metformin dissolved in drinking water: 596.00 ± 71.92, n = 17, p < 0.05) and catalase activity (μmol/min/μg protein; mean ± S.E.M., n; young adult hamsters: 3.45 ± 0.42, n = 12; young adult hamsters that were given metformin dissolved in drinking water: 4.81 ± 0.42, n = 17, p < 0.05).

Serum FSH levels remained statistically unchanged in the four groups analyzed (Table 1).

The number of seminiferous tubules (expressed per mm²) was significantly higher while the diameter of the seminiferous tubules was lower in 22-month-old hamsters than in 5-month-old animals (Table 1). Furthermore, an age-dependent increase in the percentage of seminiferous tubules with epithelial alterations and in the percentage of seminiferous tubules with premature detachment of spermatocytes and spermatids was seen in testes of 22-month-old hamsters compared to the gonads of 5-month-old animals (Table 1). These changes taking place in testes of 22-month-old hamsters are illustrated in Fig. 4. Instead, the among of seminiferous tubules with vacuolization within the cytoplasm of Sertoli cells (expressed as the percentage of seminiferous tubules with epithelial vacuolization and as the number of adluminal vacuoles per tubule) were unaltered by age, while the percentage of seminiferous tubules in stage VII of the epithelial cell cycle showed a statistically non-significant tendency to decrease with age (p = 0.055) (Table 1).

Fig. 4.

Impact of age and metformin daily oral supplementation on testicular histology in adult hamsters. Representative micrographs of testes from young adult (5-month-old) and aged adult (22-month-old) Syrian hamsters that were given metformin dissolved in drinking water (170 mg/kg/day) for two months prior to sacrifice or that received drinking water alone. Scale bars: 50 μm (panels a) and 20 μm (panels b). Asterisks show seminiferous tubules with mild disorganization of the epithelium. Arrows indicate premature detachment of spermatocytes and spermatids. Hashtags highlight seminiferous tubules with severe disorganization of the epithelium and loss of spermatocytes and spermatids. Arrowheads show vacuolization within the cytoplasm of Sertoli cells

Oral daily supplementation of both, young and old Syrian hamsters, with metformin tended to increase FSH levels, and significantly increased the percentage of seminiferous tubules with premature detachment of spermatocytes and spermatids (Table 1). The percentage of seminiferous tubules with epithelial alterations showed a statistically non-significant tendency to increase, while the among of seminiferous tubules with vacuolization within the cytoplasm of Sertoli cells significantly increased when elderly hamsters were treated with metformin (Table 1). Instead, in young hamsters supplemented with metformin, these parameters remained unchanged. On the other hand, oral supplementation of young and old Syrian hamsters for two months with a daily dose of metformin did not affect the number and the diameter of seminiferous tubules or the percentage of seminiferous tubules in stage VII of the epithelial cell cycle (Table 1). Figure 4 shows premature detachment of spermatocytes and spermatids in testes of 5-month-old and 22-month-old hamsters supplemented with metformin as well as in seminiferous tubules with vacuolization within the cytoplasm of Sertoli cells in metformin-treated aged hamsters.

DSP per testis was significantly decreased by age but unaffected by metformin treatment in both, young and aged hamsters (Table 1). Sperm production efficiency [calculated as DSP by testicular parenchyma weight (g)] showed a statistically non-significant tendency to decrease by age but was not altered by the treatment of young and old hamsters with metformin (Table 1).

Epididymal sperm concentration diminished in aged hamsters compared to young adult animals, while supplementation of young and elderly hamsters with metformin did not modify this parameter (Table 1). Sperm motility was unaffected by age or metformin (Table 1).

Effects of aging and metformin on testicular collagen deposition and the thickening of the seminiferous tubules wall in adult hamsters

Collagen deposition was evaluated by picrosirius red staining in testicular slices from 5-month-old and 22-month-old Syrian hamsters that received drinking water alone, as well as in testicular slices from young and aged Syrian hamsters that were given metformin dissolved in drinking water (170 mg/kg/day) for two months prior to sacrifice.

Picrosirius red-positive staining was found associated with the testicular capsule, blood vessels wall and the seminiferous tubules wall (Fig. 5a).

Fig. 5.

