- Email: [email protected]
Effects of Tectal Grafts on Sound Localization Deficits Induced by Inferior Colliculus Lesions in Hooded Rats
EXPERIMENTAL NEUROLOGY ARTICLE NO.
145, 16–23 (1997)
EN976439
Effects of Tectal Grafts on Sound Localization Deficits Induced by Inferior Colliculus Lesions in Hooded Rats Mark C. Zrull* and James R. Coleman†,‡ *Department of Psychology, Appalachian State University, Boone, North Carolina 28608; and †Department of Psychology and ‡Department of Physiology, University of South Carolina, Columbia, South Carolina 29208
sound localization ability (23). For example, neurons of auditory brain-stem structures receive afferent information derived from both ears (1, 4, 8, 27) and act to make comparisons of interaural time and intensity differences (5, 17, 26, 27). These binaural comparisons provide cues for the location of sound sources. The inferior colliculus (IC) is the major center of the rostral brain stem for encoding binaurally derived afferent information as well as refining the neural code of sound location cues from lower brain-stem structures (17, 21, 26). In cat, lesions of the IC and/or its brachial efferents result in deficits in sound localization behaviors (6, 15, 24). One goal of the present study was to examine the role of the IC as a mediator of sound localization behavior in the pigmented rat. Previous work has demonstrated that bilateral IC lesions impair sound detection ability in rat (33, 34). These detection deficits were mitigated by transplants of fetal tectal tissue placed into the IC lesion sites (33). Neural grafts into the IC exhibit structural features similar to those of the normal rat IC (31) and are metabolically active following sound stimulation (35). There is some evidence that prenatal tecta grafted into adult rat IC lesion sites become interconnected with host auditory structures (31, 33). These results suggest that integration of graft tissue into host pathways may contribute to physiological and behavioral recovery from IC lesion-induced deficits. Another goal of the present study was to examine the effects of tectal grafts on performance of a sound localization task following bilateral IC lesions. While the effects of neural tissue grafts on recovery of various forebrain-mediated abilities are fairly numerous (e.g., 3, 7, 9, 10, 20, 28), relatively few examples of graft-induced recovery of brain-stem-mediated behaviors exist (e.g., 22, 32–34). Preliminary experiments show that bilateral IC lesions induce sound localization impairments in rat (18, 30, 32) and neural grafts reduce the severity of these deficits (32). The present study elaborates on mitigation of hearing deficits produced by fetal tectal grafts placed into the damaged auditory brain stem (e.g., 33, 34) by examining the effects of
The goals of this research were to examine the role of the inferior colliculus (IC) in mediating sound localization behavior and the ability of tectal grafts to restore function after IC ablation in the Long-Evans rat. Previous work has suggested that the IC is a major center for processing of information used in localizing sound sources in space. Adult rats were trained on a lick suppression paradigm to discriminate the location of the second pulse in a noise burst pair presented in the horizontal interaural plane. Following baseline testing, rats received bilateral IC lesions, bilateral lesions followed in 1 week by bilateral tectal grafts, or were sham operated. Sound localization ability was then tested 15 to 30 days and 40 to 50 days following surgical procedures. Performance across experimental groups was statistically the same during baseline testing. During the first operative test period lesion-only and grafted animals showed deficits in sound localization ability relative to controls. By the second postoperative test period control and grafted animals did not differ statistically in sound localization ability and performance of both groups was superior to that of lesion-only animals. Histology revealed a similar extent of IC damage in lesion-only and lesion-graft animals and revealed the presence of implanted tectal tissue in all grafted animals. There was significant neuron loss in the dorsal nucleus of lateral lemniscus (DNLL) in lesion-only animals relative to grafted rats and sham controls. Behavioral results suggest that the IC of pigmented rat is important for sound localization ability. The sparing of DNLL neurons in grafted animals suggests that the tectal grafts may directly integrate into or aid intrinsic recovery of the host auditory pathway. r 1997 Academic Press
INTRODUCTION
The ability to locate the sources of sound in the environment can be critically important for the survival of hearing animals. Masterton and co-workers have suggested that the structures of the mammalian auditory system evolved in a manner which permits the encoding of spatial sound characteristics and facilitates 0014-4886/97 $25.00 Copyright r 1997 by Academic Press All rights of reproduction in any form reserved.
16
SOUND LOCALIZATION AND GRAFTED TECTUM
tectal grafts on sound localization deficits induced by IC lesions. METHODS
Subjects Male Long-Evans hooded rats (N 5 15, Harlan Sprague–Dawley) weighing 250 to 275 g and 60 days old at the beginning of behavioral training were subjects. Animals were group housed in an accredited vivarium (AAALAC) maintained at 22°C with a 12 h on, 12 h off light–dark cycle (lights on at 0700 h) and had free access to food throughout the experiment. Water was obtained during daily behavioral sessions (20 to 25 ml) and freely for 36 to 48 h at the end of each week. Ear canals and tympana were checked periodically throughout and at the conclusion of the experiment and were found to be free of debris and infection in all animals. Sound Localization Task Sound localization ability was examined using a lick suppression paradigm in a sound attenuating chamber (ambient noise level 30 dB re 20 µN/m2). Animals were trained to suppress licking from a water spout located at the front of a hardware cloth cage centered in the chamber when the location of the second pulse of a noise burst pair (target) differed from the source of the first pulse of the pair (cue). Cue and target pulses were filtered broadband noise (23 125 ms, 150-ms interpulse interval, 8 to 12 kHz, Coulbourn S81-02) that were shaped with a linear envelope (5-ms rise and fall times, Coulbourn S84-04), amplified (Realistic SA-150), attenuated to 50 dB (Hewlett Packard 350D), and switched to matched dome tweeters (Realistic 40-1276B). The loudspeakers were located at 45° intervals beginning at 0° azimuth (cage front) in the horizontal interaural plane surrounding the cage. Equipment was calibrated across the filtered noise frequency spectrum (Larson Labs 800B) and checked prior to each experimental session with an analog meter (Realistic 33-2050) precalibrated against the integrating meter. A sound localization session consisted of blocks of one to four cue–target noise burst pair presentations (trials). During each daily session the source of the cue pulse for all trials was always in the same auditory hemifield and the target pulse could emanate from a source in either auditory hemifield. Each block consisted of four cue-same-source target trials, one to three cue-same-source target trials followed by a cue-differentsource trial, or a single cue-different-source noise burst pair. Presentation of cue-same-target and cue-differenttarget noise burst pairs was randomized such that the probability of a cue-different-target stimulus occurring on any trial was 0.33. Sound localization behavior was shaped by gradually increasing the number of possible
17
target sound sources that differed from the cue source from one to five locations (see Fig. 1). Aversive control was used to encourage lick suppression for cue-differenttarget stimuli and consisted of a 250-ms shock presented 50 ms after the offset of a different target noise pulse. The onset and offset of a feedback light were paired with shocks. Because stimuli were switched to matched tweeters at different locations after gating, shaping, and amplification, noise burst pairs differed only in location and not in other properties. Sound localization was quantified by correcting the appropriate lick suppression rate (cue-different-target trials) using the inappropriate lick suppression rate (cue-same-target trials). Each rate was measured by monitoring voltage across the spout and cage floor during a 100-ms period beginning during the second pulse of a noise burst pair and ending prior to onset of the shock-light complex onset (cue-different-target pair) or after the second pulse (cue-same-target pair). The period was subdivided into 10 bins, and circuit closure counts ranged from 0 to 10 with low values indicating a suppression of licking. Responses were discarded when an animal broke contact with the spout at any time during a 1-s period immediately preceding a trial. Theory of Signal Detection methods (12) were applied to two measured responses: the proportion of correct lick suppression on cue-different-target trials (bin counts of less than 5), or hits (H), and the proportion of incorrect lick suppression on cue-same-target trials (bin counts of less than 5), or false alarms (FA). Correct responses were adjusted with incorrect responses and a
FIG. 1. Sound localization task performance for all animals across the 13 training sessions is shown. The number of possible target sources in cue–target noise burst pairs increased from one to five across the training sessions. The probability of a target noise burst emanating from a different source than the cue source was 0.33 on any trial. All noise bursts were presented from sources in the horizontal interaural plane. Points are shown as M 6 SD, and each session is based on 60 6 5 cue-same-target (false alarm, FA) and 36 6 5 (M 6 SD) cue-different-target (hit, H) noise burst pairs. Squares show H 2 H 3 FA, which is the localization rate.
