| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
-Tubulin1
a Department of Environmental Toxicology, University of California, Davis, California 95616
ABSTRACT
In the testis, microtubule-disrupting agents cause breakdown of the Sertoli cell cytoskeleton and sloughing of germ cells with associated Sertoli cell fragments, although the mechanism underlying this event is not understood. In this study, we investigated the effects of carbendazim and colchicine on microtubule polymerization status and posttranslational modifications of tubulin in freshly isolated rat seminiferous tubules. Soluble and polymerized tubulin pools were separated and tubulin was quantified using a competitive ELISA. Carbendazim and colchicine caused extensive microtubule depolymerization, shifting the ratio of soluble to polymerized tubulin from 40%:60% to 78%:22%, and to 84%:16%, respectively. Total tubulin levels remained relatively constant after carbendazim treatment but decreased twofold after colchicine treatment. To determine if modifications to tubulin may be associated with polymerization status, tubulin pools were analyzed by immunoblotting. Acetylated
-tubulin and ßIII-tubulin distribution in tubulin pools was not affected by treatment. Tyrosinated -tubulin (52 kDa) was localized in both tubulin pools and had decreased tyrosination in the microtubule pool after carbendazim treatment. A 47-kDa protein immunoreactive with both tyrosinated -tubulin and general -tubulin antibodies was found only in the microtubule pool. The 47-kDa protein (potentially an -tubulin isoform) lost tyrosination, yet was still present in the microtubule pool based on detection with the general -tubulin antibody, after carbendazim treatment. Similar effects were seen with colchicine, although loss of total tubulin protein was measured. Thus, decreased tyrosination of the microtubule pool of tubulin appears to be associated with depolymerization of microtubules.
Sertoli cells, spermatogenesis, testis, toxicology
INTRODUCTION
Microtubules are structural elements that make up an essential part of the cellular cytoskeleton and play a central role in cellular processes such as mitosis and organelle transport [1]. They are maintained in cells in a dynamic equilibrium of soluble subunits and polymerized microtubules, the ratio of which is thought to be specific to cell type [2] and may be regulated by a number of physiological factors [1, 3]. Microtubules are composed of heterodimers of two similar protein subunits, - and ß-tubulin, both of which exist as multiple isoforms encoded by multigene families [4–7].
Heterogeneity of - and ß-tubulin isoforms is also due to multiple posttranslational modifications that occur almost exclusively at the hypervariable C-terminus of the proteins [5, 7, 8]. These modifications include acetylation and tyrosination-detyrosination of -tubulin, and phosphorylation, polyglutamylation, and polyglycylation of - andß-tubulin [7–9]. Despite many studies investigating a specific role for the tyrosination and detyrosination of -tubulin, the functional significance of this posttranslational modification has not been conclusively determined [9]. In recent studies, detyrosination of -tubulin was suggested as a possible signal to regulate interactions of vimentin filaments with microtubules in fibroblasts [10–12].
The seminiferous epithelium of the mammalian testis is composed of Sertoli cells, developing germ cells, and peritubular myoid cells with Leydig cells in the interstitium. The Sertoli cell cytoskeleton is comprised of an extensive network of microtubules, arranged parallel to the long axis of the cell [13–17]. Sertoli cell microtubules have an unusual polarity in that the plus, or fast-growing ends, are oriented perinuclear while the slow-growing ends are in the apical region of the cell [18]. In addition to maintaining cell shape, Sertoli cell microtubules are important for assisting in the movement of spermatids from the basal to luminal surface of seminiferous tubules [18, 19]. Other prominent elements of the Sertoli cell cytoskeleton are actin microfilaments and vimentin intermediate filaments that interact with and assist microtubules in the movement of maturing germ cells.
Several microtubule-disrupting agents, e.g., colchicine, carbendazim, and vinblastine, induce pathological changes in the rat testis including the loss or sloughing of whole, flagellar-intact sperm cells plus apical portions of Sertoli cells [16, 20–23]. This sloughing of the apical seminiferous epithelium occurs concomitantly with an apparent breakdown or rupture of Sertoli cell microtubules based on immunostaining and histological studies [16, 20–23]. The rupture of Sertoli cell microtubules may be due to depolymerization based on the well-described ability of colchicine to block microtubule assembly [24, 25]. However, the biochemical basis for breakdown of the microtubular cytoskeleton and subsequent loss of the apical seminiferous epithelium after treatment with microtubule-disrupting agents is not clear.
In the present study, we have used two microtubule-disrupting agents, carbendazim and colchicine, to investigate tubulin polymerization status in the seminiferous epithelium and the potential relationship of posttranslational modifications to microtubule polymerization. An in vitro model system has been developed for short-term treatment of isolated seminiferous tubules with disrupting agents, followed by isolation of soluble and polymerized tubulin pools. The levels of soluble (unpolymerized) and polymerized tubulin and composition of the tubulin pools were analyzed using a competitive ELISA and immunoblotting techniques. Specific posttranslationally modified forms of tubulin were characterized in the tubulin pools, to look for changes that may be associated with microtubule polymerization status.
