Targeted disruption of galectin 3 in mice delays the first wave of spermatogenesis and increases germ cell apoptosis

Abstract

Galectin 3 is a multifunctional lectin implicated in cellular proliferation, differentiation, adhesion, and apoptosis. This lectin is broadly expressed in testicular somatic cells and germ cells, and is upregulated during testicular development. Since the role of galectin 3 in testicular function remains elusive, we aimed to characterize the role of galectin 3 in testicular physiology. We found that galectin 3 transgenic mice (Lgals3^−/−) exhibited significantly decreased testicular weight in adulthood compared to controls. The transgenic mice also exhibited a delay to the first wave of spermatogenesis, a decrease in the number of germ cells at postnatal day 5 (P5) and P15, and defective Sertoli cell maturation. Mechanistically, we found that Insulin-like-3 (a Leydig cell marker) and enzymes involved in steroid biosynthesis were significantly upregulated in adult Lgals3^−/− testes. These observations were accompanied by increased serum testosterone levels. To determine the underlying causes of the testicular atrophy, we monitored cellular apoptosis. Indeed, adult Lgals3^−/− testicular cells exhibited an elevated apoptosis rate that is likely driven by downregulated Bcl-2 and upregulated Bax and Bak expression, molecules responsible for live/death cell balance. Moreover, the percentage of testicular macrophages within CD45⁺ cells was decreased in Lgals3^−/− mice. These data suggest that galectin 3 regulates spermatogenesis initiation and Sertoli cell maturation in part, by preventing germ cells from undergoing apoptosis and regulating testosterone biosynthesis. Going forward, understanding the role of galectin 3 in testicular physiology will add important insights into the factors governing the development of germ cells and steroidogenesis and delineate novel biomarkers of testicular function.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-021-03757-2.

Introduction

Mammalian spermatogenesis is a complex process comprised of pre-meiotic, meiotic, and spermiogenic stages [1–3]. During the pre-meiotic stage, spermatogonia proliferate and differentiate by mitosis before transforming into meiotic spermatocytes [2]. Meiosis is a unique process of cell division, taking place only in germ cells and leading to the formation of haploid gametes [4]. After two rounds of cell division, the spermatocytes form round spermatids, which later—after undergoing several morphological and structural changes—transform into elongated spermatids [3, 5]. In mice, the first wave of spermatogenesis is initiated in a few days after birth, and then continues in a synchronized manner, up to post-natal day (P) 35 [6, 7]. Germ cell types arise in a sequential pattern during the first wave of spermatogenesis in all cords [7]. At P5 the seminiferous tubules are mainly comprised of spermatogonia and immature Sertoli cells [8]. Then, early spermatocytes emerge at P9, pachytene spermatocytes at P14 as meiosis is initiated, and round spermatids at P20 [9]. During this period, the Sertoli cells mature, morphologically diversify and become functional—as evidenced by the secretion of testicular fluid to the lumen [10, 11]. The precise regulation of the proliferation and differentiation of these somatic and germ cells during the first wave of spermatogenesis is essential to ensuring normal fertility.

Galectin 3 is a 30-kDa lectin that binds to β-galactoside [12]. Galectin 3 is involved in various cellular processes, including proliferation, differentiation, macrophage activation, apoptosis regulation, and angiogenesis. Functionally, galectin 3 is best characterized for its role in driving inflammation, fibrosis, and immune responses [13–15]. This lectin is widely expressed in different tissues and developmentally regulated. In human, murine and porcine testes galectin 3 expression is up-regulated during testicular development [16–18]. Galectin 3 is expressed in the testis and epididymis, in ejaculated spermatozoa and extracellular vesicles in human seminal plasma, which serve as vehicles to transfer the lectin to the sperm surface during post-testicular maturation. Sperm-bound galectin-3 then plays a critical role in spermatozoa—zona pellucida binding [16, 19–21].

Furthermore, galectin 3 expression can be upregulated upon treating cultured Sertoli cells with follicle stimulating hormone (FSH) or epidermal growth factor (EGF) [16]. Galectin 3 expression also increases in the atrophied seminiferous tubules after vasovasostomy in mice, which could result from testicular inflammation [17]. Indeed, we have observed galectin 3 upregulation in rat testes from animals with testicular inflammation (unpublished observations).

In a broader context of reproduction galectin 3 play a role in embryo implantation, immune regulation, trophoblast-matrix interaction and in placental angiogenesis [22]. Deficiency in galectin 3 during murine pregnancy leads to placental dysfunction, higher abortion rate and fetal growth restriction [23].

Since galectin 3 is often secreted into biological fluids, its role as an important biomarker of heart or autoimmune diseases is frequently discussed [24].

Taken together, it seems that galectin 3 is an important lectin involved in testicular development and inflammation. Yet despite these observations, the testicular function of galectin 3 is unknown. In this study, we aimed to investigate the influence of galectin 3 depletion on key parameters of testicular function such as germ cell proliferation, differentiation and apoptosis; Sertoli cell maturation; Leydig cell steroidogenesis; and testicular macrophage numbers using whole body galectin 3 knockout mice Lgals3^−/−.

Materials and methods

Animals

Homozygous galectin-3 knockout mice (Lgals3⁻/⁻) were bred on a C57/BL6 genetic background, as described previously [25]. Both Lgals3⁻/⁻ and Lgals3⁺/⁺ mice were maintained under standard specific-pathogen-free conditions. Organ collection in this study was approved by the local Animal Ethics Committee (Regierungspräsidium Giessen GI Nr. 646_M). For euthanasia, animals were deeply anaesthetized by inhalation of 5% isoflurane and sacrificed by cervical dislocation.

Testes were collected from neonatal (P5), pre-pubertal (P15), pubertal (P20), and adult (P35, P50, P100) mice. Both testes were removed, weighed, and either snap frozen in liquid nitrogen for RNA extraction and protein isolation or fixed in Bouin’s solution for paraffin embedding. Paraffin tissue Sects. (5 µm) were subjected to histological analysis (hematoxylin–eosin staining) and immunostaining.

