| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Articles |
a Department of Anatomy and Cell Biology and
b Institute of Immunology, Philipps-University, D-35037 Marburg, Germany
ABSTRACT
Macrophage migration inhibitory factor (MIF), originally described as a T-cell product, has recently been identified in several endocrine organs. In the rat testis, MIF is secreted by the Leydig cells into testicular interstitial fluid that directly contacts Sertoli and peritubular cells. To investigate whether MIF is involved in calcium-dependent signal transduction, we have isolated rat Sertoli and peritubular cells. Despite progress in understanding functional properties of MIF, the molecular mechanism of MIF action in target cells is almost completely unknown. Here we find that recombinant MIF evokes a transient increase in calcium levels in peritubular cells but not in Sertoli cells from dissociated rat testis. Concentrations in the range between 12.5 ng/ml and 120 ng/ml of recombinant MIF were found to be effective, with 50 ng/ml yielding the largest increase in intracellular calcium. Preincubation of MIF with a neutralizing monoclonal antibody specifically blocked the response. Incubation of the peritubular cells in calcium-free buffer clearly decreased the evoked response in intracellular calcium concentration. However, the calcium response was greatly decreased by thapsigargin, an inhibitor of the Ca2+ ATPase of the endoplasmic reticulum. The results strongly indicate that calcium is mobilized from reticulum stores during MIF-mediated signal transduction in the testis. In conclusion, our results provide the first characterization of MIF signal transduction in the testis and suggest that signaling from Leydig cells to peritubular cells through MIF is mediated by receptors coupled to release of intracellular calcium.
Macrophage migration inhibitory factor (MIF) was originally described as a factor derived from the T lymphocyte that recruits derived macrophages to sites of inflammation by preventing their random migration [1]. Although MIF was first thought to be synthesized exclusively by T cells, later studies revealed constitutive expression by a variety of cells and tissues, including many endocrine organs like the testis, adrenal, and pancreas [2–4]. In addition, it was shown as a protein synthesized by cells of the pituitary gland [5] and proposed as a critical mediator of septic shock [6]. This was confirmed by targeted disruption of the MIF gene in mice [7]. The MIF-/- mice were resistant to the lethal effects of high doses of lipopolysaccharide (LPS) or Staphylcococcus aureus enterotoxin B (SEB), suggesting that blockade of MIF action may be a useful therapy for sepsis.
In a previous report, we demonstrated that MIF protein and mRNA are present in the Leydig cells of normal adult rat testis, but that testicular-resident macrophages did not appear to be a source of MIF. We also found that addition of recombinant MIF (rMIF) to cultures of rat seminiferous tubules resulted in decreased secretion of inhibin, suggesting a paracrine role for MIF in Sertoli cell regulation [2]. Moreover, MIF showed a unique compensatory production in testes that were treated with the Leydig cell-ablating toxin ethane dimethane sulfonate (EDS). Testicular MIF mRNA and protein were only marginally reduced by EDS treatment, in spite of the fact that the Leydig cells were completely destroyed within 7 days. Immunohistochemistry revealed that the previously negative Sertoli cells then showed distinct MIF labeling. As Leydig cells repopulated the interstitial space, MIF expression switched from the Sertoli cells back to the Leydig cells [8].
Although a substantial body of data has emerged over the past years about the tissue distribution and pleiotropic roles of MIF outside the immune system, very little is known about its signal transduction within target cells. The identification of signal transduction pathways was certainly hampered by the fact that a MIF receptor has not yet been identified. This study was aimed to determine if MIF can evoke a calcium signaling response as a second messenger pathway. In addition, this approach was used to identify MIF target cells in the rat testis.
