Background
The incidence of diabetes in adults nears 10, and 90% of the subjects are of the insulin-dependent type 2, the seventh-leading recognised cause of death in the world [
1,
2]. Type 2 diabetes used to be diagnosed primarily in older subjects. Nowadays, type 2 diabetes, impaired glucose tolerance and obesity touch adolescents and young subjects of reproductive age in epidemic proportion [
1]. Around 25% of type 2 diabetes men exhibit hypogonadotropic hypogonadism [
3]. Anterior pituitary hormone secretion is severely perturbed in diabetic and obese subjects [
4‐
7]. Metabolic disorders impact on the general population by altering reproduction in both sexes. Yet, the type 2 diabetes-induced impact on spermatogenesis has so far received little attention.
Elevated blood glucose is a pathological feature of diabetes mellitus associated with an inadequate control of the sugar by insulin. Because glucose is controlled by insulin in different ways in cells of the body, we aim to assess the elements of the glucose metabolism that influence the signalling pathways activated by insulin receptors in the two cellular compartments within the testis because each exerts distinct functions. For this reason, glucose, InsulinT (Total), and Insulin2 concentrations and the insulin receptor (IR) beta IR-β and alpha IR-α subunits protein content were assessed in the serum, the interstitial tissue- and seminiferous tubule-enriched fractions during the normal mouse postnatal development. Next, we investigated glucose control mechanisms peculiar to the interstitium and tubules in the leptin-deficient (
ob/ob) [
8] and leptin receptor-deficient (
db/db) [
9] mice, two spontaneously diabetic and obese infertile type 2 diabetes mouse models.
The leptin-deficient (
ob/ob) or (
Lepob/Lepob) mouse model develops hyperphagia, obesity, transient hyperglycaemia, high serum insulin, elevated numbers of pancreatic beta cells, lowered LH [
6], GH [
4] in the serum but higher hypophyseal Prl protein content [
10]. Conversely, the leptin receptor-deficient
db/db mouse model exhibits hyperleptinemia, obesity, hyperglycaemia, high insulin and decreased Prl in the serum [
7] and elevated hypophyseal Prl [
10]. Humans with mutations of leptin or its receptor present the phenotype of obesity and infertility [
11].
Besides adipocytes, Leydig cells in humans and rodent [
12] and human spermatozoa express leptin [
13]. In humans, the serum leptin concentration is inversely proportional to androgen levels [
14]. The
db/db and
ob/ob mice exhibit low testosterone levels [
15]. The leptin’s inhibiting action on food intake has been ascribed to fat stores and obesity [
8]. Thirty to 40% of infertility cases are said to arise from obesity [
16] and an increase in body weight in the male was shown to impact the fertility in the couple [
17].
Glucose uptake has been shown to drop [
18] and protein synthesis [
19] and oxygen uptake ceased to be stimulated by glucose in testes in which germ cells were absent either naturally during development or in response to assaults [
20]. These studies revealed that roughly a third of the energy produced aerobically in the testis is directly contributed by glucose, the major source of energy supplied by spermatocytes and spermatids which metabolise the sugar through the Embden-Meyerh of pathway of glycolysis, acetyl CoA formation, and the citric acid cycle [
21]. The remaining of the energy produced is supplied by endogenous substrates like lipids [
22]. The lactate production is associated chiefly with the interstitial and Sertoli cells in which glucose is oxidized to CO
2 in small amounts but predominantly through the pentose cycle and pyruvate carboxylation pathways necessary for maintaining the citric acid cycle [
22].
