Biotinylated dextran amine

The experiments were carried out on thirteen (3.5–6.9 years old) rhesus (Macaca mulatta) or cynomolgus (Macaca fascicularis) monkeys (male or female, 3.0–5.0 kg; see Table 1), in accordance to the Guide for Care and Use of Laboratory Animals (ISBN 0-309-05377-3; 1996) and approved by local veterinary authorities, including the ethical assessment by the local (cantonal) Survey Committee on Animal Experimentation and a final acceptance delivered by the Federal Veterinary Office (BVET, Bern, Switzerland). The monkeys were obtained from our own colony in our animal facility (Macaca fascicularis) or were purchased (Macaca fascicularis and Macaca mulatta) from a certified supplier (BioPrim; 31450 Baziège; France), with the authorization to import the animals delivered by the Federal Veterinary Office (BVET, Bern, Switzerland). Three recent reports (Wannier et al., 2005 (link); Freund et al., 2006a (link), 2007 (link)) describe the behavioral task analyzed in the present report (‘modified Brinkman board’ task; see Fig. 1E and also http://www.unifr.ch/neuro/rouiller/motorcontcadre.htm), the surgical procedures (including transection of the CS tract in the cervical cord at C7/C8 level), the treatment with the anti-Nogo-A (n = 7) or control (n = 6) antibodies and the neuroanatomical investigations (including assessment of spinal lesion location and extent). The antibodies’ characteristics and penetration in the central nervous system have been reported elsewhere (Weinmann et al., 2006 (link); Freund et al., 2007 (link)). As previously reported in detail (Schmidlin et al., 2004 (link), 2005 (link); Wannier et al., 2005 (link); Freund et al., 2006a (link), 2007 (link)).
The present study includes the same twelve previously reported monkeys (Freund et al., 2006a (link), 2007 (link)) and a thirteenth monkey (Mk-AK), on which the experiment was completed later, and in which anti-Nogo-A antibody infusion was initiated 7 days post-lesion (Fig. 1A), in contrast to immediate infusion the day of the lesion in the other six anti-Nogo-A antibody-treated monkeys. However, in this thirteenth monkey (Mk-AK), although an osmotic pump was implanted immediately after the lesion (as was the case for all the other monkeys), only saline (NaCl 0.9%) was delivered during the first week, and delayed administration of the anti-Nogo-A antibody started 1 week post-lesion. In all cases, the antibody was delivered for a period of 4 weeks.
The monkeys’ identification codes refer to individual monkeys (Table 1 in Freund et al., 2006a (link)) and comprise, for the sake of clarity, a ‘C’ or an ‘A’ in the fourth character position, indicating whether the monkey was control antibody-treated or anti-Nogo-A antibody-treated, respectively. However, during the course of the experiments the animals had different names from which the experimenter could not deduce which antibody was infused, at least for the monkeys in which the experimenter-blind procedure was applied (Table 1).
Monkeys were housed in our animal facilities in rooms of 12 m³, each usually containing 2–4 monkeys free to move in the room and to interact with each other. In the morning, before behavioral testing, the animal keeper placed the monkeys in temporary cages for subsequent transfer to the primate chair. The monkeys had free access to water and were not food-deprived. The rewards obtained during the behavioral tests represented the first daily access to food. After the tests, the monkeys received additional food (fruits and cereals). The dexterity of each hand was assessed in all lesioned monkeys with a finger prehension task, specifically our modified Brinkman board quantitative test (Fig. 1E; see also Rouiller et al., 1998 (link); Liu & Rouiller, 1999 (link); Schmidlin et al., 2004 (link)). The tests were conducted using a Perspex board (10 cm × 20 cm) containing 50 randomly distributed slots, each filled with a food pellet at the beginning of the test (home-made behavioral apparatus). Twenty-five slots were oriented horizontally and twenty-five vertically. The dimensions of the slots were 15 mm long, 8 mm wide and 6 mm deep. Retrieval of the food pellets required fractionated finger movements, in order to produce an opposition of the index finger and the thumb, which corresponds to the precision grip. This manual prehension dexterity task was executed daily, alternatively with one and the other hand, four or five times per week for several months before and after the unilateral cervical cord lesion. A daily behavioral session typically lasted 60 min. The performance of each hand was videotaped. In the present study, two parameters were assessed: (i) the retrieval score, i.e. the number of wells from which the food pellets were successfully retrieved and brought to the mouth during 30 s, separately for the vertical and the horizontal slots; (ii) the contact time, defined as the time of contact (in s) between the fingers and the pellet, calculated for the first vertical slot and the first horizontal slot targeted by the monkey in a given daily session (see also paragraph 2 in the Results section). The contact time is comparable to the prehension time as introduced by Nishimura et al. (2007) (link) in parallel to the present study, but for a different grasping task. In our previous study, the retrieval score represented the primary outcome measure from the modified Brinkman board task (Freund et al., 2006a (link)) and was determined by the total number of pellets retrieved in 30 s. In the present study separate scores are provided for vertical and horizontal slots. The contact time data and the bivariate and trivariate statistical analyses (see below) represent secondary outcome measures from the modified Brinkman board task, newly introduced in the present report.