To measure steroid hormone concentrations, serum samples were collected from mice at P50.

Gas chromatography tandem mass spectrometry (GC–MS/MS) measurement of steroid hormones

Up to 100 µL serum was spiked with a cocktail of internal standards containing [16,16,17-2H3] testosterone (d3-T), [16,16,17-2H3] 5α-androstanediol-3α,17β (d3-AD), [16,16,17-2H3] dihydrotestosterone (d3-DHT) [26]. After equilibration, the samples were extracted with ethyl acetate and then purified by gel chromatography on Sephadex LH-20 mini columns. Heptafluorobutyric anhydride (Sigma-Aldrich) was used for derivatization. Gas chromatography was performed on an Optima® 1-MS capillary column (25 m × 0.2 mm I.D., df 0.1 µm, Macherey–Nagel, Düren, Germany) housed in a Thermo Scientific Trace 1310 Gas Chromatograph with a TriPlus RSH Autosampler coupled to a TSQ 800 triple quadrupole MS (Thermo Scientific, Dreieich, Germany). Helium (1.0 mL/min) was used as the carrier gas. The steroids of interest were eluted at a rate of 3 °C/min until the column temperature reached 250 °C. Quantification was performed in the multiple reaction monitoring mode (MRM). The following MRM or m/z ratios were measured for the analytes and their corresponding internal standards: m/z 665.1 (668.1) for T (d3-T), m/z 455.3/241.3 (458.3/244.4) for AD (d3-AD) and m/z 414.1/185.2 (417.2/188.2) for DHT (d3-DHT). The limits of the quantifications were 0.1 ng/ml for T and AD, and 0.03 ng/ml for DHT.

Immunofluorescence staining

Testicular paraffin-embedded sections were dewaxed in xylene, rehydrated in decreasing concentrations of ethanol, and boiled for antigen retrieval for 20 min in 10 mM sodium citrate buffer. After blocking with 5% goat serum in TBS containing 0.1% Tween 20 (TBST), the sections were incubated overnight with the indicated primary antibody at 4 ℃ (Table 1). The sections were then washed in TBST, and incubated with the appropriate secondary antibody at room temperature (RT) for 1 h. The nuclei were counterstained with Topro 3 (Life Technologies, Darmstadt, Germany). Images were captured under a conventional fluorescence (Axioplan 2 Zeiss microscope, Carl Zeiss, Göttingen, Germany) and confocal (Leica TCS2, Wetzlar, Germany) microscope.

Table 1.

Detailed list of antibodies and antibody dilutions used in this study

Primary antibodies	Manufacturer	Catalogue no	Dilution
Rabbit monoclonal anti-Bak	Cell signaling technology, Germany	12,105	1:1000*
Rabbit polyclonal anti-Bax	Cell signaling technology, Germany	2772	1:1000*
Rabbit monoclonal anti-Bcl-2	Cell signaling technology, Germany	2870	1:1000*
Mouse monoclonal anti-β-actin Ac-15	Sigma-Aldrich, Germany	A5441	1:5000*
Rabbit monoclonal anti-galectin 3	GeneTex, USA	GTX62084	1:500**
Rat monoclonal anti-germ cell specific antigen (TRA98)	Abcam, UK	ab82527	1:1000**
Rabbit polyclonal anti-HSD11B1	Abcam, UK	ab39364	1:100**
Rabbit polyclonal anti-Insl3	Biorbyt Ltd, UK	orb648755	1:100*
Rabbit polyclonal anti-PCNA	Abcam, UK	ab18197	1:500**
Rabbit polyclonal anti-phospho-histone H3 (Ser10)	Cell signaling technology, Germany	9701	1:200**
Rabbit monoclonal anti-Sox9	Abcam, UK	ab185230	1:250**
Rabbit monoclonal anti—StAR	Cell signaling technology, Germany	8449	1:1000*
Secondary antibodies	Manufacturer	Catalogue No	Dilution
Alexa Fluor 448 goat anti-rabbit IgG	Thermo fisher scientific, USA	A11008	1:1000**
Alexa Fluor 546 goat anti-rat IgG	Thermo fisher scientific, USA	A11081	1:1000**
HRP-labelled goat anti-rabbit IgG	MP biomedicals, USA	855,676	1:10,000*
HRP-labelled sheep anti-mouse IgG	Sigma-Aldrich, Germany	A5906	1:10,000*

Antibody dilution used for *Western blotting or **Immunofluorescence/Immunohistochemistry

Histology, immunohistochemistry, and morphometry

Tissue sections were deparaffinized in xylene, hydrated in serial dilutions of ethanol, and treated for antigen retrieval by boiling for 20 min in 10 mM sodium citrate buffer. To block endogenous peroxidase activity, the sections were incubated in 1.2% H₂O₂ in methanol for 30 min at RT. Then, the sections were blocked with 5% goat serum (Vector Laboratories, Burlingame, USA) in TBST, and incubated with the indicated primary antibody (Table 1) overnight at 4 °C. An Envision System (DAKO, Hamburg, Germany) for PCNA or Vectastain ABC Elite Kit (Vector Laboratories) for HSD11b1 was used to detect primary antibody binding, according to the manufacturer`s instructions, combined with DAB staining. The Sects. (5 µm) were stained with hematoxylin and eosin (H&E) and evaluated as previously described [1].

For comparative analyses, age-matched animals were chosen. Sections were always processed simultaneously. All images were obtained under a Leica DM LB microscope (Leica, Wetzlar, Germany).

Quantitative microscopic analyses were performed using a Leica DM LB microscope (Leica, Wetzlar, Germany). For each mouse, 100 seminiferous tubules in Lgals3^−/− and Lgals3^+/+ testicular sections at P15 and P20 were analyzed. Each tubule was classified according to lumen formation and the most differentiated germ cell stage it contained (Sper—spermatogonia, L/Z—leptotene/zygotene spermatocyte, P—pachytene spermatocyte, RS—round spermatid). The percentage of seminiferous tubules of each class was calculated for each mouse and then plotted against age.

