Animals
Thirty two young adult, male Wistar rats, weighing 360-540 g at the end of the experiment, were used in the work described here. The animals were bred in the animal house of the Nencki Institute, Warsaw, Poland. They were given free access to water and pellet food and were housed under standard humidity and temperature at 12 h light/dark cycle. Procedures involving animals and their care were conducted in conformity with the institutional guidelines of the First Local Ethics Committee in Warsaw, which are in compliance with the European Community Council Directive (86/609/EEC). Three groups of animals were tested: intact control, spinal, and spinal subjected to locomotor training.
Spinal transection and postoperative care
Twenty rats were anesthetized with Equithesin (0.4 ml/100 g b.w.) and their backs were shaved and disinfected with iodine at the incision sites. Skin and muscles were cut over the caudal thoracic segments with a fine scalpel. The position of the vertebrae was fixed by insertion of hooks into the connective tissue and muscles around the incision. A laminectomy was performed at the thoracic (T9/10) vertebrae. The dura was opened and Lidocaine (2% xylocaine) was applied on the surface of the cord. The spinal cord was then completely transected using surgical scissors and the gap between the rostral and caudal ends was enlarged by aspiration up to about 0.5 mm, washed with warm (around 36°C) 0.9% NaCl, and dried with absorbable cellulose. After careful inspection of the lesion area, the surrounding tissues were subsequently closed with surgical sutures; finally, the skin over the wound was closed with sterile stainless steel staples. About 6 ml of 0.9% NaCl was injected subcutaneously after the surgery. Enrofloxacin (Baytril 2.5%; 0.2 ml/kg) was administered subcutaneously at the end of the surgery and during 5 consecutive days in order to prevent infection. An analgesic Tolfedine (3 mg/kg, s.c.) was given during 3 postoperative days.
Immediately after surgery, the rats were placed in warm cages, covered with blankets, and inspected until they regained consciousness. They were returned to individual cages with full access to food and water as soon as they recovered after anesthesia. The animals were attended for general inspection three times daily during the first postoperative week and twice daily in the subsequent weeks, including cleaning of their bodies and manual bladder expression, if necessary. The animals had no significant health problems for weeks after spinalization, except for occasional bladder bleeding during initial post surgery days. Spontaneous micturition usually returned in the second week after surgery.
Behavioral training
During a week preceding the experiment all tested rats were accustomed, twice daily, to walking on a motor driven treadmill belt at a speed of 0.05 m/s for 5-minute periods. After getting accustomed to the treadmill locomotion, 20 animals were spinalized at the T9/10 level. After one week recovery period, 10 spinal rats were left with no exercise except for one day of testing their motor abilities during the fifth week after the surgery. The other 10 rats were then accustomed to the treadmill walking with the forelimbs and rostral trunk placed on a platform located 1 cm above the belt, while the hindlimbs were placed on the running treadmill. The experimenter secured proper position of the trunk on the platform holding the animal's body and optimized the positioning of the hindlimbs for weight support by holding and manually pressing the proximal part of the tail. When the spinal animals became accustomed to the procedure, for the next 4 weeks the locomotor training was carried out five days a week, at a speed of treadmill belt between 0.05 - 0.1 m/s. The daily training consisted of four to six walking sessions, lasting about 4 min each, separated by about 30 min rest in the home cages. The animals were rewarded with their preferred food (corn flakes) after each session. All other animals (intact and spinal non-trained) accompanied the trained rats during their daily training in the experimental room, where they were kept in cages, handled, and rewarded occasionally with corn flakes. The control group consisted initially of 12 intact rats that got accustomed to the treadmill but were never trained, although one animal was excluded from further analysis due to bad tissue preservation.
The number of sequences of steps was counted by an experimenter. At least two consecutive steps performed alternatively on both hindlimbs were classified as a sequence and taken into account for further analysis. A side view of each rat walking on the treadmill was recorded after 2nd and 3rd week of training using a Panasonic VHS 5100 video camera at 30 frames per second.