MATERIALS AND METHODS
Materials and Media
Medium used for seminiferous tubule culture was Dulbeccos modified Eagle medium (DMEM-F12, containing 15 mM Hepes buffer, L-glutamine, pyridoxine hydrochloride; Gibco BRL, Rockland, MD) supplemented with 10 µg/ml insulin, 5 µg/ml apotransferrin, and 5 µg/ml gentamycin sulfate (referred to as DMEM+). Chemicals were purchased from Sigma unless otherwise indicated.
Seminiferous Tubule Isolation and Culture
Adult male Sprague-Dawley rats (
350 g) were purchased from Charles River Laboratories (Wilmington, MA). Animals were maintained and used in accordance with the National Research Council publication, Guide for Care and Use of Laboratory Animals (copyright 1996, National Academy of Sciences). Seminiferous tubules were isolated and incubated using culture conditions modified from the method of Parvinen et al. [26]. Soluble and polymerized tubulin pools were fractionated by modifications of previously described protocols [27, 28]. Testes were removed from one rat for each experiment and placed in DMEM+; testes were decapsulated and bisected. Each testis was incubated in 20 ml of digestion medium (DMEM+ containing 1 mg/ml collagenase type IV) for 5 min at 32°C/50 rpm in a covered water bath to remove the interstitial cells. The digested tubules were washed five times with DMEM+, then the tubules were separated by pulling apart with fine forceps and cut with a razor blade into 5- to 10-mm pieces. All of the seminiferous tubule pieces were aliquotted equally into four flasks containing 75 ml DMEM+ (total volume). The four samples were preincubated for 30 min at 32°C/50 rpm under a 5% CO2/95% air atmosphere, and treated as follows: 1) no treatment (NT); 2) 0.5% dimethylformamide (DMF); 3) 100 µM carbendazim (CBZ) dissolved in DMF (final DMF concentration was 0.5%); 4) 10 µM colchicine (COL) dissolved in water. The samples were incubated at 32°C/50 rpm as above. Incubations in separate experiments were for either 30 min or 1 h, and experiments for each incubation period were repeated three to four times.
Isolation of Soluble and Polymerized Tubulin
The seminiferous tubules were pelleted in a tabletop centrifuge at 700 x g for 1 min at room temperature. Supernatants were discarded and microtubule stabilization buffer (MSB; 0.1 M Pipes, 1 mM EGTA, 1 mM MgSO4, 1 mM dithiothreitol [DTT], 5 mM GTP, 30% glycerol, 5% dimethyl sulfoxide, 1% aprotinin, pH 6.9) was added to the pellets to give a total volume of 5 ml. Tubules were homogenized (using conditions optimized to give maximal disruption of cells by visual observation) with a glass homogenizer and teflon pestle, followed by short bursts of sonication. Aliquots of homogenates were saved for determination of total protein, using a bicinchoninic acid protein assay kit (Sigma, St. Louis, MO) with BSA as the standard. Homogenates were centrifuged immediately at 130 000 x g (Beckman L-70 ultracentrifuge, Ti 70.1 rotor) for 1 h at 30°C. Supernatants containing the soluble or unpolymerized pool of tubulin, called the SN-I fraction, were stored at -80°C. Pellets were resuspended to 5 ml with ice-cold depolymerization buffer (0.1 M 2-[4 morpholino]-ethane-sulfonic acid [MES], 1 mM MgSO4, 10 mM CaCl2, 1 mM DTT, 1 mM GTP, 1% aprotinin, pH 6.9) and homogenized. Homogenates were incubated for 1 h on ice to allow depolymerization of microtubules, then centrifuged at 100 000 x g for 30 min at 4°C. The supernatants representing the polymerized or microtubule pool of tubulin, called theSN-II fraction, were stored at -80°C. The final pellets (PII) were resuspended to 5 ml with depolymerization buffer, homogenized, and stored at -80°C.
Quantitation of Soluble and Polymerized Tubulin Pools in the Seminiferous Epithelium
A competitive ELISA used for quantitatively determining the amount of soluble and polymerized tubulin in cells was based on a previously described method [28], with some modifications. This technique involved the use of two plates (96-well, high binding capacity; Corning Costar, Corning, NY): an incubation plate and a quantitation plate. Dilutions of tubulin standard or sample were preincubated with tubulin antibody in the incubation plate. The mixtures were added to the quantitation plate that was precoated with standard tubulin, and absorbance was measured. If a high concentration of tubulin was present in a given sample, less antibody would be available to bind tubulin coated on the quantitation plate; thus, the absorbance measurement would be low. The amount of tubulin in each sample was quantitated using a standard curve for known amounts of standard tubulin.