Impact of age and metformin daily oral supplementation on testicular collagen deposition in adult hamsters. Representative micrographs of testes from young adult (5-month-old) and aged adult (22-month-old) Syrian hamsters that were given metformin dissolved in drinking water (170 mg/kg/day) for two months prior to sacrifice or that received drinking water alone. Digitally processed images of a series of individual picrosirius red-stained testicular images viewed in bright-field that were automatically stitched together to obtain an entire testicular section. Scale bars: 1000 μm (panel a). Color threshold images of picrosirius red-stained testicular sections that were processed with the FIJI/ImageJ software and used to quantify collagen deposition. In all images, red corresponds to picrosirius red-positive staining. Scale bars: 1000 μm (panel b). Collagen deposition was quantified as the percentage of the total area stained with picrosirius red in the entire testicular section, in the parenchyma and in the capsule. Bar plot graph represents the mean ± S.E.M. [n (young hamsters/young hamsters + metformin/aged hamsters/aged hamsters + metformin) = 10/6/7/8]. ANOVA followed by Student-Newman–Keuls test was performed when data were normally distributed. When data were not normally distributed, the non-parametric Dunn test was performed. *p < 0.05 vs. group of young adult (5-month-old) hamsters. # p < 0.05 vs. group of aged adult (22-month-old) hamsters that received drinking water alone (panel c)

When processed color threshold images were used to quantify collagen deposition (Fig. 5b), a significant increase in the percentage of the total area stained with picrosirius red in the entire testicular section was detected in aged hamsters compared to gonads of 5-month-old animals (Fig. 5c). Furthermore, metformin daily oral supplementation markedly augmented the percentage of collagen deposition in the entire testicular section of 22-month-old hamsters (Fig. 5c). However, metformin treatment did not affect the percentage of collagen deposition in the entire testicular section of young hamsters (Fig. 5c). An identical pattern of change was seen when the percentage of the total area stained with picrosirius red was calculated in the parenchyma (Fig. 5c). Instead, a statistically non-significant tendency to increase with age was observed when the percentage of the total area stained with picrosirius red was evaluated in the capsule (Fig. 5c). Metformin supplementation did not alter the percentage of the total area stained with picrosirius red in the capsule of testes from young or old hamsters (Fig. 5c).

Whereas under standard bright-field microscopy, sections stained with picrosirius red dye showed the presence of red collagen fibers in the wall of the seminiferous tubules (Fig. 6a), under polarized light microscopy yellow–red birefringence collagen fibers were seen in the wall of all testicular sections examined (Fig. 6b). In testes of young adult hamsters that were given metformin dissolved in drinking water for two months prior to sacrifice as well as in gonads of those young animals that received drinking water alone, weak and homogeneous red birefringence collagen fibers were detected (Fig. 6b). Instead, testes of 22-month-old hamsters showed heterogeneous birefringence and the presence of gaps highlighting a major disorganization of the collagen fibers in the tubular wall compared to the gonads of 5-month-old animals (Fig. 6b). The presence of disorganized collagen fibers was also evident in testes of aged hamsters supplemented with metformin (Fig. 6b). In fact, zones of the tubular wall with high, medium–low or complete absence of birefringence can be identified in testes of aged hamsters with or without metformin treatment (Fig. 6b).

Fig. 6.

Impact of age and metformin daily oral supplementation on the organization of the tubular wall in testes of adult hamsters. Representative micrographs of testes from young adult (5-month-old) and aged adult (22-month-old) Syrian hamsters that were given metformin dissolved in drinking water (170 mg/kg/day) for two months prior to sacrifice or that received drinking water alone. Picrosirius red-stained testicular sections showing the features of the tubular wall were examined by bright-field microscopy. Scale bars: 50 μm (panel a). Picrosirius red-stained testicular sections showing the characteristic of the tubular wall were examined by polarized Light microscopy. Orange arrowheads highlight collagen fibers with homogeneous and weak birefringence. Red arrowheads feature collagen fibers with high birefringence, yellow arrowheads point out collagen fibers with medium-low birefringence, and white arrowheads show collagen fibers without birefringence. Scale bars: 50 μm (panel b). Merged images of panels a and b are shown. Scale bars: 50 μm (panel c)

Picrosirius red-stained sections observed by standard bright-field microscopy were also used to evaluate the thickness of the wall of the seminiferous tubules, which was markedly increased in aged hamsters compared to those values quantified in gonads of young adult animals (Fig. 7a and b). Metformin supplementation further increased the fibrotic thickening of the tubular wall in aged hamsters but did not modify this parameter in young animals (Fig. 7a and b).