TUNEL assay

Apoptosis was quantified by the TUNEL method (ApopTag 7100 Kit, Chemicon, Hofheim, Germany) according to the manufacturer’s instructions. Counterstaining was carried out with hematoxylin. Counting was performed in a blinded fashion. All tubules of one testicle cross-section were analyzed by counting the total number of tubules as well as the total number of tubules with at least one TUNEL-positive nucleus. Results are presented as percentage (positive tubules/all tubules × 100) (n = 3 per group).

Western blotting

Testes or cells were homogenized in ice-cold RIPA buffer (50 mM Tris pH 8.0, 1% Nonidet P 40, 0.5% deoxycholate, 0.1% SDS, 150 mM NaCl) containing a 1% protease inhibitor cocktail (Sigma-Aldrich, Steinheim, Germany). Equal amounts of protein were separated by 10% SDS–polyacrylamide gel electrophoresis and electro-blotted onto nitrocellulose membranes (GE Healthcare, Freiburg, Germany). The efficiency of the transfer was monitored by Ponceau S staining. The membranes were blocked with 5% BSA in TBST for 1 h before incubation overnight with the appropriate primary antibody (Table 1). Then, the membranes were washed and incubated with the appropriate secondary HRP-conjugated antibody (Table 1). The signals were visualized using ECL (Millipore Corporation, Billerica, USA) and analyzed by densitometry (Fusion FX, Witec AG, Luzern, Switzerland).

RNA extraction and quantitative PCR

Total RNA was isolated using an RNeasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Quantitative PCR was performed on a CFX96 Touch thermal cycler (Bio-Rad, Munich, Germany) using iTaq Universal SYBR Green Supermix (Bio-Rad, Munich, Germany). The primer sequences and amplicon sizes are provided in Table 2. Relative gene expression was calculated by using the ΔΔCt method [27].

Table 2.

List of PCR primers used in this study

Gene	Primer (5` → 3`)	Entrez gene ID	Amplicon size (bp)
Actb	F: TGACAGGATGCAGAAGGAGAT	11,461	156
	R: TACTCCTGCTTGCTGATCCAC
Amh	F: CGAGCTCTTGCTGAAGTTCC	11,705	302
	R: TGAAACAGCGGGAATCAGAG
Cyp11a1	F: CCAGTGTCCCCATGCTCAAC	13,070	74
	R: TGCATGGTCCTTCCAGGTCT
Cyp17a1	F: GCCCAAGTCAAAGACACCTAAT	13,074	159
	R: GTACCCAGGCGAAGAGAATA
Ddx4	F: GGTCCAAAAGTGACATATATACCC	13,206	140
	R: TTGGTTGATCAGTTCTCGAGT
Gapdh	F: TGACGTGCCGCCTGGAGAAA	14,433	98
	R: AGTGTAGCCCAAGATGCCCTTCAG
Hprt	F: CTGGTGAAAAGGACCTC	15,452	110
	R: CTGAAGTACTCATTATAGTCAAG
Hsd3b1	F: TGGACAAAGTATTCCGACCAGA	15,492	250
	R: GGCACACTTGCTTGAACACAG
Insl3	F: TGCAGTGGCTAGAGCAGAGA	16,336	151
	R: GTGCAGCCAGTAAGACAGCA
Lgals3	F: GATCACAATCATGGGCACAG	16,854	100
	R: ATTGAAGCGGGGGTTAAAGT
Star	F: TGCCCATCATTTCATTCATCCTT	76,205	232
	R: AAAAGCGGTTTCTCACTCTCC
Wt1	F: ACCCAGGCTGCAATAAGAGA	22,431	230
	R: CCTGGTGTGGGTCTTCAGAT

Transfection of MLTC-1 cell line with Lgals3 siRNA

A mouse Leydig tumor cell line (MLTC-1) was cultured in ATCC-formulated RPMI-1640 (Gibco, Darmstadt, Germany) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 0.1 mg/mL streptomycin. Cell cultures were maintained in a 5% CO₂ humidified atmosphere at 37 °C. The cell line was kindly provided by Prof. Thomas Linn and Dr. Qingkui Jiang (Clinical Research Unit, Centre of Internal Medicine, Justus-Liebig-University, Giessen, Germany). A commercially available, pre-designed short interfering RNA (siRNA) against Lgals3 and Silencer® select negative control siRNA was purchased from Life Technologies GmbH (Darmstadt, Germany). Transient transfection was performed using Lipofectamine^® RNAiMAX Transfection Reagent (Thermo Fisher Scientific, Waltham, USA), according to the manufacturer's instructions. Briefly, 1 × 10⁵ MLTC-1 cells were seeded in a 12-well plate and transfected at 60–80% confluency with 15 pmol siRNA and 4.5 µl Lipofectamine® RNAiMAX per well. After 48 h incubation, the cells were treated with 1 IU/ml hCG (Predalon 5000 I.E. Organon, Oberschleissheim, Germany) for 24 h. The supernatants and cell lysates were collected to measure testosterone levels and Western blot analyses, respectively.

Testosterone measurement

The levels of testosterone in the cell culture supernatants were measured using a testosterone ELISA kit (IBL International, Hamburg, Germany), according to the manufacturer’s protocol.

Flow cytometry

Decapsulated testes were incubated with 1.2 mg/ml collagenase A and 15 U/ml DNase (Roche Diagnostics, Mannheim, Germany) in PBS in a shaking water bath at 34 °C for 15 min. The enzymes were inactivated by adding ice-cold PBS, and the tubule fragments were allowed to settle for 4 min. Then, the supernatant was filtered and centrifuged at 300 × g for 10 min at 4 °C. The pellet was washed with PBS and the erythrocytes were depleted by osmotic lysis using red blood cell lysis buffer (Qiagen, Hilden, Germany) for 5 min at RT. The final cell suspension was washed in washing buffer (0.5% BSA and 2 mM EDTA in PBS) and then processed directly for flow cytometric macrophage staining. Interstitial cells (1 × 10⁶) were incubated with the following fluorochrome conjugated recombinant REAfinity antibodies: CD45 -VioGreen (clone REA737), CD11b -VioBright FITC (clone REA592), Ly6C -PE (clone REA796), and F4/80 -APC (clone REA126) (all from Miltenyi Biotech, Bergisch-Gladbach, Germany). Background staining was evaluated using the appropriate REA control antibodies (clone REA293; Miltenyi Biotech). Finally, the cells were washed in washing buffer. Data were collected for 700,000 events using a MACSQuant Analyzer 10 flow cytometer and analyzed using MACSQuantify software version 2.11 (Miltenyi Biotech).

