| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
-Tubulin Overexpression in Sertoli Cells In Vivo: I. Localization to Sites of Spermatid Head Attachment and Alterations in Sertoli Cell Microtubule Distribution1
Departments of Pathology and Laboratory Medicine3
Molelcular Microbiology and Immunology,4 Brown University, Providence, Rhode Island 02912
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-tubulin, was used to selectively disrupt microtubule networks in Sertoli cells in vivo. -Tubulin overexpression was observed to cause redistribution of Sertoli cell microtubule networks, and overexpression of a -tubulin-enhanced green fluorescent protein fusion protein was observed to localize to the site of elongate spermatid head attachment to the seminiferous epithelium.
Sertoli cells, spermatogenesis, testis, toxicology
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The role of Sertoli cell microtubules in supporting spermatogenesis has been largely deduced from outcomes of pharmacologic manipulations and toxicant exposures. Drugs, such as colchicine, Taxol, and nocodazole, which interfere with microtubule physiology, compromise the epithelial structural integrity and impair aspects of spermatogenesis, including spermatid positioning, residual body elimination, and seminiferous tubule fluid secretion [3, 11, 12, 18, 19]. Toxicants with demonstrated effects on microtubules, such as 2,5-hexanedione, also impair spermatogenesis, although the time course and trajectory of the injury process are more complicated [20–26]. Because tubulin is a ubiquitous protein, toxicant and pharmacologic manipulations of microtubules performed in vivo suffer from a lack of cell type selectivity. It is likely that such treatments impair microtubule physiology of germ cells as well as somatic cells, which confounds mechanistic interpretation of target cell vulnerability. To evaluate germ cell dependence on aspects of Sertoli cell microtubule physiology, it is necessary to selectively deliver microtubule-disrupting agents to Sertoli cells without directly inducing alterations in germ cell microtubule functions.
When delivered to the lumen of the seminiferous epithelium, adenoviral vectors are capable of targeting transgene expression to Sertoli cells in vivo [27–29]. Using this method, it is possible to alter expression of specific proteins within the context of an intact, adult testis.
Overexpression of -tubulin in dividing cells leads to microtubule disruption and mitotic arrest [30, 31]. Overexpressed -tubulin nucleates microtubules ectopically in the cytoplasm and impairs microtubule organization in the cell, similar to the effect of Taxol treatment [31]. In dividing cells, -tubulin overexpression leads to mitotic arrest and cell death [30, 31]. However, the effects of -tubulin overexpression have not been described in polarized cultures or in postmitotic, polarized epithelial cells in vivo.
As described in these companion articles, adenoviral vectors that overexpress -tubulin were delivered to Sertoli cells in vivo to test the hypothesis that selective disruption of Sertoli cell microtubules would impair spermatogenesis. Toward that end, adenoviral vectors were designed to overexpress either enhanced green fluorescent protein (EGFP) and -tubulin co-ordinately, under the control of a tetracycline-responsive promoter, or a fusion of EGFP to the C-terminus of -tubulin. The first part of these companion articles documents the cell biologic outcomes of -tubulin overexpression on the microtubule structure of the infected tissues and the behavior of the overexpressed protein in influencing the cytoskeleton. The second article describes the outcomes of those effects on the associated germ cells and describes the histopathology that resulted from the microtubule disruption.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Young adult male Fischer rats, weighing 150–175 g were obtained from Charles River Laboratories (Wilmington, MA) and housed in hanging wire cages at a constant temperature (70° ± 2C), with 35%–70% humidity and a 12-h alternating light-dark schedule. All rats were acclimatized for at least 3 days prior to experimental manipulation and treated according to the NIH Guide for the Care and Use of Laboratory Animals.
All chemicals were purchased from Sigma Chemical Co. (St. Louis, MO) unless otherwise specified. Cell culture reagents, agarose, ethidium bromide, and polymerase chain reaction (PCR) primers were obtained from Invitrogen (Carlsbad, CA). SDS-PAGE reagents were obtained from BioRad (Hercules, CA) or Roche (acrylamide, Indianapolis, IN). Anti-
-tubulin monoclonal antibody (5A6) was a gift of Dr. David Brown (University of Ottawa, Ottawa, ON, Canada). The control viral vector, AdGFP, which expresses EGFP under the control of a cytomegalovirus (CMV) promoter, was purchased from QbioGene (Montreal, Canada). AdtTA was a gift of Drs. Thomas C. Harding and James Uney (University of Bristol, Bristol, UK).