MATERIALS AND METHODS
Isolation and Culture of Peritubular Cells and Sertoli Cells
Peritubular cells and Sertoli cells were isolated according to a modified protocol of Hoeben et al. [9]. Juvenile male Wistar rats (between Day 19 and Day 21 postnatal) were obtained from Charles River (Kislegg, Germany). Testes were removed and briefly rinsed several times in medium PBS-A (500 ml Dulbecco's PBS without Ca2+ and Mg2+, but supplemented with 1000 U/ml penicillin and 1000 U/ml streptomycin; all from Gibco, Eggenstein, Germany). After decapsulation, the tissue was minced in PBS-A. Thereafter, the tissue fragments were incubated in PBS-A that included 0.25% trypsin and 10 µg/ml DNase I (Boehringer Mannheim, Germany) for 20 min at 32°C with constant shaking. The enzymatic reaction was stopped with 5 mg/ml trypsin inhibitor in PBS-A (Boehringer Mannheim) and the tubule fragments were allowed to settle for 10–20 min. The pellet was rinsed in 20 ml of PBS-A with 2.5% trypsin inhibitor, followed by six to eight washes in 30 ml PBS-A. Subsequently, the fragmented tubules were incubated for 8–10 min at 32°C in PBS-A with 1 mg/ml collagenase (Boehringer Mannheim), 1 mg/ml hyaluronidase (Sigma, Deisenhofen, Germany), and 10 µg/ml DNase I at 32°C in a shaking water bath. Thereafter, 30 ml of PBS-A was added, and tubule fragments were allowed to settle for about 10–20 min. The supernatant containing the peritubular cells was removed, supplemented with 20 ml RPMI standard medium containing 10% fetal calf serum (Greinder, Frickenhausen, Germany), and centrifuged at 50 x g for 10 min at room temperature. Peritubular cells were seeded at a density of 1.5 x 106/ml on glass coverslips in six-well multidishes (Falcon, Meylan, France). After three passages, purity of the peritubular cells was 95%, and cells were prepared for calcium measurement 3–5 days after the last passage. For the isolation of Sertoli cells, the remaining seminiferous tubule fragments were washed five to six times in PBS-A before incubation in 20 ml of 1 mg/ml hyaluronidase, 10 µg/ml DNase I in PBS-A for 50–60 min at 32°C under constant shaking. Tubule fragments were rinsed five times in PBS-A, resuspended in RPMI without fetal calf serum, and subsequently collected by centrifugation at 50 x g for 10 min. Sertoli cells were cultured in RPMI at a density of about 2.5 x 106/ml on glass coverslips in six-well multidishes. After 3 days, the contaminating germ cells were lysed by hypotonic shock treatment in 20 mM Trid-Cl (pH 7.4) for 90 sec. Sertoli cells were prepared for calcium measurement 2–4 days after hypotonic shock treatment.
Monitoring of Intracellular Free Ca2+ Concentrations
The cells grown as a monolayer on coverslips were incubated at 32°C for 30 min with 3 µM Fura-2-AM (from a 1 mM stock in dimethylsulfoxide; Molecular Probes, Leiden, Netherlands) and 3 µM pluronic F-127. Extracellular free Fura-2-AM was removed by rinsing the cells (3x) with HEPES-buffered salt solution (HBSS, pH 7.4). To allow complete hydrolysis of the intracellular Fura-2-AM, the cells were incubated a further 30 min at 32°C with 5% CO2.
The dye-loaded cells were placed into a measuring chamber and examined with a 100x oil objective and a 10x ocular in an inverted microscope (Nikon, Diaphot 400, Düsseldorf, Germany). The cells were excited at 340 nm and 380 nm using a monochromator system (Deltascan, PTI, Wedel, Germany). Emission intensities were recorded as a ratio (340 nm/380 nm) at 510 nm with a photomultiplier (Deltascan, PTI). The background-corrected ratiometric signal R was calibrated by applying the standard equation
where Rmin and Rmax were determined empirically from cells [10]. All measurements were performed between 15 and 60 min after dye loading.
Thapsigargin and BODIPY FL thapsigargin were purchased from Molecular Probes (Leiden, Netherlands). Stimuli were applied by bath perfusion, as previously described by Wennemuth et al. [11].
