A spinal cord injury (SCI) is a devastating condition characterized by a sudden loss of spinal functions distal to the level of trauma, including sensory, motor, and autonomic function. These effects are due to various traumatic or non-traumatic factors. The regenerative capacity of the injured spinal cord is extremely limited in adult mammals [1
]. Patients suffer various disabilities which greatly impact quality of life [3
]. The deficiency of spinal regenerative ability after an SCI has been attributed to various alterations in the local microenvironment, including: (1) the accumulation of many inhibitory factors by the degeneration of neurons, axon, and myelin; and (2) the lack of a sufficient number of upregulated or activated neuritogenic-associated genes and proteins [4
]. At present, although there are many attempts for the treatment of the SCI, such as IL-6, stem cell transplantation [9
], etc. few are actually available for clinical practice. Therefore, endeavors towards new strategies to revitalize neuritogenic ability following spinal injuries are being continuously made.
Botulinum toxin A (BoNT/A) is one of seven botulinum toxins (categorized based on serologic features) and is one of the most potent toxins known [11
]. The toxicity of BoNT/A is due to its selective binding to the membranous receptor located on the presynaptic membrane of the motor endplate. This leads to the cleavage of the synaptosomal-associated protein 25 kDa (or SNAP-25) which blocks the release of the neurotransmitter acetylcholine contained in the vesicles [15
]. This results in the paralysis of skeletal muscle. Patients die of severe respiratory failure due to a lack of respiratory muscle contractions [17
]. Besides the aforementioned mechanisms of BoNT/A, it has been verified that it can also inhibit or block the release of other components in synaptic vesicles of different neurons, such as the calcitonin gene related peptide (CGRP) and substance P in sensory neurons [18
]. Because of the actions of BoNT/A on neurotransmitters, it has been used as therapy to relieve hyper-reactive muscle dysfunctions (e.g., torticollis and hemifacial spasm), various types of chronic pain, hyperhidrosis, etc. [23
Structurally, BoNT/A consists of two polypeptide chains linked by a disulfide bond. The longer chain (MW = 100 kDa) is called a heavy chain (HC) of BoNT-A; it is mainly responsible for binding to target cells and mediating the endocytosis of the toxin. The other chain with a smaller molecular weight is a light chain (LC, MW = 50 kDa). The LC is a zinc-dependent proteolytic enzyme which contributes to the cleavage of SNAP-25 at the neuromuscular junction [15
]. It had been observed by some investigators that new axonal branches formed around skeletal muscles that had been paralyzed by BoNT/A for three to four months [27
]. Meanwhile, the observations in vitro showed that BoNT/A was able to promote neuritogenesis in mouse motor neuron cultures [30
]. The observation of neurite sprouting associated with BoNT/A in vivo or in vitro could not simply be attributed to the toxic role of BoNT-A. Moreover, the observations form Coffield’s lab also showed that the isolated binding domain heavy chain (BoNT/A HC) has the same role as BoNT/A [30
]. The signal pathways relating to the binding of the heavy chain and the receptor might also be a reasonable alternative explanation.
In our previous studies, the increased phosphorylation of Akt and ERK1/2 (growth- and survival-related intracellular signal proteins) was paralleled with neuritogenesis when adding BoNT/A heavy chain into neuro-2a cell cultures (data published in Chinese). In that study, an obvious increase in phosphorylated-ERK 1/2 was seen between one to five hours after adding 1 nmol/L of BoNT/A HC (p
< 0.05), and an increased protein level of phosphorylated-Akt was observed mainly at 15 and 60 min (p
< 0.05). Based on the results, it was concluded that the BoNT-A heavy chain stimulates neuritogenesis. The neuritogenic mechanism of it on Neuro-2a cells might be due to the activation of ERK1/2 and Akt phosphorylation. Meanwhile, it is reasonably presumed that BoNT/A heavy chain exerted its biological roles by modulating many intracellular signals and proteins expression, some of which may be involved in neural growth and regeneration; others are possibly related to growth inhibition. Among of these proteins, growth-associated protein-43 (GAP-43) and superior cervical ganglion 10 (SCG 10) have already been evidenced to be closely related to the axonal growth and neural regeneration [31
]. Therefore, commercial recombinant BoNT/A HC was applied in this study to explore the effect of BoNT/A HC on the expression of some selective growth-associated proteins, as well as its role in neurite outgrowth in rat lumbar SCI model in vivo. Meanwhile, the motor and sensory functions of the ipsilateral hindlimb were evaluated after BoNT/A HC treatment as a therapeutic regent.