Microscopy was performed using an Eclipse E800 microscope (Nikon Instruments, Melville, NY) equipped for epifluorescence. Digital photomicroscopy was performed using a SpotRT camera (Diagnostic Instruments, Sterling Heights, MI). Postacquisition image processing, where required, was performed using Photoshop (Adobe, San Jose, CA).
Cell Culture
TM4 cells were obtained from ATCC (Manassas, VA) and maintained as recommended by the supplier. Cells were cultured in Dulbeccco modified Eagle medium (DMEM) containing 2.5% fetal calf serum, 5% heat-inactivated horse serum, and 50 µg/ml gentamicin sulfate, and cultures were maintained at 37°C, 5% CO2.
HEK293A cells were obtained from QBioGene and maintained as recommended by the supplier. Cells were cultured in DMEM containing 10% fetal calf serum and 50 µg/ml gentamicin sulfate and maintained at 37°C, 5% CO2.
Production of Viral Vectors
Construction of replication-deficient viral vectors made use of a recently developed bacterial recombinational strategy [32]. Briefly, insertion of the transgene into the adenoviral backbone is mediated by homologous recombination between a shuttle plasmid, bearing transgenes and associated regulatory sequences, and the adenoviral genome plasmid, pAdEasy (
E1, E3), a gift of Dr. Bert Vogelstein (Johns Hopkins University, Baltimore, MD).
Tetracycline-responsive elements have been used widely in an effort to evaluate expression from plasmids or viral vectors in a controlled fashion [33–41]. In the present study, the regulatable protein expression units were used to minimize effects of -tubulin overexpression on viral packaging cells [36].
The tetracycline-responsive element is derived from the transcriptional regulatory element of Escherichia coli (TetO). Expression from the TetO utilizes a modified bacterial tet-repressor protein, tTA, which acts as a transactivator to stimulate expression from the promoter. In the absence of tetracycline, tTA drives expression of the transgene; however, binding of tetracycline to tTA inhibits promoter activity. HEK293A cells, which lack tTA, support reduced expression from the TetO at a level that is compatible with viral production [36]. High-level expression from tetracycline-responsive elements requires coexpression of a tetracycline-responsive element transactivator protein, tTA.
Production of AdBiEGFPT Construct
The -tubulin transgene, EGFP, and regulatory sequences were inserted into pShuttle (a gift of Dr. Bert Vogelstein, Johns Hopkins University, Baltimore MD), in a two-part subcloning procedure. The bidirectional tet-responsive promoter containing the EGFP marker and polyadenylation sequences was first subcloned into pShuttle by excising from the pBiEGFP plasmid (Clontech, Palo Alto, CA) using AatII and SapI (New England Biolabs, Beverly, MA), and fragments were blunt-end ligated into pShuttle. The full-length -tubulin transgene was then PCR amplified from the pH3–16 plasmid [42] (a gift of Dr. Berl Oakley, Ohio State University, Columbus, OH), which contained the full-length -tubulin insert. Primers, were designed with 5' terminal MluI sites to facilitate insertion into the shuttle plasmid construct, as follows: forward: 5'pGCGCAAGCACGCGTCGGCCACCATGCCGAGGGAAATCATCACCCTA3' and reverse: 5'pGCGCGCACGCGTGCTCACTGCTCCTGGGTGCCCC3'. The product was digested with MluI and ligated into the corresponding site in the pShuttle/BiEGFP plasmid (Fig. 1A).
|
Recombination between pAdEasy and the shuttle vector, containing the transgene of interest, was performed as described [32]. Plasmids obtained from minipreps were screened by size comparison with nonrecombinant pAdEasy plasmid in a 0.8% agarose gel, using ethidium bromide staining for visualization. Recombinants were digested with Mlu1 and restriction patterns were compared with nonrecombinant MluI-digested pAdEasy to verify insertion of the transgene.