Production of Recombinant MIF and Antibody Generation
Reverse transcription–polymerase chain reaction of peripheral blood monocytes was used to amplify the coding sequence for mouse MIF using primers 5'-CGCCATATAGCCTATGTTCATCGTGAAC-3' and 5'-CGGATCCGACTCAAGCGAAGGTGGAAC-3', which incorporated the NdeI (5') and BamHI (3') restriction sites. The cDNA was originally cloned into the pCRII vector (Invitrogen, Groningen, Netherlands), subcloned into the NdeI and BamHI sites of the pET-17b vector (Novagen, Madison, WI) and then sequenced. Following the manufacturer's protocols, protein was expressed in pLysS cells and induced with 0.4 mM isopropyl ß-D-thiogalactoside for 3 h at 25°C. Recombinant MIF was purified from the cell lysate by two-step HPLC: size-exclusion HPLC, column Bio-Sil TSK 250, and ion-exchange HPLC, column Bio-Gel TSK DEAE-5PW (Bio-Rad). Monoclonal antibodies were raised against murine MIF using standard procedures. Monoclonal antibodies from different hybridoma cell lines were tested for their ability to neutralize MIF in solution. The positively tested monoclonal anti-MIF antibody clone 5E3, identified in this manner, was used for the following immunoneutralization experiments.
Immunoneutralization
Various concentrations of rMIF (3.57–50 ng/ml) diluted in HBSS were incubated with excess anti-MIF antibody (clone 5E3) at 4°C overnight under slow and constant shaking. After centrifugation (12 000 x g, 10 min) the supernatant was heated to 32°C and added to the measuring chamber. Samples with rMIF at the same concentrations were treated accordingly.
All experiments were repeated with at least five different preparations of isolated rat testicular peritubular cells. Every dose of rMIF was applied to more than four individual glass slides per cell preparation. All assays used peritubular cells that were trypsinized and freeze–thawed or kept in culture for up to four passages after isolation were assayed. Data were analyzed with the Excel software program (version 97 SR1, Microsoft Co.) and expressed as the mean value of 6–12 individual experiments.
Statistical Analysis
Student's t-test was used for comparing differences in increases in calcium between experimental groups. All statistical analyses were performed using IGOR Pro version 3.1 Software (WaveMetrics, Inc., Lake Oswegon, OR). All values are mean ± SD (n = 12 experiments/group) unless otherwise indicated. A value of P < 0.05 was considered significantly different.
RESULTS
The stimulation of isolated rat testicular peritubular cells with 50 ng/ml rMIF caused a rapid rise of intracellular Ca2+ ions (Fig. 1). The slow, monophasic recovery continued upon the removal of rMIF by perfusion of the measuring chamber with HBSS buffer back to resting levels. A second stimulation of the same cell with rMIF applied after 2 min of recovery produced a similar increase in Ca2+ concentrations (data not shown).
|
Incubation of peritubular cells in increasing concentrations of rMIF (3.57–200 ng/ml) showed a dose-dependent response (Fig. 2). The lowest concentration caused no measurable ratio rise in peritubular cells. However, starting at concentrations of 12.5 ng/ml, a significant rise in ratio was detectable. Increasing concentrations of rMIF resulted in a faster and larger rise in intracellular calcium concentrations. Maximum responses were recorded at 50 ng/ml rMIF (mean 1102 nM ± 61.63 SD). This mean value was set as a control to compare the changes in maximum rise rates or maximum calcium concentrations after the antibody neutralization, thapsigargin treatment, and incubation in calcium-free medium.
|
Interestingly, rMIF doses exceeding 120 ng/ml did not induce any calcium fluxes in these cells. Results were not affected whether freshly isolated peritubular cells that were kept in culture for three to six passages or cells that underwent one freeze–thaw cycle were assayed. A less pronounced Ca2+ spike was observed in rat peritubular cells with more than seven passages in culture. No calcium influx in cultured rat Sertoli cells was observed using rMIF at concentrations between 3.5 ng/ml and 150 ng/ml (Fig. 1b). Consequently, the following experiments were continued with the peritubular cells only.