18
ZRULL AND COLEMAN
measure of performance was computed block by block during a session. Specifically, a localization rate was defined as: LR 5 H 2 (H 3 FA), which reflects perfect localization (LR 5 1.00) only for perfect hit rates and no false alarms (FA 5 0; see Ref. 14 for discussion). For cue-same-target trials, localization performance was defined by altering the formula for localization of cue-different-target stimuli (H 2 [H 3 FA]) to: LR 5 (1 2 FA) 2 [(1 2 FA) 3 (1 2 H)]. Note that the preceding cue-same target LR reflects correctly licking through same source target pulses (1 2 FA) corrected relative to the rate of incorrectly licking through different source target pulses (1 2 H). Surgery Lesions. Electrolytic lesions were made in each IC of 10 anesthetized (ketamine:xylazine hydrochlorides 50:10 mg/kg, ip) rats using stereotaxic procedures and blunted ear bars. A 0.5-mm tungsten electrode was lowered at 30° from vertical into each IC through a single burr hole located at the midline posterior to interaural zero (AP: 20.3 mm, L: 0.0). Current (1.2 mA DC) was passed for 12 s at 23.5, 22.5, and 21.5 mm relative to the dura. Following surgery, holes in the skull were filled with Gelfoam (Upjohn) and covered with bone wax, and the skin was sutured. Each rat was monitored regularly for 24 h after surgery. Neural grafts. Donor tissue was obtained at 18 days of gestation (E18; mating day E0) from fetuses of an anesthetized (ketamine and xylazine hydrochlorides 60:10 mg/kg, ip) pregnant Long-Evans dam. Brains were removed from rat fetuses responsive to tactile stimulation and placed in an ice-cold nutrient mixture (HAM F-10, 4500 mg/liter glucose, 200 mM L-glutamine, Sigma). The midbrain tectum was dissected from each brain, meninges and surface blood vessels were removed, and the caudolateral part of each tectum was placed in fresh ice-cold nutrient mixture. Each fetus yielded two pieces of IC precursor tissue approximately 0.5 mm3 in volume. One week following IC lesions, five rats received grafts of fetal tectal tissue. Under anesthesia (ketamine: xylazine 50:10 mg/kg, ip) and using stereotaxic procedures, graft tissue was introduced into IC lesion sites via an 0.8-mm glass cannula mounted on a stereotaxic manipulator and attached to a 1.0-ml syringe with tubing filled with isotonic saline. Grafts were single caudolateral E18 tectum lowered into the host brain at 30° from vertical through the existing burr hole (22.0
mm re dura) and placed by slow pressure injection (approximately one-third tissue volume per minute). The cannula was left in place for 5 min following injection and subsequently inspected to verify expulsion of the fetal tissue into the host brain. The implantation process was identical for the contralateral lesion site. After successful implantation, the burr hole was filled and covered, and the skin was sutured. Procedure After animals were trained to perform the sound localization task with as many as five possible different target sound sources consistently (see Fig. 1), a preoperative series of eight sound localization sessions with five possible target sound sources which differed from the cue source was conducted (Series 1). Task performance was calculated by averaging LRs across the final four sessions of the series. LRs were computed for each cue to different target angle (45°, 90°, 135°, 180°), for cue-same target trials (0° cue to target angle), and across the cue to target angles. Five rats received bilateral IC lesions 2 days after initial behavioral testing ended, and 1 week later, these animals received bilateral grafts of fetal tectal tissue into the lesion sites (LG group). One week after sound localization testing five rats received bilateral IC lesions (LO group), and the remaining five animals received all lesion procedures except invasion of the brain to avoid damaging the sinus above the midbrain (Sham group). Postoperative behavioral testing for each rat resumed 2 weeks after initial preoperative testing ended, which corresponded to 1 week following the final surgical procedure. Postoperative testing consisted of retraining for six sessions across which the number of possible target locations that differed from the cue source was gradually increased from one to five locations. Subsequently, a postoperative series of sound localization testing proceeded identically to preoperative testing across eight sessions with five possible target sources that differed from the cue location (Series 2). Task performance was calculated by averaging LRs across the final four sessions of the series. Animals were retested on the sound localization task over a period of 10 days beginning 45 days after the IC lesion or sham surgical procedure (Series 3), and LRs were averaged across the final four sessions of the series. Histology All rats were sacrificed with a lethal overdose of sodium pentobarbital (100 mg/kg, ip) at the conclusion of the final sound localization testing series. Rats were perfused intracardially with 0.1 M phosphate-buffered saline followed by 10% formaldehyde in 0.1 M phosphate buffer. The brains were removed, transferred to
SOUND LOCALIZATION AND GRAFTED TECTUM
buffered fixative containing 20% sucrose, and stored at 4°C for at least 24 h. Frozen sections were made at 50 µm on a sliding microtome, mounted from phosphate buffer onto gelatin-coated slides, and allowed to air-dry for 24 h. The tissue was dehydrated in graded alcohols, rehydrated, stained with thionin, differentiated, cleared in toluene, and coverslipped with Permount (Fisher). Sites of tectal grafts and IC lesions were reconstructed with the aid of a Bausch and Lomb Trisimplex projecting microscope, and photomicrographs were made from the Nissl processed material under brightfield conditions (Nikon Optiphot system). The dorsal nucleus of the lateral lemniscus (DNLL) was identified in LG, LO, and Sham brains using bright-field microscopy at magnifications of 503 through 2003. At 1003, borders of the DNLL were reconstructed from three sections at the rostral (25% distance from rostral tip), middle (50%), and caudal (75%) extent of the nucleus in each brain. Neurons were plotted onto the border reconstructions using a drawing tube attached to a Nikon Optiphot microscope. Cell counts were averaged across the three sections at each level of each DNLL. Thus, each brain yielded three DNLL cell counts for each side of the brain stem. RESULTS
Stereomicroscopic observation of brains prior to histological processing revealed evidence of grafted tectal tissue in all LG animals, bilateral lesions in all LO animals, and no damage to Sham brains. Thus, behavioral and histological results are reported for all 15 rats. The behavioral experiment fit a 3 3 3 (Experimental Group 3 Testing Series) split-plot factorial design in which Experimental Group was a between-subjects factor and Testing Series was a within-subjects factor. A priori comparisons for localization rates (LRs) without regard to cue to target angles were made among the experimental groups (LG, LO, and Sham) for Series 1, 2, and 3 testing within an ANOVA framework; that is, paired comparisons for the Experimental Group simple main effect were made for each Testing Series with Type I error controlled at 0.05 across the tests using Bonferroni’s adjustment (50.05/3 5 0.0167 per comparison; see Ref. 19). DNLL cell count data were compared across experimental groups using ANOVA and pair-wise comparisons. Sound Localization Task The effects of IC lesions (LO and LG rats) on localization rates (LRs) were evident at initial (Series 2) and subsequent (Series 3) postoperative testing, and tectal grafts (LG rats) affected LRs during Series 3 behavioral testing (see Fig. 2). A significant Experimental Group 3 Test Series effect on performance of the localization task was found, F(4, 24) 5 6.37, P , 0.05. LRs did not
19
FIG. 2. Localization rate (LR, M 6 SD) for cue-target noise burst pairs is shown. The Experimental Group 3 Test Series effect on detection is significant, P , 0.05. (A) Preoperative baseline series; (B) initial postoperative series, 15 to 30 days after the last surgery; (C) subsequent postoperative series, 40 to 50 days after the lesion procedure. Task performance for all groups is above chance (LR 5 0.35, binomial, P , 0.05). Means with the same * or ** in A and B are significantly different (P , 0.0167; i.e., 0.05/3, see text for explanation). Each mean is based on 120 6 8 cue-same-target (FA) and 64 6 8 cue-different-target (H) noise burst pairs.