Fig. 7.

Impact of age and metformin daily oral supplementation on the thickening of the wall of the seminiferous tubules in testes of adult hamsters. Representative micrographs of testes from young adult (5-month-old) and aged adult (22-month-old) Syrian hamsters that were given metformin dissolved in drinking water (170 mg/kg/day) for two months prior to sacrifice or that received drinking water alone. Picrosirius-red stained testicular sections showing the thickness of the tubular wall were examined by bright-field microscopy. Arrows highlight the thickness of the tubular wall. Scale bars: 20 μm (panel a). Thickening of the tubular wall was quantified using the FIJI/ImageJ software. Bar plot graph represents the mean ± S.E.M. [n (young hamsters/young hamsters + metformin/aged hamsters/aged hamsters + metformin) = 10/6/7/8] of the thickness of the tubular wall. ANOVA followed by Student-Newman–Keuls test was performed. *p < 0.05 vs. group of young adult (5-month-old) hamsters. # p < 0.05 vs. group of aged adult (22-month-old) hamsters that received drinking water alone (panel b)

Discussion

The Syrian hamster is widely used for biomedical research due to the anatomical, physiological, metabolic and pathophysiological similarities that it shares with humans in comparison with other laboratory rodents [50–54]. Reports of other authors and previous studies from our group carried out not only in Syrian hamsters but also in mice with altered longevity allow us to propose that aging increases the inflammatory and oxidative status of the testis [4, 10].

In the present work, we continued and deepened our investigations on testicular aging in Syrian hamsters. Our initial findings suggesting an age-dependent increase in testicular inflammation have now been further supported by activation of the NLRP3 pathway. The latter was evidenced through higher expression levels of NLRP3, Nek7, caspase 1, and IL1β in testes of 22-month-old hamsters than in gonads of 5-month-old animals. Previous studies from our group and others described the expression of NLRP3 in testicular macrophages, Sertoli cells, Leydig cells and peritubular cells in different species including hamster, mouse, rat and human [4, 55–58].

Confirming our own previous observations [4], the age-associated rise of testicular inflammation in Syrian hamsters took place in parallel to an increase in local oxidative stress revealed by high lipid peroxidation accompanied by an elevated catalase expression. Various oxidants including lipid peroxides are able to induce the expression of the anti-oxidant enzyme catalase which catalyzes the decomposition of hydrogen peroxide to water and oxygen [59]. Early immunohistochemical studies from our group revealed that catalase expression is predominantly associated with Sertoli cells of the human and hamster testis [29, 37]. Nevertheless, some peritubular cells and interstitial cells also showed catalase immunoreactivity [37]. In contrast, catalase immunostaining was not seen in germ cells [37]. The increase in testicular catalase expression in aged hamsters was not accompanied by evident changes in the activity of this enzyme. Such discrepancy could result from post-translational (phosphorylation and/or ubiquitination of the protein) and/or epigenetic (DNA methylation, alterations of histones) modifications that regulate catalase enzyme activity as was previously reported [60].

While neither body weight nor blood glucose showed age-associated modifications, the inflammatory-oxidative alterations described in testes of aged hamsters went along with a significant fall in the testicular weight and the gonadosomatic index. Moreover, increased expression levels of the autophagy marker p62, and diminished expression levels of the lysosome marker LAMP2 were detected in testes of old hamsters. Since p62 is continually degraded by autophagy [14, 61], high protein expression levels of p62 in testes of old hamsters suggest that late-phase autophagy is blocked and, consequently, autophagic activity is reduced in aging. On the other hand, loss of LAMP2 expression causes the accumulation of autophagosomes and autolysosomes in several tissues triggering lysosomal/autophagic disturbances [62, 63]. In this context, decreased expression of LAMP2 in testes of old hamsters further confirmed the existence of impaired autophagy at advanced ages.

As expected, circulating blood levels of LH and testosterone, as well as the expression of StAR and 17β-HSD were markedly reduced in testes of 22-month-old hamsters compared to gonads of 5-month-old animals, confirming that the steroidogenic capacity is negatively affected by aging.