The stimulatory effect of BoNT/A HC on neurite outgrowth has previously been verified in vitro. In this work, the intermittent administration of BoNT/A HC through intrathecal injection on a lumbar unilateral SCI rat model was used to determine the role of BoNT/A HC in the local expression of growth-associated proteins. It was then explored whether, in vitro, BoNT/A HC was able to promote re-growth by upregulating the expression of growth-associated proteins after an SCI. Unlike the peripheral nervous system (PNS), the central nervous system (CNS) is deficient in its ability to regenerate after injury [34
]. Based on previous research, the CNS has difficulty regenerating due to the abundance of myelin inhibitory factors and the shortage of macrophages [6
]. Another possible explanation is the limited or arrested initiation of intrinsic neuron growth programs in the injured CNS [7
]. Because the spine is part of the CNS, most patients suffer from severe spinal dysfunction and lifelong disability due to various pathogeneses when SCIs develop. Therefore, one of the principal foci of regenerative medicine after an SCI is to explore strategies and methods which can influence the improvement of the spinal microenvironment (e.g., the exclusion of myelin inhibitory factors) or revive the intrinsic regenerative ability of neuron itself [36
The human recombinant BoNT/A HC used in this study is a non-toxic peptide, unlike the original potential holotoxin. The results evident after treatment with BoNT/A HC can be explained by its ability to bind to membranes and the activation of related intracellular signals upon the attachment of the heavy chain to the receptor. A previous study from the same research team has demonstrated that the BoNT/A HC had a stimulatory role in promoting neurite outgrowth in Neuro-2a cell cultures. It is known that the basis for neurogenesis and neuritogenesis is the synthesis of proteins; even the myelin inhibitory factors that accumulate after SCI can be regarded as a secondary phospholipoprotein. Therefore, it can be assumed that there is a significant modification in protein expression after an SCI, and the administration of the BoNT/A HC following an SCI should have an input in protein expression. Some definite results were obtained from this study. First, the application of BoNT/A HC reversed the alterations caused by the SCI, such as the increased expression of some proteins when the SCI developed. When the BoNT/A HC was given at the same time as the injury established, the increase dropped towards the levels of the controls, which is consistent with our previous study (published in Chinese). Briefly, the main difference in the previous study are as foolows: (1) the SCI model was established with a modified needle inserted into the lumbar spinal cord at the same anatomical location as this study; and (2) different doses of BoNT/A heavy chain (2 μg, 4 μg, 6 μg and 8 μg) was injected one-time into the spinal cord cavity while the injury made. The second result observed in this study is that BoNT/A HC treatment magnified the expression of some proteins evident from the SCI, for example, the expression of GAP43 and SCG10 were obviously increased after BoNT/A HC treatment. In this study, the selection of GAP43 and SCG10 as markers (to detect the efficiency of BoNT/A HC on stimulating protein regeneration) is based on the changes of protein molecular weight on the two-dimensional gel and the Western blot. In fact, the levels of both GAP43 and SCG10 displayed a certain elevation shortly after the SCI (two days post-injury), but these declined as the post-injury period progressed. However, when the SCI model was treated with BoNT/A HC, both proteins exhibited a continuous increased expression until the end of the experiment.
The increased expression of GAP43 after injury has been considered by most researchers to be actively involved in axonal regeneration [37
]. GAP43 is mainly involved in the sprouting and regeneration of mature axons in their phosphorylated form. Its expression was upregulated following nerve injury. SCG10, also known as stathmin-like 2 (STMN2) protein, is another regenerative protein [42
]. It is mainly involved in axonal microtubule dynamics and protein transport [43
]. The initial increase in GAP43 and SCG10 expression after the SCI, without the application of BoNT/A HC, can be explained as a limited regenerative response after spinal cord injuries, however the response ceased when the injury continued to exist. The mechanism might be related to the intervention of myelin inhibitory factors binding to their specific receptors. The continuous enhancement of GAP43 and SCG10 expression when BoNT/A HC is applied is evidence of the recombinant peptide’s involvement in promoting nerve regeneration after injury. The study also gave an outline of the role BoNT/A HC plays in stimulating the in vivo sprouting of neuronal processes. When BoNT/A HC treatment was administered during the SCI’s development, the length and the number of neuronal process around injury site (in both its rostral and caudal parts) increased compared to measurements in the SCI-only group. The percentage of neurons with processes was also greater than that in the SCI-only group. Additionally, it was found that the spinal motor function at the ipsilateral hindlimb was improved due to the efficiency of BoNT/A HC in the upregulation of growth-associated proteins and in the promotion of neuronal process re-growth. The ineffectiveness of BoNT/A HC on improving spinal sensory function might be attributed to the difference in the expression of the heavy chain binding receptor in sensory neurons. Besides, direct damage of the sensory neurons in injury location might contribute to the difference of sensory and motor after SCI with or without BoNT/A HC treatment. Therefore, more detailed information about these should be explored after all in the future.
5. Materials and Methods
5.1. Establishing the Rat Spinal Cord Hemi-Section Injury Model
All experimental procedures were done in accordance with the National Institute of Health’s Guide for the Care and Use of Laboratory Animals, and were approved by the Ethics Committee of Animal Research at Shanxi Medical University (IACUC2017-001) on 20 January 2017. Every effort was made to minimize animal suffering and the number of animals sacrificed.
Sprague-Dawley rats (6–7 weeks old, and weighing between 200 and 220 g) were provided by Beijing Haidian Experimental Animal Farm (No. SCXK (Beijing) 2014-0013). As there is no difference regarding spinal cord injury study, male rats were used in this research. The rats were deeply anaesthetized with 1% pentobarbital sodium (at a dose of 4 mL/kg body weight) and given a laminectomy at vertebral level T9/10 to expose the spinal cord. Brieﬂy, the T9 and T10 vertebrae were first orientated using the iliac crest as an anatomical landmark. The lumbar spinal cord was exposed by removing the T9 vertebra dorsally. A unilateral (left) lumbar spinal injury was achieved by nipping against the left side of the dorsal median vein with the tip of a pair of fine forceps. Left hindlimb paralysis was regarded as the sign of success in achieving a unilateral SCI. Meanwhile, the loss of motor function (based on hindlimb grasp power measurement) and the abnormality of sensory function (based on an assay of thermal hyperalgesia) after surgery were used to evaluate the severity of spinal cord function impact. The animals were placed in a temperature-controlled (24 °C ± 1 °C) chamber under a 12 h light/dark cycle. Standard amounts of food and tap water were given daily.
5.2. BoNT/A Heavy Chain Administration and Groups
Recombinant BoNT-A HC was purchased from List Biological Laboratories Inc. (Campbell, CA, USA).
Intermittent administration of BoNT/A HC applied in the study was done via two routes: (1) local application of the BoNT/A HC directly onto injury site (4 μg/μL in 16 μL saline); and (2) administration of BoNT/A HC via a lumbar intrathecal catheter [46
]. After establishing the rat model, 2 μg/μL of BoNT/A HC were administered every week to the SCI in the BoNT/A HC treatment groups. Animals were divided into the following groups, with six rats in each group: (1) the control (or pseudo surgery) group for which the skin incisions were made, the laminae were cut, and the spinal cords were exposed but not injured; (2) the SCI-only group for which unilateral (left side) lumbar spinal cord injuries were made (to prevent the vehicle intervention on BoNT/A heavy chain, the same volume of sterile saline was applied in SCI-only animal with same method and period); and (3) the SCI with BoNT/A HC treatment, in which BoNT/A HC was administered every week, and the injury was made. This group was further divided into four groups (with six rats each) based on different periods post-injury, i.e., two days, one week, two weeks and four weeks. The groups of animals are summarized in Table 2
5.3. Two-Dimensional Gel Electrophoresis
The spinal cords were fully exposed when the experiment’s period was completed. Then, the segment of the cord containing the spinal lesion center and the rostral and caudal area (6 mm in total) was collected. Spinal protein was extracted using a pre-cooled lysis buffer (urea: 21 g, thiourea: 7.9 g, DTE: 0.5 g, Tris: 0.5 g, in 50 mL of DD H2O). The protein concentration was determined using the Bradford protein assay.
For the two-dimensional gel, two 150 mg protein samples from each group were loaded onto IPG strips (NL) for first-dimension isoelectric focusing (IEF). The IPG strips were then washed and equilibrated in a 2D equilibration buffer, and then they underwent a 12.5% sodium dodecyl sulfate polyacrylamide gel vertical electrophoresis. After that, the gel was stained with 0.1% silver nitrate. The gel was imaged and analyzed with the Bio-Rad gel imaging system. Some interesting protein dots from different group were chosen and compared upon their changes on size and density at the same level of molecular weight and the isoelectric point (pI). For statistical analysis, two gels were prepared. The selective dots were circled and assessed by the integrated optional density (IOD) using ImageJ software in a double-blinded way and two people who were not related to the study were asked to do the assessment. In this way, four values four each dot were obtained and analyzed.
5.4. SDS-PAGE and Western Blot
Twenty micrograms of protein homogenates were subjected to 15% sodium dodecyl sulfate polyacrylamide gel electrophoresis. Subsequently, the proteins were transferred onto polyvinylidene difluoride (PVDF) membranes for Western blotting. The PVDF membranes were blocked with 5% non-fat milk powder in a TBST wash buffer (Tris buffered saline containing 1% Tween-20) for 1 h at room temperature. The selective primary antibodies were applied onto the membrane and incubated at 4 °C overnight. These antibodies were the following: rabbit anti-phosphorylated GAP43 antibody (1:3000, Thermo Fisher, Waltham, MA, USA); rabbit anti-SCG10 polyclonal antibody (1:1000, Thermo Fisher, Waltham, MA, USA); and rabbit anti-GAPDH polyclonal antibody (1:5000, Bioword, Nanjing, China). The following day, the membranes were incubated with HRP-conjugated secondary antibodies (goat anti-rabbit IgG, 1:1000, Bioword, Nanjing, China) for 1 h at room temperature. The membranes were then immersed into the EasySee Western blot kit (DW101, TRAN, Beijing, China; A solution: B solution = 1; 1 plus 2 μL C solution) and made to undergo the reaction for 3 min in dark. X-ray film which corresponded to the membrane was developed in a darkroom. The protein bands were imaged and analyzed for the variation in integrated optional density (IOD), while GAPDH bands as a loading control, with Bio-Rad gel imaging system.
5.5. Immunofluorescence Staining
At the set time point, rats from each group were anesthetized lethally with 1% pentobarbital sodium. They then underwent a 30 min transcardial perfusion with ice-cold 4% paraformaldehyde for fixation purposes. A 1 cm segment of spinal cord containing the lesion center was dissected out. The samples were post-fixed in the same fixatives for another hour and subsequently immersed in a 30% sucrose solution at 4 °C overnight. After being embedded onto an OCT compound, a section of spinal cord 16 μm in thickness across the lesion site was cut dorsal-coronary-longitudinally by cryostat (1950-Cryostat, Leca, Nussloch, Germany).
The sections were penetrated with 0.1% Triton-X100 for 10 min, washed three times with 0.1 mol/L PBS buffer, and blocked with a blocking buffer (10% goat or donkey serum and 0.1% Triton X-100 in PBS) for 1 h. Primary antibodies (as previously listed) were incubated with the section at 4 °C overnight. The following day, the sections were incubated with Alexa Fluoro-594-labeled goat anti-rabbit IgG (1:500; Life Technologies, Shanghai, China) and Alexa Fluoro-488-labeled donkey anti-rabbit IgG (1:500; Life Technologies, Shanghai, China) in the dark for 1 h, at room temperature. The sections were then mounted with an antifade mountant with DAPI (2-(4-amidinophenyl)-1H-indole-6-carboxamidine, Life Technologies, Grand Island, NY, USA). These were then checked and photographed under a fluorescence microscope (Olympus IX71, Tokyo, Japan).
5.6. Measurement of Neuronal Processes
The immunofluorescence staining of SCG 10 was applied to 6–8 sections each samples. Approximately 5–7 of images were captured of each section. The imaging area was mainly at the periphery of the spinal cord lesion. The measurements included the total length of neurites, the number of neurites, and percentage of cells with neurites in all the immunofluorescence positive cells in the image. These were taken using ImageJ software. The identification of neurites was determined and confirmed according to literature [47
]. Neurites were traced from the cell body to the end of the process, and the total length was calculated for each process. Neurites were distinguished according to their soma diameter; this meant that only neurites with a length greater than the maximum diameter of the cell would be considered a real neurite.
5.7. Behavioral Evaluation
Six rats in each group were used for assessment of motor and sensory function of ipsilateral hindlimb. Alteration in sensory and motor function at the ipsilateral hindlimb of rats with an SCI or an SCI with BoNT/A HC treatment were evaluated using the thermal hyperalgesia test and the grip power test, respectively. The thermal hyperalgesia test was completed using thermal stimulation of the PWTL by the paw thermal radiation stimulation tester (SERIES 8; RWD Life Science). The response time (in seconds) for the ipsilateral paw to withdraw because of the thermal stimulus was recorded. In fact, the response time represented the latency of the stimulation of the rat’s foot (accurate to 0.1 s). To avoid tissue damage caused by long time thermal stimulation, 25 s was regarded as the maximum time of PWTL. At every time point, the average value was obtained using at least three measurements. A 10 min interval was suggested between taking these measurements. The values from three or more assessments was applied to calculate the average ± SD at a certain time point.
The grip test was performed using the rats grip tester based on the manufacturer’s instructions (YLS-13A, Zhishuduobao, Baoding, China). First, wrap two fore-limbs and one counter-lateral hindlimb with tape, leaving the ipsilateral hind-limb free. Then, the rat was placed gently on the grasping force tester, the rat’s tail was grabbed and dragged backwards. The power (in grams) of the ipsilateral hind paw in holding the grip was automatically recorded by the instrument as the maximum rat grip power; this stands for its motor function. The value was the average of three measurements, with 15 min intervals between each measurement.
5.8. Statistical Analysis
The data were expressed as mean and standard error (SEM). The protein bands or protein spots were quantified by ImageJ software. Data were analyzed using GraphPad Prism 5.0 software and one-way ANOVA was used to analyze the significance. p < 0.05 was considered statistically significant.