The report of a negative correlation between glucose and human sperm motility [
23] evidences an impact of the sugar on the gamete’s metabolism which albeit varies amongst species. For instance, in the mouse, glucose is required for hyper activated motility at the end of capacitation for successful fertilisation by epididymal spermatozoa [
24], beyond capacitation [
25] and on the sperm-oocyte fusion [
26]. As well, glucose optimises capacitation and fertilization in human sperm [
27]. Incubating sperm with glucose increases in vitro fertilization rates in human [
28]. Sperm motility is altered in subjects with insulin-dependent diabetes [
29]. The spermatozoon’s plasma membrane and the acrosome are targets for insulin [
30]. Insulin has been located in Leydig cells and spermatids in rat [
31] and in the human spermatozoon’s subacrosomal space, midpiece and tail [
32]. Not only human ejaculated spermatozoa express the mRNA and insulin protein but, in addition, insulin is secreted by the gametes through an autocrine feedback affecting its own secretion [
32]. The nuclear and mitochondrial DNA fragmentation and apoptosis in spermatozoa were higher in diabetic than in normal subjects [
33] indicating detrimental effects on the germ cells development within the testis. Moreover, sperm DNA damage was said to decline embryo quality and implantation rates [
34].
A single copy of insulin-coding gene was reported in the human and Guinea pig genome [
35] by contrast to mouse and rat in which insulin genes are part of a two-gene system [
36,
37]. The
ins2 is an ortholog to the insulin genes in other mammals including humans;
ins1 which results from a duplication of the ancestral
ins2 gene is a rodent-Murinae-specific retrogene involved in the glucose metabolic pathways [
38]. In the pancreas,
ins1 and
ins2 are transcribed and both encode proinsulin peptides which are made up of signal peptide, B chain, C-peptide, and A chain. The report that only the
ins1 gene hastens the onset of type 1 diabetes in the knockout nonobese diabetic (NOD) male mice indicates that
ins1 and
ins2 gene hold different functions [
39,
40]. The fact that in the above reports, the male
ins1-carrying NOD mice were principally affected denotes a unique impact of the insulin genes in the male. We took advantage of the
db/db and
ob/ob mouse models to identify the individual impact of
ins1 and
ins2 in the glucose metabolic pathways within the interstitial tissue and seminiferous tubules.
Insulin is an anabolic peptide hormone of a disulfide-linked 21-amino-acid chain A and a 30-amino-acid B chain secreted by the pancreatic β cells and on which depends glucose homeostasis. The hormone’s impact on the glucose metabolism varies with the target cells but its initial action takes place through binding as a monomer to a glycoprotein receptor endowed with insulin-stimulated tyrosine kinase activity located within the target cells’ plasma membrane [
41]. The insulin receptor (IR) belongs to a subfamily of receptor tyrosine kinases that encompasses the insulin-like growth factor (IGF)-I receptor and the insulin receptor-related receptor (IRR) [
42]. In somatic cells, the insulin receptor is a heterotetrameric complex made up of two disulfide bond-linked extracellular α-subunits each one linked by another disulfide bond to a transmembranous β-subunit that comprises the cytoplasmic tyrosine kinase domain and the phosphorylation sites [
43]. The tyrosine kinase activity of the β-subunit is constitutively inhibited by the α-subunit [
42]. The insulin binding to the α-subunit releases the kinase activity in the β-subunit from the inhibition causing the transphosphorylation and conformational changes that activates signalling cascades [
42].
This study shows that during development, glucose concentration augmented in the serum, the interstitium and seminiferous tubules in which glucose levels became less than in the interstitial tissue in adulthood. Total (T) Insulin and Insulin2 concentrations dropped in serum whereas in the interstitium, InsulinT rose while Insulin2 decreased in adulthood. The results show that insulin is not glucose-regulated in the interstitium in contrast to the tubules where insulin regulates glucose levels and high Insulin2 coexists with stable glucose concentrations in normal adult mouse. The db and ob mutations produced different effects on the content and fragmentation of the IR proteins and on insulin signalling. The db and ob mutation-induced downregulation of α- and β subunits decreased IR in the interstitium. Conversely, in tubules and the anterior pituitary, the α subunit was not affected but the reduced β subunits to be activated by insulin would not suffice to stimulate downstream effectors.
Materials and methods
Chemicals
Phenylmethane-sulfonyl fluoride (PMSF), leupeptin, aprotinin, and Lumi-lightPlus chemiluminescence detection kit were purchased from Roche (Laval, QC, Canada). Potassium bisperoxo (1, 10- phenanthroline) oxovanadate (V) [bpV (phen)] was obtained from Calbiochem (San Diego, CA, USA). Protein G agarose was from Expedeon (San Diego, CA, USA).