differ across the groups during Series 1 (see Fig. 2A); however, analysis of the simple main effect for the groups at Series 2 indicates that the experimental groups differed during initial postoperative testing (see Fig. 2B). Task performance of both LO (0.48 6 0.10, M 6 SD) and LG (0.53 6 0.06) rats differed from Sham controls (0.67 6 0.06), F(1, 26) 5 13.21 and F(1, 26) 5 24.33, P , 0.0167, respectively. During Series 3, the difference in task performance of LG (0.61 6 0.06) and Sham (0.70 6 0.04) rats was not statistically significant. Series 3 task performance of LO rats (0.50 6 0.06) did differ significantly from both LG animals, F(1, 26) 5 8.15, P , 0.0167, and Sham animals, F(1, 26) 5 26.95, P , 0.0167 (see Fig. 2C). Histology Lesions in LO and LG rats (see Fig. 3) extended from the posteriormost aspect to a region of the IC just caudal to the beginning of the superior colliculus (SC). The largest lesions included all IC tissue from its posterior extent through the invasion of fibers of the lateral lemniscus (LL) and destroyed the central nucleus of the IC (CNIC). Rostral to the LL afferents, large lesions included most of the IC caudal to SC. Smaller lesions encompassed CNIC and most of the external cortex of the IC (ECIC). In all LO and LG cases some of the brachium of IC was spared. Implanted tissue was identified in brains of all LG rats which were sacrificed 8 weeks after the transplantation surgery. The tectal grafts were relatively small, 0.5 to 2.0 mm3, and were restricted to the IC lesion sites (see Fig. 4). Grafts were populated by a combination of
20
ZRULL AND COLEMAN
FIG. 3. (A) Composite of reconstructions of lesion sites made from thionin-stained frontal sections of brains from rats with lesions only (LO). (B) Composite of reconstructions of lesion sites made from thionin-stained frontal sections of brains from rats with lesions and grafts (LG). Serial sections are at approximately 200-µm intervals with the maximum extent of damage shown in cross-hatch (Rat 9118 in A and Rat 9114 in B) and common damage shown in black. Bar, 500 µm. Cbl, cerebellum; IC, inferior colliculus.
FIG. 4. Drawings showing the location of E18 tectal grafts in inferior colliculus lesion sites of grafted rats. Graft tissue is shown in black, and stippling indicates regions primarily populated by macrophage. (A) Rat 9115. (B) Rat 9110. (C) Rat 9109. (D) Rat 9114. (E) Rat 9106. Bar, 500 µm.
SOUND LOCALIZATION AND GRAFTED TECTUM
neurons and glial cells and were often surrounded by macrophage cells. While Nissl material did not reveal extensive adult-like neuron populations within the grafts, an area of contact between grafted and host tissue was found in each LG brain. Counts of DNLL cells, which normally project to the IC (21, 25, 27), revealed differences among the experimental groups, F(2, 12) 5 35.35, P , 0.05. The Sham group averaged 79 6 2.3 (M 6 SEM) cells per section of DNLL. Both LG and LO brains produced significantly lower per-section DNLL cell counts than found in Sham brains. LG brains averaged 61 6 2.3 cells, t(12) 5 5.25, P , 0.0167 (77.2% of Sham cell count); and LO brains averaged 51 6 2.9 cells, t(12) 5 69.11, P , 0.0167 (64.6% of Sham cell count). The average number of cells observed in the DNLL of grafted rats was significantly greater than that observed in LO rats, t(12) 5 3.07, P , 0.0167. Figure 5 shows the paucity of DNLL cells in LO animals relative to sham-operated and grafted rats. DISCUSSION
The present study demonstrated that pigmented rats, like wild and albino rats (13, 14, 16), can perform a
21
task which requires the ability to locate the sources of target sounds in acoustic space. For all animals, localization rates improved during training as the number of possible target sources was increased, and task performance of control animals continued to improve throughout the experiment. Postlesion deficits in task performance of animals with bilaterally damaged IC were indicative of the role of this auditory midbrain structure in sound localization behavior. These results suggest that the IC in rat, as in cat (6, 15, 24), plays a significant role in the ability to localize sound. In the albino rat, the auditory cortex and medial geniculate body are not necessary for sound localization (16, 18); however, deep lesions of the IC which include its brachial efferents are reported to produce impairments in sound localization behavior (18). The present results suggest that the rat IC is involved in directional hearing, including the detection of and attention to sound in acoustic space (33, 34) as well as the ability to identify sound source locations. These behavioral results are consistent with physiological data suggesting that IC neurons encode binaurally derived intensity and time differences of spatial sound stimuli (5, 17, 19). Animals receiving bilateral grafts after lesions exhibited localization performance that did not differ statisti-
FIG. 5. Photomicrographs of thionin-stained sections through the rostrocaudal midpoint of the dorsal nucleus of the lateral lemniscus (DNLL). (A) Arrows point out the dorsal and ventral aspects of DNLL in sham-operated Rat 9112. (B) In grafted Rat 9115, the DNLL is qualitatively similar to DNLL of Sham rats with somal clusters aligned along fiber tracts of the lateral lemniscus. (C) A corresponding section from Rat 9111 with an IC lesion and no graft shows a paucity of DNLL neurons. Bar, 100 µm.
22
ZRULL AND COLEMAN
cally from the performance of rats with IC lesions only during initial postoperative testing. Subsequently, performance of grafted rats was significantly improved relative to conspecifics with IC lesions and did differ, but not statistically, from control animals. The delay in the restorative effect of tectal grafts on performance of the current sound localization task differs from the immediate postoperative differences between IC lesiononly and grafted animals observed in a simpler sound detection task (33). Sound localization requires the process of identification or knowing what stimulus has occurred; in the present study identification of the same or a different source target noise burst was required. In the previously reported study (33), whether the experimental task required simply detection or an attention process, only knowing that a sound had occurred was important. In more commonly studied forebrain systems, graft-dependent recovery from behavioral deficits can be immediate (20, 28) or delayed (7, 9), dependent upon the behavior to be recovered and the manner in which implanted tissue interacts (see Ref. 2) with the damaged host brain. While grafted rats of the present study never achieved a level of performance equal to that which occurred under preoperative conditions or that of control animals, they did localize sound stimuli at rates statistically better than rats with IC lesions only. Often, partial recovery of behavioral function is reported in experiments assessing the restorative efficacy of neural grafts (e.g., 3, 10, 28, 33). The DNLL has prominent input to the IC that is important for binaural processing of IC neurons. The IC receives bilateral input from DNLL in rat (1, 8, 25) and cat (4, 27). In rat, Li and Kelly (21) have demonstrated that input to IC from DNLL is critical for appropriate encoding of the laterality of sound source in IC neurons. Interaural intensity difference (IID) functions in IC neurons shift after temporary DNLL blockade with kynurenic acid and return to normal after recovery of DNLL. In the present study animals with bilateral IC lesions exhibit substantial retrograde loss of DNLL neurons, which suggests that the IC is the principal target of these neurons, and behavioral findings in the present study show that animals with bilateral lesions are deficient on a sound localization task. These results imply not only the role of damaged IC, but demonstrate the loss of DNLL neurons known to perform functions essential to binaural encoding required for localization of sound in space (21, 27). Grafted animals in the present study showed behavioral recovery not observed in lesioned animals not grafted. There was considerable sparing of DNLL neurons in these animals relative to those with lesions alone. These results suggest that DNLL sparing is due to the presence of the neural grafts. Although grafts were of modest size (with macrophage still observed at
this stage), grafts of IC show long-term survival and viability (31, 35). The recovery in localization ability in these animals may be due to graft tissue providing targets for DNLL neurons or permitting the DNLL cells to reach alternate target neurons in the host brain, perhaps through mechanisms such as growth factor release. It remains to be determined whether behavioral recovery is due to a direct or indirect impact of grafts on the damaged host auditory pathway. Previous work examining the efficacy of neural grafts to restore lost function after neocortical damage suggests that graft-induced survival of thalamic afferents may contribute to behavioral recovery through direct or novel, indirect connections (10, 28). In summary, the present experiment shows that bilateral lesions of rat IC and accompanying retrograde degeneration of the DNLL impair sound localization behavior. These deficits are partially remediated by bilateral implantation of caudal tectal grafts. The mechanisms for observed behavioral recovery may involve direct DNLL–graft connections, the opening of alternative DNLL–host connections, or other alterations involving host auditory brain-stem pathways. ACKNOWLEDGMENTS We acknowledge the contributions from Grants DBS-9200624 and SBR00285 (J.R.C.), the Deafness Research Foundation (J.R.C.), and the Department of Psychology at South Carolina through its Tindall Award (M.C.Z.).