DSP and epididymal sperm concentration significantly diminished in aged hamsters accompanying not only the fall in testicular weight but also a significant decrease in the tubular diameter and an evident increase in the number of seminiferous tubules per unit area. These observations are in agreement with previous reports [64, 65]. We also observed a non-statistically significant (p = 0.055) decrease in the percentage of seminiferous tubules in stage VII of the epithelial cell cycle, in line with a previous report in which the spermatogonia proliferation index was found to be significantly lower in aging hamsters specifically in stages VII–VIII of the seminiferous epithelium cycle [65]. These changes took place concomitantly with an age-dependent increase in the percentage of seminiferous tubules showing epithelial alterations and in the percentage of seminiferous tubules with premature detachment of spermatocytes and spermatids, revealing a clear decline of the spermatogenic performance in aged hamsters.

Collagen deposition has previously been associated with testicular aging [3]. Now, by picrosirius red-positive staining we showed the existence of collagen fibers in the testicular capsule, blood vessel walls as well as in the wall of the seminiferous tubules in testes of adult hamsters. Furthermore, a significant increase in the percentage of collagen deposition in testes of aged hamsters compared to gonads of 5-month-old animals has been seen. Because statistically non-significant changes were detected in the capsule, the increase in collagen deposition in the entire testicular section of aged hamsters seems to take place mainly at the expense of an augment in the percentage of the total area of each testicular section picrosirius red-stained in the parenchyma. Supporting this assumption, aged-related thickening of the tubular wall has become evident.

We also detected a statistically significant fall in the expression of the phosphorylated form of H2AX (γH2AX) in testes of 22-month-old hamsters compared with gonads of 5-month-old animals. Phosphorylation of the DNA-damage repair marker H2AX is crucial for activating the DNA damage response promoting cell survival. While acute stress increases γH2AX levels [66], chronic oxidative stress, just like what happens in the aging testis, seems to decrease the total levels of H2AX protein and, therefore, might prevent normal accumulation of its phosphorylated form, and the subsequent formation of the γH2AX loci [67, 68]. Our results showing a decreased expression of γH2AX in testes of old hamsters are in agreement with a previous report in testicular cells of 18-month-old C57BL/6 mice [69], and could be derived from the chronic elevated lipid peroxidation detected in gonads of these elderly animals. Oppositely, Zhao et al. [70] reported accumulation of γH2AX in testes of 24-month-old rats, while comparable γH2AX signals were described between testes of young (9–18 months of age) and old (7–15 years of age) dogs [71].

Total protein expression of the widely used marker of proliferation PCNA showed a statistically significant decrease in aged hamster testes compared with gonads of young animals. Similarly, Hussein et al. [72] described a marked decrease in the mean number of positive PCNA stained testicular cells with advancing age in albino rats. In hamster testis, PCNA staining is observed in the seminiferous tubules and the interstitium [73]. In this context, it has been reported that the number of Leydig cells and Sertoli cells as well as the spermatogonia proliferation index decrease during aging in the Syrian hamster probably as a result of an imbalance between proliferation and cell death [65, 74]. However, more PCNA-positive germ cells were identified during aging in other species [75, 76], presumably as a result of the activation of a mechanism to compensate for impaired spermatogenesis and to maintain spermatogenic output.

In parallel with PCNA expression decrease, PTEN expression was markedly higher in testes of old hamsters than in young adult animals. Overexpression of PTEN has been associated not only with inhibition of cell proliferation but also with aging protection [24]. In this context, Pten transgenic mice present extended lifespan [24].

Overall, our results further confirm that testicular aging in Syrian hamsters is linked to the development of local inflammatory-oxidative-fibrotic processes that take place in parallel to diminished autophagy, reduced steroidogenesis, impaired spermatogenesis, decreased cell proliferation, and altered DNA damage response.

Metformin is a synthetic dimethyl biguanide used as an oral antidiabetic drug in patients with type 2 diabetes mellitus [3, 77] that also mitigates the progression of various aging-related diseases. After sixty years of its clinical use, metformin can be considered a safe drug. The fact that this anti-hyperglycemic drug is widely used worldwide, relatively inexpensive and has no or only mild and well-known side effects [78], has encouraged the development of numerous investigations by the scientific community to elucidate its potential as an anti-aging therapy.