Construction of the EGFP--Tubulin Fusion Vector, AdTetTEGFP
To visualize -tubulin localization in Sertoli cells in vivo, a C-terminal fusion of EGFP to -tubulin was constructed. Production of the shuttle plasmid vector containing the tetracycline-regulatable EGFP--tubulin fusion protein (TEGFP) was performed using a three-part subcloning process. Briefly, the full-length -tubulin sequence [42] was subcloned into the pEGFP-N1 fusion vector (Clontech) by PCR amplification from the pH3–16 plasmid, using the following primer pair: forward: 5'GCGCGCAAGCTTCGCAACGCCGGTGCCTGAGGAGCGATGCC3'; reverse: 5'CGCGCGGTACCTTTGACTGCTCCTGGGTCGCCCCAGGAGATGTAGTCT3'. A mutation was introduced into the reverse primer to suppress the stop codon, and HindIII and KpnI restriction sites were introduced into the forward and reverse primers, respectively. The -tubulin fusion product was excised from pEGFP-N1 using NotI and NheI and ligated into corresponding restriction sites in pTRE2 (Clontech) (Fig. 1B). The transgene, along with the tet-responsive element and SV40 polyadenylation sequence derived from pTRE2, were excised with XmnI and AseI. Blunt ends were generated using T4 DNA polymerase (New England Biolabs), and the fragment was ligated into the EcoRV site of the pShuttle plasmid. Plasmid DNA was screened for insertion of transgene coding sequence by size comparison with pAdEASY in a 0.8% TAE gel.
Transfection of HEK293A Cells for Viral Packaging and Propagation
HEK293A cells were transfected with Pac1-linearized recombinants for the purposes of viral production. DNA was digested with PacI enzyme (New England Biolabs), and 10 µg were electroporated into 3.6 x 106 cells in a 0.8-ml volume of DMEM, using a 0.4-cm gap cuvette, 0.260 kV, 0.65 kV/cm field strength, and 970 µF capacitance setting. Following electroporation, 0.7 ml of each cell suspension was plated in a 100-mm dish, and 0.1 ml was plated in one 35-mm well of a 6-well plate. Cells were supplemented with complete growth medium and incubated at 37°C 5% CO2 until plaque formation was observed.
Culture dishes were left undisturbed for the duration of the packaging period, and all observations were performed using the aliquots plated on 6-well plates. After 5 days, the 6-well plates were examined for the presence of plaques. No plaques were visible on the bottom of the wells; however, by epifluorescence, numerous patches of GFP-positive cells were observed. Plates were returned to the incubator for another 7 days, after which virus-containing cells were harvested from the 100-mm dishes by scraping. The contents of each 100-mm dish were placed into a 14-ml conical tube (Falcon, BD Biosciences, Bedford, MA) and subjected to three cycles of freeze thawing in a dry ice-methanol bath, with vigorous vortexing between freezing cycles.
Viral stocks were propagated by diluting 5 ml of the crude lysate with 5 ml PBS++ (PBS + 0.5 mM MgCl2 and 0.68 mM CaCl2) and overlaying 150-mm plates containing 80% confluent HEK293A cells. The inoculum was adsorbed for 1 h at 37°C, and then cells were supplemented with 20 ml complete growth medium. Cells were harvested by scraping when >80% were cytopathic, which typically occurred between 72 and 96 h post infection. The contents were placed into 50-ml conical tubes and frozen at -80°C until used for bulk propagation. Viral stock was purified and concentrated by CsCl centrifugation, and stocks were dialyzed against a solution containing 10 mM tris pH 8.0, 2 mM MgCl2, 5% sucrose. Dialyzed viral stock was then sterilized by passage through a 0.2-µm syringe filter and frozen at -80°C until used. Titer determination was made using standard methods [43].
Measuring -Tubulin Fold-Induction in Cultures of TM4 Cells at 24 and 48 h Post Infection
Cultures of TM4 cells were plated at a density of 2.5 x 105 cells/well in 6-well plates and supplemented with 2 ml of complete TM4 growth medium. Cells were allowed to adhere overnight and then infected at multiplicities of infection (MOIs) of 10 and 100. After 24 or 48 hours post infection, the proportion of cells expressing the transgene was determined by epifluorescence, cells were lysed and -tubulin levels assessed by Western blotting. Briefly, cells were washed once with PBS++ and overlaid with 0.5 ml PBS++ containing diluted viral stock or PBS++ alone (control). The following treatments were performed: PBS++ alone, AdtTA alone, and AdtTA + AdBiEGFPT. All treatments were performed at MOIs of 10 and 100, and experiments were performed in triplicate. Cultures were returned to the incubator for 1 h and supplemented with 2 ml complete medium and returned to the incubator for either 24 or 48 h.