Statistical analyses

The data represent the means ± SEM. Comparisons between two groups were made by two-tailed t test. P values < 0.05 were considered statistically significant. All tests were performed using GraphPad Prism 5 software (GraphPad Software, San Diego, USA).

Results

Loss of galectin 3 delays the first wave of spermatogenesis in Lgals3^−/− mice

To investigate the effects of endogenous galectin 3 on testicular development, we harnessed Lgals3^−/− mice [25]. These mice have been previously extensively studied in the context of fibrosis, inflammation, heart failure, atherosclerosis or cancer but not for the effects of galectin 3 deficiency on male reproduction and fertility [24]. In our first analyses, testis weights were determined from mice at different time points after birth and in adulthood. Testicular weights increased progressively with age in mice of all genotypes but were significantly decreased in adult Lgals3^−/− mice (P50 and P100) compared to age-matched Lgals3^+/+ controls (Fig. 1a).

Fig. 1.

The first wave of spermatogenesis is delayed in Lgals3^−/− mice. a Testicular weight (mg) collected from Lgals3^+/+ and Lgals3^−/− mice at postnatal (P) days 5, 15, 20, 35, 50 and 100; (n = 4, * P < 0.05). b Representative histology of mouse testes from Lgals3^+/+ (i, iii, v) and Lgals3^−/− mice (ii, iv, vi) at P5 (i, ii), P15 (iii, iv) and P20 (v, vi). (C and D) The percentage of tubule cross sections was quantified according to the most differentiated germ cell stage visible (G gonocytes, L leptotene spermatocyte, P pachytene spermatocyte, RS round spermatid, S Sertoli cell, Sper spermatogonia, Z zygotene spermatocyte) at P15 (c) and P20 (d); (n = 3, *P < 0.05, **P < 0.01, ***P < 0.001). The data represent the means ± SEM

To characterize testicular development and the onset of spermatogenesis in Lgals3^−/− testes, we performed a histological analysis of the testes at P5, P15 and P20. At P5, the seminiferous tubules from Lgals3^+/+ and Lgals3^−/− testes were mainly composed of Sertoli cells and spermatogonia (Fig. 1b i–ii). We also observed gonocytes (G) in both genotypes (Fig. 1b i–ii). By P15, meiosis progressed such that pachytene spermatocytes (P) were now present in the Lgals3^+/+ testes. However, most of the differentiated germ cells in Lgals3^−/− mice were composed of zygotene spermatocytes (Z), with numerous tubules containing only spermatogonia (Fig. 1b iii–iv). Finally, by P20, we observed signs that the next step of spermatogenesis had occurred in Lgals3^+/+ mice, as evidenced by the presence of round spermatids (RS) in the seminiferous tubules (Fig. 1b v) indicating complete passage through meiosis. By contrast, most tubules in age-matched Lgals3^−/− testes contained only pachytene spermatocytes, a mid-meiotic stage (Fig. 1b vi) As such, it seems that loss of galectin 3 delays the first wave of spermatogenesis in Lgals3^−/− mice.

To confirm these findings, we aimed to quantify spermatogenesis in testes from Lgals3^+/+ and Lgals3^−/− animals at P15 and P20 (Fig. 1c, d). To do so, we determined the percentage of seminiferous tubules containing the most advanced stage of spermatogenesis in each section. At P15, we found that the percentage of tubules containing pachytene spermatocytes was significantly lower in Lgals3^−/− testes as compared to Lgals3^+/+ testes (9% vs. 23%, respectively; Fig. 1c). Subsequently, at P20, the percentage of tubules containing round spermatids and pachytene spermatocytes in Lgals3^−/− testes was also significantly lower compared with Lgals3^+/+ testes (3.6% vs. 17% and 53% vs. 82%, respectively; Fig. 1d). Conversely, the percentage of tubules containing less-differentiated stages of spermatogenesis (such as leptotene/zygotene spermatocytes or spermatogonia) was increased in Lgals3^−/− testes compared to Lgals3^+/+ testes (14% vs. 3.7% and 18% vs. 0%, respectively; Fig. 1d). Taken together, these data illustrate that the first wave of spermatogenesis in Lgals3^−/− testes is delayed.

Germ cell number is reduced in neonatal Lgals3^−/− testes, independently of cellular proliferation

We next wanted to investigate the underlying causes of this apparent delay of the first wave of spermatogenesis in Lgals3^−/− testes. To do so, we assayed germ cell numbers and germ and somatic cell proliferation by immunostaining with the germ cell marker Tra98 and the mitosis marker phospho-histone H3 Ser10 (PH3) during early postnatal testis development. We found that Lgals3^−/− testes had nearly a twofold lower number of germ cells at P5 compared to Lgals3^+/+ control testes (Fig. 2a–g).

Fig. 2.

