QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online July 23, 2003

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Zottoli, S. J. Articles by Freemer, M. M. Search for Related Content PubMed PubMed Citation Articles by Zottoli, S. J. Articles by Freemer, M. M.

Recovery of C-starts, equilibrium and targeted feeding after whole spinal cord crush in the adult goldfish Carassius auratus

S. J. Zottoli^* and M. M. Freemer

Department of Biology, Williams College, Williamstown, MA 01267, USA

View larger version (25K): [in a new window]   Fig. 1. Schematic diagram of the behavioral testing apparatus used to elicit and analyze goldfish C-starts. Light is projected from above and fish images are captured from below the tank by both a video camera and a matrix camera. Goldfish are stimulated with a vibratory stimulus created by lifting the whole tank with a solenoid. The computer simultaneously triggers the solenoid and the matrix camera, which starts saving silhouettes at a rate of 500 silhouettes s-1 (i.e. every 2 ms). These fish silhouettes are stored in the expanded memory of a controller and then loaded onto the computer hard drive for analysis.

View larger version (25K): [in a new window]   Fig. 2. Method of analysis of the goldfish C-starts. (A) Superimposition of silhouettes of a goldfish during a C-start. Silhouettes were taken every 2 ms but, for clarity, only those occurring at 14-ms intervals starting at the beginning of the response are shown. The lightest image is the silhouette at the beginning of the response and it has been arbitrarily oriented with the nose (black circle) upwards. The darkest image is the fish silhouette at approximately 104 ms after the stimulus delivery. (B) Superimposition of midlines determined from silhouette images. The silhouettes were reduced to a midline a single pixel thick using a thinning algorithm. For clarity, every fourth midline is shown (i.e. every 8 ms). The first midline is at the start of the response. (C) The rostral portion of each midline, corresponding to the rostral 40% of the midline (Nissanov, 1991). Regression lines are shown in 2 ms increments. The regression line that begins stage 2 and the line 70 ms after the start are labeled. The angle at the beginning of stage 2 is formed between the regression line at start and the regression line at the stage 2 latency. The escape trajectory angle (ETA) is formed between the regression line at start and the regression line 70 ms later. The straight-line distance that the center of mass travels during 70 ms after the start is delineated as well.

View larger version (81K): [in a new window]   Fig. 3. Completeness of the spinal cord crush wound. (A) Schematic diagram of the dorsal view of a goldfish brain to show the location of the crush site at the spinomedullary level (SML). Biocytin was applied to a cut wound (1-2 mm caudal to the SML; not shown) 8 days after a crush wound at the SML to determine if any fibers were spared damage (see text for details). For reference, the Mauthner cells (M-cells) and their axons (M-axons) are superimposed on the diagram. Ce, cerebellum; FL, facial lobe; VL, vagal lobe; C, caudal; R, rostral. (B,C) Photomicrographs of 60-µm transverse tissue sections. The brain midline is centered on and the ventral edge of the brain is located at the bottom edge of each photomicrograph. (B) Section 1.5 mm rostral to the crush wound. The ventral area shown here has no biocytin-filled fibers and is representative of the entire medulla oblongata at this level. In fact, this fish had no biocytin-filled fibers rostral to the wound site, indicating that all fibers were damaged. (C)Section near the site of the SML crush. A few biocytin-filled fibers can be seen at this level. Scale bar, 100 µm for B and C.

View larger version (16K): [in a new window]   Fig. 4. Fish survival and behavioral classification of goldfish after whole spinal cord crush. (A) Behavioral status of goldfish 190 days after spinal cord crush. Of the 25 surviving fish, 12 regained equilibrium and C-starts (48%), two regained equilibrium but not C-starts (8%), six did not gain equilibrium and had no apparent body abnormalities (24%; a C-start could be elicited in one of these fish), while five had body abnormalities (20%) that are believed to have hindered behavioral recovery. (B) Distribution of the number of experimental fish that died over the 190-day postoperative interval. Twenty of the 45 experimental fish died during the course of this study. The distribution of the number of fish that died reveals two clusters: one at 3-52 days and another at 91-162 dayspostoperatively.