Lysates were prepared by overlaying monolayers with 1 ml lysis buffer (25 mM tris, pH 6.8, 0.5% SDS) containing protease inhibitors (complete EDTA-free protease inhibitor cocktail, Roche). Lysis was performed for 1 min at room temperature, and then lysates were collected, frozen in liquid nitrogen, and stored at -80°C.
Western Blotting and Densitometric Analysis
Protein concentration was measured using a modification of the Lowry method (DC protein assay, BioRad). Lysates were passed three times through a 26-gauge needle to reduce viscosity, and 20 µg of protein was mixed with 2 x sample buffer (4% SDS, 20% glycerol, 0.005% bromophenol blue, 10% ß-mercaptoethanol) and boiled for 5 min prior to loading. The proteins were separated by SDS-PAGE through a 6% stacking gel and 12% separating gel, as described [44]. Following electrophoresis, proteins were electrophoretically transferred to a polyvinylidine membrane (Immobilon P, Millipore, Bedford, MA) and processed for Western blotting.
Membranes were blocked in TBST (100 mM Tris-HCl pH 7.5, 0.9% NaCl, 0.1% Tween-20) containing 5% nonfat dry milk for 30 min at room temperature, with gentle agitation, and subsequently incubated with primary antibody diluted 1:2500 in blocking solution for 1 h, as described above. Washes were performed for 15 min, with several changes of TBST. Secondary antibody, horseradish peroxidase-conjugated goat-anti-mouse IgG (Amersham, Piscataway, NJ) was diluted 1:2500 in blocking solution and applied to membranes. Incubations and washes were performed as described. -Tubulin immunoreactivity was evaluated using the ECL detection kit (Amersham).
Densitometry was performed using NIH Image (Scion Corp., Frederick, MD); optical density (OD) measurements taken from scanned autoradiographic film were calculated and mean values normalized to those obtained with PBS treatment. This was performed by dividing the mean OD values of the -tubulin bands obtained in each of the overexpression regimens by the mean OD value obtained from cells treated with PBS++ alone. For infections with AdBiEGFPT/AdtTA or AdTetTEGFP/AdtTA, average OD values, which represent the sum of endogenous and overexpressed -tubulin or overexpressed -tubulin fusion protein, respectively, were adjusted to reflect the proportion of EGFP-positive cells observed at the time of cell lysis. Approximately 100 cells per infection regimen were counted and the proportion of GFP-positive cells determined by epifluorescence. All infections and counts were performed in triplicate. The fraction of GFP-positive cells was then used to estimate the fold-induction by dividing the mean OD ratio (mean OD obtained from infected cells/mean OD of PBS-treated cells) by the fraction infected.
Delivery of Adenoviral Vectors to the Seminiferous Epithelium in Vivo
Adenoviral vectors were delivered to the lumen of the seminiferous epithelium via retrograde perfusion of the rete testis [45]. Briefly, animals were anesthetized with 50 mg/kg sodium pentobarbital or a combination of 75 mg/kg ketamine and 13 mg/kg xylazine, and the testis was exposed via abdominal incision. A 30-gauge needle attached to a 1-ml syringe (BD Biosciences) into which a piece of 0.23-cm diameter Tygon tubing had been inserted (Norton Performance Plastics, Akron, OH) was carefully inserted, and the distribution of the perfusate was monitored by observing the flow pattern of the tracking dye. In successful injections, dye was observed to enter the ductuli efferentes and individual seminiferous tubules; otherwise it was noted to distribute in a diffuse pattern between seminiferous tubules, indicative of delivery to the interstitial spaces.
Between 50 and 100 µl of adenoviral stock, containing 1% blue food dye, was introduced into the testis at a rate of 25–50 µl/min, with the aid of an Eppendorf Femtojet microinjection pump (Eppendorf, Hamburg, Germany), set at 50 hPa. Equal volumes of AdtTA plus one of the following were injected into adult testes as described: AdGFP (expressing EGFP alone), AdBiEGFPT (expressing EGFP and -tubulin as separate transcripts), or AdTetTEGFP (expressing the -tubulin-EGFP fusion protein). The titers of AdGFP, AdBiEGFPT, and AdTetTEGFP were adjusted to 5 x 109 plaque-forming units (PFu)/ml, and that of AdtTA was 5 x 1010 PFU/ml, yielding a transactivator:target vector ratio of 10:1. Total titer delivered was between 1.4 x 109 and 2.75 x 109 PFU pertestis, comparable with titers reported previously [27, 28]. Control infections with AdGFP included AdtTA to normalize delivered titer and control for potential effects of AdtTA. Following perfusion, testes were returned to the scrotum, the abdominal incision was closed, and the animal was allowed to recover for 48 h post infection.