Antibodies
Mouse monoclonal IgG anti-IR-β subunit from Millipore (Etobicoke, ON, Canada). Rabbit polyclonal IgG anti-IR-α subunit from Biorbyt (Cambridge, UK). Mouse monoclonal anti-IR-α subunit was from ThermoFisher, Rockford, IL, USA). HRP-conjugated anti-mouse IgG and HRP-conjugated anti-rabbit IgG were obtained from Jackson Immunoresearch Laboratories (West Grove, PA, USA).
Animals
The mouse model provides invaluable opportunities to explore consequences of altering the coding of specific genes on precise tissue functions. However, the small size of mouse testis allowed obtaining interstitial tissue- and seminiferous tubule-enriched fractions in small amounts. Mice were first anaesthetized (urethane, 1 g/kg IP, Sigma, St-Louis, MO, USA) before decapitation next, testes and epididymides were harvested. Blood was collected, allowed to clot; serum was obtained by centrifugation at 1500 rpm (GS-6R Beckman Centrifuge, JH-3.8 Rotor) 20 min and stored at -80C. Animal use protocol was approved by University of Montreal Animal Care Committee (Protocol number 12–126).
Normal mouse
Studies on development were carried out on the same male mice of BALB/cJ background that we used and described in our earlier study [
15]. Five animals were used per age group.
Diabetic and obese mice
These studies were carried out on the same mice as the ones we used and described in our earlier study [
15]. Twenty male mice aged of 10 weeks with the leptin receptor (B6.BKS(D)-
Leprdb/J homozygote (
db/
db) Stock Number 00697) mutation, 25 male mice aged of 10 weeks with the leptin (B6.Cg-
Lepob/J homozygote (
ob/
ob) Stock Number 00632) mutation both experimental group on the C57BL/6 J genetic background and ten wild type (wt) mice were used to identify the consequences of diabetes and obesity resulting from distinct mutations of specific genes on selected testicular functions. Mice were purchased from Jackson Lab (Bar Harbor, ME, USA). They were housed at RT with food and water ad libitum and exposed to a 12 h: 12 h light-dark cycle.
Isolation of seminiferous tubule-enriched fractions
Different anatomical and functional characteristics set apart the interstitium and the seminiferous tubules in the testis. Yet, in most studies, assays are performed on whole testis extracts rather than on interstitium- and/or seminiferous tubule-enriched fractions as was done in the present study. We showed that exposure to enzymes significantly alters the detection of the phosphorylated and glycosylated proteins forms within tissue samples [
44]. For this reason, the interstitial tissue-enriched and seminiferous tubule-enriched fractions were obtained without a beforehand enzymatic digestion using the technical approach detailed elsewhere [
15,
44]. Briefly, seminiferous tubules were mechanically teased apart from the interstitium with Dumont fine tweezers from freshly decapsulated testes in cold phosphate buffered saline (PBS: 137 mM NaCl, 3 mM KCl, 8 mM Na
2HPO
4, 1.5 mM KH
2PO
4, pH 7.4) containing 2 mM PMSF, 1 mM EGTA, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 4 mM Na
3VO
4, 80 mM NaF and 20 mM Na4P
2O
7 with 10 μM bpV (phen). The resulting seminiferous tubule-interstitium solution was centrifuged 15 min at 400 rpm, (GS-6R Beckman Centrifuge, JH-3.8 Rotor) at 4C after having been allowed to decant. The interstitial tissue- (ITf) and seminiferous tubule-enriched (STf) fractions were centrifuged 10 min at 1000 rpm (GS-6R Beckman Centrifuge, JH-3.8 Rotor) at 4C. The enriched fractions were characterised under the light microscope [
15,
45].
Isolation of epididymal spermatozoa
The isolation of epididymal spermatozoa was carried out as described before [
46]. Briefly, epididymides from
db/db,
ob/ob and WT mice were diced in cold PBS with proteases and phosphatase inhibitors, filtered through a 74 mm mesh, and centrifuged at 2000 rpm for 15 min in a GS-6R Beckman centrifuge (JH-3.8 Rotor) at 4C to recover spermatozoa. Gametes were resuspended in10 mM Tris-HCl, pH 8, containing 1 mM EDTA for 5 min to lyse epithelial and blood cells [
47], washed twice, and diluted 1:1 in cold PBS with proteases and phosphatase inhibitors. Cells were sonicated in a Fisher Sonic Dismembrator (model 300; Fisher, Farmington, NY) during three 30 s intervals.