In this study, to evaluate the impact of metformin in testes of aged Syrian hamsters, testicular fragments were incubated in the presence or absence of 50 µM metformin, which is considered a clinically relevant concentration [79]. Interestingly, we detected protective effects of metformin against inflammation and oxidative stress in testes of aged hamsters evidenced by a reduction in the expression levels of the NLRP3 inflammasome and IL1β, along with a decline in the degree of lipid peroxidation and the expression/activity of the anti-oxidant enzyme catalase. These changes took place concomitantly with an enhanced autophagy revealed by a lower expression of p62, marker of decreased autophagy, and a higher expression of LAMP2, the indicator of efficient lysosomal activity and autophagic flux.

While proliferation and steroidogenesis were unaffected in testes of aged hamsters incubated in the presence of metformin, phosphorylation of H2AX significantly increased suggesting the activation of the DNA damage repair system.

While in vitro studies provide controlled conditions and a high-throughput screening, in vivo studies allow long-term observations and physiological relevance, being then essential as complementary approaches in advancing the understanding of biology and medicine. Therefore, we decided to contrast our in vitro findings suggesting advantages of metformin therapy on the physiological status of the aging testis, with in vivo experiments. Unexpectedly, in vivo metformin treatment showed opposite results compared to those found in our previous in vitro experiments.

Body weight, blood glucose levels, testicular weight, and the gonadosomatic index were lower in 22-month-old hamsters orally supplemented with a daily dose of metformin equivalent to that used for type 2 diabetes mellitus therapy in humans. Furthermore, the percentage of seminiferous tubules showing vacuolization within the cytoplasm of Sertoli cells and the number of adluminal vacuoles per tubule significantly increased in testes of aged hamsters supplemented with metformin, which could lead to spermatogenesis disruption. In this context, the increase in the percentage of seminiferous tubules with premature detachment of spermatocytes and spermatids in hamsters supplemented with metformin dissolved in drinking water highlights a severe disorganization of the seminiferous epithelium.

Collagen deposition, a hallmark of testicular aging, was further increased in testes of 22-month-old hamsters supplemented with metformin mainly as a result of augmented picrosirius red staining in the parenchyma. In addition, tubular wall fibrotic events became evident in metformin-treated hamsters. These fibrotic changes have previously been related to damage of the testicular tissue and the irreversible loss of male reproductive function [80].

On the other hand, oral metformin supplementation increased the expression of inflammatory markers and indicators of oxidative state, while this antidiabetic drug further impaired autophagy and steroidogenesis. In addition, the expression of γH2AX was reduced by metformin reflecting an altered DNA damage response.

In brief, results from our in vivo experiments support a deleterious impact of metformin in the histomorphology and physiology of the aged testis. Although this negative effect is not fully reflected in the efficiency of the spermatogenesis, it remains to be evaluated whether longer periods of metformin supplementation could finally lead to a poor/arrested spermatogenesis.

The contradictory results from in vitro and in vivo studies are most likely due to fundamental discrepancies, on the one hand, in the experimental conditions and, on the other hand, in the biological complexity [81]. Because in vitro studies cannot replicate the complex processes taking place in vivo, they can lead to different outcomes. In addition, metabolism and distribution of drugs in a living organism with high organization significantly impact their effectiveness and toxicity [82].

Unlike the effects described in old hamsters, daily oral administration of metformin for two months showed fewer detrimental effects in testes of young adult hamsters. In this context, oxidative stress and the percentage of seminiferous tubules showing premature detachment of spermatocytes and spermatids increased, while body and testicular weights, the gonadosomatic index, and blood LH and testosterone levels decreased with metformin treatment in young adult animals. Therefore, some effects of metformin administration such as testicular redox imbalance and impaired steroidogenesis along with reduction of body and gonadal weights seem to take place independently of age. Nevertheless, the physiological changes that occur during testicular aging would markedly raise the metabolic targets of metformin compared to those present in the young testis and, therefore, increase considerably the risk of harmful actions of metformin in the aging male gonad.