REFERENCES 1. 2.
3.
4.
5.
6.
7.
Adams, J. C. 1979. Ascending projections to the inferior colliculus. J. Comp. Neurol. 183: 519–538. Bjo¨rklund, A., O. Lindvall, O. Isacson, P. Brundin, K. Wictorin, R. E. Strecker, D. J. Clarke, and S. B. Dunnett. 1987. Mechanisms of action of intracerebral neural implants: Studies on nigral and striatal grafts to the lesioned striatum. Trends Neurosci. 10: 509–516. Brundin, P., O. G. Nilsson, R. E. Strecker, O. Lindvall, B. Astedt, and A. Bjo¨rklund. 1986. Behavioural effects of human fetal dopamine neurons grafted in a rat model of Parkinson’s disease. Exp. Brain Res. 65: 235–240. Brunso-Bechtold, J. K., G. C. Thompson, and R. B. Masterton. 1981. HRP study of the organization of auditory afferents ascending to the central nucleus of inferior colliculus. J. Comp. Neurol. 197: 705–722. Calford, M. B., D. R. Moore, and M. E. Hutchings. 1986. Central and peripheral contributions to coding of acoustic space by neurons in inferior colliculus of cat. J. Neurophysiol. 55: 587– 603. Casseday, J. H., and W. D. Neff. 1975. Auditory localization: Role of auditory pathways in brain stem of the cat. J. Neurophysiol. 38: 842–858. Cassel, J. C., C. Kelche, and B. Will. 1990. Time-dependent effects of intrahippocampal grafts in rats with fimbria–fornix lesions. Exp. Brain Res. 81: 179–190.
SOUND LOCALIZATION AND GRAFTED TECTUM 8.
9.
10.
11. 12. 13.
14.
15.
16. 17.
18.
19. 20.
21.
22.
Coleman, J. R., and W. J. Clerici. 1987. Sources of projections to subdivisions of inferior colliculus in rat. J. Comp. Neurol. 262: 215–226. Dunnett, S. B., A. T. Wareham, and E. M. Torres. 1993. Septal grafts improve acquisition of an operant timing task after fimbria-fornix lesions in rats: Role of the interval between lesion and transplantation surgeries. Restorative Neurol. Neurosci. 5: 263–274. Dunnett, S. B., C. N. Ryan, P. D. Levin, M. Reynolds, and S. T. Bunch. 1987. Functional consequences of embryonic neocortex transplanted to rats with prefrontal cortex lesions. Behav. Neurosci. 101: 489–503. Faye-Lund, H., and K. K. Osen. 1985. Anatomy of the inferior colliculus in rat. Anat. Embryol. 171: 1–20. Green, D. M., and J. A. Swets. 1973. Signal Detection Theory and Psychophysics. Krieger, Huntington, NY. Heffner, H. E., and R. S. Heffner. 1985. Hearing in two cricetid rodents: Wood rat (Neotoma floridana) and grasshopper mouse (Onychomys leucogaster). J. Comp. Psychol. 99: 275–288. Heffner, H. E., and R. S. Heffner. 1985. Sound localization in wild Norway rats (Rattus norvegicus). Hearing Res. 19: 151– 155. Jenkins, W. M., and R. B. Masterton. 1982. Sound localization: Effects of unilateral lesions in central auditory system. J. Neurophysiol. 47: 987–1016. Kelly, J. B. 1980. Effects of auditory cortical lesions on sound localization by the rat. J. Neurophysiol. 44: 1161–1174. Kelly, J. B., S. L. Glenn, and C. Beaver. 1991. Sound frequency and binaural response properties of single neurons in rat inferior colliculus. Hearing Res. 56: 273–280. Kelly, J. B., and P. W. Judge. 1985. Effects of medial geniculate lesions on sound localization by the rat. J. Neurophysiol. 53: 361–372. Kirk, R. E. 1982. Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole, Monterey, CA. Labbe, R., A. C. Firl, Jr., E. J. Mufson, and D. G. Stein. 1983. Fetal brain transplants: Reduction of deficits in rats with frontal cortex lesions. Science 217: 470–472. Li, L., and J. B. Kelly. 1992. Inhibitory influences of the dorsal nucleus of the lateral lemniscus on binaural responses in the rat’s inferior colliculus. J. Neurosci. 12: 4530–4539. Lund, R. D., and S. D. Hauschka. 1976. Transplanted neural tissue develops connections with the host rat brain. Science 193: 582–584.
23. 24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
23
Masterton, B., H. Heffner, and R. Ravizza. 1969. The evolution of human hearing. J. Acoust. Soc. Am. 45: 966–985. Masterton, R. B., J. A. Jane, and I. T. Diamond. 1968. The role of brainstem auditory structures in sound localization. II. Inferior colliculus and its brachium. J. Neurophysiol. 31: 96–108. Mercha´n, M. A., E. Saldan˜a, and I. Plaza. 1994. Dorsal nucleus of the lateral lemniscus in the rat: Concentric organization and tonotopic projection to the inferior colliculus. J. Comp. Neurol. 342: 259–278. Sally, S. L., and J. B. Kelly. 1992. Effects of superior olivary complex lesions on binaural responses in rat inferior colliculus. Brain Res. 572: 5–18. Shneiderman, A., D. L. Oliver, and C. K. Henkel. 1988. Connections of the dorsal nucleus of the lateral lemniscus: An inhibitory parallel pathway in the ascending auditory system? J. Comp. Neurol. 276: 188–208. Stein, D. G., C. Palatucci, D. Kahn, and R. Labbe. 1988. Temporal factors influence recovery of function after embryonic brain tissue transplants in adult rats with frontal cortex lesions. Behav. Neurosci. 102: 260–267. Yin, T. C. T., J. C. K. Chan, and D. R. F. Irvine. 1986. Effects of interaural time delays of noise stimuli on low-frequency cells in the cat’s inferior colliculus. I. Responses to wideband noise. J. Neurophysiol. 58: 562–583. Zrull, M. C., C. L. Miner, C. A. Parker, C. L. Kluttz, and J. Rochelle. 1995. Neonatal inferior colliculus lesions affect soundguided navigation. Soc. Neurosci. Abstr. 21: 671. Zrull, M. C., and J. R. Coleman. 1991. Structural features in whole grafts of the rat inferior colliculus. Hearing Res. 55: 117–132. Zrull, M. C., and J. R. Coleman. 1993. Inferior colliculus lesions and grafts: Deficits and recovery of sound locating ability. Soc. Neurosci. Abstr. 19: 1741. Zrull, M. C., and J. R. Coleman. 1996. Effects of tectal grafts on sound detection deficits induced by inferior colliculus lesions in hooded rats. Cell Transplant. 5: 293–304. Zrull, M. C., J. R. Coleman, and C. Bennett. 1991. Tectal grafts promote recovery from behavioral deficits after inferior colliculus lesions. Soc. Neurosci. Abstr. 17: 1138. Zrull, M. C., T. Wessinger, and J. R. Coleman. 1992. Differential sound-evoked metabolic activity in fetal tectum grafted into damaged adult rat inferior colliculus. Restorative Neurol. Neurosci. 4: 285–289.