Tubulin standard (ICN, Costa Mesa, CA) was solubilized in MEM buffer (0.1 M MES, 1 mM EGTA, 1 mM MgSO4, pH 6.9; final concentration 3 x 10-3 µg/µl), and 100-µl aliquots were coated on quantitation plates overnight at 4°C. Incubation plates were blocked by adding 150 µl of blocking buffer (0.1 M MES, 1 mM MgSO4, 0.1% BSA, 0.05% Tween-20, pH 6.9) per well and incubating for 1 h (all incubations were at room temperature). Incubation plates were washed three times with Tris-buffered saline (TBS)-Tween (0.05% Tween-20 in TBS, pH 7.4); all subsequent washes were three times with TBS-Tween. Tubulin standard was solubilized in MSB (50 µg/ml); then standard, SN-I, and SN-II fractions were diluted (in 0.1 M MES, 1 mM MgSO4, 10 mM CaCl2, 0.2% BSA, 0.05% Nonidet P-40); and 60-µl aliquots were added in duplicate to the wells of the incubation plate. Mouse ß-tubulin monoclonal antibody (Roche, Indianapolis, IN) was diluted 1:100 in TBS-Tween, and 60-µl aliquots were added to the wells containing standard or samples and incubated for 1 h.
Tubulin-coated quantitation plates were washed, and nonspecific protein binding was blocked by adding 200 µl of blocking buffer to the wells, incubating for 45 min, and washing again. Next, 100-µl aliquots were transferred from the wells of the incubation plate to the wells of the quantitation plate and incubated for 1 h. The quantitation plate was washed three times using a Nunc Immuno-wash 12 (Nalge Nunc Int., Naperville, IL). Aliquots (100 µl) of goat anti-mouse-horseradish peroxidase (diluted 1:25 000 in TBS-Tween) were added to the plate, incubated for 1 h, and then washed. Substrate (100 µl of 0.05 M citrate, 0.4 mM 2'2-azino-bis[3-ethylbenzthiazoline-6-sulfonic acid], 1.6 mM H2O2, pH 4.0) was added to the quantitation plate and incubated approximately 40 min to allow color to develop. Absorbance was measured using an EL 340 Microplate Bio Kinetics Reader (Bio-Tek Instruments, Winooski, VT) with a 405-nm filter. Wells containing substrate only served as a baseline for comparison.
Seminiferous Tubule ATP Measurements
To estimate the viability of the seminiferous tubules during the incubation periods, ATP levels were determined as previously described [29] using the luciferin/luciferase system (LKB Wallac; PerkinElmer, Foster City, CA) and a model TD 20e luminometer (Turner Designs, Sunnyvale, CA). Samples were taken from seminiferous tubules before addition of CBZ or COL and after 1-h incubations. Proteins were precipitated with 70% ethanol, and ATP levels were measured in the ethanolic supernatant.
Immunoblot Analysis
The composition of soluble and polymerized tubulin pools was analyzed by immunoblotting. Equal volumes of 2x Laemmli sample buffer were added to SN-I, SN-II, and PII fractions (obtained after 1 h exposure to disrupting agents) and heated at 95°C for 5 min. In representative experiments, the protein concentrations measured in initial homogenates of seminiferous tubules (before separation of tubulin pools) for no treatment, DMF, CBZ, and COL samples were 7.7, 7.5, 7.2, and 7.7 mg total protein/ml, respectively. Gels were loaded with: initial homogenates and SN-I fractions, 1.3 µl; SN-II and PII fractions, 10 µl. Samples were separated by SDS-PAGE on 7.5% polyacrylamide gels [30]. Molecular weight markers were purchased from Amersham-Pharmacia (Piscataway, NJ). Following electrophoresis, proteins were transferred to nitrocellulose membranes (0.45 µm; Bio-Rad Laboratories, Richmond, CA) using a semidry transfer apparatus (Bio-Rad). The membranes were blocked with 5% nonfat dry milk and 2% BSA in TBS-Tween for 1 h and 15 min.
Primary antibodies diluted in TBS-Tween were incubated with membranes for 1 h. TBS-Tween without primary antibody was used as a control for all antibodies. The antibodies (and dilutions) used were mouse monoclonal antibodies anti--tubulin (clone B512, 1:3000), anti-acetylated tubulin (clone 611B1, 1:2000 dilution), anti-tyrosinated tubulin (clone TUB-1A2, 1:10 000), anti-ß-tubulin isotype III (1:400; all from Sigma); and mouse monoclonal anti-ß-tubulin (1:400; Roche). Alkaline phosphatase-conjugated secondary antibody and nitroblue tetrazolium/5-bromo-4-chloro-3-indoylphosphate toluidinium substrate (Promega Biotech, Madison, WI) were used to detect bound primary antibodies.