After the monkeys reached a level of performance corresponding to a plateau (usually after 30–60 days of initial training), we used the score from 30–50 daily sessions to establish a pre-lesion behavioral score for each monkey. A unilateral cervical cord lesion was performed in thirteen monkeys as follows. Intramuscular injection of ketamine (Ketalar® Parke-Davis, 5 mg/kg, i.m.) was delivered to induce anesthesia followed by an injection of atropine (i.m.; 0.05 mg/kg) to reduce bronchial secretions. In addition, before surgery, the animal was treated with the analgesic Carprofen (Rymadil®, 4 mg/kg, s.c.). A continuous perfusion (0.1 ml/min/kg) through an intravenous catheter placed in the femoral vein delivered a mixture of 1% propofol (Fresenius®) and a 4% glucose solution (1 volume of propofol and 2 volumes of glucose solution) to induce a deep and stable anesthesia. The animal's head was placed in a stereotaxic headholder, using ear bars covered at their tip with local anesthetic. The surgery was carried out under aseptic conditions, with continuous monitoring of the following parameters: heart rate, respiration rate, expired CO₂, arterial O₂ saturation and body temperature. In early experiments, an extra intravenous bolus of 0.5 mg of ketamine diluted in saline (0.9%) was added at potentially more painful steps of the surgical procedure (e.g. laminectomy) whereas, in later experiments, ketamine was added to the perfusion solution and delivered throughout surgery (0.0625 mg/min/kg). The animal recovered from anesthesia 15–30 min after the propofol perfusion was stopped, and was treated post-operatively with an antibiotic (Ampiciline 10%, 30 mg/kg, s.c.). Additional doses of Carprofen were given daily (pills of Rymadil mixed with food) for about 2 weeks after the surgery. Following the cervical cord lesion, the animal was kept alone in a separate cage for a couple of days in order to perform a careful survey of its condition. The details of surgical procedures and lesioning are available in previous reports (Schmidlin et al., 2004 (link), 2005 (link); Wannier et al., 2005 (link); Freund et al., 2006a (link), 2007 (link)).
After lesion, and following the period of recovery lasting generally 30–40 days, a post-lesion level of performance corresponding to a plateau was established, based on a block of ten behavioral sessions (usually the last ten sessions conducted). For the retrieval score, functional recovery was expressed quantitatively as the ratio (expressed as a percentage) of the post-lesion average retrieval score value to the pre-lesion average score value. Because contact time was measured only for the first vertical and first horizontal slots targeted by the monkey, in order to minimize the impact of outliers the pre-lesion and post-lesion contact time was assessed as the median value (Fig. 3C and D). Considering that good performance is reflected by a short contact time (in the pre-lesion condition), post-lesion performance (recovery) was expressed quantitatively as the ratio (expressed as a percentage) of the pre-lesion median contact time to the post-lesion median contact time. For measures of both recovery of score and contact time, if the calculated values exceeded 100% (i.e. post-lesion performance was better than pre-lesion performance), the recovery was considered to be complete and therefore expressed quantitatively as 100%.
Besides the new behavioral parameter of contact time introduced here, the present study also comprises a new analysis regarding the lesion size. In our previous reports (Freund et al., 2006a (link), 2007 (link)), the extent of the lesion was expressed as a percentage of the corresponding hemi-cord surface, as assessed from a 2-D reconstruction of the lesion in the transverse plane (see Fig. 1B and C). These standard values of lesion extent have been considered here again in Figs 2 and 4 (see also Table 1). The present study expands upon these data by further calculating the estimated volume of the cervical lesion in order to consider the extent of the lesion in 3-D. After completion of the post-lesion behavioral analysis (see below), the monkeys were killed and prepared for histology as follows. Each monkey was pre-anaesthetized with ketamine (5 mg/kg, i.m.) and given an overdose of sodium pentobarbital (Vetanarcol; 90 mg/kg, i.p.). Transcardiac perfusion of saline (0.9%) was followed by paraformaldehyde (4% in phosphate buffer 0.1 m, pH 7.4), and 10, 20 and 30% solutions of sucrose in phosphate buffer. The brain and the spinal cord were dissected and stored overnight in a solution of 30% sucrose in phosphate buffer. Frozen sections (50 μm thick) of the cervical cord (approximately segments C6-T3) were cut in the parasagittal longitudinal plane and collected in three series for later histological processing (see below).
Using an ad hoc function of the Neurolucida software (based on the Cavalieri method; MicroBrightField, Inc., Colchester, VT, USA), the volume of the cervical lesion (in mm³) was extrapolated from the reconstructions of the lesion on consecutive histological longitudinal sections of the cervical cord (see Table 1). The volume measurement of the cervical lesion was conducted on one out of three series of sagittal sections (50 μm thick), treated immunocytochemically with the SMI-32 antibody (Covance, Berkeley, CA, USA), as previously reported (Liu et al., 2002 (link); Beaud et al., 2008 (link); Wannier-Morino et al., 2008 (link)). The epitope recognized by the SMI-32 antibody lies on nonphosphorylated regions of neurofilament protein and is only expressed by specific categories of neurons (Campbell & Morrison, 1989 (link); Tsang et al., 2006 (link)). The other two series of sections were processed to visualize biotinylated dextran amine (BDA; Invitrogen, Molecular Probe, Eugene, OR, USA) and fluorescein dextran amine (Invitrogen, Molecular Probe, Eugene, OR, USA) staining, resulting from injections of BDA in the contralesional motor cortex and fluorescein dextran amine in the ipsilesional motor cortex (see Freund et al., 2006a (link), 2007 (link)). Measurements of volume of the cervical lesion were also conducted on sections processed for BDA but the lesion contour was not as well defined as on the SMI-32-stained sections, where a clear scar region could be distinguished from a penumbra lesion at the periphery of the lesion (yellow and red outlines in Fig. 1D). The scar region was characterized by a dense fibrous tissue or granulous tissue forming a central zone of the lesion where the SMI-32 staining was absent. The lesion volume data presented (Table 1, Fig. 5) and considered for statistical analysis (Table 2) are the measurements corresponding to the scar as seen on the SMI-32-stained sections.
Because of the limited number of animals, two independent statistical tests were used to compare the group of control antibody-treated monkeys (n = 6) with the group of anti-Nogo-A antibody-treated monkeys (n = 7). The first test (based on a linear Fisher discriminant analysis) takes into account one of the two parameters reflecting the size of the lesion (i.e. the extent of hemi-cord lesion or the volume of the lesion) and one of the four parameters reflecting the percentage of functional recovery (score for vertical slots, score for horizontal slots; contact time for vertical slots or contact time for horizontal slots), and thus is aimed at assessing the overlap or segregation between the two groups of data (Figs 2E and F, and 4C and D). The test provides maximal separation between the groups (see Everitt, 2005 ) in the form of a linear function of the observed variables such that the ratio of the between-groups variance to its within-group variance is maximized. We used the R package to get the two lines plotted in each of Figs 2E and F, and 4C and D. Line 1 (dashed line) yields maximal separation and the projected samples are provided on the orthogonal line 2 (solid line). For better visualization, line 2 was proportionally enlarged and positioned vertically on the right side of the graph (green arrows). With respect to the statistics, the sample size does not permit an assumption of normality so we considered the statistical problem of separation of the projected samples using the nonparametric Mann–Whitney U-test. The obtained results are summarized in Table 2 (row A, bivariate analysis).
The second statistical test (the trivariate analysis) examined the three-dimensional data produced by differences in ‘recovery of scores’ (number of pellets retrieved, as illustrated in Fig. 2), ‘recovery of contact time’ (time to grasp first pellet, as illustrated in Fig. 4) and ‘lesion extent’, using a nonparametric multivariate rank test (Oja & Randles, 2004 ). This test includes all three parameters and can be considered an index of overall functional recovery. We assumed two independent random samples from bivariate distributions F(x-c1) and F(x-c2) located at centers c1 and c2, and tested the null hypothesis that there was no effect of treatment (i.e. c1 = c2 versus the alternative c1 is different from c2). Data were transformed to make the test affine-invariant, to ensure a consistent performance over all possible choices of coordinate system, and then projected onto a sphere where a rank test was performed. As the law for this test is still unknown, we used Monte-Carlo simulations to compute the P-value. The obtained results are summarized in Table 2 (row B, trivariate analysis). A complete description of these bivariate and trivariate statistical analyses, applicable also to other types of lesions and to other behavioral tests of manual dexterity in primates, will be reported elsewhere in a methodological report. The same two statistical analyses (bivariate and trivariate tests) were applied in a similar way as above for the estimated volume of the lesion (Table 2, rows C and D).