To investigate the specificity of the MIF induced effect, different MIF neutralizing monoclonal antibodies were used. The ability of these antibodies to bind MIF in solution was prior tested by immunoprecipitation using sepharose-coupled protein G and Western blot analysis. The pretreatment of the MIF-containing samples with the neutralizing monoclonal anti-MIF antibody 5E3 was able to block the MIF effect on peritubular cells (mean 27.36 nM ± 4.77 SD). For the next step the 5E3 antibody was removed from the measuring chamber. Thereafter the ability of the cells to respond to the MIF stimulation was shown after perfusion of the measuring chamber with HBSS supplemented with 50 ng/ml rMIF. However, the rise of intracellular calcium ions was clearly slower and the recovery time was prolonged (Fig. 3). The other four antibodies that did not show MIF immunoneutralization failed to prevent the MIF-induced calcium current.
|
Involvement of Intra- and Extracellular Calcium Stores
To investigate the involvement of intracellular calcium stores, we used thapsigargin, a potent inhibitor with high affinity for the Ca2+-dependent ATPase of the endoplasmic reticulum (ER). The distribution of Ca2+-dependent ATPase was visualized by incubation of the peritubular cell with BODIPY FL thapsigargin (10 µM) for 5 min. Fluorescence images (excitation/emission 494/520 nm) showed a diffuse localization with bright intensity that indicates a high density of ER in peritubular cells (Fig. 4). To deplete the ER as an intracellular calcium store, we used 10 µM thapsigargin as an inhibitor of the Ca2+-dependent ATPase. This caused an increase in the recordable calcium concentrations (Fig. 5). After 60 sec, cells with depleted ER calcium stores were perfused with rMIF (50 ng/ml). Only a weak increase in intracellular calcium concentration was measurable (mean 175.33 nM ± 44.58 SD) (Fig. 5).
|
|
In addition, we perfused the cells with calcium-free HBSS buffer to remove calcium from the external milieu. After an incubation time of 5 sec in calcium-free medium, the testicular peritubular cells were stimulated with 50 ng/ml rMIF. Only a very low increase in the fluorescence ratio was measurable (mean 165.2 nM ± 66.4 SD). By re-adding 2 mM calcium to the buffer a large increase in intracellular Ca2+ concentration was detectable (Fig. 6).
|
Generally, no significant differences were observed when peritubular cells that were either trypsinized and freeze–thawed or kept in culture for up to four passages after isolation were assayed.
DISCUSSION
Many studies over the past years have shown that the cytokine MIF is a pluripotent mediator of the physiological and pathophysiological regulation in diverse tissues. Despite this progress the mechanism of MIF action at the molecular level is almost completely unknown. This is mainly due to the fact that a conventional membrane receptor for MIF has not yet been identified. Ony Mitchell et al. [12] reported an activation of p44/p42 MAP kinase and cytoplasmic phospholipase A2 by MIF in NIH/3T3 fibroblast cells. However, some evidence indicates that the inhibitory effect of MIF on the migration of macrophages is Ca2+ dependent [13–16]. All of these studies find that the removal of calcium from the external medium inhibits the ability of MIF to modulate macrophage migration. In rat testis, MIF was found to be secreted by the Leydig cells and accumulate in considerable amounts into the interstitial fluid [2]. These results led us to ask whether the cell types that have direct access to testicular interstitial fluid possess mechanisms that involve Ca2+ fluxes in MIF signal transduction.
Our findings show that MIF changes the concentration of intracellular free Ca2+ in isolated rat testicular peritubular cells, whereas Sertoli cells do not react. Migration inhibitory factor, at concentrations of 12.5 ng/ml evokes a transient increase in Ca2+ concentrations in rat testicular peritubular cells. The peak amplitude of the Ca2+ spike was dose dependent. The highest rise in intracellular Ca2+ was observed after stimulation with 50 ng/ml of rMIF. Higher doses led to a clearly less pronounced amplitude that disappeared completely at rMIF doses of more than 150 ng/ml. A potential explanation for this phenomenon would be that there are two receptors or one receptor with two sites, a high-affinity agonist site and a lower-affinity antagonist site comparable to the fMetLeuPhe receptor of neutrophils [17]. However, this assumption could not be tested because a receptor for MIF has not been identified. Nevertheless, the MIF concentrations effective in vitro approximate the physiological concentration determined by a specific ELISA at 14.7 ng/ml in rat interstitial fluid [2]. To understand better the molecular transduction pathway for MIF, we have used Ca2+-free medium and peritubular cells pretreated with thapsigargin, a selective inhibitor of the ER Ca2+ ATPase [18]. The addition of rMIF to thapsigargin-pretreated cells evoked a much reduced rise in Ca2+ concentrations. The same observation was made after incubation of the peritubular cells in Ca2+-free medium. Of note, readdition of calcium to the external medium again resulted in a high MIF-induced Ca2+ influx. This is typical for a signal transduction pathway termed capacitative or store-operated Ca2+ entry. The initial phase of the response is usually caused by the activation of phospholipase C-mediated generation of diacyglycerol and inositol 1,4,5-triphosphate (IP3), which provokes a subsequent release of Ca2+ from the ER. This depletion of intracellular Ca2+ stores is followed by the capacitative Ca2+ entry, a vast influx of extracellular Ca2+ into the cytoplasm by the activation. This kind of influx is mediated by so called calcium release activated channels (Icrac) [19]. The phase of Ca2+ mobilization from extracellular sources is required for refilling intracellular stores and for cellular processes that depend on a sustained increase in Ca2+, such as the transcriptional regulation in T cells of interleukin (IL)-2 [20] or the IL-8-mediated Ca2+ flux in leukocytes [21].