Statistics
The Student t-test was used to compare the means obtained for soluble and polymerized tubulin isolated from the four different samples. P values <0.05 were considered to be statistically significant. The molecular masses for immunoreactive bands indicated on immunoblots represent the means from three to seven experiments.
RESULTS
Depolymerization of Microtubules in Isolated Seminiferous Tubules After Treatment With CBZ or COL
In order to investigate the effect of CBZ and COL on the polymerization status of microtubules in rat seminiferous epithelium, an in vitro model system was developed for short-term culture of seminiferous tubules, followed by separation and quantitation of soluble tubulin and microtubule pools. The isolation protocol included digestion with collagenase to dissociate interstitial tissue from seminiferous tubules followed by several wash steps to remove dissociated cells and sperm in the lumen. The procedure for separation of soluble and polymerized tubulin is outlined in Figure 1. Integral to this approach was stabilization of the microtubule pool so that no net change in polymerization state occurred during the incubation or isolation protocols, allowing any changes caused by disrupting agents to be evaluated. This was verified by including a control in which seminiferous tubules were not treated, and a control for the vehicle (DMF) used to solubilize CBZ, in every experiment. After incubation with CBZ or COL, the effect on polymerization state was determined using a competitive ELISA to quantitate the tubulin from soluble and polymerized pools.
|
The results from this set of experiments are illustrated in Figure 2. The amount of tubulin measured in the soluble SN-I fraction from no treatment seminiferous tubules was the same at both the 30-min and 1-h incubation periods, representing between 38% and 40% of total tubulin detected (Fig. 2A). The corresponding values for polymerized tubulin (SN-II) from no treatment tubules were 62% and 60%, respectively (Fig. 2B). The ratio of tubulin in SN-I and SN-II fractions from DMF-treated tubules, 38% soluble to 62% polymerized, was essentially the same as for the no treatment control group (differences between SN-I andSN-II fractions from these two groups were not statistically significant). In contrast, incubation of seminiferous tubules with CBZ or COL resulted in a marked depolymerization of microtubules (Fig. 2, A and B). Carbendazim incubation for 30 min caused substantial depolymerization compared to controls, from 39% to 69% soluble tubulin (P value relative to DMF control was 0.02). Carbendazim incubation for 1 h resulted in extensive depolymerization of microtubules compared to controls, indicated by the increase in soluble tubulin to 78% (P value relative to DMF control groups was 0.0006). The corresponding values for polymerized tubulin (SN-II) in seminiferous tubules after incubation with CBZ (for 30 min and 1 h) were 31% and 22%, respectively (Fig. 2B). Colchicine treatment for both 30 min and 1 h time points resulted in extensive depolymerization, as the level of soluble tubulin detected was 81% and 84%, respectively (P values relative to no treatment control groups were both 
0.0004; Fig. 2A). Treatment of seminiferous tubules with CBZ for 30 min resulted in submaximal depolymerization of microtubules, while the COL effect was essentially maximal after 30 min.
|
The amount of tubulin in PII fractions was also measured by the ELISA to determine if microtubule fractions were completely depolymerized. The tubulin values calculated for all four samples were low (ranging from 0.5–4%; data not shown), suggesting that depolymerization of Sertoli cell microtubules was essentially complete based on ELISA measurements.
The total amount of tubulin isolated from seminiferous tubules was calculated to determine if tubulin protein was lost following exposure to the depolymerizing agents (Fig. 3). Although there was variability observed from experiment to experiment, total tubulin (expressed as µg tubulin/mg seminiferous tubule protein) measured in no treatment, DMF, and CBZ groups was comparable after the 30-min and 1-h incubation periods (differences between these three groups were not statistically significant). This indicated that tubulin levels remained relatively constant following exposure to CBZ. In contrast, COL treatment for 30 min caused an approximately twofold decrease in the amount of total tubulin measured (Fig. 3; P value for comparison of COL to untreated controls was 0.028). The same trend was evident after 1-h exposures to COL, suggesting that a loss of tubulin had occurred. Based on these results, CBZ and COL cause depolymerization of microtubules in the seminiferous epithelium. Furthermore, CBZ treatment does not appear to cause significant loss of total tubulin protein, while COL treatment does result in a decrease in total tubulin.
|
Viability of Seminiferous Tubules During Treatment and Specificity of Antibody for ELISA
The level of ATP in no treatment, DMF-, CBZ-, and COL-treated tubules was measured to estimate the viability of seminiferous tubules during the incubation period. The ATP levels did not decrease in treated samples when compared to the controls after 1-h incubations (data not shown). These results indicate that the seminiferous tubules were metabolically healthy during the incubation periods used in this study.