Histological Evaluation
GFP expression was weakly detectable by epifluorescent transillumination of unfixed tissue within 24 h post infection and maximal by 36–48 h post infection (data not shown). Because of the difficulty encountered in preserving GFP expression through histological procedures at the early time points when expression levels were low, evaluation was performed 48 h post infection when sufficient GFP preservation was possible to facilitate discrimination between infected and uninfected seminiferous tubules. Moreover, preliminary examination of the time course of injury following infection with -tubulin overexpressing vectors indicated that the 36- to 48-h time point was optimal for obtaining mechanistic data about the injury process. By 48 h, appreciable histological alterations were observed; however, by 96 h, the histological architecture was severely disrupted in seminiferous tubules infected with AdBiEGFPT/AdtTA or AdTetTEGFP/AdtTA, and in many cases it was impossible to ascertain the stage of spermatogenesis of an infected cross-section. This gross disruption was not observed with control infections, however. For these reasons, the 48-h time point was selected for evaluation of histopathological changes and microtubule disruption.
Testes were fixed by puncturing the tunica albuginea with a 22-gauge needle and submerging in 10% neutral buffered formalin [46] for 24 h, after which they were transferred to PBS pH 7.4, containing 6.8% sucrose. Cross-sections (1 mm thick) were embedded in glycol methacrylate histological embedding resin (Technovit 8100, Heraeus Kulzer GmBH, Wehrheim, Germany), as described by the manufacturer. Serial sections (2 µm) were used for paired fluorescence and histological analysis. Sections prepared for light microscopic analysis were stained with periodic acid-Schiff/hematoxylin (PAS/H) and serial sections prepared for epifluorescent analysis were mounted in antifade (0.1% p-phenylenediamine, 10% PBS, 90% glycerol) to prevent quenching of the GFP signal.
Sections that were processed for tubulin immunofluorescence were first incubated in cold acetone (-20°C) for 5 min and washed for 10 min in several changes of PBS. Nonspecific binding was blocked by preincubation with PBS+ (PBS containing 0 .1% BSA and 5% normal goat serum) for 30 min at room temperature. Sections were incubated for 1 h with primary antibody (5A6) and diluted 1:2500 in PBS+. Slides were washed three times on a rocking platform at room temperature for 5 min each. Secondary antibody, Alexa594-conjugated goat-anti-mouse IgG (Molecular Probes, Eugene OR), was diluted 1:1000 and applied as described.
Histological evaluation was also performed in cryosectioned tissue. Tissues were fixed as described above, and 1-mm slices were incubated at 4°C for 48 h in PBS containing 50% sucrose. Tissues were then embedded in OCT (Miles, Inc., Indianapolis, IN) and frozen in liquid nitrogen. Sections (8 µm) were cut, mounted in antifade, and examined for the presence of GFP by epifluorescence.
To determine which stages are most susceptible to infection, 2 µm serially sectioned, glycol methacrylate-embedded tissues were examined by epifluorescence and light microscopy. GFP-positive seminiferous tubules were identified by epifluorescence, and the seminiferous tubule was staged by light-microscopic examination of the PAS/H-stained serial section. Three testes (n = 3 animals) per treatment were examined.
To determine whether infection altered timing of spermatogenesis, histological cross-sections from testes that were infected for 48 h were examined by bright-field microscopy, and seminiferous tubule cross-sections were assigned to one of four stage categories: I-III, IV-VI, VII-VIII, and IX-XIV. All seminiferous tubule cross-sections in each testis were scored, three testes (n = 3 animals) per treatment group were examined, and evaluations were performed on randomly coded slides.
Statistical Evaluation
Statistical analyses were performed using StatView software (Abacus Concepts, Berkeley, CA). Stage specificity of expression for the different infection regimens, and the comparison of stage distributions in testes infected with different viral constructs was analyzed by one-way ANOVA, followed by multiple pair-wise comparisons using the Fisher protected least significant difference test. For all analyses, the criterion for significance was P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Two adenoviral vectors that expressed -tubulin were constructed: one that expressed -tubulin and an EGFP reporter co-ordinately (AdBiEGFPT) under the control of a bidirectional, tetracycline-responsive element and another that expressed EGFP fused at the C-terminus of -tubulin (AdTetTEGFP) under the control of a tetracycline-responsive element (Fig. 1). Infection of TM4 cells with AdTetTEGFP/AdtTA (expressing the EGFP--tubulin fusion protein) or AdBiEGFPT/AdtTA (expressing both EGFP and -tubulin separately) yielded high-level expression in cultured TM4 cells (Fig. 2). The distribution of -tubulin in these cells was predominately diffusely cytoplasmic; however, clusters of -tubulin reactivity were noted in some cells (Fig. 2, C and D), consistent with the observation that -tubulin can self-assemble into polymeric structures [31].
|
To determine the level of -tubulin induction in cells at different time points post infection, TM4 cells, infected with a range of MOIs, were lysed and Western blots for -tubulin were performed. At 24 h post infection with AdtTA + AdBiEGFPT, -tubulin content was elevated 7- to 10-fold above baseline levels in cells infected at MOIs of 10 and 100. Experiments were performed in triplicate, and representative blots from 24- and 48-h time points are illustrated (Fig. 3, A and B). Densitometric analysis of blots revealed that viral infection did not itself affect the -tubulin level of infected cells because cells infected with AdtTA alone did not exhibit alterations in -tubulin levels. Coinfection with AdtTA and AdBiEGFPT, however, greatly increased -tubulin levels measured in cultures infected at MOIs of 10 and 100 (Fig. 3C).
|
To evaluate the ability of TEGFP to function similarly to endogenous -tubulin, cultures of mitotic Sertoli cells (TM4 cells) were infected and tubulin patterns assessed. Similar to uninfected cells, cells infected with AdTetTEGFP exhibited normal spindle formation, compared with uninfected TM4 cells (Fig. 4A). In addition, TEGFP showed centrosomal incorporation, suggesting that -tubulin does indeed substitute for endogenous -tubulin in dividing cells (Fig. 4B). Many infected cells, however, did become multinucleate (Fig. 4C), suggesting that -tubulin overexpression induces mitotic arrest, consistent with earlier reports [30, 31].
|
Control Infection with AdGFP/AdtTA: Effects on Histological Architecture and Tubulin Distribution
In fixed, glycol methacrylate-embedded tissue, cross-sections of seminiferous tubules yielded a pattern of GFP localization typical of Sertoli cell infection. Bright, GFP-positive nuclei were observed near the base of seminiferous tubules, and GFP was noted to radiate toward the lumen, consistent with the known distribution of Sertoli cell cytoplasm (Fig. 5B'). Moreover, dark regions were observed in the length of the Sertoli cell cytoplasm in a pattern consistent with the expected positions of embedded germ cells (Fig. 5B').
|
Infection with control (AdGFP/AdtTA) was not observed to alter the distribution of tubulin in infected cells. Similarly staged uninfected and control-infected tubules are shown in Figures 5, A and A'. During this stage of spermatogenesis, tubulin distribution in infected Sertoli cells occurred along the trunk and radiated into cytoplasmic extensions surrounding germ cells, similar to uninfected cells (Fig. 5, C, C', D, and D') [2, 5].
Infection with AdBiEGFPT: Effects on Histological Organization and Tubulin Distribution
In contrast to control infections, overexpression of -tubulin dramatically altered the organization of the Sertoli cell microtubule cytoskeleton. Figure 6, A and A', illustrates two seminiferous tubules at similar stages of spermatogenesis from the same testis. One tubule was uninfected (A) and the other was infected with AdBiEGFPT (A'). Strong transgene expression was observed with this vector (Fig. 6, B vs. B'). By immunofluorescence, it appeared that -tubulin expression altered the distribution of tubulin in infected cells. Figure 6, C and C' illustrates the effect of infection with AdtTA and AdBiEGFPT on the histology and tubulin distribution in two seminiferous tubules at similar stages of spermatogenesis in the same testis. Expression of -tubulin shifted the position of microtubules toward the luminal surface. Whereas most tubulin staining in the uninfected seminiferous tubule localized basally and along the trunk of the cell in the uninfected seminiferous tubule, the staining in the infected seminiferous tubule was localized throughout the seminiferous epithelium and was particularly abundant in the luminal region of the seminiferous epithelium (Fig. 6, C vs. C' and D').