Materials
The primary polyclonal antibody against BDNF (N-20; sc-546) and the respective control peptide (N20P 546) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Thanks to Dr. David Kaplan (HSC, University of Toronto), we also used an antibody against BDNF produced in his laboratory. Monoclonal anti-synaptophysin (MAB5258-50UG) was from Chemicon, and monoclonal anti-MAP-2 (M4403, clone HM-2, ascites fluid) was from Sigma. Other immunoreagents including Vector M.O.M. Kit for monoclonal antibodies, fluorescein conjugate with avidin DCS used for the amplification of fluorescent signal, standard and Elite Vectastain ABC detection kits, and secondary anti-rabbit antibody conjugated with Texas Red were all purchased from Vector Laboratories (Burlingame, CA, USA). Secondary antibodies conjugated with AlexaFluor were from Molecular Probes. All other chemicals and reagents were from Sigma, except for PFA (Merck, Germany), DPX (Park, UK), alcohols, and xylene (POCh, Poland).
Immunohistochemistry
Tissue processing
Eighteen rats were subjected to immunohistochemistry (6 intact, 6 spinal non-trained, and 6 spinal and trained). The rats were anesthetized with lethal doses of sodium pentobarbital (80 mg/kg, i.p.) and perfused for 2-3 min via the ascending aorta with 200 ml 0.1 M phosphate-buffered saline (PBS), pH 7.4, and, subsequently for the next 20 min, with 400 - 500 ml of ice-cold fixative (2% paraformaldehyde plus 0.2% parabenzoquinone in 0.1 M PB). Spinal cords were removed from the vertebral columns and were postfixed in the fixative for 1.5 h at RT. The tissue was then cryoprotected overnight in 10% sucrose in 0.1 M PB at 4°C followed by 20% and 30% sucrose, until the tissue sank. The lumbar segments of the spinal cord were frozen with pre-cooled heptane (temp. around -30°C), placed on tissue holders, surrounded by the Jung tissue-freezing medium (Leica), and sectioned with a cryostat. Forty-micrometer transverse sections were collected free-floating in PBS, pH 7.4, to perform single-immunolabeling and complementary cresyl violet staining. Consecutive sections were collected to neighboring wells to assure that patterns of BDNF and synaptophysin labeling were analyzed on adjacent tissue areas. Five to six 40-μm sections per rat, representing L3 and L4 segments, were taken for analysis. For double-labeling studies, 16-μm glass mounted (BDNF/synaptophysin; BDNF/MAP-2) and 40-μm free-floating (BDNF/MAP-2) transverse sections were collected. Glass mounted sections were frozen at -20°C, whereas free-floating sections were collected and kept in anti-freeze medium until used.
Within each experiment, immunohistochemical processing of tissue sections from all groups was carried out simultaneously. The conditions of all procedures (dilutions of reagents and antibodies, washings, incubation time and temperature, blocking of nonspecific staining, and reaction development regimen), were rigorously maintained throughout the assays and were identical for the sections from all tested groups.
BDNF immunostaining
Two polyclonal anti-BDNF antibodies (from Santa Cruz and Dr Kaplan's) were used throughout the experiment. Santa Cruz polyclonal anti-BDNF antibody was extensively characterized in our previous experiments [6, 8]. To confirm the effects evaluated with the Santa Cruz antibody, we also used the antibody kindly provided by Dr D. Kaplan. Both antibodies recognized mature BDNF protein but were raised against two different peptides from the carboxyl terminus of BDNF. Prior to labeling, sections were washed in PBS with 0.2% Triton X-100, pH 7.4 (PBST), incubated in a solution of 0.3% H2O2 in water for 20 min to quench endogenous peroxidase activity, washed extensively in PBST and, finally, incubated with 3% normal goat serum (NGS) in PBST for 60 min to reduce non-specific staining. The sections were then incubated overnight at 4°C with anti-BDNF rabbit polyclonal antibody (Santa Cruz, 1:1000 or Kaplan's 1:3000, in PBST + 1% NGS). The sections were then rinsed in PBST prior to 1 h incubation at room temperature with the respective biotinylated secondary antibodies from the ABC kit. Subsequently, after extensive washings with PBST, sections were incubated for 1 h with AB complex containing avidin-HRP conjugate. The sections were then washed with PBST and the antigenic sites were revealed by treating with 0.05% DAB and 0.01% H2O2. The reaction was terminated by addition of PBST and by subsequent PBS washings. The sections were mounted on gelatin-subbed slides, dehydrated in ascending alcohol concentrations, cleared through xylene, and covered with DPX resin.