The specificity of the ß-tubulin antibody used in the ELISA was verified by immunoblotting (Fig. 4). The SN-I, SN-II, and PII fractions from all four treatment groups were separated by SDS-PAGE, transferred to nitrocellulose, and probed with the monoclonal antibody used for the ELISA. Only one band was detected with the ß-tubulin antibody in all of the samples tested, at the expected size of 50 kDa (only no treatment tubule fractions are shown in Fig. 4, as they are representative of all four samples). Furthermore, it is evident that essentially all of the ß-tubulin present in the initial seminiferous tubule homogenates has been measured in the SN-I and SN-II fractions, as there was only a faintly detectable band in the final pellets (Fig. 4, lane 4).
|
Acetylated -Tubulin and ßIII-Tubulin in Soluble and Polymerized Tubulin Pools Were Unaffected by CBZand COL Treatment
To determine if specific isoforms or posttranslationally modified forms of tubulin may be associated with polymerization status of microtubules in the seminiferous epithelium, the composition of tubulin pools was analyzed. Tubulin fractions obtained after CBZ or COL treatment were probed with antibodies against acetylated -tubulin and against the ßIII-tubulin isoform (Fig. 5). Acetylated-tubulin has previously been identified in many tissues [31, 32], while class ßIII-tubulin isoforms are located only in neurons [33, 34] and the mammalian testis [6]. Acetylated -tubulin and the ßIII-tubulin isoform were detected, both in soluble (SN-I) and microtubule (SN-II) fractions (Fig. 5). However, no changes in protein level or distribution were observed when comparing DMF- to CBZ-treated tubules, based on staining with the respective antibodies. Likewise, no changes were observed in the tubulin pools obtained from COL-treated fractions compared to the no treatment controls (data not shown).
|
Interestingly, the PII fractions from all four samples contained only a trace of ßIII-tubulin (Fig. 5, lanes 9 and 12), in agreement with results using the antibody against ß-tubulin (Fig. 4), yet a significant amount of acetylated-tubulin was detected (Fig. 5, lanes 3 and 6). In an earlier study, an antibody to acetylated -tubulin strongly labeled the flagella of spermatids, both at the light and electron microscopic level [31]. Thus, the acetylated -tubulin detected in PII fractions might be due to reactivity with tubulin derived from axonemal microtubules in sperm tails. These microtubules are very stable and are not susceptible to depolymerizing agents or depolymerization conditions [35, 36] such as the ones used during the isolation procedure.
A Diminished Level of Tyrosinated -Tubulinin Microtubule Fractions after CBZ and COL Treatment
To determine if changes in tyrosination level of -tubulin are associated with polymerization status, an antibody against tyrosinated -tubulin was used to probe the tubulin pools (Fig. 6). Tyrosinated -tubulin (52 kDa) was present in SN-I (lanes 2, 5, 8, and 11) and SN-II fractions (lanes 3, 6, 9, and 12) from all four samples. However, the amount of tyrosinated -tubulin in SN-II fractions from treated samples appeared to be less than that detected in the corresponding controls (compare lanes 3 to 6 and 9 to 12). This is in agreement with our finding that incubation with CBZ and COL causes depolymerization, evidenced by a concomitant decrease in the amount of tubulin in SN-II fractions (Fig. 2). Likewise, the soluble SN-I fractions appear to have an increased amount of tyrosinated tubulin compared to controls, supporting the corollary finding that CBZ and COL increase the soluble tubulin pool via depolymerization. The PII fractions from all samples also contained tyrosinated -tubulin. Again, this may be due to the presence of sperm axonemal microtubules in the final pellet, because tyrosinated tubulin has previously been localized in sperm tails [31]. The band apparent in the SN-I fractions at 36 kDa (Fig. 6) was repeatedly detected in all four samples with the anti-tyrosinated tubulin antibody and was unaffected by treatment. This band may correspond to a proteolytic fragment or it may represent nonspecific antibody binding.