145, 16–23 (1997)
EN976439
Effects of Tectal Grafts on Sound Localization Deficits Induced by Inferior Colliculus Lesions in Hooded Rats Mark C. Zrull* and James R. Coleman†,‡ *Department of Psychology, Appalachian State University, Boone, North Carolina 28608; and †Department of Psychology and ‡Department of Physiology, University of South Carolina, Columbia, South Carolina 29208
sound localization ability (23). For example, neurons of auditory brain-stem structures receive afferent information derived from both ears (1, 4, 8, 27) and act to make comparisons of interaural time and intensity differences (5, 17, 26, 27). These binaural comparisons provide cues for the location of sound sources. The inferior colliculus (IC) is the major center of the rostral brain stem for encoding binaurally derived afferent information as well as refining the neural code of sound location cues from lower brain-stem structures (17, 21, 26). In cat, lesions of the IC and/or its brachial efferents result in deficits in sound localization behaviors (6, 15, 24). One goal of the present study was to examine the role of the IC as a mediator of sound localization behavior in the pigmented rat. Previous work has demonstrated that bilateral IC lesions impair sound detection ability in rat (33, 34). These detection deficits were mitigated by transplants of fetal tectal tissue placed into the IC lesion sites (33). Neural grafts into the IC exhibit structural features similar to those of the normal rat IC (31) and are metabolically active following sound stimulation (35). There is some evidence that prenatal tecta grafted into adult rat IC lesion sites become interconnected with host auditory structures (31, 33). These results suggest that integration of graft tissue into host pathways may contribute to physiological and behavioral recovery from IC lesion-induced deficits. Another goal of the present study was to examine the effects of tectal grafts on performance of a sound localization task following bilateral IC lesions. While the effects of neural tissue grafts on recovery of various forebrain-mediated abilities are fairly numerous (e.g., 3, 7, 9, 10, 20, 28), relatively few examples of graft-induced recovery of brain-stem-mediated behaviors exist (e.g., 22, 32–34). Preliminary experiments show that bilateral IC lesions induce sound localization impairments in rat (18, 30, 32) and neural grafts reduce the severity of these deficits (32). The present study elaborates on mitigation of hearing deficits produced by fetal tectal grafts placed into the damaged auditory brain stem (e.g., 33, 34) by examining the effects of
The goals of this research were to examine the role of the inferior colliculus (IC) in mediating sound localization behavior and the ability of tectal grafts to restore function after IC ablation in the Long-Evans rat. Previous work has suggested that the IC is a major center for processing of information used in localizing sound sources in space. Adult rats were trained on a lick suppression paradigm to discriminate the location of the second pulse in a noise burst pair presented in the horizontal interaural plane. Following baseline testing, rats received bilateral IC lesions, bilateral lesions followed in 1 week by bilateral tectal grafts, or were sham operated. Sound localization ability was then tested 15 to 30 days and 40 to 50 days following surgical procedures. Performance across experimental groups was statistically the same during baseline testing. During the first operative test period lesion-only and grafted animals showed deficits in sound localization ability relative to controls. By the second postoperative test period control and grafted animals did not differ statistically in sound localization ability and performance of both groups was superior to that of lesion-only animals. Histology revealed a similar extent of IC damage in lesion-only and lesion-graft animals and revealed the presence of implanted tectal tissue in all grafted animals. There was significant neuron loss in the dorsal nucleus of lateral lemniscus (DNLL) in lesion-only animals relative to grafted rats and sham controls. Behavioral results suggest that the IC of pigmented rat is important for sound localization ability. The sparing of DNLL neurons in grafted animals suggests that the tectal grafts may directly integrate into or aid intrinsic recovery of the host auditory pathway. r 1997 Academic Press
INTRODUCTION
The ability to locate the sources of sound in the environment can be critically important for the survival of hearing animals. Masterton and co-workers have suggested that the structures of the mammalian auditory system evolved in a manner which permits the encoding of spatial sound characteristics and facilitates 0014-4886/97 $25.00 Copyright r 1997 by Academic Press All rights of reproduction in any form reserved.
16
SOUND LOCALIZATION AND GRAFTED TECTUM
tectal grafts on sound localization deficits induced by IC lesions. METHODS
Subjects Male Long-Evans hooded rats (N 5 15, Harlan Sprague–Dawley) weighing 250 to 275 g and 60 days old at the beginning of behavioral training were subjects. Animals were group housed in an accredited vivarium (AAALAC) maintained at 22°C with a 12 h on, 12 h off light–dark cycle (lights on at 0700 h) and had free access to food throughout the experiment. Water was obtained during daily behavioral sessions (20 to 25 ml) and freely for 36 to 48 h at the end of each week. Ear canals and tympana were checked periodically throughout and at the conclusion of the experiment and were found to be free of debris and infection in all animals. Sound Localization Task Sound localization ability was examined using a lick suppression paradigm in a sound attenuating chamber (ambient noise level 30 dB re 20 µN/m2). Animals were trained to suppress licking from a water spout located at the front of a hardware cloth cage centered in the chamber when the location of the second pulse of a noise burst pair (target) differed from the source of the first pulse of the pair (cue). Cue and target pulses were filtered broadband noise (23 125 ms, 150-ms interpulse interval, 8 to 12 kHz, Coulbourn S81-02) that were shaped with a linear envelope (5-ms rise and fall times, Coulbourn S84-04), amplified (Realistic SA-150), attenuated to 50 dB (Hewlett Packard 350D), and switched to matched dome tweeters (Realistic 40-1276B). The loudspeakers were located at 45° intervals beginning at 0° azimuth (cage front) in the horizontal interaural plane surrounding the cage. Equipment was calibrated across the filtered noise frequency spectrum (Larson Labs 800B) and checked prior to each experimental session with an analog meter (Realistic 33-2050) precalibrated against the integrating meter. A sound localization session consisted of blocks of one to four cue–target noise burst pair presentations (trials). During each daily session the source of the cue pulse for all trials was always in the same auditory hemifield and the target pulse could emanate from a source in either auditory hemifield. Each block consisted of four cue-same-source target trials, one to three cue-same-source target trials followed by a cue-differentsource trial, or a single cue-different-source noise burst pair. Presentation of cue-same-target and cue-differenttarget noise burst pairs was randomized such that the probability of a cue-different-target stimulus occurring on any trial was 0.33. Sound localization behavior was shaped by gradually increasing the number of possible
17
target sound sources that differed from the cue source from one to five locations (see Fig. 1). Aversive control was used to encourage lick suppression for cue-differenttarget stimuli and consisted of a 250-ms shock presented 50 ms after the offset of a different target noise pulse. The onset and offset of a feedback light were paired with shocks. Because stimuli were switched to matched tweeters at different locations after gating, shaping, and amplification, noise burst pairs differed only in location and not in other properties. Sound localization was quantified by correcting the appropriate lick suppression rate (cue-different-target trials) using the inappropriate lick suppression rate (cue-same-target trials). Each rate was measured by monitoring voltage across the spout and cage floor during a 100-ms period beginning during the second pulse of a noise burst pair and ending prior to onset of the shock-light complex onset (cue-different-target pair) or after the second pulse (cue-same-target pair). The period was subdivided into 10 bins, and circuit closure counts ranged from 0 to 10 with low values indicating a suppression of licking. Responses were discarded when an animal broke contact with the spout at any time during a 1-s period immediately preceding a trial. Theory of Signal Detection methods (12) were applied to two measured responses: the proportion of correct lick suppression on cue-different-target trials (bin counts of less than 5), or hits (H), and the proportion of incorrect lick suppression on cue-same-target trials (bin counts of less than 5), or false alarms (FA). Correct responses were adjusted with incorrect responses and a
FIG. 1. Sound localization task performance for all animals across the 13 training sessions is shown. The number of possible target sources in cue–target noise burst pairs increased from one to five across the training sessions. The probability of a target noise burst emanating from a different source than the cue source was 0.33 on any trial. All noise bursts were presented from sources in the horizontal interaural plane. Points are shown as M 6 SD, and each session is based on 60 6 5 cue-same-target (false alarm, FA) and 36 6 5 (M 6 SD) cue-different-target (hit, H) noise burst pairs. Squares show H 2 H 3 FA, which is the localization rate.