While metformin has been reported to exert beneficial effects in the female reproductive system [83–86], literature concerning metformin effects in the testis is controversial. A recent article described that metformin disrupts germinal epithelial integrity in the testes of healthy adult male Wistar Albino and causes spermatogenic cell loss in a dose-dependent manner [87]. However, most of the studies in which the impact of metformin in the testis was investigated, used samples from patients showing diabetes and/or metabolic syndrome or, alternatively, male gonads from rodent models with induced diabetes or kept under a high fat diet (HFD), with these studies yielding contradictory results [88–98].

In summary, previous findings as well as our current work reveal that the effects of metformin on testicular function are still far from being conclusive and, consequently, further investigations will be required in diverse aging models sharing similarities in genetics, physiology, and behavior with humans. The dose used, administration route, the timing of intervention, duration of treatment, and the existence of comorbidities (i.e. diabetes, metabolic syndrome and/or obesity), among others, seem to be critical issues that can significantly influence the outcomes leading to beneficial or harmful effects on the testes of metformin-treated individuals.

Overall, it is currently suggested that prescribing metformin to healthy individuals might delay aging and extend Lifespan. Therefore, healthy non-diabetic middle-aged individuals may now consider taking metformin for prophylactic or anti-aging purposes. However, our data show that daily metformin administration in a dose equivalent to that usually received by patients with type 2 diabetes mellitus exerts not only negative effects in the testicular histology and the inflammatory-oxidative-fibrotic state in healthy (non-obese/non-diabetic) aged hamsters but also other harmful actions such as decreasing blood glucose levels which could affect the energy source required for testicular metabolism. Therefore, the anti-aging therapeutic potential of metformin to reverse or ameliorate testicular age-related alterations earlier proposed in the literature should be seriously reconsidered and, if necessary, disregarded.

Supplementary information

Below are the links to the electronic supplementary materials.

Abbreviations

17β-HSD

17β-hydroxysteroid dehydrogenase

ANOVA

One-way analysis of variance

Co

Control

CO₂

Carbon dioxide

COX2

Cyclooxygenase 2

DHT

Dihydrotestosterone

DMSO

Dimethyl sulfoxide

DNA

Deoxyribonucleic acid

DSBS

Double-strand breaks

DSP

Daily sperm production

ECM

Extracellular matrix

FSH

Follicle stimulating hormone

Gapdh

Glyceraldehyde 3-phosphate dehydrogenase

HFD

High fat diet

IL18

Interleukin 18

IL1β

Interleukin 1β

IU

International unit

LAMP2

Lysosome-associated membrane protein 2

LH

Luteinizing hormone

MDA

Malondialdehyde

NEK7

NIMA-related kinase 7

NIMA

Never in mitosis gene A

NLRP3

Nucleotide-binding oligomerization domain, leucine rich repeat and pyrin domain containing 3

PCNA

Proliferating cell nuclear antigen

PFA

Paraformaldehyde

PTEN

Phosphatase and tensin homolog on chromosome 10

qPCR

Quantitative polymerase chain reaction

RIA

Radioimmunoassay

ROS

Reactive oxygen species

RT

Reverse transcription

S.E.M

Standard error of the mean

StAR

Steroidogenic acute regulatory

TBARS

Thiobarbituric Acid Reactive Substances

Author contribution

A.C.G: methodology, conceptualization, formal analysis and investigation, writing—original draft preparation, writing—review and editing. S.P.R: conceptualization, formal analysis and investigation, funding acquisition, writing—review and editing. T.M.P: methodology, writing—review and editing. R.S.C: conceptualization, formal analysis and investigation, writing—review and editing. V.G.D.R: methodology, writing—review and editing. D.J.C: methodology, writing—review and editing. J.M.R.G: methodology, writing—review and editing. M.S.B: methodology, writing—review and editing. M.S.T: methodology, conceptualization, writing—review and editing. M.E.M: conceptualization, formal analysis and investigation, funding acquisition, supervision, writing—review and editing. M.B.F: conceptualization, formal analysis and investigation, writing—original draft preparation, funding acquisition, supervision, writing—review and editing. 

Funding

This study was supported by grants from the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET project number PIP 2021 963), the Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT, project numbers BID PICT 2018 01221, BID PICT 2019 2479, BID PICT 2020 3633, and BID PICT 2021 0277), and René Barón, Williams, Lucio Cherny and Honorio Bigand Foundations of Argentina.

Data availability

Data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.