|
Infection with AdTetTEGFP: Effects on Histological Architecture and Tubulin Distribution
To visualize coassociation of overexpressed -tubulin with the tubulin immunostaining pattern, seminiferous tubules were infected with an adenovirus that expressed -tubulin fused to EGFP. Infection with AdTetTEGFP elicited similar histological and cytoskeletal alterations to that of AdBiEGFPT. Figure 7, A and A' illustrates two seminiferous tubules at similar stages of spermatogenesis from the same testis, one that was uninfected (Fig. 7A) and one that was infected with AdBiEGFPT (Fig. 7A'). Although similar, the nuclear architecture of the elongating spermatids suggests that the infected seminiferous tubule was at a slightly later stage of spermiogenesis. Tubulin immunolocalization at this stage was grossly disorganized, compared with uninfected seminiferous tubules. The majority of tubulin immunostaining occurred along the trunk of the Sertoli cell and at the luminal surface, coincident with the points of step 19 spermatid attachment (Fig. 7C). In the infected seminiferous tubule, tubulin distribution in the trunk of the Sertoli cell was diminished and was intensely redistributed along the length of step 19 spermatid heads (Fig. 7, C' and D'). This pattern coincided with the localization of the TEGFP in infected Sertoli cells (Fig. 7B').
|
In AdTetTEGFP/AdtTA-infected tissue, TEGFP localization occurred throughout the cell in both diffusely cytoplasmic and focal patterns. Foci of TEGFP localized to areas surrounding spermatid heads in stages that contain highly condensed elongates. Figure 8 illustrates the pattern observed in TEGFP localization within the seminiferous epithelium. These areas often also contained abnormally shaped spermatids (Fig. 8), which were also associated with foci of TEGFP.
|
Stage Specificity of Expression in Vivo
The stage specificity of infection observed in this study was different from that previously reported [27, 28]. The stage distribution of infected seminiferous tubules was broad and variable, but the majority of infected seminiferous tubules was observed to be between stages VII and XIV. Figure 9 describes the stage distribution of GFP-positive tubule cross-sections for the three infection regimens. There were no significant differences among the three treatments in the stages most prone to infection. At 48 h, more than 80% of infected seminiferous tubules were between stages VII and XIV, suggesting that they became infected between stages V and XII. Moreover, there were differences in the level of -tubulin expression observed in seminiferous tubules of different stages. The highest levels of -tubulin expression (as evidenced by levels of EGFP expression) were noted in stages IX–XIV, and the lowest levels were observed between stages I–III (data not shown).
|
To determine whether there was a block to spermatogenic progression brought about by -tubulin overexpression, manifested as deficits and/or overrepresentation in sequential stages, seminiferous tubules in cross-sections of testes taken from the three different treatment groups were staged. These values were then compared with those from uninfected testes. No differences in the distribution of stages were observed between uninfected and any of the infected tissues (Figure 10).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-Tubulin is the protein primarily responsible for nucleating microtubules in eukaryotic cells. In dividing cells, -tubulin is distributed in the cytoplasm and pericentrosomal matrix [47–49]. Soluble -tubulin-containing complexes aggregate into ring structures that accumulate at the centrosomes as the cell enters mitosis and nucleate microtubules that form mitotic spindles [50–52]. The factors that determine the spatial distribution of microtubules in polarized cells have not been well established. Unlike most cultured cell models, polarized epithelial cells have microtubule-organizing centers oriented at the apex of the cell, away from the basally positioned nucleus. Moreover, the microtubule-organizing center is more loosely distributed and not associated with any particular organelle [1, 7, 8, 53–58].
The involvement of the centrosome in nucleating microtubules at the apical surface varies among the different models studied [4, 55]. In Sertoli cells, however, microtubule regrowth does not occur from the basally positioned centrosome. Following nocodazole treatment, bundles of microtubules re-emerge from the apical surface and are distributed longitudinally in the cell, with their plus-ends pointed toward the nucleus; however, no microtubules are observed to arise from the perinuclear centrosome [4]. Considerable effort has been made toward determining the organization of the microtubule cytoskeleton in Sertoli cells in vivo; however, the composition and precise location of nucleating complexes have not been established. Although the role of -tubulin in patterning microtubule distribution in Sertoli cells has not been defined, its role in many other polarized cell types in vitro and in vivo has been established. For that reason, it was of interest to determine the effect of -tubulin overexpression on the organization of the microtubule cytoskeleton of Sertoli cells in vivo and determine the effects of microtubule disruption in Sertoli cells on spermatogenesis.