18
ZRULL AND COLEMAN
measure of performance was computed block by block during a session. Specifically, a localization rate was defined as: LR 5 H 2 (H 3 FA), which reflects perfect localization (LR 5 1.00) only for perfect hit rates and no false alarms (FA 5 0; see Ref. 14 for discussion). For cue-same-target trials, localization performance was defined by altering the formula for localization of cue-different-target stimuli (H 2 [H 3 FA]) to: LR 5 (1 2 FA) 2 [(1 2 FA) 3 (1 2 H)]. Note that the preceding cue-same target LR reflects correctly licking through same source target pulses (1 2 FA) corrected relative to the rate of incorrectly licking through different source target pulses (1 2 H). Surgery Lesions. Electrolytic lesions were made in each IC of 10 anesthetized (ketamine:xylazine hydrochlorides 50:10 mg/kg, ip) rats using stereotaxic procedures and blunted ear bars. A 0.5-mm tungsten electrode was lowered at 30° from vertical into each IC through a single burr hole located at the midline posterior to interaural zero (AP: 20.3 mm, L: 0.0). Current (1.2 mA DC) was passed for 12 s at 23.5, 22.5, and 21.5 mm relative to the dura. Following surgery, holes in the skull were filled with Gelfoam (Upjohn) and covered with bone wax, and the skin was sutured. Each rat was monitored regularly for 24 h after surgery. Neural grafts. Donor tissue was obtained at 18 days of gestation (E18; mating day E0) from fetuses of an anesthetized (ketamine and xylazine hydrochlorides 60:10 mg/kg, ip) pregnant Long-Evans dam. Brains were removed from rat fetuses responsive to tactile stimulation and placed in an ice-cold nutrient mixture (HAM F-10, 4500 mg/liter glucose, 200 mM L-glutamine, Sigma). The midbrain tectum was dissected from each brain, meninges and surface blood vessels were removed, and the caudolateral part of each tectum was placed in fresh ice-cold nutrient mixture. Each fetus yielded two pieces of IC precursor tissue approximately 0.5 mm3 in volume. One week following IC lesions, five rats received grafts of fetal tectal tissue. Under anesthesia (ketamine: xylazine 50:10 mg/kg, ip) and using stereotaxic procedures, graft tissue was introduced into IC lesion sites via an 0.8-mm glass cannula mounted on a stereotaxic manipulator and attached to a 1.0-ml syringe with tubing filled with isotonic saline. Grafts were single caudolateral E18 tectum lowered into the host brain at 30° from vertical through the existing burr hole (22.0
mm re dura) and placed by slow pressure injection (approximately one-third tissue volume per minute). The cannula was left in place for 5 min following injection and subsequently inspected to verify expulsion of the fetal tissue into the host brain. The implantation process was identical for the contralateral lesion site. After successful implantation, the burr hole was filled and covered, and the skin was sutured. Procedure After animals were trained to perform the sound localization task with as many as five possible different target sound sources consistently (see Fig. 1), a preoperative series of eight sound localization sessions with five possible target sound sources which differed from the cue source was conducted (Series 1). Task performance was calculated by averaging LRs across the final four sessions of the series. LRs were computed for each cue to different target angle (45°, 90°, 135°, 180°), for cue-same target trials (0° cue to target angle), and across the cue to target angles. Five rats received bilateral IC lesions 2 days after initial behavioral testing ended, and 1 week later, these animals received bilateral grafts of fetal tectal tissue into the lesion sites (LG group). One week after sound localization testing five rats received bilateral IC lesions (LO group), and the remaining five animals received all lesion procedures except invasion of the brain to avoid damaging the sinus above the midbrain (Sham group). Postoperative behavioral testing for each rat resumed 2 weeks after initial preoperative testing ended, which corresponded to 1 week following the final surgical procedure. Postoperative testing consisted of retraining for six sessions across which the number of possible target locations that differed from the cue source was gradually increased from one to five locations. Subsequently, a postoperative series of sound localization testing proceeded identically to preoperative testing across eight sessions with five possible target sources that differed from the cue location (Series 2). Task performance was calculated by averaging LRs across the final four sessions of the series. Animals were retested on the sound localization task over a period of 10 days beginning 45 days after the IC lesion or sham surgical procedure (Series 3), and LRs were averaged across the final four sessions of the series. Histology All rats were sacrificed with a lethal overdose of sodium pentobarbital (100 mg/kg, ip) at the conclusion of the final sound localization testing series. Rats were perfused intracardially with 0.1 M phosphate-buffered saline followed by 10% formaldehyde in 0.1 M phosphate buffer. The brains were removed, transferred to
SOUND LOCALIZATION AND GRAFTED TECTUM
buffered fixative containing 20% sucrose, and stored at 4°C for at least 24 h. Frozen sections were made at 50 µm on a sliding microtome, mounted from phosphate buffer onto gelatin-coated slides, and allowed to air-dry for 24 h. The tissue was dehydrated in graded alcohols, rehydrated, stained with thionin, differentiated, cleared in toluene, and coverslipped with Permount (Fisher). Sites of tectal grafts and IC lesions were reconstructed with the aid of a Bausch and Lomb Trisimplex projecting microscope, and photomicrographs were made from the Nissl processed material under brightfield conditions (Nikon Optiphot system). The dorsal nucleus of the lateral lemniscus (DNLL) was identified in LG, LO, and Sham brains using bright-field microscopy at magnifications of 503 through 2003. At 1003, borders of the DNLL were reconstructed from three sections at the rostral (25% distance from rostral tip), middle (50%), and caudal (75%) extent of the nucleus in each brain. Neurons were plotted onto the border reconstructions using a drawing tube attached to a Nikon Optiphot microscope. Cell counts were averaged across the three sections at each level of each DNLL. Thus, each brain yielded three DNLL cell counts for each side of the brain stem. RESULTS
Stereomicroscopic observation of brains prior to histological processing revealed evidence of grafted tectal tissue in all LG animals, bilateral lesions in all LO animals, and no damage to Sham brains. Thus, behavioral and histological results are reported for all 15 rats. The behavioral experiment fit a 3 3 3 (Experimental Group 3 Testing Series) split-plot factorial design in which Experimental Group was a between-subjects factor and Testing Series was a within-subjects factor. A priori comparisons for localization rates (LRs) without regard to cue to target angles were made among the experimental groups (LG, LO, and Sham) for Series 1, 2, and 3 testing within an ANOVA framework; that is, paired comparisons for the Experimental Group simple main effect were made for each Testing Series with Type I error controlled at 0.05 across the tests using Bonferroni’s adjustment (50.05/3 5 0.0167 per comparison; see Ref. 19). DNLL cell count data were compared across experimental groups using ANOVA and pair-wise comparisons. Sound Localization Task The effects of IC lesions (LO and LG rats) on localization rates (LRs) were evident at initial (Series 2) and subsequent (Series 3) postoperative testing, and tectal grafts (LG rats) affected LRs during Series 3 behavioral testing (see Fig. 2). A significant Experimental Group 3 Test Series effect on performance of the localization task was found, F(4, 24) 5 6.37, P , 0.05. LRs did not
19
FIG. 2. Localization rate (LR, M 6 SD) for cue-target noise burst pairs is shown. The Experimental Group 3 Test Series effect on detection is significant, P , 0.05. (A) Preoperative baseline series; (B) initial postoperative series, 15 to 30 days after the last surgery; (C) subsequent postoperative series, 40 to 50 days after the lesion procedure. Task performance for all groups is above chance (LR 5 0.35, binomial, P , 0.05). Means with the same * or ** in A and B are significantly different (P , 0.0167; i.e., 0.05/3, see text for explanation). Each mean is based on 120 6 8 cue-same-target (FA) and 64 6 8 cue-different-target (H) noise burst pairs.
