Abstract

BACKGROUND

The vitamin D receptor (VDR) is expressed in human spermatozoa, and VDR-knockout mice and vitamin D (VD) deficiency in rodents results in impaired fertility, low sperm counts and a low number of motile spermatozoa. We investigated the role of activated VD (1,25(OH)2D3) in human spermatozoa and whether VD serum levels are associated with semen quality.

METHODS

Cross-sectional association study of semen quality and VD serum level in 300 men from the general population, and in vitro studies on spermatozoa from 40 men to investigate the effects of VD on intracellular calcium, sperm motility and acrosome reaction. All men delivered samples for routine semen analysis and blood for measurements of follicle stimulating hormone, Inhibin B, 25-hydroxy-VD, albumin, alkaline phosphatase, calcium and parathyroid hormone (PTH).

RESULTS

In the association study, 44% were VD insufficient (<50 nM), and VD was inversely correlated with PTH (P < 0.0005). VD serum levels correlated positively with sperm motility and progressive motility (P < 0.05), and men with VD deficiency (<25 nM) had a lower proportion of motile (P = 0.027), progressive motile (P = 0.035) and morphologically normal spermatozoa (P = 0.044) compared with men with high VD levels (>75 nM). 1,25(OH)2D3 increased intracellular calcium concentration in human spermatozoa through VDR-mediated calcium release from an intracellular calcium storage, increased sperm motility and induced the acrosome reaction in vitro.

CONCLUSIONS

1,25(OH)2D3 increased intracellular calcium concentration, sperm motility and induced the acrosome reaction in mature spermatozoa, and VD serum levels were positively associated with sperm motility, suggesting a role for VD in human sperm function.

Introduction

The vitamin D receptor (VDR) and the vitamin D (VD) metabolizing enzymes are expressed in the human ejaculatory duct, germ cells and mature spermatozoa (Blomberg Jensen et al., 2010). The importance of VD for male reproduction has been demonstrated in several animal studies (Inpanbutr et al., 1996; Kinuta et al., 2000; Bouillon et al., 2008). One of the VDR knockout mice models is characterized by an infertile phenotype (Kinuta et al., 2000; Bouillon et al., 2008), and VD deficiency in rodents leads to reduced sperm counts, impaired sperm motility and lower fertility rates in females inseminated with semen from VD deficient males (Kwiecinski et al., 1989). In humans, VDR and all the VD metabolizing enzymes are co-expressed during the late stages of spermatogenesis in the neck of mature spermatozoa (Nangia et al., 2007; Blomberg Jensen et al., 2010). Optimal sperm function may thus depend on a direct effect of VD; however, it could also be influenced indirectly through calcium homeostasis since the impaired fertility in animal models was partly restored solely by normalization of serum calcium levels (Uhland et al., 1992). VD has widespread biological functions, including an essential role for systemic calcium homeostasis (Bouillon et al., 2008) and a role for calcium in the maturation of human spermatozoa is well documented and highlighted by the 2–3-fold higher calcium concentration in human epididymal and prostate fluid compared with serum. VD has been hypothesized to be important for the proposed trans-epithelial calcium transfer in the epididymis (Blomberg Jensen et al., 2010).

The effects of VD rely on activation of cholecalciferol (inactive vitamin D3), which normally starts in the skin, where UV-B irradiation converts 7-hydrocholesterol to VD3. It is subsequently activated by the hepatic 25-hydroxylases (CYP2R1, CYP27A1) and the renal 1α-hydroxylase (CYP27B1), before the active 1,25(OH)2D3 (calcitriol) binds with high affinity to the VDR until inactivated by CYP24A1 (Prosser and Jones, 2004). 1,25(OH)2D3 exerts both genomic and non-genomic actions through binding to VDR (Norman, 1998), but the precise effects of VD in human spermatozoa remains to be shown. VD status in humans is generally monitored by measuring serum 25-hydroxyvitamin D levels, and there is a high frequency of VD insufficiency (<50 nM) among otherwise healthy adults in the western world (Holick and Chen, 2008). We hypothesized that VD deficiency in humans could aggravate semen quality as it does in rodents. Here, we show a positive association between VD serum level and sperm motility, and a direct effect of 1,25(OH)2D3 on intracellular calcium concentration [Ca2+]i, motility and acrosome reaction in mature spermatozoa.

Materials and Methods

Cross-sectional association study

From an ongoing monitoring study of semen quality of young men from the general Danish population (Jørgensen et al., 2010a,b), 308 men participating January to December 2007 were retrospectively investigated for an association between serum VD levels and semen quality. Eight were excluded due to use of anabolic steroids (n = 3), failure to deliver a semen sample (n = 2), obstructive azoospermia (n = 1), lack of blood sampling (n = 1) and Gilbert Meulengracht disease (n = 1), leaving 300 men in the study. The men delivered one semen sample, one blood sample (stored at −20°C until analysis), had a physical examination performed and answered a comprehensive questionnaire. The questionnaire included information on age, previous or current diseases, any known history of fertility, medication, etc.

Assessment of VD, reproductive hormones and other biochemical variables

Measurements of 25-OH VD rely on determination of both 25-OH Vitamin D2 and D3 and were conducted by isotope dilution liquid chromatography tandem mass spectrometry (LC-MS/MS). The method was tested by the use of 25-OH Vitamin D2 (MW 412.7, Fluka No. 17937), 25-OH Vitamin D3 (MW 400.7 Sigma-Aldrich No. H 4014) and an internal standard [2H6]-25-OH Vitamin D3 (26,26,26,27,27,27-hexadeutero-25-hydroxycholecalciferol, Mw: 406.7 Synthetica, Norway) dissolved in 50% ethanol/acetic acid, 200 mmol/l, pH 4.0. VD standard (NIST, USA, No. SRM 972) was used as primary calibrator, while working calibrators were 25-OH Vitamin D2 and D3 added to human serum to obtain four calibrators with D2 concentration up to 200 nmol/l, and D3 concentration up to 300 nmol/l. 25-OH Vitamin D3 and D2 were analyzed after liquid-liquid extraction. Internal standard (300 µl) was added to serum, control or calibrator (300 µl), and vortex-mixed for 2 min with 1 ml n-heptane. After mixing, the heptane phase was transferred to a heating block, and evaporated to dryness at 75°C under a stream of nitrogen. The residue was dissolved in 300 µl MeOH:H2O (80:20), and applied to the LC-MS/MS instrument (sample vol. 50 µl). The LC-MS/MS system consisted of Waters Alliance 2795 high-performance liquid chromatography interfaced to Waters Micromass Quattro Micro API tandem quadrupole mass spectrometer (Waters, Milford, MA, USA). Chromatographic separation was achieved with a Waters analytical column (Atlantis dC18, 3 µm, 2.1 × 50 mm, part no. 186001291) equilibrated with MeOH:H2O (80:20) containing 0.1% formic acid. Applying a linear gradient of 100% MeOH containing 0.1% formic acid over 3 min eluted the VD metabolites. Flow rate was 0.4 ml/min, and total cycle time on LC-MS/MS was 8 min. The tandem mass spectrometer used positive electrospray ionization at an operating voltage of 3.5 kV, and a desolvation temperature of 350°C. The instrument was operated in MRM (multiple-reaction monitoring) mode, with the following transitions: m/z+ 413.15/395.2 for 25-OH VD2, m/z+ 401.15/383.15 for 25-OH VD3, and m/z+ 407.2/389.3 for [2H6]-25-OH VD3. The inter-assay coefficients of variation (CV) for 25-OH VD3 were 2.2 and 2.8% at 30 and 180 nmol/l, respectively, and for 25-OH VD2 were 7.6 and 4.6% at 43 and 150 nmol/l, respectively.

Serum FSH levels were determined using a time-resolved immunofluorometric assay (Delfia; Wallac, Turku, Finland) and Inhibin-B by a specific two-sided enzyme linked immunoassay (Bio-Innovation Ltd, Oxford, UK). The intra- and inter-assay variation for measurements of FSH and Inhibin B were 3 and 4.5%, 15 and 18%, respectively. Furthermore, S-albumin (CV 1.8%), S-Calcium (CV 1.2%), and S-Alkaline Phosphatase (CV 1.3%) were measured on Cobas c501, and S-Parathyroid hormone (PTH) (CV 3.4%) on Cobas e601. All procedures performed as recommended by the manufacturer (Roche). Albumin-corrected calcium was calculated (Total S-calcium + (0.0166 × (39,9 g/l − S-albumin (g/l)). All serum assessments were done at the same time to reduce the influence of inter-assay variation.

Semen analysis

The participants produced a semen sample by masturbation. Self-reported information of duration of ejaculation abstinence was obtained. Trained technicians conducted semen analysis and a detailed description of assessment of semen samples has previously been published (Jørgensen et al., 2001, 2002). Semen volume was estimated by weighing. For sperm motility assessment, duplicates of 10 µl of well-mixed semen were placed on a glass slide, examined on a heating stage kept at 37°C, under a microscope at ×400 magnification, and spermatozoa were classified as progressive motile (WHO class A+B), non-progressive motile (class C) or immotile (class D). The average of the two motility assessments was used. For the assessment of the sperm concentration the samples were diluted in a solution of 0.6 mol/l NaHCO3 and 0.4% (v/v) formaldehyde in distilled water, and subsequently assessed using Bürker–Türk hemocytometer. Only spermatozoa with tails were counted (Jørgensen et al., 2001, 2002). Finally, smears were prepared, Papanicolaou stained and spermatozoa morphology assessed according to strict criteria (Menkveld et al., 1990).

Semen samples for in vitro studies

Semen samples for the in vitro studies were obtained from 40 men from the general population, who were investigated between October 2009 and December 2010. Their semen samples were assessed as described above. To investigate the effect of VD on sperm motility, the semen samples were exposed to 1 nM 1,25(OH)2D3 for 45 min and motility scored as described earlier. Motility assessment was conducted in 17 samples (in duplicates) under non-capacitating conditions (in their seminal plasma), and 3 samples were also analyzed in capacitating conditions. Seven consecutively included men were used to investigate a dose–response relationship between VD and sperm motility.

Calcium measurements and the acrosome reaction were performed on motile spermatozoa isolated after percoll density gradient centrifugation (Supra sperm gradients, (ORIGO)) following the manufacturer's instructions. Briefly, spermatozoa were separated from seminal plasma by centrifugation on a discontinuous percoll gradient using a two-step gradient comprising 2 ml layers of 55 and 80% percoll respectively. Semen was placed on the top of the gradient and centrifuged at 300g for 25 min. The spermatozoa at the base of the 80% fraction were collected and washed with a 5 ml volume of capacitating medium (CaCl2 1,5 mM, KCl 4.3 mM, NaHCO3 31 mM, HEPES 15 mM, Human Serum Albumin 5 mg/ml, Glucose 4.5 mM, Sodium Pyruvate 0.8 mM, Synthetic Serum Replacement, 0.5 ug/ml human recombinant insulin (Sperm preparation media (ORIGO)) and finally resuspended in capacitating media and used for the following assays of calcium measurements and acrosome reaction. All men were part of the above mentioned ongoing surveillance program of semen quality among young Danish men and included after approval from the local ethical committee (No. H-4-2010-016).

Effect of VD on calcium levels in human spermatozoa

Semen samples were separated by percoll density gradient centrifugation to ensure viable and motile spermatozoa and resuspended in capacitating media for 1.5 h. Afterwards, loaded with 3 µM of the fluorescent probe (intracellular calcium binding) fura-2/AM (Invitrogen, USA) for 30 min, centrifuged (10 min at 300 g) and washed with Krebs–Ringer buffer (120 mM NaCl, 15 mM NaHCO3, 5 mM KCl, 1 mM CaCl2, 5 mM Na2HPO4, 1 mM MgCl2 pH 7.3). In a Ca-free Krebs–Ringer solution CaCl2 was substituted for 5 µM EGTA (calcium-free with EGTA). Subsequently, 80 µl of the sperm sample was diluted in 920 µl Krebs–Ringer buffer and loaded on polylysine-coated chambers to minimize movement of the spermatozoa (Lab-Tek; Nalge Nunc International, USA). A Zeiss Axiovert 135 microscope equipped with a Zeiss Achrostigmat 40 × 1.3 NA objective was used to acquire images from the fluorescent probe. Excitation was obtained by a Polychrome Villuminator from Till Photonics (Germany) and images were acquired using a Cool Snap CCD camera (Photometrics, USA) from Robert Scientific (Malaysia). For measurements of [Ca2+]i, the excitation wavelengths were 338 and 380 nm, measuring emission above 510 nm using a cutoff filter. Calculations of [Ca2+]i were conducted by MetaFluor software from Molecular Devices and using a Kd of 160 nM (Gromada et al., 1996). After background subtraction, ratio images were formed using MetaFluor software. Chemicals, 1α,25(OH)2D3 (1,25(OH)2D3), 25-OH D3, progesterone, thapsigargin, nifedipine and ionomycin were purchased from Sigma-Aldrich (USA), PL were carried out in triplicates C-inhibitor U73122 (1-[6-[[17fi-3-methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-lH-pyrrole-2,5-dione) from Biomol, PA, U.S.A. The non-genomic VDR antagonist 1β,25-(OH)2D3 was a gift from LEO Pharma (Denmark).

All experiments started by using low magnification to ensure visibility of several spermatozoa simultaneously. When a color change was observed in most cells, higher magnification was used. All measurements of [Ca2+]i were based upon evaluation of at least three spermatozoa simultaneously (see exact number in figure legends), and were repeated in at least three separate experiments. We used 10 different semen donors to ensure that all experiments were reproduced on semen samples from at least two different men.

Acrosome reaction

Acrosome status was assessed according to a modified version of the method described by Cross et al. (1986). Motile spermatozoa were isolated by percoll gradient centrifugation, washed twice with 5 ml 0.9% sodium chloride followed by centrifugation at 300g for 10 min. Hereafter, the motile spermatozoa were resuspended with or without 1 nM 1,25(OH)2D3 for 1.5 h at room temperature in 0.9% sodium chloride or resuspended in capacitating media for 1.5 h at 37°C and 5% CO2 before the sperm was smeared on a glass slide and a cytospin was created. The smear and cytospin were fixed in 96% ethanol for 5 min after air drying, washed in distilled water and stained for minimum 2 h with Pisum sativum agglutinin labeled with fluorescein isothiocyanate (Sigma-Aldrich) in Dulbecco phosphate buffered saline at 4°C. Finally, the slides were washed in distilled water and mounted with Vectrashield (H-1080 Vectar Laboratories) and 200 spermatozoa were counted on each slide using a fluorescence microscope at a magnification of 400×. The mean percentage of acrosome-reacted spermatozoa of the two slides was calculated. When more than half the head of a spermatozoon was brightly and uniformly fluorescent, the acrosome was considered to be intact. Spermatozoa without fluorescence or with a fluorescing band limited to the equatorial segment were considered to be acrosome-reacted. Spermatozoa with a fluorescing band limited to the equatorial segment were named eq-AR. The person assessing the slides was blinded to the treatment of the sample.

Statistics

The cross-sectional cohort of 300 men were stratified according to their 25(OH)D2+3 (25-OH-VD) serum level: <25 nmol/l (deficient), 25–50 nmol/l (insufficient), 50–75 nmol/l (sufficient) and >75 nmol/l (high). For a basic description median and 5–95 percentiles were used. Between-group differences were tested by the non-parametric Kruskal–Wallis test. Correlations between 25-OH-VD (as a categorical variable) and both semen variables and reproductive hormone levels were examined using the non-parametric Spearman's rank test. To test for the effect of 25-OH-VD regression analyses were also performed. 25-OH-VD categories were first entered as a covariate in models having the semen variables or the reproductive hormones as dependent variables. Secondly, 25-OH-VD and 25-OH-VD-squared entered the model as continuous variables. In all regression models, semen volume, sperm concentration, total sperm count, FSH and Inhibin-B were natural logarithm transformed and motility variables logit transformed before analysis to obtain normal distribution of the residuals. The following covariates were tested: duration of ejaculation abstinence, hour-of-day of blood sampling, smoking, BMI, age, medication, fever, time from ejaculation to motility assessment, morphology observer, history of cryptorchidism, birthweight, season, albumin-corrected calcium, covariates with a significance level of 15% or below were included as stated for each analysis in Table I. The results from the in vitro studies were tested using paired t-test (two-tailed). For all analyses, P < 0.05 was considered statistically significant. Analyses were performed using PASW GradPack 18.0 (SPSS Inc., Chicago, IL, USA).

Table I

Characteristics of 300 normal men stratified according to VD serum levels.

Variable Total n 25(OH)D2+3 concentration (nmol/l)
 
P-value 
Median (5–95 percentile) or number (%)  <25 25–50 50–75 >75 
Included men (n300 (100%) 36 (12.0%) 98 (32.7%) 125(41.7%) 41 (13.7%)  
Age (years) 19.0 (18.4–21.8) 19.2 (18.0–24.9) 19.0 (18.4–21.0) 19.0 (18.5–21.9) 19.0 (18.2–23.5) 0.58 
Height (m) 1.81 (1.70–1.94) 1.81 (1.70–1.97) 1.80 (1.69–1.90) 1.82 (1.72–1.95) 1.81 (1.64–1.93) 0.13 
Weight (kg) 75.0 (57.7–95.3) 74.5 (55.2–126.1) 73.3 (55.1–92.5) 75.4 (59.5–93.1) 77.7 (57.8–97.1) 0.10 
BMI (kg/m222.8 (18.7–28.5) 22.8 (17.8–36.7) 21.8 (17.7–28.6) 22.9 (19.1–26.6) 23.0 (18.7–29.6) 0.29 
Testis size (ml) 24 (14–30) 25 (13–33) 23 (14–29) 25 (14–30) 25 (13–30) 0.21 
Duration of abstinence (h) 64 (37–142) 67 (34–172) 61 (38–158) 66 (34–132) 62 (41–112) 0.15 
Daily smokers (%) 131 (44%) 20 (56%) 42 (43%) 52 (42%) 17 (43%) 0.53 
Daily use of medicine (%) 40 (13%) 4 (11%) 12 (12%) 18 (14%) 6 (15%) 0.93 
Fever last 3 months (%) 20 (7%) 0 (0%) 4 (4%) 13 (10%) 3 (7%) 0.09 
Variable Total n 25(OH)D2+3 concentration (nmol/l)
 
P-value 
Median (5–95 percentile) or number (%)  <25 25–50 50–75 >75 
Included men (n300 (100%) 36 (12.0%) 98 (32.7%) 125(41.7%) 41 (13.7%)  
Age (years) 19.0 (18.4–21.8) 19.2 (18.0–24.9) 19.0 (18.4–21.0) 19.0 (18.5–21.9) 19.0 (18.2–23.5) 0.58 
Height (m) 1.81 (1.70–1.94) 1.81 (1.70–1.97) 1.80 (1.69–1.90) 1.82 (1.72–1.95) 1.81 (1.64–1.93) 0.13 
Weight (kg) 75.0 (57.7–95.3) 74.5 (55.2–126.1) 73.3 (55.1–92.5) 75.4 (59.5–93.1) 77.7 (57.8–97.1) 0.10 
BMI (kg/m222.8 (18.7–28.5) 22.8 (17.8–36.7) 21.8 (17.7–28.6) 22.9 (19.1–26.6) 23.0 (18.7–29.6) 0.29 
Testis size (ml) 24 (14–30) 25 (13–33) 23 (14–29) 25 (14–30) 25 (13–30) 0.21 
Duration of abstinence (h) 64 (37–142) 67 (34–172) 61 (38–158) 66 (34–132) 62 (41–112) 0.15 
Daily smokers (%) 131 (44%) 20 (56%) 42 (43%) 52 (42%) 17 (43%) 0.53 
Daily use of medicine (%) 40 (13%) 4 (11%) 12 (12%) 18 (14%) 6 (15%) 0.93 
Fever last 3 months (%) 20 (7%) 0 (0%) 4 (4%) 13 (10%) 3 (7%) 0.09 

For conversion of 25(OH)D from nmol/l to ng/ml, divide by 2.496; For conversion of PTH from pg/ml to pmol/l divide by 11.1. Kruskal–Wallis test was used to test for difference between groups.

Results

Cross-sectional association study

Table I summarizes the basic characteristics of the study subjects and Table II their semen variables, reproductive hormones, VD and other serum variables. The men were examined throughout the year, and median 25-OH-VD-levels differed according to season: 52 nM in spring (March–May, 31%), 66 nM in summer (June–August, 23%), 53 nM in autumn (September–November, 26%) and 34 nM in winter (December–February, 20%). The levels were significantly lower during winter compared with spring (P < 0.0005), summer (P < 0.003) and autumn (P < 0.003).

Table II

Semen and blood variables stratified according to VD serum levels.

Variable Total n 25(OH)D2+3 concentration (nmol/l)
 
Correlation spearman Adjusted regression 
Median (5–95 percentile)  <25 25–50 50–75 >75 
Inhibin B (pg/ml) 189 (91–341) 195 (97–349) 175 (87–298) 194 (102–362) 192 (69–363) 0.37 0.72 
FSH (U/l) 2.4 (0.9–6.2) 2.5 (0.6–7–8) 2.7 (0.8–6.5) 2.3 (0.9–6.0) 2.2 (1.0–6.4) 0.38 0.67 
Semen volume (ml) 3.3 (1.5–6.5) 3.3 (1.6–7.3) 3.1 (1.4–6.5) 3.6 (1.7–6.4) 3.2 (1.2–8.2) 0.96 0.85 
Total sperm number (mill) 164 (25–605) 195 (31–602) 146 (24–521) 196 (20–700) 139 (24–766) 0.89 0.85 
Sperm concentration (mill/ml) 55 (7–175) 56 (5–209) 47 (7–195) 59 (6–184) 56 (7–137) 0.92 0.68 
Sperm motility (%) 67 (37–86) 62 (31–85) 66 (36–86) 67 (34–86) 70 (46–88) 0.049* 0.025* 
Progressive motile sperm (%) 61 (29–82) 56 (25–81) 59 (27–82) 61 (28–82) 64 (41–84) 0.021* 0.020* 
Morphologically normal (%) 7.0 (1.0–16.8) 6.0 (2.0–14.4) 7.0 (1.0–17.2) 7.0 (0.5–19.0) 8.0 (1.5–17–0) 0.07 0.06 
D vitamin total (nmol/l) 53 (18–91) 20 (9–24) 36 (26–48) 62 (51–73) 86 (76–107) – – 
PTH (pg/ml) 28 (13–53) 33 (18–55) 29 (15–59) 27 (13–47) 24 (11–52) <0.0005** <0.0005** 
Albumin (g/l) 51 (47–58) 50 (47–55) 52 (47–60) 51 (47–57) 50 (48–55) 0.61 0.68 
Calcium (mmol/l) 2.6 (2.4–2.8) 2.5 (2.4–2.8) 2.6 (2.4–3.1) 2.5 (2.4–3.0) 2.5 (2.3–2.7) 0.10 0.35 
Alb.cor. calcium (mmol/l) 2.37 (2.23–2.58) 2.37 (2.24–2.56) 2.38 (2.29–2.74) 2.37 (2.22–2.64) 2.36 (2.13–2.46) 0.09 0.14 
Alkaline phosphatase (U/l) 59 (38–93) 56 (31–85) 57 (35–97) 61 (39–93) 61 (40–104) 0.015* 0.09 
Variable Total n 25(OH)D2+3 concentration (nmol/l)
 
Correlation spearman Adjusted regression 
Median (5–95 percentile)  <25 25–50 50–75 >75 
Inhibin B (pg/ml) 189 (91–341) 195 (97–349) 175 (87–298) 194 (102–362) 192 (69–363) 0.37 0.72 
FSH (U/l) 2.4 (0.9–6.2) 2.5 (0.6–7–8) 2.7 (0.8–6.5) 2.3 (0.9–6.0) 2.2 (1.0–6.4) 0.38 0.67 
Semen volume (ml) 3.3 (1.5–6.5) 3.3 (1.6–7.3) 3.1 (1.4–6.5) 3.6 (1.7–6.4) 3.2 (1.2–8.2) 0.96 0.85 
Total sperm number (mill) 164 (25–605) 195 (31–602) 146 (24–521) 196 (20–700) 139 (24–766) 0.89 0.85 
Sperm concentration (mill/ml) 55 (7–175) 56 (5–209) 47 (7–195) 59 (6–184) 56 (7–137) 0.92 0.68 
Sperm motility (%) 67 (37–86) 62 (31–85) 66 (36–86) 67 (34–86) 70 (46–88) 0.049* 0.025* 
Progressive motile sperm (%) 61 (29–82) 56 (25–81) 59 (27–82) 61 (28–82) 64 (41–84) 0.021* 0.020* 
Morphologically normal (%) 7.0 (1.0–16.8) 6.0 (2.0–14.4) 7.0 (1.0–17.2) 7.0 (0.5–19.0) 8.0 (1.5–17–0) 0.07 0.06 
D vitamin total (nmol/l) 53 (18–91) 20 (9–24) 36 (26–48) 62 (51–73) 86 (76–107) – – 
PTH (pg/ml) 28 (13–53) 33 (18–55) 29 (15–59) 27 (13–47) 24 (11–52) <0.0005** <0.0005** 
Albumin (g/l) 51 (47–58) 50 (47–55) 52 (47–60) 51 (47–57) 50 (48–55) 0.61 0.68 
Calcium (mmol/l) 2.6 (2.4–2.8) 2.5 (2.4–2.8) 2.6 (2.4–3.1) 2.5 (2.4–3.0) 2.5 (2.3–2.7) 0.10 0.35 
Alb.cor. calcium (mmol/l) 2.37 (2.23–2.58) 2.37 (2.24–2.56) 2.38 (2.29–2.74) 2.37 (2.22–2.64) 2.36 (2.13–2.46) 0.09 0.14 
Alkaline phosphatase (U/l) 59 (38–93) 56 (31–85) 57 (35–97) 61 (39–93) 61 (40–104) 0.015* 0.09 

*P < 0.05, **P < 0.01, VD categories is used as covariate in both analyses reported in Table I. Relevant confounders used in the adjusted regression analyses: Semen volume: duration of abstinence, season and medication; Total sperm number: duration of abstinence, fever, season and medication; Sperm concentration: duration of abstinence, alb.cor. calcium and fever; Motility and Progressive motility: time from ejaculation to motility assessment and alb.cor. calcium; Morphology: duration of abstinence and season; FSH, albumin and alb.cor. calcium: BMI; Inhibin B, calcium and alkaline phosphatase: none.

Correlation coefficients were positive between serum 25-OH-VD and all semen variables and Inhibin-B, although only significantly correlated for progressive motility and total motility (AB Motility: R = 0.134, P = 0.021; ABC Motility: R = 0.114, P = 0.049; Table II). The correlations remained significant for the percentage of progressive motile spermatozoa (AB) (R = 0.124, P = 0.031) and borderline significant for percentage of motile spermatozoa (ABC) (R = 0.112, P = 0.053) when using VD serum levels or VD squared as a continuous variable instead of VD as a categorical variable. 25-OH-VD was inversely correlated with PTH (R = −0.266, P < 0.0005) and positively associated with alkaline phosphatase (R = 0.140, P = 0.015). Calcium and albumin-corrected calcium were not correlated significantly with VD categories (R = −0.099, P = 0.086), but were significantly inversely correlated when using VD as a continuous variable (R = −0.114, P = 0.044), although the calcium level did not differ significantly between the high status and VD deficient group. Interestingly, we found a positive correlation between calcium level (with and without albumin correction) and both progressive and total motility (AB Motility: R = 0.127, P = 0.027; ABC Motility: R = 0.127, P = 0.028).

Figure 1 shows motility, progressive motility and morphology levels according to four categories of serum 25-OH-VD levels. For all three semen variables the between-group differences were only significant when comparing the highest versus the lowest 25-OH-VD as shown in Fig. 1. Adjustment for relevant confounders (Semen volume: duration of abstinence, season and medication; Total sperm number: duration of abstinence, fever, season and medication; Sperm concentration: duration of abstinence, alb.cor. calcium and fever; Motility and Progressive motility: time from ejaculation to motility assessment and alb.corr. calcium; Morphology: duration of abstinence and season; FSH, albumin and alb.cor. calcium: BMI; Inhibin B, calcium and alkaline phosphatase: none) did not change the results markedly for sperm motility, progressive motility or morphology, but the significance level increased for motility (ABC: P = 0.025, AB: P = 0.020) and morphology being borderline significant (P = 0.060). When 25-OH-VD entered the regression model as a continuous squared term the associations remained positive, and was reinforced with lower P-values (ABC: P = 0.026, AB: P = 0.023), whereas the significance level for morphology increased to P = 0.109 (Supplementary data, Fig. S1). Regression analyses did not show any associations between 25-OH-VD and semen volume, total sperm concentration, total sperm count, Inhibin-B or FSH, height, weight, BMI and testis size neither when entered as a categorical or a continuous variable.

Figure 1

Serum levels of 25-OH D2+3 and semen quality variables. (A) Association with sperm motility (ABC-motility). (B) Association with progressive motile spermatozoa (AB-motility). (C) Association with normal sperm morphology. Results are mean and error bars indicate 95% CI. Note broken Y-axis in A and B. * marks significant results with P < 0.05.

Figure 1

Serum levels of 25-OH D2+3 and semen quality variables. (A) Association with sperm motility (ABC-motility). (B) Association with progressive motile spermatozoa (AB-motility). (C) Association with normal sperm morphology. Results are mean and error bars indicate 95% CI. Note broken Y-axis in A and B. * marks significant results with P < 0.05.

VD and intracellular calcium in human spermatozoa

Addition of 1,25(OH)2D3 (100 pM–10 µM) to the motile spermatozoa induced a rapid increase in intracellular calcium concentration ([Ca2+]i) (Fig. 2A–C and Supplementary data, Video S1). The increase was 5–10-fold from baseline calcium level and started within 5 s in the neck region and propagated up in the post-acrosomal region and down through the proximal midpiece (Fig. 2A–C). This effect was reproducible from 100 pM, and the threshold for the observed all-or-nothing response was between 10–50 pM 1,25(OH)2D3 (Fig. 2C). The steep increase in [Ca2+]i was followed by a short plateau phase (up to 50 s for <10 nM 1,25(OH)2D3) until [Ca2+]i restored back to baseline concentration (Fig. 2C). The amplitude did not increase with higher 1,25(OH)2D3 concentrations, but the duration of the plateau phase was longer (50–300 s). The maximum amplitude (max amplitude) of the calcium increase induced by 1 nM 1,25(OH)2D3 varied between the investigated men. It was 5–10-fold higher than baseline calcium level and ∼40–80% of the max amplitude elicited by 10 µM progesterone (Supplementary data, Fig. S2). The effect of progesterone could not be abrogated by pretreatment with 1,25(OH)2D3 (Fig. 2C). The nuclear VDR antagonist (ZK159222, 0.1–100 µM) was unable to abrogate or suppress the effect of 1,25(OH)2D3 treatment (data not shown), but the VD response was blocked completely by pretreatment with 1–40 µM of the non-genomic VDR antagonist 1β,25(OH)2D3 (Fig. 2D).

Figure 2

VD and [Ca2+]i in human spermatozoa. (A) Fura-2 loaded spermatozoa, where color indicates [Ca2+]i level (low to high: blue, green, yellow and red). (B) Rapid change in calcium levels after addition of 1 nM 1,25D3. (C) Corresponding changes in [Ca2+]i after addition of 1 nM 1,25(OH)2D3 and 10 µM progesterone, n = 7. (D) Effect of 1β,25(OH)2D3, 1,25(OH)2D3 and 10 µM progesterone, n = 7. Each trace represents the recorded calcium level in a single spermatozoa.

Figure 2

VD and [Ca2+]i in human spermatozoa. (A) Fura-2 loaded spermatozoa, where color indicates [Ca2+]i level (low to high: blue, green, yellow and red). (B) Rapid change in calcium levels after addition of 1 nM 1,25D3. (C) Corresponding changes in [Ca2+]i after addition of 1 nM 1,25(OH)2D3 and 10 µM progesterone, n = 7. (D) Effect of 1β,25(OH)2D3, 1,25(OH)2D3 and 10 µM progesterone, n = 7. Each trace represents the recorded calcium level in a single spermatozoa.

Pretreatment with nifedipine 10 µM did not affect the response to 1,25(OH)2D3. The increase in [Ca2+]i was initiated rapidly with a maximum amplitude 7-fold higher than baseline, which was followed by 4 min long repolarization phase. However, nifedipine treatment caused an aberrant response to progesterone. The rapid initiation and steep increase in calcium seemed unaffected, but the plateau phase was terminated rapidly by a slow repolarization phase, and the calcium level was restored to baseline level 9 min after initiation of progesterone treatment (Figs 2C, D, 3A and Supplementary data, Fig. S2). Treatment with thapsigargin 4 µM caused a slow increase in baseline [Ca2+]i in all spermatozoa (Fig. 3B). Nine minutes after thapsigargin exposure it reached a stable plateau 3-fold higher than baseline calcium concentration. Addition of 1,25(OH)2D3 (9 min after initiation of thapsigargin treatment) to these spermatozoa was unable to elicit a substantial calcium response, while lower concentrations of thapsigargin (<1 µM) were unable to abrogate the effect of 1,25(OH)2D3.

Figure 3

VD and intracellular calcium in human spermatozoa. (A) Spermatozoa pretreated with 10 µM nifedipine before treatment with 1,25(OH)2D3 (1 nM in all experiments) and 10 µM progesterone, n = 20. (B) Addition of 4 µM thapsigargin resulted in a slow increase in calcium, and 1,25(OH)2D3 was added after 9 min exposure to thapsigargin, n = 8. (C) Spermatozoa pretreated with 2 µM PLC-inhibitor U73122 and the corresponding changes in intracellular calcium after addition of 1,25(OH)2D3, n = 13. (D) Effect of 1,25(OH)2D3 in calcium-free media without EGTA, n = 15. (E) Effect of 1,25(OH)2D3, progesterone and ionomycin in calcium-free media with 5 µM EGTA, n = 14. (F): Effect of 1,25(OH)2D3 in calcium-free media with EGTA followed by addition of calcium, n = 3. (A–B) Data presented as mean with s.e.m. (C–F) Each trace represents calcium levels in single spermatozoa. Notice different scale in E and F.

Figure 3

VD and intracellular calcium in human spermatozoa. (A) Spermatozoa pretreated with 10 µM nifedipine before treatment with 1,25(OH)2D3 (1 nM in all experiments) and 10 µM progesterone, n = 20. (B) Addition of 4 µM thapsigargin resulted in a slow increase in calcium, and 1,25(OH)2D3 was added after 9 min exposure to thapsigargin, n = 8. (C) Spermatozoa pretreated with 2 µM PLC-inhibitor U73122 and the corresponding changes in intracellular calcium after addition of 1,25(OH)2D3, n = 13. (D) Effect of 1,25(OH)2D3 in calcium-free media without EGTA, n = 15. (E) Effect of 1,25(OH)2D3, progesterone and ionomycin in calcium-free media with 5 µM EGTA, n = 14. (F): Effect of 1,25(OH)2D3 in calcium-free media with EGTA followed by addition of calcium, n = 3. (A–B) Data presented as mean with s.e.m. (C–F) Each trace represents calcium levels in single spermatozoa. Notice different scale in E and F.

To test the mechanism downstream to the VDR, we used a phospholipase C (PLC) inhibitor U73122, to investigate whether VDR signaling was mediated by a PLC-dependent IP3 production, which may induce an intracellular calcium release. After exposure to 10 µM U73122, the majority (>80%) of the spermatozoa increased their baseline [Ca2+]I more than 5-fold, most were immotile and were unresponsive to treatment with VD. To avoid bias from dying spermatozoa, we changed the concentration of U73122 to 2 µM, where most of the spermatozoa remained low and did not fluctuate in [Ca2+]I and were motile. 1,25(OH)2D3 was unable to induce a rapid increase in [Ca2+]i in spermatozoa exposed to 2 µM U73122. The increase in [Ca2+]i was instead very slow with long-lasting (several minutes) increase unlike the normal rapid response, where the max amplitude was achieved within 1min. The maximum amplitude was also lower, while the plateau phase was longer caused by a delayed repolarization (Fig. 3C). PLC inhibition was unable to abrogate the effect of progesterone (data not shown). We observed a marked response to 1,25(OH)2D3 in a calcium-free media ([Ca2+] <5 μM) without EGTA (Fig. 3D). The response to 1,25(OH)2D3 was aberrant in this media. Here, we observed a longer plateau phase and a delayed repolarization phase compared with the normal calcium containing media. However, 1,25(OH)2D3 or progesterone could not induce an increase in [Ca2+]i in calcium-free media with 5 µM EGTA, despite the spermatozoa being motile, low in baseline [Ca2+]i and exposed shortly (<20 min) to low concentrations of EGTA (Fig. 3E). Ionomycin (500 nM) induced a low but reproducible increase in [Ca2+]i in the calcium-free media with EGTA (Fig. 3E). The amplitude of the response to ionomycin was low (<3-fold from baseline), indicating that the intracellular Ca2+ stores could be partially emptied. Addition of 1 mM Ca2+ to the calcium-free media after treatment with either 1,25(OH)2D3 or progesterone resulted in an increase in [Ca2+]I, which was comparable to the increase observed in the normal calcium containing Krebs–Ringer solution although the plateau phase was longer in the spermatozoa exposed initially to a calcium-free media (Fig. 3F). 25-OH D3 (100 nM) was unable to elicit an increase in [Ca2+]i. Resuspending the motile spermatozoa in non-capacitating media resulted in a lower amplitude to 1 nM 1,25(OH)2D3 treatment, but a comparable response pattern to the effects observed in capacitating media (data not shown). In both media few spermatozoa were unresponsive to 1,25(OH)2D3 treatments.

VD induces sperm motility and the acrosome reaction in vitro

We investigated a dose–response relationship between 1,25(OH)2D3 and the induction sperm motility (ABC motility). We tested the effect of 1,25(OH)2D3 in the concentration range from 10−15 to 10−3 M and found a non-linear dose–response relationship. We found a modest increase in median sperm motility between 2 and 7% in the concentration range from 10−14 to 10−9 M (Fig. 4A). The concentration range from 10−12 to 10−9 M induced the highest increase in motility, and the strongest induction was found for 10−9 M with a median increase of 7% (P = 0.006). Concentrations above 10−7 M resulted in a decreased motility compared with control (10−5 M, median 15% decrease). We extended this by studying semen samples from 17 men with and without exposure to 10−9 M 1,25(OH)2D3 for 30 min at room temperature between 1 and 3 h after ejaculation. Their mean sperm motility increased significantly (P = 0.010) after VD exposure with 2.9%, 95% CI [1%;5%]. Most men had a modest increase in sperm motility although a subset (n = 3) of the men had a slight decrease in motility (Fig. 4B). 1,25(OH)2D3 (10−9 M) significantly (P = 0.024) increased the number of acrosome-reacted spermatozoa with an average increase of 6%, 95% CI [1%;10%] (median, control group 12% and VD group 21%), while the percentage of spermatozoa with an equatorial band (eq.AR) increased from a median of 4.7–5.3% (P = 0.214; Fig. 4C). Six men were investigated in non-capacitating conditions (NaCl), and there was no significant difference between VD treated and the control group (4.2 versus 3.7%).

Figure 4

VD induces sperm motility and the acrosome reaction. (A) Dose–response relationship between 1,25(OH)2D3 and sperm motility (ABC) under non-capacitating conditions. Data presented as mean and error bars indicate s.e.m. (B) Motility assessment in 17 normal men with and without addition of 1 nM 1,25(OH)2D3, thick horizontal bars indicate mean. (C) Acrosome-reacted spermatozoa isolated after percoll gradient centrifugation treated with or without 1 nM 1,25(OH)2D3, under capacitating conditions. EqAR: percentage of spermatozoa with equatorial band following PSA staining. AR: percentage acrosome-reacted spermatozoa following PSA staining. Data presented as mean and error bars represents 95% CI.

Figure 4

VD induces sperm motility and the acrosome reaction. (A) Dose–response relationship between 1,25(OH)2D3 and sperm motility (ABC) under non-capacitating conditions. Data presented as mean and error bars indicate s.e.m. (B) Motility assessment in 17 normal men with and without addition of 1 nM 1,25(OH)2D3, thick horizontal bars indicate mean. (C) Acrosome-reacted spermatozoa isolated after percoll gradient centrifugation treated with or without 1 nM 1,25(OH)2D3, under capacitating conditions. EqAR: percentage of spermatozoa with equatorial band following PSA staining. AR: percentage acrosome-reacted spermatozoa following PSA staining. Data presented as mean and error bars represents 95% CI.

Discussion

The positive association between VD serum level and sperm motility supports the observed in vitro effects of activated VD, and indicates that VD may have a role in human sperm function. The associations to sperm motility and progressive motility were modest, but significant and were best described using a squared term, indicating that the potential effect of VD is probably most pronounced in men with VD deficiency as suggested in the animal deficiency models (Kwiecinski et al., 1989; Uhland et al., 1992). Calcium alone restored the fertility potential in VD deficient rodents (Uhland et al., 1992) and calcium was positively associated with sperm motility, so we tried to separate the effect of VD and calcium by controlling for calcium in the regression analyses. The associations to sperm motility were strengthened, indicating a direct effect of VD on human spermatozoa.

Serum VD level has a relatively short half-life and is primarily influenced by the magnitude and duration of UV-B radiation (Holick, 2008). To oppose bias, the cohort was recruited throughout the year in different seasons with variable sun exposure (Lips, 2006). Forty-four percent of our healthy young men had VD levels below 50 nM, comparable with previous studies conducted on healthy Danes from the general population, indicating that this cohort appear representative for young Danish men (Mosekilde et al., 2005; Frost et al., 2010; Hey, 2010). Other potential confounders besides calcium such as, BMI, weight, season, birthweight, fever, medication, duration of abstinence, time from ejaculation to motility assessment, morphology observer, medication, smoking and history of genital disease were accounted for in the statistical analyses, and none of these factors explained our findings. A recent study reported on a positive association between VD serum levels and sperm motility, but it was only borderline significant (Ramlau-Hansen et al., 2011). However, this study had some limitations, primarily caused by few men with VD insufficiency or VD deficiency in the cohort. Furthermore, they stratified the men in tertiles rather than dividing them according to clinical relevant VD status groups. Therefore, men in the lowest tertile had VD serum levels ranging from 8 to 62 nM (only 6% having <25 nM), so they compared men with mild VD insufficiency with men having a higher level of VD and found only a borderline statistically significant correlation. In addition, they adjusted sperm motility for eight different confounders (not including calcium) without reporting the significance level for any of covariates. Thus, there is no discrepancy between these recently published findings and ours, and we believe that these results are supportive for a weak positive association between VD serum levels and sperm motility, although the strongest associations are found in men with VD insufficiency or deficiency.

The lower number of motile, especially progressive motile and morphologically normal spermatozoa in VD deficient men are not of clinical importance in men with a normal spermatogenic function. However, it could be an important reversible factor, if these findings also holds true in infertile men, because VD supplements alone was able to restore semen quality in VD deficient rodents (Kwiecinski et al., 1989). Male infertility in humans is a heterogeneous disease, and one may argue that VD is not an etiological factor, but instead a marker of general health (Skakkebaek et al., 1994). On the other hand, our in vitro data showed direct effects of VD in mature spermatozoa, which is supported by the testicular histology in one of the VDR KO mice; characterized by several atrophic tubules and few tubules with complete spermatogenesis, and that normalization of serum calcium alone cannot normalize testis histology (Kinuta et al., 2000). Moreover, mice with induced cryptorchidism improved histology (Johnson score) after treatment with activated VD (Hirai et al., 2009). We found no association between VD serum level and total sperm number in our cohort. This was supported by comparable levels of Inhibin-B and FSH in all VD groups and in accordance with the results published recently (Ramlau-Hansen et al., 2011; Jorgensen et al., 2010a). However, VD deficient and VDR KO mice have low sperm number (Uhland et al., 1992; Sood et al., 1995), and the lack of an association between VD serum levels and sperm production in the human studies could be influenced by several factors: long duration of human spermatogenesis (70–80 days), short VD half-life in serum and serum calcium level. VD deficient rodents develop hypocalcemia and their sperm counts could be restored by calcium supplements alone (Uhland et al., 1992; Kinuta et al., 2000).

Presence of VDR and the VD metabolizing enzymes in the neck of mature spermatozoa indicates that VD may have a function in the female reproductive tract. The concentration of VD progenitors in both female and male reproductive tract has to our knowledge not been shown, although the concentration of 1,25(OH)2D3 in the follicular fluid during in vitro fertilization has been measured and suggested to be cyclic with the highest concentration at ovulation (Potashnik et al., 1992). VD is metabolized locally in the male reproductive tract, and the concentrations of 1,25(OH)2D3 that induced an effect in the in vitro studies are close to the physiological concentrations in serum (≈1 × 10−10 M), other somatic tissues, and to the dissociation constant for VDR ≈1–7 × 10−10 M (Norman and Silva, 2001; Pilz et al., 2010). The rapid increase in [Ca2+]i was initiated in the neck of the spermatozoa, where VDR co-localize with the redundant nuclear envelope (RNE). RNE is thought to be a calcium storage, and the Ca2+ release following 1,25(OH)2D3 treatment was rapid; initiated within seconds and is presumably a non-genomic effect (Ho and Suarez, 2003). Spermatozoa are transcriptionally silent and a non-genomic effect of 1,25(OH)2D3 was confirmed by using the non-genomic VDR antagonist 1β,25(OH)2D3, which abrogated the response to 1,25(OH)2D3 completely (Norman et al., 2004; Galeraud-Denis et al., 2007).

The effects downstream to the VDR were elucidated by showing that nifedipine was unable to affect the VD response, unlike the progesterone response, in which the plateau phase was terminated rapidly by inhibition of the voltage-gated L-type calcium channel in accordance with an earlier publication (Kirkman-Brown et al., 2003). The VD-mediated calcium release from an intracellular calcium storage was supported by showing that thapsigargin diminished the effect of 1,25(OH)2D3, although there exist some controversy regarding the concentrations needed to block the SERCA pumps in human spermatozoa (Bedu-Addo et al., 2008). Naaby-Hansen el al. (2001) demonstrated by double staining that the calcium-binding protein calreticulin reside in the neck of human spermatozoa where it co-localize with the 1,4,5-trisphosphate receptor (IP3R), and this putative calcium storage has been suggested to be important during capacitation (Ho and Suarez, 2003). We have previously shown that the VDR and all the VD metabolizing enzymes are expressed at the neck of human spermatozoa (Blomberg Jensen et al., 2010), and since 1,25(OH)2D3 is a known non-genomic inducer of PLC in other mesenchymal derived tissues, we investigated whether PLC inhibition was able to abrogate the calcium release induced by 1,25(OH)2D3 (Civitelli et al., 1990). The PLC-inhibitor U73122 caused a delayed and diminished response to 1,25(OH)2D3 indicating that activated VDR may signal partly through PLC activation and hence increase IP3, which through activation of IP3R-gated calcium channels in RNE releases calcium from an intracellular storage (Ho and Suarez, 2003; Suarez, 2008). To validate the calcium release from intracellular stores, we investigated the response to VD and progesterone in a calcium-free media without EGTA. 1,25(OH)2D3 increased [Ca2+]i rapidly, indicating that the calcium increase originated from intracellular stores. We also noticed that the phase of repolarization was aberrant resulting in an extended plateau phase, and we suggest that spermatozoa may have a calcium sensing receptor that record changes in extracellular calcium concentration. When using a Ca-free Krebs–Ringer solution with a low EGTA concentration, we noticed that the effect of both 1,25(OH)2D3 and progesterone was abrogated, and ionomycin elicited only a small Ca2+ rise when added to the solution. It has previously been suggested that the IP3-gated calcium storage in the RNE can be rapidly depleted by EGTA. This may explain the low amplitude induced by ionomycin in the calcium-free media, which ultimately could be caused by calcium release from other compartments, such as mitochondria (Norman, 1998; Jimenez-Gonzalez et al., 2006; Publicover et al., 2007; Costello et al., 2009).

The VD-induced calcium increase in human spermatozoa indicates that VD may be involved in the induction of motility, which is supported by the cross-sectional study, the animal models and the proposed function of the calcium storage in the neck of human spermatozoa (Bedu-Addo et al., 2008). A role for hyperactivation remains to be shown, but 1,25(OH)2D3 was able to induce a modest increase in sperm motility, which is in accordance with a previous study showing that VD increased calcium on sperm lysates, induced motility and increased acrosin activy (Aquila et al., 2009). We found that 1,25(OH)2D3 induced a significant higher proportion of motile spermatozoa to undergo the acrosome reaction (under capacitating conditions), which may be caused by propagation of the calcium increase from the neck to the head of the spermatozoa.

In conclusion, we show here that the positive association between VD serum levels and sperm motility is supported by novel functional findings, which indicate that VD may contribute to optimal sperm function. VDR and the VD metabolizing enzymes are expressed in mature spermatozoa, and in vitro experiments revealed that VD through non-genomic VDR activation increased intracellular calcium from an intracellular calcium storage in the neck of human spermatozoa, induced sperm motility and the acrosome reaction. Further studies are warranted to investigate the role of VD in spermatozoa from infertile men.

Supplementary data

Supplementary data are available at http://humrep.oxfordjournals.org/.

Funding

The study was economically supported by Åse og Ejnar Danielsens Foundation, Novo-Nordisk Foundation, Center for Healthy Ageing Foundation (Nordea Fonden), Rigshospitalet and the Danish Agency for Science, Technology and Innovation (grant no. 271070678).

Acknowledgements

We thank Ana Ricci Nielsen, Lene Andersen, Kirsten Jørgensen and all the technicians in the molecular-, histological and hormone laboratories at University Department of Growth and Reproduction for skillful technical assistance. We thank Tore Duvold from Leo Pharma for providing the non-genomic antagonist and the Fertility Clinic at Rigshospitalet for performing the percoll gradient centrifugation.

References

Aquila
S
Guido
C
Middea
E
Perrotta
I
Bruno
R
Pellegrino
M
Ando
S
Human male gamete endocrinology: 1alpha, 25-dihydroxyvitamin D3 (1,25(OH)2D3) regulates different aspects of human sperm biology and metabolism
Reprod Biol Endocrinol
 , 
2009
, vol. 
7
 pg. 
140
 
Bedu-Addo
K
Costello
S
Harper
C
hado-Oliveira
G
Lefievre
L
Ford
C
Barratt
C
Publicover
S
Mobilisation of stored calcium in the neck region of human sperm—a mechanism for regulation of flagellar activity
Int J Dev Biol
 , 
2008
, vol. 
52
 (pg. 
615
-
626
)
Blomberg Jensen
M
Nielsen
JE
Jorgensen
A
Rajpert-de Meyts
E
Kristensen
DM
Jorgensen
N
Skakkebaek
NE
Juul
A
Leffers
H
Vitamin D receptor and vitamin D metabolizing enzymes are expressed in the human male reproductive tract
Hum Reprod
 , 
2010
, vol. 
25
 (pg. 
1303
-
1311
)
Bouillon
R
Carmeliet
G
Verlinden
L
van
EE
Verstuyf
A
Luderer
HF
Lieben
L
Mathieu
C
Demay
M
Vitamin D and human health: lessons from vitamin D receptor null mice
Endocr Rev
 , 
2008
, vol. 
29
 (pg. 
726
-
776
)
Civitelli
R
Kim
YS
Gunsten
SL
Fujimori
A
Huskey
M
Avioli
LV
Hruska
KA
Nongenomic activation of the calcium message system by vitamin D metabolites in osteoblast-like cells
Endocrinology
 , 
1990
, vol. 
127
 (pg. 
2253
-
2262
)
Costello
S
Michelangeli
F
Nash
K
Lefievre
L
Morris
J
Hado-Oliveira
G
Barratt
C
Kirkman-Brown
J
Publicover
S
Ca2+-stores in sperm: their identities and functions
Reproduction
 , 
2009
, vol. 
138
 (pg. 
425
-
437
)
Cross
NL
Morales
P
Overstreet
JW
Hanson
FW
Two simple methods for detecting acrosome-reacted human sperm
Gamete Research
 , 
1986
, vol. 
15
 (pg. 
213
-
226
)
Frost
M
Abrahamsen
B
Nielsen
T
Hagen
C
Andersen
M
Brixen
K
Vitamin D status and PTH in young men: a cross-sectional study on associations with bone mineral density, body composition and glucose metabolism
Clin Endocrinol (Oxf)
 , 
2010
, vol. 
73
 (pg. 
573
-
580
)
Galeraud-Denis
I
Lambard
S
Carreau
S
Relationship between chromatin organization, mRNAs profile and human male gamete quality
Asian J Androl
 , 
2007
, vol. 
9
 (pg. 
587
-
592
)
Gromada
J
Frokjaer-Jensen
J
Dissing
S
Glucose stimulates voltage- and calcium-dependent inositol trisphosphate production and intracellular calcium mobilization in insulin-secreting beta TC3 cells
Biochem J
 , 
1996
, vol. 
314
 
Pt 1
(pg. 
339
-
345
)
Hey
H
Hvilken dosis D-vitamin?
Ugeskr Laeger
 , 
2010
, vol. 
172
 pg. 
2482
 
Hirai
T
Tsujimura
A
Ueda
T
Fujita
K
Matsuoka
Y
Takao
T
Miyagawa
Y
Koike
N
Okuyama
A
Effect of 1,25-dihydroxyvitamin d on testicular morphology and gene expression in experimental cryptorchid mouse: testis specific cDNA microarray analysis and potential implication in male infertility
J Urol
 , 
2009
, vol. 
181
 (pg. 
1487
-
1492
)
Ho
HC
Suarez
SS
Characterization of the intracellular calcium store at the base of the sperm flagellum that regulates hyperactivated motility
Biol Reprod
 , 
2003
, vol. 
68
 (pg. 
1590
-
1596
)
Holick
MF
Sunlight, UV-radiation, vitamin D and skin cancer: how much sunlight do we need?
Adv Exp Med Biol
 , 
2008
, vol. 
624
 (pg. 
1
-
15
)
Holick
MF
Chen
TC
Vitamin D deficiency: a worldwide problem with health consequences
Am J Clin Nutr
 , 
2008
, vol. 
87
 (pg. 
1080S
-
1086S
)
Inpanbutr
N
Reiswig
JD
Bacon
WL
Slemons
RD
Iacopino
AM
Effect of vitamin D on testicular CaBP28K expression and serum testosterone in chickens
Biol Reprod
 , 
1996
, vol. 
54
 (pg. 
242
-
248
)
Jimenez-Gonzalez
C
Michelangeli
F
Harper
CV
Barratt
CL
Publicover
SJ
Calcium signalling in human spermatozoa: a specialized ‘toolkit’ of channels, transporters and stores
Hum Reprod Update
 , 
2006
, vol. 
12
 (pg. 
253
-
267
)
Jørgensen
N
Andersen
AG
Eustache
F
Irvine
DS
Suominen
J
Petersen
JH
Andersen
AN
Auger
J
Cawood
EH
Horte
A
, et al.  . 
Regional differences in semen quality in Europe
Hum Reprod
 , 
2001
, vol. 
16
 (pg. 
1012
-
1019
)
Jørgensen
N
Carlsen
E
Nermoen
I
Punab
M
Suominen
J
Andersen
AG
Andersson
AM
Haugen
TB
Horte
A
Jensen
TK
, et al.  . 
East-West gradient in semen quality in the Nordic-Baltic area: a study of men from the general population in Denmark, Norway, Estonia and Finland
Hum Reprod
 , 
2002
, vol. 
17
 (pg. 
2199
-
2208
)
Jørgensen
N
Liu
F
Andersson
AM
Vierula
M
Irvine
DS
Auger
J
Brazil
CK
Drobnis
EZ
Jensen
TK
Jouannet
P
, et al.  . 
Serum inhibin-b in fertile men is strongly correlated with low but not high sperm counts: a coordinated study of 1,797 European and US men
Fertil Steril
 , 
2010a
, vol. 
94
 (pg. 
2128
-
2134
)
Jørgensen
N
Meyts
ER
Main
KM
Skakkebaek
NE
Testicular dysgenesis syndrome comprises some but not all cases of hypospadias and impaired spermatogenesis
Int J Androl
 , 
2010b
, vol. 
33
 (pg. 
298
-
303
)
Kinuta
K
Tanaka
H
Moriwake
T
Aya
K
Kato
S
Seino
Y
Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads
Endocrinology
 , 
2000
, vol. 
141
 (pg. 
1317
-
1324
)
Kirkman-Brown
JC
Barratt
CL
Publicover
SJ
Nifedipine reveals the existence of two discrete components of the progesterone-induced [Ca2+]i transient in human spermatozoa
Dev Biol
 , 
2003
, vol. 
259
 (pg. 
71
-
82
)
Kwiecinski
GG
Petrie
GI
DeLuca
HF
Vitamin D is necessary for reproductive functions of the male rat
J Nutr
 , 
1989
, vol. 
119
 (pg. 
741
-
744
)
Lips
P
Vitamin D physiology
Prog Biophys Mol Biol
 , 
2006
, vol. 
92
 (pg. 
4
-
8
)
Menkveld
R
Stander
FS
Kotze
TJ
Kruger
TF
van Zyl
JA
The evaluation of morphological characteristics of human spermatozoa according to stricter criteria
Hum Reprod
 , 
1990
, vol. 
5
 (pg. 
586
-
592
)
Mosekilde
L
Nielsen
LR
Larsen
ER
Moosgaard
B
Heickendorff
L
[Vitamin D deficiency. Definition and prevalence in Denmark]
Ugeskr Laeger
 , 
2005
, vol. 
167
 (pg. 
29
-
33
)
Naaby-Hansen
S
Wolkowicz
MJ
Klotz
K
Bush
LA
Westbrook
VA
Shibahara
H
Shetty
J
Coonrod
SA
Reddi
PP
Shannon
J
, et al.  . 
Co-localization of the inositol 1,4,5-trisphosphate receptor and calreticulin in the equatorial segment and in membrane bounded vesicles in the cytoplasmic droplet of human spermatozoa
Mol Hum Reprod
 , 
2001
, vol. 
7
 (pg. 
923
-
933
)
Nangia
AK
Hill
O
Waterman
MD
Schwender
CE
Memoli
V
Testicular maturation arrest to testis cancer: spectrum of expression of the vitamin D receptor and vitamin D treatment in vitro
J Urol
 , 
2007
, vol. 
178
 (pg. 
1092
-
1096
)
Norman
AW
Receptors for 1alpha,25(OH)2D3: past, present, and future
J Bone Miner Res
 , 
1998
, vol. 
13
 (pg. 
1360
-
1369
)
Norman
AW
Silva
FR
Structure function studies: identification of vitamin D analogs for the ligand-binding domains of important proteins in the vitamin D-endocrine system
Rev Endocr Metab Disord
 , 
2001
, vol. 
2
 (pg. 
229
-
238
)
Norman
AW
Mizwicki
MT
Norman
DP
Steroid-hormone rapid actions, membrane receptors and a conformational ensemble model
Nat Rev Drug Discov
 , 
2004
, vol. 
3
 (pg. 
27
-
41
)
Pilz
S
Frisch
S
Koertke
H
Kuhn
J
Dreier
J
Obermayer-Pietsch
B
Wehr
E
Zittermann
A
. Effect of Vitamin D Supplementation on Testosterone Levels in Men
Horm Metab Res
 , 
2010
 
Dec 10. [Epub ahead of print]
Potashnik
G
Lunenfeld
E
Levitas
E
Itskovitz
J
Albutiano
S
Yankowitz
N
Sonin
Y
Levy
J
Glezerman
M
Shany
S
The relationship between endogenous oestradiol and vitamin D3 metabolites in serum and follicular fluid during ovarian stimulation for in-vitro fertilization and embryo transfer
Hum Reprod
 , 
1992
, vol. 
7
 (pg. 
1357
-
1360
)
Prosser
DE
Jones
G
Enzymes involved in the activation and inactivation of vitamin D
Trends Biochem Sci
 , 
2004
, vol. 
29
 (pg. 
664
-
673
)
Publicover
S
Harper
CV
Barratt
C
[Ca2+]i signalling in sperm–making the most of what you've got
Nat Cell Biol
 , 
2007
, vol. 
9
 (pg. 
235
-
242
)
Ramlau-Hansen
CH
Moeller
UK
Bonde
JP
Olsen
J
Thulstrup
AM
Are serum levels of vitamin D associated with semen quality? Results from a cross-sectional study in young healthy men
Fertil Steril
 , 
2011
, vol. 
95
 (pg. 
1000
-
1004
)
Skakkebaek
NE
Giwercman
A
de
KD
Pathogenesis and management of male infertility
Lancet
 , 
1994
, vol. 
343
 (pg. 
1473
-
1479
)
Sood
S
Reghunandanan
R
Reghunandanan
V
Marya
RK
Singh
PI
Effect of vitamin D repletion on testicular function in vitamin D-deficient rats
Ann Nutr Metab
 , 
1995
, vol. 
39
 (pg. 
95
-
98
)
Suarez
SS
Control of hyperactivation in sperm
Hum Reprod Update
 , 
2008
, vol. 
14
 (pg. 
647
-
657
)
Uhland
AM
Kwiecinski
GG
DeLuca
HF
Normalization of serum calcium restores fertility in vitamin D-deficient male rats
J Nutr
 , 
1992
, vol. 
122
 (pg. 
1338
-
1344
)