The viral constructs utilized in this study expressed the EGFP reporter gene either co-ordinately with -tubulin, or as a fusion to the C-terminus of -tubulin. Expression from tetracycline-responsive, bidirectional promoters has been demonstrated to result in efficient expression of both transgenes in vitro [39, 59]. Indeed, high-level expression of both -tubulin and EGFP were observed in cultured TM4 cells. Moreover, coadministration of adenoviral vectors that express tTA (a "tet-off" transactivator protein required for expression from tet-responsive elements), with vectors that express the transgene of interest, has been demonstrated to result in efficient expression of transgenes in vivo [35, 37, 41].
Expression of EGFP fused to the C-terminus of -tubulin has been previously used to study various functions of the centrosome in dividing cells [60, 61] and has been observed to behave similarly to endogenous -tubulin. Other -tubulin fusion proteins, including C-terminal fusion of c-myc with -tubulin, have also been observed to result in efficient incorporation into centrosomes, consistent with the role of -tubulin in dividing cells and suggesting that the fusion construct behave analogously to endogenous -tubulin [31, 62].
In this study, the human -tubulin cDNA sequence was used to perturb the rat Sertoli cell cytoskeleton. Human -tubulin has been demonstrated to function normally in a wide array of cell types, ranging from hamster to yeast [31, 63]. The -tubulin fusion product used in this study was observed to behave in a manner consistent with previous reports, suggesting that the overexpressed -tubulin functions similarly to endogenous -tubulin in Sertoli cells in vivo [60]. Indeed, that the TEGFP fusion protein incorporates normally into spindles of mouse TM4 cells supports the notion that the protein functions as anticipated.
The levels of -tubulin observed in cultures of TM4 cells are relatively modest, compared with those observed in previous studies aimed at understanding the effects of -tubulin overexpression. Overexpression in yeast was observed to be lethal at 160-fold of endogenous levels and well-tolerated at levels of 50-fold above endogenous, and the mechanism of injury was observed to be via mitotic arrest [30]. Presumably this pathway of injury would not occur in Sertoli cells, which are postmitotic in adult animals.
These measurements, although important indices of promoter activity, are unlikely to be predictive of expression levels in individual cells in vivo. Perfusion of the rete testis leads to high-level exposure of Sertoli cells in the immediate vicinity of the incoming fluid, but titers would be expected to decline as fluid flows downstream of the initial site of exposure and viral particles are taken up by cells. Thus, the level of exposure likely occurs along a gradient distributed spatially over the length of the perfused seminiferous tubule segment. Moreover, because expression is a function of the number of copies of viral DNA present in infected cells, intracellular expression levels are expected to exhibit great variability between seminiferous tubule cross-sections in a single testis.
The localization of TEGFP to spermatid heads suggests that spermatid heads may serve as points of -tubulin aggregation in the seminiferous epithelium in general. No data about the distribution of -tubulin, or its likely role in nucleating microtubules in Sertoli cells have been presented in the literature to date. The low abundance of endogenous -tubulin in most cells and the complexity of the seminiferous epithelial architecture make such evaluations extremely difficult. Numerous attempts were made during the course of these experiments to localize endogenous -tubulin in vivo, using different reagents and conditions, and no unambiguous pattern was obtained.
Although it is unwise to model the behavior of endogenous proteins solely on information gleaned from models of overexpressed proteins, such data may provide clues to the role of proteins for which information about their endogenous roles is scant. That -tubulin may have a role in the establishment of the cytoskeletal connection between the ectoplasmic specialization and the Sertoli cell microtubule network merits further investigation.
|
ACKNOWLEDGMENTS |
|---|
-tubulin transgene; Dr. David Brown, University of Ottawa, for the gift of the monoclonal antitubulin antibody, 5A6; Drs. Thomas Harding and James Uney, University of Bristol, Bristol, UK, for the gift of the AdtTA viral stock; and Dr. Daniel Johnson for his expert advice in the development of the viral delivery system. Thanks are also due to Susan Hall for her assistance with histology and in devising the injection apparatus. |
FOOTNOTES |
|---|
2 Correspondence. Kim Boekelheide, Department of Pathology and Laboratory Medicine, Brown University, Providence, Rhode Island 02912. FAX: 401 863 9008; Kim_Boekelheide{at}Brown.edu
Received: 26 September 2002.
First decision: 13 November 2002.
Accepted: 3 March 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