Synaptophysin immunostaining
Immunofluorescent staining was performed on free-floating sections. After three 5-min rinses of the sections in PBS, nonspecific binding was blocked by incubating sections for 1 h with M.O.M. Blocking Reagent from Vector M.O.M. Kit for monoclonal antibodies. The sections were then briefly washed 3 times in PBS. The following steps were performed strictly according to the Vector protocol. Briefly, the sections were pre-incubated with the M.O.M. Diluent for 5-min at room temperature. The excess of the Diluent was tapped off and monoclonal anti-synaptophysin antibody (Chemicon) diluted 1:1000 in the M.O.M. Diluent, was applied on sections. After 30-min incubation followed by three 2-min rinses in PBS, the sections were incubated with biotinylated secondary antibody for 10 min. After three further rinses, the sections were incubated for 20 min with avidin DCS-fluorescein conjugate (Vector, 1:25). The sections were then washed three times in PBS and mounted onto glass slides, dried, and mounted with the Vectashield Mounting Medium for fluorescence.
Double immunolabeling of synaptophysin and BDNF
The sections were elaborated for synaptophysin, strictly as for a single immunolabeling, and after three 5-min rinses in PBS the sections were incubated overnight at 4°C with anti-BDNF rabbit polyclonal antibody (Santa Cruz, 1:1000). Next day, the sections were again washed 3 times (5 min each) in PBS and incubated with anti-rabbit secondary antibody conjugated with Texas Red (Vector, 1:200), washed three times (5 min each) in PBS, mounted onto glass slides, air-dried, and coverslipped with the Vectashield Mounting Medium.
Double immunolabeling of BDNF and MAP-2
The sections were washed in PBST and incubated with a solution mixture of 3% normal goat serum (NGS) and 3% normal horse serum (NHS) in PBST for 60 min, in order to reduce non-specific staining. The sections were incubated overnight at 4°C with anti-BDNF rabbit polyclonal antibody (Santa Cruz, 1:1000) combined with anti-MAP-2 mouse monoclonal antibody (Sigma AP-14 1:50), diluted with 1% NGS+1% NHS solution mixture in PBST. The sections were then rinsed in PBST prior to 1-hour incubation at room temperature with the respective secondary antibodies linked to Alexa Fluor 488 (green fluorescence dye, for BDNF labeling; 1:200) or Alexa Fluor 594 (red fluorescence dye, for MAP-2 labeling; 1:200). After several washes, the sections were mounted onto glass slides, air-dried, and coverslipped with the Vectashield Mounting Medium.
Control of immunolabeling specificity
We performed a series of controls to validate the specificity of immunohistochemical profiles observed in our study. 1) Immunolabeling specificity was routinely examined by omitting a primary antibody in the incubation mixture. Under these conditions, no immunolabeling was ever detected. 2) In the initial series of experiments control immunostaining with the Santa Cruz antibodies preincubated with blocking peptides was performed (Figure 1). A non-specific staining was absent except for a faint background labeling, detected predominantly in the outermost part of the funiculi. 3) The specificity of staining in fluorescent labeling assays was verified in two ways, first by omission of the primary antibodies, and, second, by omission of the secondary antibodies. These tests resulted in lack of fluorescent staining. In double immunolabeling approach the labeling was controlled for each antibody separately, by omission of the secondary antibody while all other steps remained unchanged. These tests proved that none of the staining obtained in our experiments was due to non-specific fluorescence or filter bleed-through.
Synaptic zinc histochemistry
Tissue preparation
Fourteen animals were used in order to visualize synaptic zinc. Since the quality of sections from the lumbar segments of one rat was not satisfactory, this animal was excluded from the study, thus the final analysis was performed on five intact, four spinal non-trained, and four spinal trained rats.