differ across the groups during Series 1 (see Fig. 2A); however, analysis of the simple main effect for the groups at Series 2 indicates that the experimental groups differed during initial postoperative testing (see Fig. 2B). Task performance of both LO (0.48 6 0.10, M 6 SD) and LG (0.53 6 0.06) rats differed from Sham controls (0.67 6 0.06), F(1, 26) 5 13.21 and F(1, 26) 5 24.33, P , 0.0167, respectively. During Series 3, the difference in task performance of LG (0.61 6 0.06) and Sham (0.70 6 0.04) rats was not statistically significant. Series 3 task performance of LO rats (0.50 6 0.06) did differ significantly from both LG animals, F(1, 26) 5 8.15, P , 0.0167, and Sham animals, F(1, 26) 5 26.95, P , 0.0167 (see Fig. 2C). Histology Lesions in LO and LG rats (see Fig. 3) extended from the posteriormost aspect to a region of the IC just caudal to the beginning of the superior colliculus (SC). The largest lesions included all IC tissue from its posterior extent through the invasion of fibers of the lateral lemniscus (LL) and destroyed the central nucleus of the IC (CNIC). Rostral to the LL afferents, large lesions included most of the IC caudal to SC. Smaller lesions encompassed CNIC and most of the external cortex of the IC (ECIC). In all LO and LG cases some of the brachium of IC was spared. Implanted tissue was identified in brains of all LG rats which were sacrificed 8 weeks after the transplantation surgery. The tectal grafts were relatively small, 0.5 to 2.0 mm3, and were restricted to the IC lesion sites (see Fig. 4). Grafts were populated by a combination of
20
ZRULL AND COLEMAN
FIG. 3. (A) Composite of reconstructions of lesion sites made from thionin-stained frontal sections of brains from rats with lesions only (LO). (B) Composite of reconstructions of lesion sites made from thionin-stained frontal sections of brains from rats with lesions and grafts (LG). Serial sections are at approximately 200-µm intervals with the maximum extent of damage shown in cross-hatch (Rat 9118 in A and Rat 9114 in B) and common damage shown in black. Bar, 500 µm. Cbl, cerebellum; IC, inferior colliculus.
FIG. 4. Drawings showing the location of E18 tectal grafts in inferior colliculus lesion sites of grafted rats. Graft tissue is shown in black, and stippling indicates regions primarily populated by macrophage. (A) Rat 9115. (B) Rat 9110. (C) Rat 9109. (D) Rat 9114. (E) Rat 9106. Bar, 500 µm.
SOUND LOCALIZATION AND GRAFTED TECTUM
neurons and glial cells and were often surrounded by macrophage cells. While Nissl material did not reveal extensive adult-like neuron populations within the grafts, an area of contact between grafted and host tissue was found in each LG brain. Counts of DNLL cells, which normally project to the IC (21, 25, 27), revealed differences among the experimental groups, F(2, 12) 5 35.35, P , 0.05. The Sham group averaged 79 6 2.3 (M 6 SEM) cells per section of DNLL. Both LG and LO brains produced significantly lower per-section DNLL cell counts than found in Sham brains. LG brains averaged 61 6 2.3 cells, t(12) 5 5.25, P , 0.0167 (77.2% of Sham cell count); and LO brains averaged 51 6 2.9 cells, t(12) 5 69.11, P , 0.0167 (64.6% of Sham cell count). The average number of cells observed in the DNLL of grafted rats was significantly greater than that observed in LO rats, t(12) 5 3.07, P , 0.0167. Figure 5 shows the paucity of DNLL cells in LO animals relative to sham-operated and grafted rats. DISCUSSION
The present study demonstrated that pigmented rats, like wild and albino rats (13, 14, 16), can perform a
21
task which requires the ability to locate the sources of target sounds in acoustic space. For all animals, localization rates improved during training as the number of possible target sources was increased, and task performance of control animals continued to improve throughout the experiment. Postlesion deficits in task performance of animals with bilaterally damaged IC were indicative of the role of this auditory midbrain structure in sound localization behavior. These results suggest that the IC in rat, as in cat (6, 15, 24), plays a significant role in the ability to localize sound. In the albino rat, the auditory cortex and medial geniculate body are not necessary for sound localization (16, 18); however, deep lesions of the IC which include its brachial efferents are reported to produce impairments in sound localization behavior (18). The present results suggest that the rat IC is involved in directional hearing, including the detection of and attention to sound in acoustic space (33, 34) as well as the ability to identify sound source locations. These behavioral results are consistent with physiological data suggesting that IC neurons encode binaurally derived intensity and time differences of spatial sound stimuli (5, 17, 19). Animals receiving bilateral grafts after lesions exhibited localization performance that did not differ statisti-
FIG. 5. Photomicrographs of thionin-stained sections through the rostrocaudal midpoint of the dorsal nucleus of the lateral lemniscus (DNLL). (A) Arrows point out the dorsal and ventral aspects of DNLL in sham-operated Rat 9112. (B) In grafted Rat 9115, the DNLL is qualitatively similar to DNLL of Sham rats with somal clusters aligned along fiber tracts of the lateral lemniscus. (C) A corresponding section from Rat 9111 with an IC lesion and no graft shows a paucity of DNLL neurons. Bar, 100 µm.
22
ZRULL AND COLEMAN
cally from the performance of rats with IC lesions only during initial postoperative testing. Subsequently, performance of grafted rats was significantly improved relative to conspecifics with IC lesions and did differ, but not statistically, from control animals. The delay in the restorative effect of tectal grafts on performance of the current sound localization task differs from the immediate postoperative differences between IC lesiononly and grafted animals observed in a simpler sound detection task (33). Sound localization requires the process of identification or knowing what stimulus has occurred; in the present study identification of the same or a different source target noise burst was required. In the previously reported study (33), whether the experimental task required simply detection or an attention process, only knowing that a sound had occurred was important. In more commonly studied forebrain systems, graft-dependent recovery from behavioral deficits can be immediate (20, 28) or delayed (7, 9), dependent upon the behavior to be recovered and the manner in which implanted tissue interacts (see Ref. 2) with the damaged host brain. While grafted rats of the present study never achieved a level of performance equal to that which occurred under preoperative conditions or that of control animals, they did localize sound stimuli at rates statistically better than rats with IC lesions only. Often, partial recovery of behavioral function is reported in experiments assessing the restorative efficacy of neural grafts (e.g., 3, 10, 28, 33). The DNLL has prominent input to the IC that is important for binaural processing of IC neurons. The IC receives bilateral input from DNLL in rat (1, 8, 25) and cat (4, 27). In rat, Li and Kelly (21) have demonstrated that input to IC from DNLL is critical for appropriate encoding of the laterality of sound source in IC neurons. Interaural intensity difference (IID) functions in IC neurons shift after temporary DNLL blockade with kynurenic acid and return to normal after recovery of DNLL. In the present study animals with bilateral IC lesions exhibit substantial retrograde loss of DNLL neurons, which suggests that the IC is the principal target of these neurons, and behavioral findings in the present study show that animals with bilateral lesions are deficient on a sound localization task. These results imply not only the role of damaged IC, but demonstrate the loss of DNLL neurons known to perform functions essential to binaural encoding required for localization of sound in space (21, 27). Grafted animals in the present study showed behavioral recovery not observed in lesioned animals not grafted. There was considerable sparing of DNLL neurons in these animals relative to those with lesions alone. These results suggest that DNLL sparing is due to the presence of the neural grafts. Although grafts were of modest size (with macrophage still observed at
this stage), grafts of IC show long-term survival and viability (31, 35). The recovery in localization ability in these animals may be due to graft tissue providing targets for DNLL neurons or permitting the DNLL cells to reach alternate target neurons in the host brain, perhaps through mechanisms such as growth factor release. It remains to be determined whether behavioral recovery is due to a direct or indirect impact of grafts on the damaged host auditory pathway. Previous work examining the efficacy of neural grafts to restore lost function after neocortical damage suggests that graft-induced survival of thalamic afferents may contribute to behavioral recovery through direct or novel, indirect connections (10, 28). In summary, the present experiment shows that bilateral lesions of rat IC and accompanying retrograde degeneration of the DNLL impair sound localization behavior. These deficits are partially remediated by bilateral implantation of caudal tectal grafts. The mechanisms for observed behavioral recovery may involve direct DNLL–graft connections, the opening of alternative DNLL–host connections, or other alterations involving host auditory brain-stem pathways. ACKNOWLEDGMENTS We acknowledge the contributions from Grants DBS-9200624 and SBR00285 (J.R.C.), the Deafness Research Foundation (J.R.C.), and the Department of Psychology at South Carolina through its Tindall Award (M.C.Z.).