View larger version (27K): [in a new window]   Fig. 5. Time course of behavioral recovery after whole spinal cord crush at the spinomedullary level (SML). The percentage of fish that recovered a particular behavior (# fish that recovered a behavior/total number that ultimately recovered the behavior) was calculated for each 10-day interval up to 190 postoperative days. Although there was a great deal of variability between fish as to when a particular behavior returned, return of pectoral and/or pelvic fin movement was generally first and was followed by targeted feeding, partial equilibrium, full equilibrium and C-starts. Eight sham-operated control fish targeted food, maintained full equilibrium and displayed C-starts throughout the postoperative interval (not shown).

View larger version (18K): [in a new window]   Fig. 6. Probability of eliciting C-starts from sham-operated control and experimental goldfish to a vibratory stimulus over six 25-day postoperative intervals (T1-T6). The mean probability and S.D. are shown for experimental fish that recovered C-starts and for eight sham-operated control fish. The first experimental responses occurred in two fish between 51 and 75 days postoperatively. C-starts could be elicited from 11 of the 12 fish by 150 dayspostoperatively. One other fish regained C-starts on the 190th postoperative day (not shown).

View larger version (28K): [in a new window]   Fig. 7. Diverse trajectories of recovered C-starts from three experimental fish. Regression lines are shown for C-starts of three fish. Each regression line represents the position of a fish every 2 ms for a total of 104 ms, as described in Fig. 2. (A) A double C-start from fish 135, 120 days postoperatively (1st recovered response was elicited on day 64). (B) Two C-starts from fish 138, 120 days and 134 days postoperatively (1st recovered response was elicited on day 92). (C) Three C-starts from fish 112. Two were elicited 142 days postoperatively (C1,C2) and the third on 198 days postoperatively (1st recovered response was elicited on day 72).

View larger version (23K): [in a new window]   Fig. 8. In general, experimental fish displayed C-starts that were more restricted in their extent than those elicited preoperatively or by sham-operated control animals. Each regression line represents 2 ms as described in Fig. 2. Three consecutive startle responses preoperatively (A1-3) of fish 152 are compared with three responses (B1-3) elicited between 95 and 109 days postoperatively. The most obvious difference is the angle achieved 70 ms after the delivery of the stimulus. The mean angle of the trials in A is 134.1° while that in B is 53.8°.

View larger version (27K): [in a new window]   Fig. 9. The distribution of C-start kinematic parameters in control and experimental fish. All data up to 150 days postoperatively (T1-T6) are combined on the graph. Latency from the stimulus to the response (A), the straight-line distance that the center of mass traveled during 70 ms after the start of the response (B) and the velocity of the straight-line center of mass movement (C) are presented. Although there is no overlap between experimental and control latency values, there is substantial overlap with other parameters.

View larger version (13K): [in a new window]   Fig. 10. The occurrence of C-starts with no second stage was greater in experimental fish than in sham-operated control fish. Each regression line represents 2ms as described in Fig. 2. (A) A C-start with no second stage from a control fish. (B) A C-start from fish 112, 114 days postoperatively. (C) A C-start from fish 135, 106 days postoperatively.

View larger version (114K): [in a new window]   Fig. 11. Whole spinal cord re-crush at the spinomedullary level (SML) of fish that had recovered C-starts, full equilibrium and targeted feeding resulted in the loss of these behaviors. Photographs were taken from the side of the aquarium. (A) Goldfish 107, 200 days postoperatively. This fish was upright, was able to feed from the surface and displayed C-starts in response to the vibratory stimulus. (B) After spinal cord re-crush at the SML, this fish was on its side and was unable to target food from the water surface. In addition, vibratory stimuli did not elicit C-starts. This fish was 11 cm in body length.