Testicular germ cell numbers are reduced in neonatal Lgals3^−/− testes. a–d Identification of germ cell numbers in Lgals3^+/+ and Lgals3^−/− testes using Tra98 immunofluorescence staining (red) at P5 (a–d). Phosphorylated histone 3 staining (PH3, green; b, c, d) and Tra98 staining were used to distinguish between dividing germ and somatic cells (a, c, d). Part d shows the magnified areas identified by the white boxes in part (c). The thin arrow indicates germ cells; the thick arrow indicates somatic cells, both positive for PH3 (a cell proliferation marker). Nuclei were counterstained with Topro 3 (blue). e–g The numbers of germ cells (e), PH3-positive germ cells (f) and somatic cells (g) were normalized to the total cross-section area (mm²) in both Lgals3^+/+ and Lgals3^−/− testes. h Ddx4 mRNA expression (germ cell marker) of P5 and P15 testes was normalized to Actb and Gapdh; (n = 3–4, *P < 0.05). The data represent the means ± SEM

We also monitored Ddx4 (VASA) expression, which serves as a germ cell marker during germ cell differentiation [28]. Here we found that Ddx4 mRNA expression was downregulated in neonatal (P5) and prepubescent (P15) Lgals3^−/− testes compared to control Lgals3^+/+ testes (Fig. 2h), which strongly supports a reduction in germ cell numbers in Lgals3^−/− testes. By contrast, the number of dividing germ cells (PH3- and Tra98-positive) and somatic cells (PH3-positive) in Lgals3^−/− testes was unaffected (Fig. 2f, g).

Next we studied the expression of proliferating cell nuclear antigen (PCNA)—a marker of cell proliferation. At P5, PCNA staining revealed no overt differences in the proliferation of testicular cells between Lgals3^+/+ and Lgals3^−/− mice (Supplementary Fig. 1). By contrast, several seminiferous tubules were found without proliferating cells in Lgals3^−/− testes at P15 and P20 (Supplementary Fig. 1). Together, these findings point to a reduced number of gonocytes and spermatogonia that is independent of proliferation, in the neonatal Lgals3^−/− testis.

Germ cells are less frequent in the seminiferous tubules of neonatal and pubertal Lgals3^−/− testes

The PCNA staining results described above showed in several tubules no sign of cell proliferation in P15 and P20 Lgals3^−/− testes. To decipher a possible influence of endogenous galectin 3 on germ cell development, we stained testes from neonatal (P5), pre-pubertal (P15) and pubertal (P20) Lgals3^+/+ and Lgals3^−/− mice with the Sertoli cell marker Sox9 and the germ cell marker Tra98. Our quantitative analyses revealed that in P5 Lgals3^−/− testes, the number of seminiferous tubules without germ cells was significantly increased as compared to Lgals3^+/+ testes (46% vs. 13%, respectively). In Lgals3^+/+ testes some tubules were entirely free of germ cells (Fig. 3a–e), an observation not seen from P15 onwards when all seminiferous tubules showed germ cells. Conversely, 8% of tubules at P15 and ~ 15% of tubules at P20 in Lgals3^−/− mice had no germ cells (Fig. 3e). By P35, however, all seminiferous tubules in Lgals3^−/− testes contained germ cells. These results indicate that the number of seminiferous tubules containing germ cells in Lgals3^−/− mice is reduced during the neonatal and prepubescent periods.

Fig. 3.

Testes from Lgals3^−/− mice exhibit a higher number of seminiferous tubules without germ cells and altered maturation of Sertoli cells. a–d Representative immunofluorescent labeling of germ cells with Tra98 (red; a, c, d) and Sertoli cells with Sox9 (green; b–d) in testes from Lgals3^+/+ and Lgals3^−/− mice at P5. Part d shows the magnified areas identified by the white boxes in part (c). The white stars indicate the seminiferous tubules without germ cells. The enumeration of tubules containing germ cells was performed on P5, P15 and P20 (i). At least 100 seminiferous tubules were counted in each testis; (n = 3, *P < 0.05). e Representative HE staining of testes from Lgals3^+/+ and Lgals3^−/− mice at P15. The lower panel (iii, iv, v, vi) shows the magnified areas from (i) and (ii) containing seminiferous tubules with (iii, v) or without (iv, vi) a lumen in Lgals3^+/+ (i) and Lgals3^−/− (ii) testes, respectively. f–g The percentage of seminiferous tubules without (f) and with (g) a lumen was quantified by counting at least 100 seminiferous tubule cross sections in each testis. (H) Relative Amh mRNA expression in testes from Lgals3^+/+ and Lgals3^−/− mice at P5, P15, P20 and P35 was normalized to Actb and Gapdh; (n = 3, *P < 0.05, **P < 0.01). The data represent the means ± SEM

Lgals3^−/− testes display defects in Sertoli cell maturation

Sertoli cell maturation and concomitant fluid secretion results in the formation of the lumen of the seminiferous tubules during testicular development [29]. To characterize the effects of Lgals3 gene disruption on Sertoli cell maturation, we analyzed lumen formation at P15 and performed a quantitative analysis of the Sertoli cell maturation marker anti-Müllerian hormone (Amh) [11]. The percentage of tubules containing a lumen in Lgals3^−/− testes was two-fold lower compared to in Lgals3^+/+ control tubules (Fig. 3f–g). Moreover, our results showed that Amh mRNA was highly expressed in the neonatal period (P5) in testes from both Lgals3^+/+ and Lgals^−/− mice (Fig. 3h) and then decreased at P15 before reaching extremely low levels at P20 and P35 in both strains. These findings confirm the progressive maturation process of Sertoli cells (Fig. 3h). Interestingly, although slightly decreased compared to P5, the levels of Amh mRNA expression in Lgals3^−/− testes at P15 were three-fold higher as compared to Lgals3^+/+ control testes, indicating a delay to Sertoli cell maturation in Lgals3^−/− testes (Fig. 3h).

We next studied the effects of galectin 3 on Sertoli cell number, and the ratio between Sertoli and germ cells in the developing testes. To do so, we performed an immunostaining using Sox9 and Tra98 antibodies at P15. Although not quantified, the immunostaining revealed no obvious difference in Sox9 expression between Lgals3^−/− and Lgals3^+/+ testes at P15 (Supplementary Fig. 2).

Wilms’ tumor gene (Wt1) is considered a stable marker of Sertoli cells during testicular development [11]. Our analyses of Wt1 mRNA expression showed that the expression levels in Lgals3^−/− testes were comparable to those in Lgals3^+/+ testes at P15 (Supplementary Fig. 2). This finding suggests that the number of Sertoli cells is unaffected in Lgals3^−/− mice. As such, while Lgals3^−/− testes display altered Sertoli cell maturation, the overall number of Sertoli cells is unaffected.