The rats were injected intraperitoneally with 2% sodium selenite in deionized water (20 mg/kg). After 60 min, animals were deeply anesthetized with sodium pentobarbital (100 mg/kg, i.p.) and injected into the heart with 0.2 unit of heparin (in 0.2 ml PB). The rats were perfused transcardially with 100 ml of ice-cold 0.1 M PB, at a flow rate of 20 ml/min, followed by 200 ml of ice-cold 4% paraformaldehyde in 0.1 M PB. After perfusion, spinal cords were removed from the vertebral column and cut onto segments which were postfixed for 4 h in 4% paraformaldehyde and then cryoprotected in 30% sucrose. The lumbar segments of the spinal cord were embedded in tissue-freezing medium (Leica) in separate molds to preserve their shape, frozen by immersion in cold heptane (-70°C), and stored at -70°C. Transverse sections (20 μm) were obtained on a cryostat, mounted onto gelatine/chrome-alum-coated glass slides, air-dried, and stored at -70°C until histochemical staining.
Synaptic zinc visualization
Spinal cord sections from the lumbar segments obtained from the three groups of rats described above were processed for histochemical staining simultaneously to avoid possible variations in reaction conditions. The histochemical development was performed according to Danscher [24], with slight modification by Czupryn and Skangiel-Kramska [25]. The frozen sections mounted on glass slides were allowed to dry at room temperature and then were progressively rehydrated with descending alcohol solutions (96% ethanol for 15 min, followed by 70% ethanol for 2 min and by 50% ethanol for 2 min), dipped in water, and finally in 0.5% gelatin solution. The slides were then immersed in freshly-prepared developing solution containing 37 mM silver lactate, 0.5 M hydroquinone and 40% arabic gum in 2 M citrate buffer (pH 3.5), and were incubated in the dark for 40 min at 26°C. We assessed the reaction time at this temperature in our preliminary experiments in order to optimize staining intensity. After washing for 20 min in 37°C running tap water, the sections were rinsed twice in deionized water, immersed for 12 min in 5% thiosulfate solution, and then rinsed again in de-ionized water. Finally, the sections were postfixed for 60 min in 70% ethanol, dehydrated in ascending series of alcohols, and coverslipped with Permount (Fisher Scientific).
Sections analysis and quantification of results
The sections processed for different staining were examined using a Nikon Eclipse 80i microscope equipped with a monochromatic CCD camera Evolution VF (Media Cybernetics, Inc., Silver Spring, MD, USA). Image-Pro Plus 5.0 (Media Cybernetics, Inc., Silver Spring, MD, USA) digitizer and software and Neurolucida (MicroBrightField, Inc., Williston, VT, USA) were used for data analysis. The light source was stabilized during image acquisition to maintain the same illumination level at each imaging session, and the settings of the camera and the lamp were constant [26]. For bright field microscopy, shading correction was applied. Brightness and contrast were adjusted to obtain images as close as possible to those observed directly under the microscope. Figures were assembled using Adobe Photoshop software.
Analysis of BDNF labeling
Microscopic images for measurement of BDNF by densitometry were captured during one session, to ensure the same illumination level. At least three sections representing L3 and L4 segments from each rat were chosen for morphometry and densitometry of BDNF immunoreactive profiles of the perikarya, processes, and fibers. Profiles of the motor nuclei of lamina IX were outlined manually according to the rat brain atlas [27] and the densitometric analysis was performed within these outlines (Figure 2A). Only cell bodies with sharp, well-defined edges were taken into account, outlined, and evaluated for each section. The densitometric signal of the background was collected in the area occupied by corticospinal tract fibers and was subtracted from the BDNF signal of the cell bodies in every section. The images collected for densitometric analysis by means of Image-Pro Plus 5.0 were captured at a single focus plane.
Spinal sections were processed with two anti-BDNF antibodies, obtained from Santa Cruz (SC) and from Dr Kaplan. Both antibodies effectively labeled perikarya and processes. However, labeling with Kaplan's antibody produced punctuate labeling which marked perikaryonal edges better than SC antibody. Therefore, it was the antibody of choice for detailed analysis of BDNF IR in the perikarya. The SC antibody strongly labeled the dense network of processes and fibers, enabling detailed analysis of the fiber network.
For fiber network analysis, a skeletonization method using Image-Pro Plus 5.0 software was applied. After thresholding, extracted BDNF-immunopositive processes and objects were smoothed and subjected to "thinning filtering" procedure, which reduced the image to the skeleton. The skeletonized image was than compared to the original one to confirm whether this transformation did not bring substantial artifacts. This analysis allowed measuring (i) the density of extracted BDNF IR processes and fibers, (ii) their length, and (iii) the area occupied by BDNF IR processes and fibers in the tested area. This procedure was sensitive for extracting all strongly-labeled objects (without the cell bodies, which were subtracted in this analysis). The boundaries of the spinal laminae were verified microscopically at Nissl stained cross-sections.
To confirm the results obtained with skeletonization technique, we tracked BDNF IR processes and fibers with Neurolucida 7 software. This analysis was performed in part of the ventro-lateral motor nucleus. An area of about 50 000 μm2 was delineated and all clearly visible BDNF IR processes were tracked along the Z-axis at multiple focal planes. One representative section from the L4 segment per rat was taken for the analysis. Changes in diameter of tracked processes were also taken into account. This analysis allowed measuring (i) the density of BDNF IR processes, (ii) their length, and (iii) volume.
Analysis of synaptophysin immunofluorescence
All images were captured at identical exposure times during one microscopic session in order to ensure the same illumination level. Three sections per rat, separated from each other by at least 240 μm, were chosen for analysis. Analysis of synaptophysin immunofluorescence (IF) was confined to large neurons within lamina IX of the L3-L4 segments of the spinal cord as only their cell bodies had sharp, well-defined edges and visible nuclei. The mean numbers of large neurons per rat selected for further analysis were: 28 in the intact, 33 in spinal, and 25 in spinal trained animals. The Image-Pro Plus software was used to encircle the perimeter of each neuron manually (Figure 2B). Mean optical density within the outlined areas of 4 μm in width was measured. Averaged optical density of neuronal perimeters was calculated for each animal and these results were used for statistical analysis. Additionally, digital images were captured by a confocal inverted microscope Leica DM IRE2 with optical slice of 1 μm, using a HCX PL APO 63× oil-immersion objective lens.
Analysis of double immunofluorescence
Digital images were captured by a confocal inverted microscope (Leica DM IRE2) with optical slice of 0.5 μm, using a HCX PL APO 63× oil-immersion objective lens. To detect and evaluate the contacts and overlap of structures labeled with BDNF/synaptophysin and BDNF/MAP2, optical slices of 0.5 -1.0 μm were digitally merged.
Analysis of synaptic zinc staining
Images of individual sections, stained for synaptic zinc within the ventral quadrants of the L3-L4 spinal segments, were examined under a microscope and the staining patterns were evaluated by two investigators. Typically, 3-4 sections per rat were subsequently acquired for quantitative analysis. Measurements of the relative optical density (R.O.D.) values representing gray level within sections in arbitrary units were performed after importing digital images from Image-Pro Plus 5.0 acquisition system into MCID M4 image analysis system (Imaging Research Inc., Saint Catherine, Ontario, Canada). Zinc staining levels were analyzed within the ventrolateral and ventral funiculi in four rectangular defined areas (7300 μm2 each), and then R.O.D. values were averaged for each section. The area of the pyramidal tract, a region consistently showing low staining level, was taken as a reference structure, and the averaged reference R.O.D. values for each section were calculated. A ratio of R.O.D. values for each section was determined by calculating the mean value from the area of interest divided by values obtained from the reference area, and finally the averaged mean ratio for each rat was calculated.
Statistical analysis
The Kruskall-Wallis analysis followed by the Dunn post-hoc test or one-way Anova followed by the Tukey-Kramer test were used for statistics. The Sign test was also applied to analyze the progress in the locomotor capability of the spinal trained rats. The level of significance was set at p < 0.05.