Protein quantification
Proteins in samples were assayed using materials from BioRad (BioRad, Mississauga, ON, Canada).
Electrophoresis and western blot analyses
Twenty to thirty μg total homogenate of sample were loaded on polyacrylamide gels, separated by 10% SDS-PAGE, transferred onto nitrocellulose membranes and subjected to western blotting as previously described [
48]. In all western blot experiments, the membranes were first stained with Ponceau red to ensure equal loading. Next, membranes were blocked 1 h at 37C with 5% skimmed milk in TRIS-buffered saline (TBS: 137 mM NaCl, 27 mM KCl, 25 mM Tris-HCl pH 7.4) next, incubated with the different antibodies. The antibody dilutions were prepared in 5% skimmed milk-TBS: polyclonal anti-IR-α (2 μg/ml), monoclonal anti-IR-α (5 μg/ml), monoclonal anti-IR-β (1.25 μg/ml). Next, membranes were washed in TBS containing 0.05% Tween 20 and incubated 1 h with a corresponding secondary antibody conjugated to HRP at RT. The antigen-antibody complexes were detected by chemiluminescence. The intensity of the immunoreactive bands was quantified by laser scanning with the public Scion Image Software (Scioncorp, MD, USA).
Serum and tissue glucose measurements
Serum, STf and ITf glucose content was measured using an enzymatic (Mutarotase-GOD) calorimetric technique (Autokit Glucose Wako, Wako, TX, USA)) according to the manufacturer’s instructions. Tissue factions were prepared as described by Koya et al. [
49] with some modifications. Briefly, STf and ITf were sonicated in 6 N perchloric acid while in an ice bath. The acid homogenates were centrifuged at 14,000
g and the supernatant used for glucose determination. Ten μl serum or testicular fraction homogenates were mixed with 1.5 ml colour reagent and next incubated 10 min. The absorbance of samples and standards was measured at 505 nm against the blank.
Insulin measurements in serum and tissues
Insulin content in serum and tissue fractions was measured with commercially available ELISAs. STf and ITf were homogenized with a tissue grinder in an acid ethanol solution (180 mM HCl in 70% ethanol; 0.01 ml /mg tissue) while on ice [
50]. Tissue lysates were sonicated (Fisher Sonic Dismembrator) 3 X 15 s before being centrifuged 5 min at 10,000
g at 4C. The supernatant was recovered for insulin determination. Total insulin (Insulin T) levels were measured with an ELISA kit from ALPCO Diagnostics (Salem NH, USA) and insulin 2 levels were measured with Ins2 ELISA kit from Aviva Systems Biology (San Diego, CA, USA).
Immunoprecipitation
One milligram protein from serum was incubated either with 2 μl rabbit polyclonal anti-IR-α (100 μg/ml) or sample buffer (used as control) and left overnight on a rotating drum at 4C. The following day, samples were incubated with 50 μl protein-G agarose for 3 h at 4C and centrifuged 5 min at 14,000 g at 4C next, the supernatant was discarded. The pellets were washed with 10 mM TRIS, pH 7.4, 150 mM NaCl, 1% Triton X-100, and 1 mM EDTA and the supernatant discarded. Lastly, pellets were resuspended in 50 μl of 2 x SDS-PAGE loading buffer heated 10 min at 50C and centrifuged 2 min (14,000 g) at 4C. The supernatant was transferred to a new tube, boiled 5 min before western blotting. Membranes were incubated overnight at 4C with either polyclonal anti-IR-α or monoclonal anti-IR-α.
Data and statistical analysis
The statistical analyses were done with Stata software (Stata Corporation, College Station, TX, USA). The data were evaluated with the Student’s t test or analysis of variance (ANOVA) followed by Tukey honest significant difference (HSD) test according to the number of groups to be compared.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.