Lgals3^−/− testicular cells exhibit an increased rate of apoptosis accompanied by mild atrophy

To determine the underlying causes of testicular atrophy in adult Lgals3^−/− mice, we performed TUNEL staining to detect the level of cellular apoptosis. Quantifications of TUNEL-positive cells revealed higher numbers of apoptotic cells in adult Lgals3^−/− testes compared to Lgals3^+/+ testes at P5, P35, P50 and P100 (Fig. 4a–c). Due to the fact that ~ 15% of tubules at P20 in Lgals3^−/− mice contained no germ cells analysis of TUNEL positive cells had a bias and was consequently not included.

Fig. 4.

Testicular cells in Lgals3^−/− mice experience increased apoptosis due to an imbalance of anti- and pro-apoptotic proteins. a, b TUNEL staining (arrows) of testicular sections from Lgals3^+/+ mice and Lgals3^−/− mice at P50. c The numbers of apoptotic germ cells were quantified in 100 tubular cross sections in testes from Lgals3^+/+ and Lgals3^−/− mice at P5, P15, P35, P50 and P100. d A scheme is illustrating how galectin 3 can stabilize the mitochondrial membrane by binding to Bcl-2 and inhibiting cytochrome c release. Bcl-2 inhibits activity of pro-apoptotic proteins Bak and Bax. Western blot e and densitometric analysis of Bcl-2 f Bak g and Bax h expression in testes from Lgals3^+/+ and Lgals3^−/− mice at P50 is shown; (n = 3, *P < 0.05; **P < 0.01; ***P < 0.001). β-actin was used as a loading control. The data represent the means ± SEM

Galectin 3 can stabilize the mitochondrial membrane by binding to the anti-apoptotic protein Bcl-2; it can also inhibit cytochrome c release by blocking the activity of the pro-apoptotic proteins Bak and Bax [30–32] (Fig. 4d). We thus investigated the expression of these pro- and anti-apoptotic proteins in Lgals3^−/− and Lgals3^+/+ testes at P50. Bcl-2 was downregulated, while Bak and Bax were upregulated in Lgals3^−/− testes, as compared to Lgals3^+/+ testes (Fig. 4e–h). These data indicate that an increase in the number of apoptotic cells in adult Lgals3^−/− mouse testes results from an imbalance in anti-/pro-apoptotic protein expression. This perturbation might contribute to mild testicular atrophy in Lgals3^−/− mice.

Gonadal Insl3 and steroidogenic enzymes and circulating testosterone levels are upregulated in Lgals3^−/− mice

Production of galectin 3 in granulosa cells is regulated by luteotropic hormone (LH), which indicates an involvement of galectin 3 in steroidogenesis [33]. Similar to granulosa cells, the predominant function of testicular Leydig cells is sex hormone production (testosterone) through a steroidogenic process [34]. To investigate whether Lgals3 gene ablation affects Leydig cell function, we analyzed the gene and protein expression of Insulin-like factor 3 (Insl3) and steroidogenic enzymes (Cyp11a1, Hsd3b1 and Cyp17a1) in adult Lgals3^−/− and Lgals3^+/+ mice by quantitative RT-PCR and Western blotting.

Compared to Lgals3^+/+ mice, the Cyp11a1 and Hsd3b1 mRNA expression levels were significantly upregulated in adult Lgals3^−/− mice at P35 and P100 (Fig. 5a); Cyp17a1 mRNA expression was significantly increased in Lgals3^−/− mice testes at P50 and P100 (Fig. 5a).

Fig. 5.

Expression of steroidogenic enzymes and Insl3 is upregulated in Lgals3^−/− testes. a–b The relative mRNA expression of Cyp11a1, Hsd3b1, Cyp17a1 (A) and Insl3 b in Lgals3^+/+ and Lgals3^−/− testes was analyzed by real-time RT-PCR at P35, P50 and P100. The right panel shows the main pathway of steroidogenesis in rodents, with the corresponding enzymes highlighted by a dotted frame. c Western blot analysis of StAR and Insl3 protein expression in testes from Lgals3^+/+ and Lgals3^−/− mice at P50; (n = 3, *P < 0.05, ***P < 0.001, ****P < 0.0001). β-actin was used as a loading control. (D) Testosterone, androstandiol-3a, 17b, and dihydrotestosterone (DHT) concentrations in the sera from Lgals3^+/+ and Lgals3^−/− mice at P50, determined by GC–MS/MS; (n = 3–4, *P < 0.05, **P < 0.01). e–f StAR expression and testosterone levels in galectin 3-depleted MLTC-1 cells stimulated with 1 IU/ml hCG for 24 h. StAR protein expression in the cell lysates e and the testosterone concentrations in conditioned media (F) were determined by Western blotting and ELISA, respectively. Representative data are shown; (n = 3–4, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). The data represent the means ± SEM

Expression of Insl3 in adult testes is an indicator of the number and differentiation status of the Leydig cells [35]. In Lgals3^−/− mice testes, Insl3 mRNA expression was increased at P35, P50 and P100 (Fig. 5b). In agreement with the mRNA data, Insl3 and the steroidogenic acute regulatory protein StAR were also elevated in Lgals3^−/− testes compared to control Lgals3^+/+ testes at P50 (Fig. 5c). By using Leydig cell staining with an antibody against 11β-hydroxysteroid dehydrogenase type 1 (HSD11B1) no differences in the cell numbers between transgenic and WT animal groups at P50 were apparent (Supplementary Fig. 3).

Next, we investigated whether the increased expression of steroidogenic enzymes affected the concentration of steroids in the serum. To do so, we used a GC–MS/MS method to measure testosterone, androstandiol-3a, 17b (AD) and dihydrotestosterone (DHT) levels in sera taken from Lgals3^+/+ and Lgals3^−/− mice. Here, testosterone and AD levels were increased in adult Lgals3^−/− mice as compared to Lgals3^+/+ mice (Fig. 5d); no significant change was found in DHT sera levels (Fig. 5d). Taken together, these results suggest that sex hormone production is increased in Lgals3^−/− mice.