|
In addition to the tyrosinated tubulin stained at 52 kDa, there was a 47-kDa band detected in no treatment and DMF-treated SN-II fractions with the tyrosinated tubulin antibody (Fig. 6, lanes 3 and 9). In contrast, the 47-kDa band was not detectable in SN-II fractions obtained after CBZ or COL treatment (Fig. 6, lanes 6 and 12). This protein also was not detectable in SN-I or PII fractions from any of the samples. The 47-kDa protein potentially represents a unique -tubulin isoform based on staining by the tyrosinated -tubulin antibody. The 47-kDa tyrosinated protein does not simply appear to be a proteolytic breakdown product of -tubulin lacking the N-terminus or an antigen bound nonspecifically by antibody, as it is present in the two control samples (Fig. 6, lanes 3 and 9) and is absent only in the CBZ- and COL-treated samples (Fig. 6, lanes 6 and 12). Thus, both CBZ and COL treatment cause a decrease in -tubulin tyrosination level that appears to be associated with a change in polymerization status in the seminiferous epithelium. Moreover, this apparent decrease in tyrosination of -tubulin is a specific effect, because no changes in acetylated -tubulin or in the ßIII-tubulin isoform were detected after treatment with disrupting agents (Fig. 5). Similarly, no changes were detected by immunoblotting in the amounts of MAP2c, cytoplasmic dynein, kinesin, or vimentin proteins in SN-I or SN-II fractions after CBZ or COL treatment (data not shown).
Decreased Tyrosination of -Tubulin after CBZ Treatment of Seminiferous Tubules Was Not Due to Lossof Tubulin Protein
The only obvious changes observed in seminiferous tubule fractions obtained after CBZ or COL treatment appeared to be a decrease in the amount of 52-kDa tyrosinated -tubulin and the loss of tyrosination of a 47-kDa band from the SN-II fractions (Fig. 6). Loss of tyrosination could be the result of detyrosination of -tubulin. Alternatively, it could be due to loss of tubulin protein, either by a general loss of total protein or by specific loss of tubulin due to either proteolysis or to decreased polymerized tubulin after treatment with disrupting agents (see Fig. 2). The current results do not support the possibilities that loss of total tubulin or proteolysis of tubulin after CBZ treatment is the cause of decreased tyrosinated -tubulin (see Figs. 3 and 6). However, to explore these possibilities further, a general -tubulin antibody (referred to as anti--tubulin) that reacts with -tubulin isoforms [32, 37, 38] was used in immunoblotting experiments.
Microtubule (SN-II) fractions from all four samples were probed with both anti--tubulin and anti-tyrosinated -tubulin antibodies (Fig. 7). The 52-kDa and 47-kDa bands were detected with the anti--tubulin antibody, in both DMF- and CBZ-treated samples (Fig. 7A, lanes 3 and 4). Therefore, the 47-kDa protein may potentially be an -tubulin isoform. Furthermore, staining of duplicate lanes with anti-tyrosinated -tubulin antibody (Fig. 7A) confirms that tyrosinated -tubulin (52 kDa) is decreased and the tyrosinated 47-kDa protein (identified in lanes 3 and 4 with anti--tubulin) is not detectable in SN-II fractions obtained after CBZ treatment (lane 6). Importantly, these results demonstrate that the 47-kDa protein is present in SN-II fractions after CBZ treatment, even though it is not detectable with the anti-tyrosinated tubulin antibody (Figs. 6 and 7). The total amount of transferred proteins that was probed with both -tubulin antibodies is demonstrated by Ponceau S staining (Fig. 7A, lanes 1 and 2). It is apparent that the amount of protein in the CBZ sample is equivalent to the amount of protein transferred in the DMF sample. Thus, the absence of the tyrosinated 47-kDa protein in SN-II fractions after CBZ treatment is not due to loss of microtubule protein, but rather appears be a consequence of detyrosination of this protein. Moreover, the 47-kDa protein was not detectable in the SN-I fractions (Fig. 7B, lanes 3–6) when probed with either the general -tubulin or tyrosinated -tubulin antibodies. Therefore, loss of the tyrosinated 47-kDa protein from the polymerized fraction after CBZ treatment is not simply due to depolymerization of microtubules and a shift in distribution to the soluble tubulin pool. Together, the results provide evidence that the identified 47-kDa protein is present in microtubule pools after CBZ treatment, at essentially the same level as in controls, yet it is no longer tyrosinated.
|
The -tubulin (52-kDa) and 47-kDa bands were detected by the general anti--tubulin antibody in both no treatment and COL-treated SN-II fractions (Fig. 7A, lanes 9 and 10). Loss of the tyrosinated 47-kDa band was observed in COL-treated samples probed with the tyrosinated tubulin antibody (lane 12). However, the amount of tubulin immunoreactive with anti--tubulin was also much decreased compared to the control (compare lanes 9 and 10), indicating that COL treatment decreased not only tyrosination but also the amount of the 47-kDa protein. The amount of transferred protein in the COL-treated samples also appears to be less than in the untreated control sample (Fig. 7A, lanes 7 and 8). Similar to results with CBZ-treated tubules, the 47-kDa band was not detectable in COL SN-I fractions with either the general or tyrosinated -tubulin antibodies (data not shown). Based on these results and on the finding that the total amount of tubulin measured in tubulin pools is decreased (Fig. 3), loss of the tyrosinated 47-kDa protein from polymerized pools after COL treatment appears to be due more to a decreased amount of total tubulin than to detyrosination.