REFERENCES 1. 2.
3.
4.
5.
6.
7.
Adams, J. C. 1979. Ascending projections to the inferior colliculus. J. Comp. Neurol. 183: 519–538. Bjo¨rklund, A., O. Lindvall, O. Isacson, P. Brundin, K. Wictorin, R. E. Strecker, D. J. Clarke, and S. B. Dunnett. 1987. Mechanisms of action of intracerebral neural implants: Studies on nigral and striatal grafts to the lesioned striatum. Trends Neurosci. 10: 509–516. Brundin, P., O. G. Nilsson, R. E. Strecker, O. Lindvall, B. Astedt, and A. Bjo¨rklund. 1986. Behavioural effects of human fetal dopamine neurons grafted in a rat model of Parkinson’s disease. Exp. Brain Res. 65: 235–240. Brunso-Bechtold, J. K., G. C. Thompson, and R. B. Masterton. 1981. HRP study of the organization of auditory afferents ascending to the central nucleus of inferior colliculus. J. Comp. Neurol. 197: 705–722. Calford, M. B., D. R. Moore, and M. E. Hutchings. 1986. Central and peripheral contributions to coding of acoustic space by neurons in inferior colliculus of cat. J. Neurophysiol. 55: 587– 603. Casseday, J. H., and W. D. Neff. 1975. Auditory localization: Role of auditory pathways in brain stem of the cat. J. Neurophysiol. 38: 842–858. Cassel, J. C., C. Kelche, and B. Will. 1990. Time-dependent effects of intrahippocampal grafts in rats with fimbria–fornix lesions. Exp. Brain Res. 81: 179–190.
SOUND LOCALIZATION AND GRAFTED TECTUM 8.
9.
10.
11. 12. 13.
14.
15.
16. 17.
18.
19. 20.
21.
22.
Coleman, J. R., and W. J. Clerici. 1987. Sources of projections to subdivisions of inferior colliculus in rat. J. Comp. Neurol. 262: 215–226. Dunnett, S. B., A. T. Wareham, and E. M. Torres. 1993. Septal grafts improve acquisition of an operant timing task after fimbria-fornix lesions in rats: Role of the interval between lesion and transplantation surgeries. Restorative Neurol. Neurosci. 5: 263–274. Dunnett, S. B., C. N. Ryan, P. D. Levin, M. Reynolds, and S. T. Bunch. 1987. Functional consequences of embryonic neocortex transplanted to rats with prefrontal cortex lesions. Behav. Neurosci. 101: 489–503. Faye-Lund, H., and K. K. Osen. 1985. Anatomy of the inferior colliculus in rat. Anat. Embryol. 171: 1–20. Green, D. M., and J. A. Swets. 1973. Signal Detection Theory and Psychophysics. Krieger, Huntington, NY. Heffner, H. E., and R. S. Heffner. 1985. Hearing in two cricetid rodents: Wood rat (Neotoma floridana) and grasshopper mouse (Onychomys leucogaster). J. Comp. Psychol. 99: 275–288. Heffner, H. E., and R. S. Heffner. 1985. Sound localization in wild Norway rats (Rattus norvegicus). Hearing Res. 19: 151– 155. Jenkins, W. M., and R. B. Masterton. 1982. Sound localization: Effects of unilateral lesions in central auditory system. J. Neurophysiol. 47: 987–1016. Kelly, J. B. 1980. Effects of auditory cortical lesions on sound localization by the rat. J. Neurophysiol. 44: 1161–1174. Kelly, J. B., S. L. Glenn, and C. Beaver. 1991. Sound frequency and binaural response properties of single neurons in rat inferior colliculus. Hearing Res. 56: 273–280. Kelly, J. B., and P. W. Judge. 1985. Effects of medial geniculate lesions on sound localization by the rat. J. Neurophysiol. 53: 361–372. Kirk, R. E. 1982. Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole, Monterey, CA. Labbe, R., A. C. Firl, Jr., E. J. Mufson, and D. G. Stein. 1983. Fetal brain transplants: Reduction of deficits in rats with frontal cortex lesions. Science 217: 470–472. Li, L., and J. B. Kelly. 1992. Inhibitory influences of the dorsal nucleus of the lateral lemniscus on binaural responses in the rat’s inferior colliculus. J. Neurosci. 12: 4530–4539. Lund, R. D., and S. D. Hauschka. 1976. Transplanted neural tissue develops connections with the host rat brain. Science 193: 582–584.
23. 24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
23
Masterton, B., H. Heffner, and R. Ravizza. 1969. The evolution of human hearing. J. Acoust. Soc. Am. 45: 966–985. Masterton, R. B., J. A. Jane, and I. T. Diamond. 1968. The role of brainstem auditory structures in sound localization. II. Inferior colliculus and its brachium. J. Neurophysiol. 31: 96–108. Mercha´n, M. A., E. Saldan˜a, and I. Plaza. 1994. Dorsal nucleus of the lateral lemniscus in the rat: Concentric organization and tonotopic projection to the inferior colliculus. J. Comp. Neurol. 342: 259–278. Sally, S. L., and J. B. Kelly. 1992. Effects of superior olivary complex lesions on binaural responses in rat inferior colliculus. Brain Res. 572: 5–18. Shneiderman, A., D. L. Oliver, and C. K. Henkel. 1988. Connections of the dorsal nucleus of the lateral lemniscus: An inhibitory parallel pathway in the ascending auditory system? J. Comp. Neurol. 276: 188–208. Stein, D. G., C. Palatucci, D. Kahn, and R. Labbe. 1988. Temporal factors influence recovery of function after embryonic brain tissue transplants in adult rats with frontal cortex lesions. Behav. Neurosci. 102: 260–267. Yin, T. C. T., J. C. K. Chan, and D. R. F. Irvine. 1986. Effects of interaural time delays of noise stimuli on low-frequency cells in the cat’s inferior colliculus. I. Responses to wideband noise. J. Neurophysiol. 58: 562–583. Zrull, M. C., C. L. Miner, C. A. Parker, C. L. Kluttz, and J. Rochelle. 1995. Neonatal inferior colliculus lesions affect soundguided navigation. Soc. Neurosci. Abstr. 21: 671. Zrull, M. C., and J. R. Coleman. 1991. Structural features in whole grafts of the rat inferior colliculus. Hearing Res. 55: 117–132. Zrull, M. C., and J. R. Coleman. 1993. Inferior colliculus lesions and grafts: Deficits and recovery of sound locating ability. Soc. Neurosci. Abstr. 19: 1741. Zrull, M. C., and J. R. Coleman. 1996. Effects of tectal grafts on sound detection deficits induced by inferior colliculus lesions in hooded rats. Cell Transplant. 5: 293–304. Zrull, M. C., J. R. Coleman, and C. Bennett. 1991. Tectal grafts promote recovery from behavioral deficits after inferior colliculus lesions. Soc. Neurosci. Abstr. 17: 1138. Zrull, M. C., T. Wessinger, and J. R. Coleman. 1992. Differential sound-evoked metabolic activity in fetal tectum grafted into damaged adult rat inferior colliculus. Restorative Neurol. Neurosci. 4: 285–289.