Lgals3 depletion in cultured Leydig tumor cells upregulates StAR and testosterone production

Our data thus far suggest that galectin 3 affects the steroidogenic activity of Leydig cells. To understand how galectin 3 affects Leydig cell function, we determined the level of testosterone secretion in a cultured Leydig cell line (MLTC-1) in which we depleted Lgals3 with siRNA. The efficiency of Lgals3 siRNA silencing is shown in Supplementary Fig. 4. We stimulated testosterone production by incubating the cells with hCG [36], and found that StAR protein levels and testosterone secretion were upregulated in both Lgals3-depleted and control cells (Fig. 5e–f). However, Lgals3 silencing led to a stronger increase in StAR expression and testosterone production under hCG stimulation compared to control cells (Fig. 5e–f). These results indicate an inhibitory effect of endogenous galectin 3 on testicular steroidogenesis.

Loss of Lgals3 results in a depletion of testicular macrophages

Galectin 3 is considered a chemoattractant for macrophages under physiological conditions, and can also accelerate M2 macrophage infiltration in tumor environments [37, 38]. Moreover, the macrophages possess a unique function in the maintenance of testicular immune privilege and support spermatogonial differentiation and Leydig cell steroidogenesis [39, 40]. Importantly for our research, testicular macrophages express galectin 3 [18]. In our final analyses, we thus investigated whether the macrophage population was affected in Lgals3^−/− testes. To do so, we performed flow cytometry using an extended antibody panel (CD45, Ly6C, F4/80 and CD11b) to target testicular macrophages (Supplementary Fig. 5). Within the population of testicular CD45⁺ leukocytes, we defined resident macrophages as Ly6C⁻F4/80⁺CD11b⁺. The percentage of testicular macrophages was significantly decreased by 11% in adult Lgals3^−/− mice compared to control Lgals3^+/+ (Fig. 6a–b). In sum, these data support that the endogenous galectin 3 regulates the number of testicular macrophages and their depletion may lead to disturbances in the spermatogenesis (via affecting spermatogonia) and steroidogenesis.

Fig. 6.

The percentage of testicular macrophages in the population of CD45⁺ leukocytes is decreased in Lgals3^−/− mice. a–b Testicular macrophage frequency was assessed by flow cytometry in Lgals3^+/+ and Lgals3^−/− testes at P50. A representative dot plot a and the corresponding quantification b is shown; (n = 4–6, *P < 0.05). The data represent the means ± SEM

Overall, our results indicate that endogenous galectin 3 plays an important regulatory role in the development of germ cells, maturation of Sertoli cells, steroidogenesis and maintenance of the macrophages niche, all comprising essential testicular functions.

Discussion

Galectin 3 is widely expressed in testicular cells including Sertoli, peritubular, Leydig, and germ cells and its upregulation is associated with testicular maturation [16, 18]. However, the role of endogenous galectin 3 in male gonadal development is unknown. Our results show for the first time that mice lacking galectin 3 expression experience a delay to the first wave of spermatogenesis, as evidenced by a decrease in the number of germ cells during the neonatal period and defects in Sertoli cell maturation. Moreover, we found evidence for increased Leydig cell marker and steroidogenic enzyme expression, elevated testosterone production, an increased rate of germ cell apoptosis, and lower numbers of testicular macrophages in adult Lgals3^−/− testes. Overall, these changes are accompanied by mild testicular atrophy. Together, these findings all point to an important role of endogenous galectin 3 in testicular physiology.

Our initial analyses identified a significant reduction in testicular weight in adult Lgals3^−/− mice compared to age-matched WT controls. We propose that the delay to the first wave of spermatogenesis in these transgenic animals may be responsible for this decrease in organ mass. Similar observations have been shown for regulatory genes, such as Rhox13, TR4, Rnf138, and Tap73 where a delay of the first wave of spermatogenesis is accompanied by a reduced testicular weight that persists into adulthood in mice [41–44].

In addition, Lgals3^−/− mice produce lower numbers of germ cells during the neonatal period. This is similar to mice lacking a functional Taf4b gene that are also characterized by reduced germ cell numbers during late embryogenesis and the neonatal period, which subsequently led to a delay to meiotic initiation [45]. Beside its influence on germ cell development or spermatogenic differentiation, galectin 3 seems to have a more complex multilayered influence on reproduction. In this regard, in human seminal plasma galectin 3 is associated with extracellular vesicles, which are responsible for immunosuppression and regulation of sperm function in the female reproductive tract [19]. Furthermore, the acrosomal region of capacitated human spermatozoa shows strong galectin 3 labelling, which mediates zona pellucida binding and affects fertilization in vitro [21].

Moreover, in our study we observed altered Sertoli cell maturation kinetics in Lgals3^−/− testes, whereby the lumen formation and the loss of Amh mRNA expression was delayed compared to WT testes. In maturating Sertoli cells AMH expression is progressively reduced reaching almost undetectable levels during transition to adulthood [46]. Galectin 3 expression in Sertoli cells is regulated by several factors, including EGF, TNF, FSH, and adult germ cells, which suggests that galectin 3 expression may affect Sertoli cell function [16]. Our data support that loss of galectin 3 in Sertoli cells could cause a delay to Sertoli cell maturation during the neonatal period, followed by a decrease in the number of germ cells. Abnormal development, proliferation and/or maturation of Sertoli cells in early life could lead to testicular dysfunction in adult life [11]. Such disruption in the Sertoli cell maturation has been associated with testicular dysgenesis syndrome [11, 46]. Conversely, the absence of germ cells in animal models and in human patients results in a delay to Sertoli cell maturation, but has no effect on their final maturation (reviewed in [11]).

We also observed an increase in the number of tubules without germ cells in both neonatal and pubertal Lgals3^−/− testes, which mirrors the reduced number of germ cells in the neonatal period. Chiarini-Garcia et al. reported that undifferentiated spermatogonia can migrate to ensure the uniform distribution of germ cell progeny at the basement of the seminiferous tubules [47]. A study by Jahnukainen et al. showed that after irradiation, spermatogenesis can recover in single testicular lobules in primate testes, indicating that spermatogenic stem cells are capable of re-colonizing their respective loop of the seminiferous tubules [48]. Therefore, we hypothesize that re-colonization by undifferentiated spermatogonia may explain our observation that many tubules were seen without germ cells in neonatal and prepubescent Lgals3^−/− testes, while they were eventually replenished in adulthood. However, testicular weight is affected in a long term and persists into adulthood, an observation that can be explained by the higher number of apoptotic germ cell in Lgals3^−/− testis.

In female mice, galectin 3 expression is increased during luteolysis and is associated with a loss of progesterone synthesis in vivo, which can be reversed by administration of hCG [33]. In this line, galectin 3 is downregulated in hCG-stimulated human granulosa cells [33]. In male mice, galectin 3 is expressed in Leydig cells [49]. Interestingly, we found an upregulated expression of the Leydig cell marker Insl3 and steroidogenic enzymes (Cyp11a1, Hsd3b1 and Cyp17a1) in adult Lgals3^−/− testes with concomitant elevated serum levels of testosterone and DHT. To determine the relationship between galectin 3 and testosterone production, we knocked down galectin 3 in the MLTC-1 Leydig cell line using Lgals3 siRNA. Following hCG stimulation, StAR expression and testosterone secretion were found increased in these cells. These data indicate that Lgals3 has an inhibitory effect on hCG-stimulated testosterone production in Leydig cells, although in our study we cannot exclude that high levels of LH may also play a role in the increased androgen levels in Lgals3^−/− mice.

We also revealed an increased number of apoptotic cells in adult Lgals3^−/− testes that is likely due to a perturbed Bcl-2, Bax and Bak expression ratio. Data from previous studies have suggested that galectin 3 can form heterodimers with Bcl-2 and thus stabilize the mitochondrial membrane [31, 50]. Bcl-2 also inhibits the oligomerization of Bax/Bak, which is associated with mitochondrial membrane permeabilization and cell apoptosis [51, 52]. Diao et al. recently reported that Lgals3 downregulation by siRNA in pituitary tumor cells leads to an increase in Bax and caspase 3 expression, and thus an elevated apoptotic rate [53]. Moreover, increased galectin 3 expression levels serve as a poor prognosis marker in hepatocellular carcinoma, bladder cancer, and pancreatic carcinoma — all related to the anti-apoptotic properties of galectin 3 [54–58]. Thus, the increased numbers of apoptotic cells in Lgals3^−/− testes in this study seems to be the result of an imbalanced anti- and pro-apoptotic protein expression due to the absence of endogenous galectin 3, an effect leading to testis shrinkage and lower number of germ cells in adulthood.

Sano et al. found that galectin 3 acts as a chemoattractant for macrophages under physiological conditions [38]. In contrast, the galectin 3 inhibitor lactose can restrain the infiltration of macrophages and neutrophils and induce pancreatic edema in acute pancreatitis [59]. Here, we found a decreased percentage of testicular macrophages within the population of CD45⁺ leukocytes in adult Lgals3^−/− mice. Interestingly, subpopulations of testicular macrophages called “peritubular macrophages” due to their close proximity to peritubular cells near the seminiferous tubule wall, express factors such as CSF1 which are involved in spermatogonial proliferation and differentiation, a finding that can explain reduced numbers of spermatogonia in macrophage depleted testis [39]. It can therefore be hypothesized that the reduced numbers of testicular macrophages in Lgals3^−/− testis at least partially contribute to the mild testicular atrophy seen in adulthood in these mice. On the other hand, interstitial macrophages, which are intimately connected to Leydig cells produce factors, e.g. 25-hydroxycholesterol, which could influence the steroidogenesis of Leydig cells [60]. 25-hydroxycholesterol from testicular macrophages is involved in the differentiation of Leydig cells to fully functional steroidogenic cells that occurs during post-natal maturation of the testis [61], a process in which macrophage numbers increase [62]. Here, we investigated the total number of macrophages in adult testis without further division into subpopulations. However, based on our in vitro results using a galectin-3 depleted MLTC-1 cell line, the increased expression of StAR protein followed by an enhanced secretion of testosterone points to a direct influence of this lectin on Leydig cell steroidogenesis, independent of an interaction with macrophages. Furthermore, the testosterone concentration in the serum is increased in Lgals3^−/− mice which also suggests a compensatory mechanism to maintain normal spermatogenesis.

Furthermore, these findings are in line with other studies demonstrating an increased expression of Star, Cyp11a1 and Cyp11b1 in the thymus followed by higher levels of corticosterone in sera and thymus of galectin 3 deficient mice [63].

Critically, it needs to be noted that our approach to utilize a whole body knockout of Lgals3^−/−does not fully allow to determine an impact of galectin 3 at the local testicular level. Yet, the broad expression of galectin-3 in most testicular cells would render it almost very difficult to deplete this lectin from the male gonad using cell-specific knock outs.

In conclusion, our data show that Lgals3^−/− mice experience a delay of the first wave of spermatogenesis. The resulting impairment to testicular development ultimately causes mild testicular atrophy due to an increased rate of germ cell apoptosis and a decreased number of macrophages in adult testes. Increased expression of steroidogenic enzymes in Lgals3^−/− testes, as well as functional in vitro data indicate that galectin 3 is involved in androgen biosynthesis. Our results thus underline an essential role for galectin 3 in archetypical testicular functions.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

The financial support of the China Scholarship Council to Tao Lei and of the Medical Faculty of Justus-Liebig University is gratefully acknowledged. This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG) BL1115/2-1 and Heisenberg Program (BL1115/3-1-BL1115/7-1) to S.M.B.

Data availability

The datasets generated during and analysed during the current study are available from the corresponding author on reasonable request.

Compliance with ethical standards

Conflict of interests

The author(s) declare(s) that they have no competing interests.

Ethics approval

All procedures involving animals were carried out in strict accordance with guidelines for care and use of experimental animals of the German law of welfare. Organ collection from mice was approved by the local Animal Ethics Committee (Regierungspräsidium Giessen GI Nr. 646_M).