DISCUSSION
In the present study we investigated the effects of two microtubule disrupting-agents on microtubule polymerization status and -tubulin tyrosination in isolated rat seminiferous epithelium. Our results demonstrate that CBZ and COL caused extensive microtubule depolymerization and decreased tyrosination of -tubulin in microtubule pools. The in vitro model described for incubation of seminiferous tubule tissue has proven to be an excellent system for isolating and quantitatively measuring tubulin pools. The protocols for culture and fractionation were sufficient to maintain an equilibrium of soluble tubulin to polymer in isolated tubules under control conditions, with no net loss of tubulin protein. The competitive ELISA described is highly specific, quantitative, and relatively easy to use compared to other methods for quantitation of tubulin [28].
The ratio of soluble to polymerized tubulin in seminiferous tubules was shifted from 40% soluble, 60% polymerized in control, DMF-treated tubules, to 78% soluble, 22% polymerized in CBZ-treated tubules. Similarly, the ratio was reversed by COL, from 38% soluble, 62% polymerized with no treatment, to 84% soluble, 16% polymerized in treated tubules. These values represent the first quantitative measurements of tubulin pools in the mammalian seminiferous epithelium, both with and without treatment with disrupting agents. The percentage of tubulin in soluble and polymerized form varies widely among different tissue types, from 30% polymerized in rat liver, to 90% polymerized in human platelets [1–3, 27]. Our results fall well within this range and are similar to an early estimation that approximately 40%–60% of tubulin in mouse 3T3 cells is in the polymerized state [39]. Moreover, the reported results provide a biochemical measurement to explain the loss of Sertoli cell microtubules previously observed after treatment of rat testis with CBZ or COL [16, 20–23, 40].
Seminiferous tubules contain a mixed population of cell types, primarily Sertoli cells and germ cells in varying states of differentiation ranging from spermatogonia to mature testicular sperm. However, the measurements of soluble and polymerized tubulin pools presented in this study represent the values primarily for the Sertoli cell cytoskeleton for the following reasons. Sertoli cells contain a very prominent microtubule cytoskeleton [13–17]. Microtubules involved in germ cell mitosis, meiosis, and in the manchette of the elongating spermatid represent a relatively small proportion compared with the proportion of microtubules localized in Sertoli cells. Sperm cells also contain a significant amount of tubulin in the axoneme of the flagellum that develops during spermiogenesis. Spermatids are attached to Sertoli cells via specialized adhesion junctions, called ectoplasmic specializations [19], and would be included in the seminiferous tubule isolates. However, the contribution of either sperm or spermatid axonemal tubulin to the total measured tubulin can be estimated to be relatively small because the flagellum is an extremely stable structure [35, 36] and would be resistant to the depolymerization conditions described. The finding that the final pellets from all four samples contain acetylated and tyrosinated -tubulins (52 kDa) tends to support this assumption because both these modified tubulins have been previously localized in sperm flagella [31].
In mammals, there are approximately six -tubulin and seven ß-tubulin isoforms [5–7, 9, 41], some of which are expressed ubiquitously, while others are expressed specifically in one or a few tissues. The 47-kDa protein identified in polymerized tubulin fractions may potentially represent a new -tubulin isoform, based on staining with both general -tubulin and tyrosinated -tubulin antibodies. The findings that this 47-kDa protein specifically loses tyrosination after exposure to CBZ, yet remains tyrosinated under two control conditions, support this notion and support that this protein is not a proteolytic breakdown product of-tubulin. Alternatively, this may be an unknown protein that is tyrosinated and shares enough sequence similarity with -tubulin that it is recognized by two -tubulin antibodies, although this appears unlikely because -tubulin is the only known protein to undergo tyrosination and detyrosination [9]. Future studies will be required to characterize further the 47-kDa protein.
Colchicine treatment of seminiferous tubules resulted in loss of approximately half of the total tubulin measured compared to controls. This may account for the decreased level of tyrosinated -tubulin (52 kDa) and the observed loss of the tyrosinated 47-kDa protein (Figs. 6 and 7), because very little of this protein was detected in COL-treated SN-II fractions with the general -tubulin antibody (Fig. 7A). This loss of tubulin after COL treatment could be due to an autoregulated decrease in tubulin mRNA stability. An increase in tubulin subunits, in this case due to the depolymerization of microtubules, is known to trigger a decrease in tubulin mRNA stability and a decrease in mRNA levels within 1–2 h [42, 43]. This explanation for the decrease in tubulin appears unlikely, however, because the half-life of tubulin protein is approximately 36 h [44]. Alternatively, the loss of protein observed after COL exposure may reflect an initial loss of tissue during early stages of germ cell sloughing. The results obtained with the general anti--tubulin antibody in immunoblotting experiments support this explanation, because there was a relative decrease in-tubulin detected in COL-treated samples compared to controls (Fig. 7). However, loss of tissue should be accompanied by a decrease in ATP levels that was not observed. Nonetheless, one or a combination of these possibilities might account for the loss of total protein observed after COL-induced depolymerization of seminiferous tubule microtubules.