Further careful study is warranted to determine the involvement of the different types of calcium channels in MIF signal transduction. Testicular peritubular cells are contractile myoid-like cells that surround the seminiferous epithelium in several layers in most mammals. The contractile ability of peritubular cells results in peristaltic propulsions of the seminiferous epithelium. This mechanism is involved in the vectorial transport of the fluid and immotile spermatozoa to the rete testis and finally into the epididymis. So far, the involvement of endothelin receptors and prostaglandins, both in a calcium-mediated fashion, as well as oxytocin have been described to be involved in this phenomenon [22–24]. Our results are suggestive of a role for Leydig cell–derived MIF in affecting peritubular myoid cell function. This could either involve contractility or the regulation of the proliferation of peritubular cells as the latter is also known to be Ca2+-dependent [25]. Despite the observation that Sertoli cells in tubule cultures respond to MIF stimulation with a decreased secretion of inhibin [2], no calcium flux was observed in these cells. However, whether the effect on Sertoli cells was direct or indirectly mediated via the attached peritubular cells was not assessed.
In conclusion, these are the first data on the intracellular signal transduction of MIF. Due to the general importance of elevations of intracellular free calcium as a second messenger and the pleiotropic functions of MIF in many tissues, our findings raise the possibility that agents disrupting calcium homeostasis may be effective counterregulators of MIF in pathophysiology. In addition, it reinforces the role of MIF as a paracrine mediator in Leydig cell–seminiferous tubule interaction.
ACKNOWLEDGMENTS
The authors are indebted to Mrs. Andrea Dersch for skillful technical assistance, to Dr. D. Babcock for critical reading of the manuscript, and to Dr. B. Müller for statistical analysis.
FOOTNOTES
First decision: 31 August 1999.
1 This study was supported in part by grants of the Fonds der Chemischen Industrie, the Kempkes-Stiftung, Marburg, and the Deutsche Forschungsgemeinschaft (We 2344/1-1, Me 1323/2-1). 
2 Correspondence: G. Wennemuth, Department of Anatomy and Cell Biology, Philipps-University, Robert-Koch-Str. 6, D-35037 Marburg, Germany. FAX: 49 6421 2865783; wennemut{at}mailer.uni-marburg.de
Accepted: February 1, 2000.
Received: July 1, 1999.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  T. K. Monsees, S. Blocher, F. Heidorn, A. Winkler, W.-E. Siems, W. Muller-Esterl, J. Hayatpour, W. Miska, and W.-B. Schill Expression and Location of the Bradykinin B2 Receptor in Rat Testis Biol Reprod, December 1, 2002; 67(6): 1832 - 1839. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Arndt, G. Wennemuth, P. Barth, M. Nain, Y. Al-Abed, A. Meinhardt, D. Gemsa, and M. Bacher Release of Macrophage Migration Inhibitory Factor and CXCL8/Interleukin-8 from Lung Epithelial Cells Rendered Necrotic by Influenza A Virus Infection J. Virol., August 12, 2002; 76(18): 9298 - 9306. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