The reported decrease in tyrosinated -tubulin in the microtubule pool after CBZ treatment appears to be a result of detyrosination of -tubulin rather than loss of tubulin. The tyrosination-detyrosination of -tubulin occurs via a regulated cycle [8, 45–47] involving two enzymes. Tubulin tyrosine carboxypeptidase catalyzes the detyrosination of microtubules by removing a terminal tyrosine residue from the C-terminus of -tubulin. Another enzyme, tubulin tyrosine ligase, catalyzes the tyrosination of -tubulin subunits by adding a tyrosine residue back on the C-terminal end. Detyrosination occurs on intact microtubules [9, 47, 48] and detyrosinated microtubules have been associated with stable microtubules [9]. However, there is considerable evidence that detyrosination in itself does not cause microtubule stability [46, 48–51], and detyrosinated microtubules may simply accumulate in older microtubule populations [9]. Because loss of tyrosination only occurred in the polymerized pools of tubulin after CBZ treatment, the detyrosination step likely occured on intact microtubules. It is conceivable that the decreased tyrosination of the microtubule pool may contribute to depolymerization of these microtubules or to a less stable subset of microtubules.
An alternative explanation for how decreased tyrosination of -tubulin is coupled to depolymerization of microtubules is that inhibition of microtubule assembly simply results in a decreased number of newly assembled, tyrosinated microtubules. Because both COL and CBZ bind tubulin subunits, and COL binding is thought to alter the ability of tubulin to be incorporated into microtubules [25], this hypothesis suggests that when polymerization is prevented, tyrosination is decreased in the absence of the microtubule substrate.
The interesting findings that the 47-kDa protein loses tyrosination and remains associated with the microtubule pool in the presence of depolymerizing agents raises the possibility that this protein may be a less readily depolymerized tubulin isoform that is involved in the tyrosination-detyrosination cycle, and that tyrosination-detyrosination may have a regulatory role in polymerization. To elucidate the precise relationship of detyrosination and the depolymerization of microtubules induced by CBZ will require further investigation.
FOOTNOTES
First decision: 21 November 2000.
1 This work was supported by NIH grants RO1 ESO 7832 and PO1 ESO 5707 (Center for Environmental Health Science). 
2 Correspondence: Liane M. Correa, Department of Environmental Toxicology, University of California, One Shields Avenue, Davis, CA 95616. FAX: 530 752 3394; lmcorrea{at}ucdavis.edu
Accepted: January 16, 2001.
Received: October 18, 2000.
REFERENCES
- and ß-tubulin isotypes during the differentiation of testis and muscle cells. J Cell Biol 1988; 106:2023–2033-tubulin by specific antibodies. Methods Enzymol 1991; 196:264–274[Medline]-tubulin mRNAs encode five distinct isotypes: testis-specific expression of two sister genes. Mol Cell Biol 1986; 6:2409–2419 chain of tubulin. Biochemistry 1977; 16:2189–2194[CrossRef][Medline]-tubulin: a mechanism for subcellular differentiation of microtubules. J Cell Biol 1987; 105:251–264This article has been cited by other articles:
|
|
 |
|
  S. Koduri, S. A. Hild, L. Pessaint, J. R. Reel, and B. J. Attardi Mechanism of Action of l-CDB-4022, a Potential Nonhormonal Male Contraceptive, in the Seminiferous Epithelium of the Rat Testis Endocrinology, April 1, 2008; 149(4): 1850 - 1860. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Moffit, B. H. Bryant, S. J. Hall, and K. Boekelheide Dose-Dependent Effects of Sertoli Cell Toxicants 2,5-Hexanedione, Carbendazim, and Mono-(2-ethylhexyl) phthalate in Adult Rat Testis Toxicol Pathol, August 1, 2007; 35(5): 719 - 727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Markelewicz Jr., S. J. Hall, and K. Boekelheide 2,5-Hexanedione and Carbendazim Coexposure Synergistically Disrupts Rat Spermatogenesis Despite Opposing Molecular Effects on Microtubules Toxicol. Sci., July 1, 2004; 80(1): 92 - 100. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. Correa, M. Nakai, C. S. Strandgaard, R. A. Hess, and M. G. Miller Microtubules of the Mouse Testis Exhibit Differential Sensitivity to the Microtubule Disruptors Carbendazim and Colchicine Toxicol. Sci., September 1, 2002; 69(1): 175 - 182. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE |
