MS #10359 Recovery of visual behaviors in adult hamsters with the peripheral nerve graft to the sectioned optic nerve. Hitoshi Sasaki, Tetsu Inoue, Hiroyuki Iso* and Yutaka Fukuda Dept. of Physiology, Osaka University Medical School 2-2 Yamadaoka, Suita 565-0871 JAPAN * Dept. of Psychology, Hyogo College of Medicine 1-1 Mukogawa, Nishinomiya 663-8131 JAPAN Key Words: Optic Nerve Regeneration, Peripheral Nerve Transplantation, Superior Colliculus, Restoration of Visual Pathway, Recovery of Function, Open Field, Classical Conditioned ingResponse, Shuttle -Box Aavoidance Running Head: Recovery of visual behaviors in PN grafted hamsters Address correspond to: Hitoshi Sasaki, Dept. of Physiology, Osaka University Medical School 2-2 Yamadaoka, Suita 565-0871 JAPAN E-mail: H.Sasaki@Physiol.com Tel: +81-66879 3611 Fax: +81-66879-3619 ABSTRACT In adult hamsters, the autologous peripheral nerve (PN) was grafted to the sectioned optic nerve to make a bridge to the superior colliculus (SC). Three behavioral tasks were used to test functional recovery of the restored retinocollicularretino-collicular pathway. Firstly, change of spontaneous ambulating activity to a decrease in environmental luminance was examined in an open field. PN grafted hamsters showed a significant increase to 186% in ambulating activity just after light-off, though it was lower than that in normal hamsters (489 %). Secondly, a classical conditioning of total body movements was tested using an increase in luminance as a conditioned stimulus (CS) paired with foot-shocks. In normal hamsters the magnitude of movements during CS increased in acquisition period, then decreased in extinction period on both 2nd and 3rd sessions, while the magnitude remained unchanged in a blind control. PN grafted hamsters showed an increase in the magnitude only on the 3rd session , although it was statistically nearly significant (p=.0619). Following section of the grafted nerve, the conditioned response disappeared completely. And thirdly, a shuttle box avoidance task was examined using a flickering light as CS. Normal hamsters showed improved avoidance scores, while blind controls did not. PN grafted hamsters showed a slight increase in the score, which was similar to that in one-eyed control. Anterogradely transported labeling of WGA-HRP, injected into the vitreous body of the grafted eye, was observed in the graft and the superficial layers of SC. These results confirm and extend our previous finding that PN grafted hamsters can restore some visual function and further suggest that the extent of recovered visual function is as good as in one-eyed animals. INTRODUCTION It has been shown that loss of function caused by damages in CNS can be restored by transplantation of the living tissues. The functional recovery after transplantation of fetal or neonatal tissues has been reported in many sites of the CNS, including the suprachiasmatic nucleus (Sawaki et al. 1984), the nigro-striatale pathway (Nikkhah et al. 1995, cf. Björklund et al. 1992) , the olfactory pathways (Amemori et al. 1988) and the neocortex (Bermudez-Rattoni et al. 1987, Plumet et al. 1990). Recent works have shown that transplantation of the peripheral nerve (PN) graft restores impaired neural pathway in adult CNS, such as the spinal cord (Richardson et al. 1982), and the retinocollicular pathway (cf. Aguayo et al. 1991), indicating a strong potential of the CNS to regenerate even in the adult mammals (cf. Brecknell and Fawcett 1996). When a segment of PN is grafted to the sectioned optic nerve, axotomized retinal ganglion cells (RGCs) can regrow their axons into the graft. This has been shown in many kinds of species such as hamsters, rats, mice and cats (So & Aguayo 1985, Vidal-Sanz et al. 1987, Inoue et al. 1995, Fukuda et al. 1991, Watanabe et al. 1993). It has been further shown histologically that regenerating RGC axons make synaptic contacts with neurons in the superior colliculus (SC), which are normal targets of the RGC axons (Vidal-Sanz et al. 1987, Carter et al. 1989, 1991, Sawai et al. 1996). Neuronal activity to light can be recorded electrophysiologically from not only presynaptic but also postsynaptic SC neurons (Keirstead et al. 1989, Sauve et al. 1995). In addition, EEG desynchronization was induced in the visual and sensorimotor corticies of the PN grafted rats after a flash of light stimulus (Sasaki et al. 1996). Pupillarly light responses and visually evoked responses in the occipital cortex can also be recorded in the PN grafted rats (Thanos 1992). In a previous study, we have shown that shuttle box avoidance behavior using light as a cue stimulus can be restored in the PN grafted hamsters (Sasaki et al. 1993). Discrimination between vertically and horizontally oriented bars in a Y-maze task was reported to be possible in the PN grafted rats (Thanos et al. 1997). Although these findings suggest that the neural activities conducted along the restored visual pathways can successfully be utilized for visually guided behaviors in the PN grafted animals, there still remain some possibility that some task-specific uncontrolled factors might cause artificial results in each behavioral task (Balsam and Tomie 1985). This possibility can be effectively reduced if we obtain the consistent results in different behavioral tasks. Importance of using different stimuli and responses in the behavioral study has been also suggested (Schneider 1969). The purpose of the present study is to determine whether the functional recovery of vision is possible or not in the PN grafted hamsters by using three different visual behaviors in hamsters; spontaneous ambulating activity in an open field, reflexive classical conditioning task and operant behaviors in a shuttle box avoidance learning. We will show that the recovery of visual behaviors is possible in the PN grafted hamsters in all of the three behavioral tasks and further that the degree of functional recovery of vision is as good as in one-eyed animals. A part of these data has been reported in a previous publication and a preliminary report form (Sasaki et al. 1993, Sasaki and Iso 1996). MATERIALS AND METHODS Animals A total of 114 adult male golden hamsters (Mesocricetus auratus) were used in the present experiment. These hamsters, weighing approximately 100 g, were purchased from Japan SLC (Shizuoka, Japan). The animals were housed individually and maintained on a 12:12 light-dark cycle (lights on at 0600 h) with free access to food and water. PN Graft Operation Under anesthesia with chloral hydrate (40 mg/kg, i.p.), the left optic nerve was sectioned intraorbitally at 8-9 weeks of age. A segment of the common peroneal nerve with its branches was removed from the right leg, then sutured to the sectioned optic nerve by using 10-0 strings (Etilon, Ethicon Inc., New Jersey, USA). The grafted nerve was laid over the skull and the other end of the graft was anchored to the temporal muscle. Four to eight weeks after the first surgery, animals were reanesthetized and distal end of the graft was inserted into the ipsilateral SC through a small hole of aspirated visual cortex. The remaining optic nerve was sectioned to facilitate regrowth of regenerating RGC axons into the SC and making synapses to the SC neurons. Two to 8 months after the second operation, behavioral tests started. One optic nerve was sectioned in an one-eyed group, and or both optic nerves were sectioned in an one-eyed or a blind control group, respectively at 7-9 weeks of age. After at least one week of recovery from surgery, behavioral tests started. Sham operated hamsters received the same operation as in blind group except transection of the optic nerve at 8 weeks of age. Open Field Task Animals Twelve adult hamsters were used for recording ambulation in an open field. Three animals received the PN graft operation to make a bridge from the RGCs to the SC. After 12 weeks of recovery from the operation, behavioral test started. One animal served for the sham-operated control. The test started after 1 week of recovery in the sham-operated hamsters. Eight intact animals (7-9 W, 90-152g) served for two control groups; one is light on-off condition (n=4) and the other is light-off condition (Normal / D-D, n=4). Another animal served for a sham operated control, and was tested after 1 week of recovery. As the result in the sham operated animal was the same as in intact animals, the data was combined with those in normal animals with the light on-off condition (Normal / L-D, n=5), as will be described in the following procedures. Apparatus A translucent rectangular chamber (60 cm wide x 60 cm depth x 30 cm height) was used to record animal's ambulating activity. A black and white checkerboard pattern of 2 cm x 2 cm was set beneath the translucent floor (Fig. 1). The chamber was put in an enclosure covered by a black sheet to maintain inside dark and to prevent an external sound noise. Light Conditions A light An electric bulb (60W) set at 120 cm above the floor of the chamber was used as a light source. Illumination level of the light was 60 lx (T-1M, MINOLTA, Osaka, Japan) at the center of the floor, above a background level of 0.1 lx. A heat preventing filter (PEERLESS AGE, PEERLESS Inc., Tokyo, Japan) was placed on the top of the chamber to eliminate heat emission from the light source while a little reduction to the luminous flux. The filter reduced transmission of electric waves with wavelength above 800 nm more than 80 %, while 62% for those at wavelength of 600 nm. Recordings of Ambulating Activity Spontaneous ambulating activity was detected by four pairs of infrared beam sensors (EE-SPW331 & EE-SPW301, OMRON, Kyoto, Japan) with a flip-flop operation. Numbers of the infrared beam cutting was were counted by a programmable controller (C20P, OMRON) interfaced to a personal computer (Macintosh II, Apple Inc.). Procedures In each trial, hamster was put at the central part of the floor of the chamber. The aAmbulatinon g activity was recorded for a three min period in a trial, then the animal was returned to the home cage in the same test room. Following a three min of rest period, the next trial begun in the same manner as previous one. These procedures were repeated for 6 times under the light-on condition in groups Normal / L-D (n=5, a half of four normal, and Å@the one sham operated animals), and PN Grafted / L-D (n=3, the three PN grafted animals hamsters). Effects of changes in luminance were tested on trials 7 and 8 with the light-off. In group Normal / D-D (n=4, the other half of the four normal animals hamsters), the ambulation ambulating activity was recorded all for the 8 trials under with a light-off condition. Classical Aversive Conditioning Animals Twenty-nine adult hamsters were used in the classical conditioning present experiment. Four PN grafted hamsters received behavioral test at 5 to 8 months after the graft operation. Thirteen hamsters underwent bilateral optic nerve sections and served for the blind control. After one week of recovery, they received four training sessions. Twelve normal hamsters received bilateral optic nerve section after three sessions of training. All of the The four PN grafted hamsters also underwent blind operation by sectioning the graft over the skull after the third 3rd session. After one week of recovery, the 4th session was started performed. Apparatus A translucent acrylic experimental chamber (25 x 25 x 25 cm) with a grid floor was placed in a sound-proof room (RION, Tokyo, Japan). A 40W light bulb placed at 60 cm above the floor of the chamber (300 lx above background level of 0.1 lx) and turned on for 5 sec was used as a conditioned stimulus (CS). An unconditioned stimulus (US) was scrambled electrical shock (AC 200V, 0.66 mA) from the grid floor for 1 sec. Recordings of Response A total body movements of the animal were recorded by a piezoelectric accelerometer (GH-313A, GA-245, KYENCE Inc., Osaka, Japan) attached to the chamber, and the signal through a low pass filter set at 60 Hz (FV-624, NF Circuit, Osaka, Japan) was digitized by an A/D converter with a sampling rate at 1 KHz (MacLab, AD Instruments, NSW, Australia) and were stored on a computer (Macintosh LC II, Apple Inc.) . Classical Conditioning Procedures A daily session consisted of 10 trials of habituation to the apparatus, 10 trials of CS habituation, 30 trials of acquisition (CS-US pairing) and 10 trials of extinction. An inter-stimulus interval (ISI) was 5 sec, and inter-trial intervals (ITIs) were random with a mean of 45 sec. Four daily training sessions were performed for all hamsters. Data Processing From the digitized values of the movements, integrated values were calculated during Pre-CS and a CS periods for 4.5 sec, preceding or starting from the onset of CS, excluding 0.5 sec just before the US onset, in order to eliminate contamination of the unconditioned response completely respectively. The integration was also done during a 4.5 sec Pre-CS period, just preceding the CS onset. The ratio for the activity change was calculated by a following equation: % Activity = (activity during CS) / (activity during Pre-CS) x 100 Shuttle Box Avoidance Conditioning Animals Seventy-three adult hamsters were used. Nine hamsters received the PN graft to make a bridge from RGCs to the ipsilateral SC. Seventeen hamsters received bilateral and 7 hamsters received unilateral optic nerve section at 7-9 weeks of age, and 40 intact hamsters at 7-9 weeks of age served as normal control groups. These hamsters were divided into 11 groups, according to different operations, US intensities, number of daily training trials, and size of shuttle-box (Table 1). Two to three months after the surgery, PN grafted hamsters were trained in the shuttle-box (groups PN-100-30/50 and PN-300-50-s). The same test was performed for the blind animals after one week (groups B-100-30 and B-300-50-s) or 12 weeks of recovery (groups B-100-50). Apparatus A two-way shuttle-box (21 x 29 x 19 cm) with a 2.5 cm high center-barrier was used. A smaller size of box (21 x 12 x 7 cm) was also used in four groups (Table 1). Crossing responses were detected by two pairs of infra-red beam sensors (EE-SPW331 & EE-SPW301, OMRON, Osaka, Japan) which were positioned at 6 cm from the center of shuttle box and 4 cm above the grid floor. An unconditioned stimulus (US) was scrambled electric shock delivered through the grid floor with a 300 Kohm resistance on which the animal was staying. Three different intensities of US (AC 100-300 V) were used (Table 1). The shuttle-box was installed in a sound attenuated box of which inside was always illuminated by a dim light (5 W, 2.9 lx). A small electric fan (AC 100V) ventilated the air inside the box, and it also provided a masking noise. Visual Stimulus A flickering (2.5 Hz) light of a light an electric bulb (60 W, 310 lx) at 2.5 Hz was used as CS in most hamsters except two hamsters in PN-100-30/50 group, which were tested using a continuous light of a fluorescent tube (6 W) as CS. In a preliminary study, effects of the flicker CS and the continuous CS a continuous light CS of a fluorescent tube (6 W) were compared in 16 normal hamsters trained with 100 V-US and 30 trials of training per day. There was no statistically significant difference between a mean avoidance scores of these two groups the flicker CS group and that of the continuous CS group. Heat emission A light beam from the light source was covered with a heat-preventing filter (PEERLESS AGE, PEERLESS Inc.). Procedures and Data Analysis The hamsters were trained with an avoidance conditioning procedure with a 10 sec CS-US interval and with random ITIs with a mean of 60 sec (ranging 45-75 sec). After a 5 min of habituation period, the first trial begun with CS onset. Each trial terminated either by the crossing response or 30 sec after CS onset. If the crossing response occur within 10 sec after CS onset, the hamster can avoid the foot-shocks. The crossing response during an inter-trial interval (ITI) was counted as an inter-trial response (ITR). In three groups (N-100-30, B-100-30 and O-100-30) and two PN grafted hamsters (HON 1 and HON 46) in group PN-100-30/50 were trained 30 trials daily while the other hamsters were trained with a daily 50-trials session (Table 1). A total number of 10 daily sessions were performed in most hamsters except four PN grafted hamsters in (group PN-300-50-s, ) which received 16 sessions of training. In the preliminary studies, avoidance score was similar in hamsters which received 30 or 50 daily trials, so that these data were combined together in Figure 3A. Histology For two PN grafted hamsters, which showed improvement of the avoidance scores, 5 ul of WGA-HRP (20%, TOYOBO, Tokyo, Japan) was injected into the vitreous body of the grafted eye with a Hamilton microsyrlinge by pressure (for 10 min). After survival of 48 hr, the animals were reanesthetized with an overdose of chloral hydrate. The hamsters were perfused, in a conventional way, with saline, followed by fixative (1.25% glutaraldehyde and 1% paraformaldehyde) and by 10% sucrose in 0.1 M phosphate buffer at pH 7.4. The brain was removed and cut at 80 um of frontal sections using a cryostat. The sections were reacted with tetrametylbenzidine (TMB) (Mesulam 1978) with some modifications. Usually we soaked and rinsed the preparation with a 9% solution of sodium nitroferricyanide for 20 min at 4 degree C for stabilization and elimination of artifactual non-specific precipitates. As we have found this process is accompanied by a noticeable diminution of terminal labelings, we passed this processed with a few second, expecting to obtain better sensitivity for a few terminals in our preparation. Serial sections were examined with both a light- and a dark- field microscopy. RESULTS Ambulating Activity in Open Field Changes of ambulating activity of the hamsters upon a reduction of environmental illumination level were examined in the open field task. Figure 2 shows the changes of ambulating activity in PN grafted and normal hamsters. In four intact and one sham operated hamsters, the mean ambulating activity decreased gradually over 6 trials under the light condition (Normal/L-D, Fig. 2A). The activity significantly increased to 489 % (paired t-test, t(5) =4.054, p<.01) on the 7th trial, just after light turned off in the normal group (Normal/L-D, Fig. 2B). Though the activity decreased slightly on the 8th trial (314 %), the level was still higher than the mean level in the preceding trials. The mean ambulating activity of three PN grafted hamsters also decreased gradually from 4th to 6th trials under the light condition (PN Grafted/L-D, Fig. 2A), though the mean level was higher than that of the normal hamsters (Normal/L-D, Fig. 2A). The mean ambulation activity significantly increased to 186% (paired t-test, t(3) =3.192, p<.05) just after the light turned off (Fig. 2B). The degree of increment in ambulating activity was lower than that obtained in the normal hamsters. Though the ambulation activity decreased slightly on the 8th trial as in the normal group, the activity was still higher than that on trial 6 (144%, Fig. 2B). These results suggest that ambulating activity of the PN grafted hamsters can be modulated by an alteration in the environmental illumination level. As for a dark control, we recorded 8 trials under dark condition in four other intact hamsters (Normal/D-D). The mean ambulation in activity of the dark control hamsters decreased gradually over the 8 trials. There was no increment on trials 7 and 8 (Fig. 2B). Classical Conditioning of Total Body Movements In the second task, functional recovery of vision was tested in a classical aversive conditioning using a light as CS and electrical foot shocks as US. Figure 3 shows percent changes of total body movements during three conditioning periods; habituation, acquisition and extinction for four sessions. There was no change in any of three groups during habituation period. In 12 normal hamsters, the activity increased during acquisition trials, then decreased in extinction trials on the 2nd session. These changes were observed more markedly on the 3rd session. Following bilateral sections of the optic nerves, the activity decreased to the initial baseline level on the 4th session. On the other hand, in 13 blind hamsters, the mean activity remained unchanged throughout the four sessions. In four PN grafted hamsters, no consistent changes were observed in the first two sessions. However, the activity increased during acquisition and extinction trials on the 3rd session. The increased activity disappeared after section of the graft on the 4th session . In the normal group, there were statistically significant differences between the mean activities in habituation and acquisition trials on the second and third sessions (paired t-test, session 2; t(11) = 2.376, p<.05, session 3; t(9) = 4.700, p<.01). In the blind group, there was no significant change throughout the sessions. In the PN grafted group, a mean activity reached at 107.5 % on the third session, showing a significant trend of difference (t(3) = 2.913, p=.0619). The score in the early half of the extinction period was higher (extinction trials 1-5; 142%) than that in the later half of the extinction period (extinction trials 6-10; 104%). However, there were no statistically significant dDifference between the mean activities in habituation and extinction trials could not attain to a statistically significant level. Shuttle Box Avoidance Conditioning In the third task, functional recovery of vision was tested in the shuttle box avoidance task using flashing light as a cue stimulus and electrical foot shocks as an aversive stimulus. At first we have examined effects of three factors which may have influence on performance in normal hamsters. For this purpose, five normal groups were examined with different US intensities (AC 100, 200 or 300V), number of daily trials (30 or 50 trials), and size of the shuttle box (normal or small). Each of the five normal groups showed similar increase in the avoidance score for 10 sessions. Although the avoidance score in group N-300-50-s was slightly higher than those in other groups after 3rd session, as expected, there was no statistically significant difference between any of these five groups. A mixed type ANOVA was performed on Figure 4. Only the main effect of session was significant (F(9,270)=31.316, p<.0001) and the main effect of group and the interaction between session and group were not significant. Figure 5 shows the avoidance scores in normal, blind , PN grafted and one-eyed hamsters. The increment of the avoidance score in the normal group was gradual and attained around 40 % of avoidance at the 10th session. On the other hand, none of blind groups did improve the avoidance scores, regardless of number of daily trials, US intensity or box size (Fig. 5). Four blind hamsters in B-300-50-s did not show any improvement in avoidance score even after 12 weeks of blind operation (data not shown). There was no difference in the score between blind groups after 1 week (B-100-30) and 12 weeks of recovery (B-100-50). A mean avoidance score of five PN grafted hamsters (PN-100-30/50) showed gradual increment during the training sessions, although the scores were lower than those in the intact group (Fig. 5). A mean avoidance score in the one-eyed hamsters (O-100-30) also increased gradually and the scores were lower than those in intact hamsters, but they were significantly higher than those in blind hamsters. A mixed type ANOVA was performed for the avoidance scores in the blind, one-eyed and PN grafted groups. The main effects of group and session, and the interaction between group and session were all significant (F(4/19)=4.307, p=.0120, F(9/171)=5.562, p<.0001, and F(36/171)=2.181, p=.0005, respectively). Post-hockh analysis (Fisher's PLSD, p<.05) showed significant differences between PN graft and each of three blind groups, between one-eyed and B-100-50, and between one-eyed and B-300-50-s. There was some decrement in avoidance latency in normal hamsters over 10 sessions. However, we could not find any significant differences in avoidance latency among normal, blind and PN grafted groups. A mean ITR during these 10 training sessions was less than 1.5 per 10 sec and it did not change significantly throughout the sessions. There were no significant differences in the ITRs among normal, blind and PN grafted hamsters. In our preliminary study, effects of deficits in motor activity caused by the PN graft operation on the avoidance learning were tested in three hamsters, of which peripheral nerve was sectioned unilaterally. After one week of recovery, the control group received avoidance training with 100V US and daily session of 30 trials. No difference in the performance of avoidance was observed between the normal and the control group with the peripheral nerve lesion. The mean avoidance score of the control reached at 60% (ranging 433.3-83.3%) in the 10th session. This score was rather higher than that in the normal group, suggesting that no significant motor deficits interfered with the performance in this task. Mean latency of the avoidance in the control decreased from 11.4+- 0.1 ms to 8.7+-1.1 ms during 10 sessions of training. Mean ITR was less than 1.1 per 10 sec in all of the animals. Histology In two PN grafted hamsters, we have ensured the restored connection of the axotomized RGC to the SC neurons. During the avoidance training sessions for 16 days, both of these hamsters showed an increase of the avoidance scores, though the increase was slight but was consistent (Fig. 6). WGA-HRP was injected into the vitreous body in these two PN grafted hamsters. Anterogradely transported labeling of WGA-HRP was observed in the graft and the superficial layers of SC in these two hamsters. Fig. 7 shows an example of labeling in the SC in one of these hamsters. The outline of the SC of the grafted side was largely distorted and pull upward by the graft (Fig. 7A). Accordingly, shapes of the layers of SC were also distorted especially at the lateral side near the entrance of the graft. The graft was tightly connected to SC and the border of the graft and SC was unclear. The labeling was most dense in the graft, and some labeled fibers could be seen in the stratum griseum superficiale (SGS) and the stratum opticum (SO) (Fig. 7B). The labeling area extended 0.56 mm rostro-caudally. The terminal labeling was not clear in our preparation, because of sparse terminals. In another hamster, a the similar results was were observed. DISCUSSION The pPurpose of the present study was to determine whether the visual inputs conducted through the restored retinocollicularretino-collicular pathway can be used for modulation of visual behaviors in the PN grafted animals. We tested this using three behavioral tasks, which measure different aspects of behaviors, and using different change of light. In all of three tests, PN grafted hamsters showed some improvements in visually guided behaviors. This finding is consistent with the preceding reports in showing that the PN grafted animals can learn visual behaviors (Kittlerova 1992, Sasaki et al. 1993, Thanos et al. 1997), and strongly confirms the functional recovery of vision in the PN grafted animals, because there seems little possibility that there still remain some uncontrolled factors which will cause misleading results. One might argue that whether there is any correlation between performance of the grafted animals on different tasks. However, it is generally agreed that experience of aversive shock will modulate the following behaviors. If two behavioral tests were done successively in the same animal using aversive stimuli, results of the second test should be affected by the preceding aversive experience, thus the results should be difficult to be interpreted. It is also well known that repeated exposures to a CS alone retard the development of conditioned response during subsequent CS-US pairing (Reiss and Wagner 1972). This effect, called as "latent inhibition", makes it difficult to compare results of the open field using change of light to those of conditioning with a light-CS. Therefore, correlation between performance on different tasks was not analyzed in the present study. Recovery of Spontaneous Ambulating Activity to Decreased Illumination Level In the present study, the PN grafted hamsters showed increased ambulating activity just after the light turned off. All hamsters showed habituation of exploratory activity with trails. Decrement in the environmental luminance induced dishabituation and lead to an abrupt increase in the activity in normal hamsters. On the contrary, no significant increase can be observed if the luminance was kept constant (Normal/ D-D), thus confirming that the increased activity in the PN grafted hamsters is caused by the changes in luminance in the environment. These findings are well consistent with the previous study in normal rats (Thinus-Blanc and Foreman 1993). In the PN grafted hamsters, the activity increased just after decreasing luminance. These results suggest that the luminance change can be detected by the PN grafted hamsters. Recovery of Visually Classical Conditioned Behavior to Increased Illumination Level In the classical conditioning study using a light turned on as CS, PN grafted hamsters showed increased locomotor activity conditioned responses (CRs). The increased activity CRs disappeared after section of the graft, confirming that the change CRs was were induced by the visual signals conducted through the grafted nerve. Although it has been shown that unavoidable electrical shocks will cause the freezing behavior in rats and hamsters (cf. Dean and Redgrave 1984, Garcia et al. 1998), we did not observe any sign of freezing behavior in the present experiment. Hamsters showed jumping, rotating or scratching behavior when they received the short-lasting electrical foot shocks and almost all normal hamsters showed increased locomotor activity during the acquisition period. Similar increase in activity has been reported in the Wistar rats using the same apparatus (Iso 1998). A question arises as to whether the increased locomotor activity is the conditioned response (CR) due to conditioning or not. To assess this point, we repeated habituation, acquisition and extinction trials in each daily session. In spite of the fact that Although extinction procedure retards the following interfered with acquisition of the CR, normal hamsters showed a marked increase in the activity during acquisition period on the 2nd and 3rd sessions. And the activity decreased to near the baseline level in the extinction periods. These results indicate that the changes of the activity are due to Pavlovian fear conditioning. In addition, after sections of both optic nerves, no more increase of the activity could be observed, supporting the idea that the change in behavioral activity is caused by the light-on CS. PN grafted hamsters showed an increased activity at somewhat late as compared to intact hamsters. The increased activity in the intact group was observed on the 2nd and 3rd acquisition periods, while in the PN grafted group, it was observed only on the 3rd acquisition period. Magnitude of activity change was also less than that in the intact group. The increased activity was nearly significant (P=.0619), it may be due to the retardation caused by the repeated extinction procedure and due to a small number of animals in this group (n=4). These findings show that the PN grafted hamsters have also developed the CRs, though the acquisition was slow. It seems reasonable to assume that the slower acquisition in the PN grafted group is due to decreased CS inputs, because it is generally accepted that attenuated CS intensity decreases the establishment of CRs (Moon et al. 1994). Recovery of Operant Behavior to Using Flashing Light as a Cue Stimulus In the present avoidance experiment, the PN grafted hamsters showed some avoidance behavior using a flashing light as CS a cue stimulus. A mean avoidance rate in the PN grafted hamsters was lower than that in the intact hamsters, though it was significantly higher than that in the blind controls. These findings are consistent with the previous findings in rats and hamsters (Kittlerova 1992, Sasaki et al. 1993). The finding that avoidance rate was lower in the PN grafted hamster than in normal controls might be due to motor deficits caused by removal of common peroneal nerve. In our preliminary study, however, the avoidance rates in the PN sectioned hamsters were similar to those in intact hamsters, thus excluding the possibility that some the motor deficits interfered with the avoidance score. The shuttle box avoidance learning might be a rather difficult task for hamsters. The level of avoidance score at 10th session in normal hamsters was around 40%. This is extremely low as compared to that in rats or mouse (Sasaki et al. 1993, Masu et al. 1995). The reason why acquisition is slow in hamsters could be ascribed to their behavioral characteristics. Hamsters frequently showed rearing behaviors spontaneously or after receiving electrical shocks from the grid floor. On the contrary, rats showed rare rearing behavior in the shuttle box and readily easy to move forward when they received the shock from the floor. Such behavioral property in hamsters seems to interfere with acquisition of the avoidance task. This hypothesis is supported by the fact that the avoidance task by the rearing behavior can easily be acquired in hamsters (Iso 19998), in contrast to the fact that the same task is difficult in rats (Shimai and Imada 1978). Degree of Functional Recovery of Vision It is rather surprising that the one-eyed hamsters showed lower performance than those in normal hamsters in the avoidance task. We have repeated the experiments in rats and obtained the similar results as in hamster (data not shown). Thus, the results are not specific to the species. There are two possibilities to explain these poorer performance in one-eyed animals. Firstly, deficits in decreased range of total visual area in the one-eyed hamsters might result in the difficulty in locating themselves to the next compartment to avoid the shock. This possibility can be excluded by the fact that the escape latency from the foot shock was not different in both one-eyed and normal groups (3.33 +- 0.44 sec vs. 3.21 +- 0.32 sec, respectively). Secondly, the poorer performance in one-eyed animals can be explained by reduced inputs to the CNS. Reduced total visual inputs seems to be parallel to reduced intensity of CS. Decreased performance was observed by using lower CS intensity (cf. Moon et al. 1994). The level of the avoidance behavior in the PN grafted hamsters was similar to that in one-eyed hamsters. On the other hand, the number of RGCs which restored the retinocollicularretino-collicular connections was only a few percent of the total number of RGCs (Vidal-Sanz et al. 1987), suggesting that amount of the visual inputs which can be used as cue stimulus in the PN grafted hamsters is much less than that in one-eyed hamsters. This apparent discrepancy between a small amount of visual inputs and a comparable level of the visual behavior in the PN grafted hamsters may be ascribed to a non-liner relationship between stimulus intensity and avoidance ratio. It seems possible to assume that in a certain range of intensity, two visual cues with different intensity will induce the same behavior, if they are above the behavioral threshold. Recovery of Visually Guided Behaviors Anatomical and electrophysiological findings show that the RGC axon can regrow and make synaptic contacts with neurons in the superficial layers of SC (Carter et al. 1991) and the collicularcollicular neurons can be activated by a visual stimulus presented to the grafted retina (Keirstead et al. 1989, Sauve et al. 1995). Both findings strongly suggest that the visual behaviors can be restored by the PN graft bridging from the retina to SC. Although in our histological data, the terminal like labeling, which is commonly observed in the normal retinocollicularretino-collicular projections, could not be well identified, the regenerated RGC axons could be fully traced in the superficial layers of SC, where the normal RGC axons terminate. In addition, after lesion of the grafted nerves, the visual behavior returned to the baseline level in open field and classical aversive conditioning. From these pieces of evidence, we can conclude that the restoration of the visual behaviors in the PN grafted hamsters is due to the PN graft operation, because there is no other pathway for the visual inputs to reach the CNS than via the PN graft. In the present study, we have tested visual behaviors in the PN grafted hamsters using three different visual stimulus, reduced luminance, increased luminance or flickering light in the three different behavioral tasks. Some recovery of visual behaviors could be observed in these three tasks, suggesting that both ON and OFF visual pathways are functioning in the PN grafted hamsters. The visual performance levels of the PN grafted hamsters were lower than those of normal hamsters in all of three tasks used in the present study. This result is well consistent with the previous finding that the visual threshold is increased in the PN grafted rats, as measured by using behavioral arousal and EEG desynchronization (Sasaki et al. 1996). However, the degree of functional recovery was as good as in the one-eyed hamsters, in spite of the fact that the number of restored RGC fibers in the PN grafted hamsters was much less than that in the one-eyed hamsters. Pathways for the Visual Behaviors In the preset study, the PN grafted hamsters showed some improved performances in the three different behavioral tasks. The performance of the grafted animals was similar to that of the one-eyed animals in the avoidance test. These facts suggest that the visual signals conducted along the regenerated RGC axons, transmitted further from the SC to the higher-order neuronal circuits, which are involved in these behaviors. It has been reported that the amygdala is involved in fear conditioning (Davis, 1992). Lesion of the amygdala impaired the avoidance behavior in rats (Werka et al. 1978, Werka and Zielinski, 1998) and also blocked fear-potentiated startle response in the fear conditioning using visual CS in rats (Campeau and Davis, 1995). In addition, lesions restricted to the central nucleus of the amygdala produced increased exploration in an open-field test (Werka et al., 1978). These facts strongly suggest that the amygdala is involved in the conditioned and the open field behaviors in the present study. On the other hand, two visual information processing systems have been widely accepted, one is the geniculo-striate and the other is the tecto-thalamo-cortical system (Schneider, 1969). It has been known that the SC projects to the lateral posterior nucleus (LP) of the thalamus (Mooney 1984), then the LP projects to the peristriate cortex in rats (Hughes, 1977, Perry 1980). From the peristriate cortex, the visual signals can conduct along the ventral pathway as in primates, to reach the amygdala through the temporal cortex (Mcdonald and Mascagni 1996). Thus, it seems to be reasonable to assume that the visual signals are transmitted from the SC to the amygdala, which plays important roles both in the fear conditioning and the open field tasks. The fact that the PN grafted animals showed some improved performance in these behavioral tasks strongly suggests that visual signals conducting through the restored retinocollicular pathway can be transmitted from the SC to the peristriate visual cortex through the LP, and further to the amygdala. CONCLUSION The present study provides strongly evidenced to show that the PN grafted hamsters can exhibit show some visually guided behaviors by using their restored retinocollicularretino-collicular pathway. Both reduction and increase in the luminance can be detected by the PN grafted hamsters. Moreover, the degree of functional recovery was similar to that of hamsters with one optic nerve transected. As one of the optic nerve has been completely sectioned in the PN grafted hamsters, the recovery of visually guided behaviors in the PN grafted hamsters seems to be fairly well, although the performance in the PN grafted hamsters is still lower than that in normal hamsters. Other visual functions, such as detection of moving spots and shape discrimination should be tested in future studies. ACKNOWLEDGMENT We thanks to Dr. Y. Tsukamoto for his assistance to take photographs of histological preparations. REFERENCE Amemori, T., V. Valouskova, T. Zigova, J. Galik, E. Racekova, J. Bures. 1988. Functional recovery after olfactory bulbectomy in rats: effect of embryonal brain grafts. Physiol. Bohemoslov. 37: 385-394. Aguayo, A. J., M. Rasminsky, G. M. Bray, S. Carbonetto, L. McKerracher, M. P. Villegas-Perez, M. Vidal-Sanz, D. A. Carter. 1991. Degenerative and regenerative responses of injured neurons in the central nervous system of adult mammals. Philos. Trans. R. Soc. Lond. B Biol. Sci. 331: 337-343. Balsam, P. D. and Tomie, A. 1985. (Eds.) Context and learning. Hillsdale, Erlbaum, NJ. Björklund A 1992 Dopaminergic transplants in experimental parkinsonism: cellular mechanisms of graft-induced functional recovery. Curr. Opin. Neurobiol., 2: 683-689. Brecknell J. E. and J. W. Fawcett. 1996. Axonal regeneration. Biol. Rev. Camb. Philos. Soc., 71: 227-255. Bermudez-Rattoni, F., J. Fernandez, M. A. Sanchez, R. Aguilar-Roblero, R. Drucker-Colin. 1987. Fetal brain transplants induce recuperation of taste aversion learning. Brain Res. 416 :147-152. Campeau, S, and M. Davis. 1995. Involvement of the central nucleus and basolateral complex of the amygdala in fear conditioning measured with fear-potentiated startle in rats trained concurrently with auditory and visual conditioned stimuli. J. Neurosci., 15: 2301-11. Carter, D. A., G. M. Bray, A. J. Aguayo. 1989. Regenerated retinal ganglion cell axons can form well-differentiated synapses in the superior colliculus of adult hamsters. J. Neurosci. 9: 4042-4050. Carter, D. A., A. J. Aguayo, G. M. Bray. 1991. Retinal ganglion cell terminals in the hamster superior colliculus: an ultrastructural study. J Comp Neurol., 311: 97-107. Davis, M. 1992. The role of the amygdala in conditioned fear. In The Amygdala: Neurobiological aspects of emotion, memory, and mental dysfunction (J. P. Aggleton, Ed.), pp. 255-305. Wiley-Liss, Inc., New York. Fukuda, Y., Watanabe, M., Sawai, H. Axonal regeneration of cat retinal ganglion cells after peripheral nerve transplant. Invest. Ophthalmol. Vis. Sci. 32: 1133, 1991. Hughes, H. C. 1977. Anatomical and neurobehavioral investigations concerning the thalamo-cortical organization of the rat's visual system. J. Comp. Neurol. 175: 311-336. Inoue, T., H. Sasaki, Y. Fukuda. 1995. Axonal regeneration of retinal ganglion cells by peripheral nerve transplant in adult mice. Society for Neurosci. Abs. 21 (1) p309. Iso, H. 19998. Rat's behavior in the course of aversive conditioning measured by a stabilimeter. Jap. J. Psychol. 1998 (in press). Keirstead, S.A., M. Rasminsky, Y. Fukuda, D.A. Carter, A.J. Aguayo, and M. Vidal-Sanz. 1989. Electrophysiologic responses in hamster superior colliculus evoked by regenerating retinal axons. Science 246: 255-257. Kittlerova, P. 1992. Regeneration of retinal ganglion cell axons through the sciatic nerve isograft into the superior colliculus - electrophysiological and behavioral study. MSc thesis, Charles University. Masu, M., Iwakabe, H., Tagawa, Y., Miyoshi, T., Yamashita, M., Fukuda, Y., Sasaki, H., Hiroi, K., Nakamura, Y., Shigemoto, R., Takada, M., Nakamura, K., Nakao, K., Katsuki, M. and Nakanishi, S. 1995. Specific deficit of the ON response in visual transmission by targeted disruption of the mGluR6 gene. Cell, 80: 757-765. Moon, Y., M. M. Ghoneim, I. Gormezano. 1994. Nitrous oxide: sensory, motor, associative, and behavioral tolerance effects in classical conditioning of the rabbit nictitating membrane response. Pharmacol. Biochem. Behav. 47: 523-529. Mooney, R.D., S.E. Fish, and R.W. Rhoades. 1984. Anatomical and functional organization of pathway from superior colliculus to lateral posterior nucleus in hamster. J. Neurophysiol. 51:407-431 . Nikkhah, G., M.G. Cunningham, R. McKay, A. Björklund. 1995. Dopaminergic microtransplants into the substantia nigra of neonatal rats with bilateral 6-OHDA lesions. II. Transplant-induced behavioral recovery. J. Neurosci. 15: 3562-3570. Perry, V.H. 1980. A tectocortical visual pathway in the rat. Neuroscience 5:915-927. Plumet, J., J. Cadusseau, M. Roger. 1990. Fetal cortical transplants reduce motor resulting from neonatal damage to the rat's frontal cortex. Neurosci. Lett., 109: 102-106. Reiss, S. and A.R. Wagner. 1972. CS habituation produces a "latent inhibition effect" but no "conditioned inhibition." Learning and Motivation, 3: 237-245. Richardson, P.M., U.M. McGuinness, A. J. Aguayo. 1982. Peripheral nerve autografts to the rat spinal cord: studies with axonal tracing methods. Brain Res. 237:147-162. Sasaki, H., T. Inoue, Y. Fukuda, H. Iso, and Y. Hayashi. 1993. Light-dark discrimination after sciatic nerve transplantation to the sectioned optic nerve in adult hamsters. Vision Res. 33: 877-880. Sasaki, H., and H. Iso. 1996. Conditioned emotional responses in hamsters with restored visual pathway by peripheral nerve graft. Int. J. of Psychol. 31: 309. Sasaki H. , P. Coffey, M.P. Villegas-Perez, M. Vidal-Sanz, M. J. Young, R. D. Lund and Y. Fukuda. 1996. Light induced EEG desynchronization and behavioral arousal in rats with restored retinocollicular projection by peripheral nerve graft. Neurosci. Lett. 218: 45-48. Sauve, Y., H. Sawai, M. Rasminsky. 1995. Functional synaptic connections made by regenerated retinal ganglion cell axons in the superior colliculus of adult hamsters. J. Neurosci. 15: 665-75. Sawai, H., M. Sugioka, K. Morigiwa, H. Sasaki, K-F. So, and Y. Fukuda. 1996. Functional and morphological restoration of intracranial brachial lesion of the retinocollicular pathway by peripheral nerve autografts in adult hamsters. Exp. Neurol. 137: 94-104. Sawaki, Y., I. Nihonmatsu, and H. Kawamura. 1984. Transplantation of the neonatal suprachiasmatic nuclei into rats with complete bilateral suprachiasmatic lesions. Neurosci. Res. 1: 67-72. Schneider, G. E. 1969. Two visual systems. Science 163:895-902. Shimai, S., and H. Imada. 1978. The effects of shock intensity on discriminated rearing avoidance conditioning in rats. Jap. Psychol. Res. 20: 143-147. So, K. -F., and A. J. Aguayo. 1985. Lengthy regrowth of cut axons from ganglion cells after peripheral nerve transplantation into the retina of adult rats. Brain Res. 328: 349-354. Taylor, A.M., G. Jeffery and A.R. Lieberman. 1986. Subcortical afferent and efferent connections of the superior colliculus in the rat and comparisons between albino and pigmented strains. Exp. Brain Res. 62:131-142. Thanos, S. 1992. Adult retinofugal axons regenerating through peripheral nerve grafts can restore the light-induced pupilloconstriction reflex. Eur. J. Neurosci. 4:691- 699. Thanos, S., R. Naskar, and P. Heiduschka 1997. Regenerating ganglion cell axons in the adult rat establish retinofugal topography and restore visual function. Exp. Brain Res. 114: 483-91. Thinus-Blanc, C., and N. Foreman. 1993. A rapid test of rodents' vision using a modified open field apparatus. Physiol. Behav. 53: 1003-1006. Vidal-Sanz, M., G. M. Bray, M. P. Villegas-Perez, S. Thanos, and A. J. Aguayo. 1987. Axonal regeneration and synapse formation in the superior colliculus by retinal ganglion cells in the adult rat. J. of Neurosci. 7: 2894-2909. Watanabe, M., H. Sawai, and Y. Fukuda. 1993. Number, distribution and morphology of retinal ganglion cells with axons regenerated into peripheral nerve graft in adult cats. J. Neurosci. 13: 2105-2117. Werka, T., J. Skar, and H. Ursin. 1978. Exploration and avoidance in rats with lesions in amygdala and piriform cortex. J. Comp. Physiol. Psychol. 92:672-81. Werka, T. and K. Zielinski. 1998. CS modality transfer of two-way avoidance in rats with central and basolateral amygdala lesions. Behav. Brain Res. 93:11-24. LEGENDS Figure 1. Schematic drawings of an open field apparatus for testing visual stimulus detection. A translucent acrylic chamber (60 cm x 60 cm x 30 cm) with four pairs of infrared beam sensors (S) was put on a black and white checker board pattern (2 cm x 2 cm) in a sound-attenuating box. A heat preventing filter (F) was set between an electric bulb (60W) and the chamber. Further details are described in the text. Figure 2. A: changes of mean ambulating activity in four normal and one sham operated hamsters (Normal/L-D), three PN grafted hamsters (PN Grafted/L-D) and four normal hamsters (Normal/D-D). After 6 trials, the light turned off in the groups Normal/L-D and PN Grafted/L-D, while Normal/D-D received 8 trials under dark condition. The activity decreased during first 6 trials in all groups. Just after reducing illumination on trial 7, the activity increased in both Normal/L-D and the PN Grafted/L-D groups, but did not in Normal/D-D control. B: percent changes of ambulating activity (mean with SE ) indicating changes from light-on (6th) to light-off (7th and 8th) trials. The activity in Normal/L-D showed a statistically significant increase on both 7th and 8th trials (** p<.01). On the other hand, there was no change in the Normal/D-D. The activity increased significantly on the 7th trial in PN Grafted/L-D (* p<.05). Figure 3. Percent changes of activity in a 5-sec ISI period to that of the preceding pre-CS period for four sessions of classical aversive conditioning task. In each session, habituation (HABI), acquisition (ACQ) and extinction (EXT) trials were performed. After 3rd session, the optic nerves were sectioned bilaterally in the normal or the grafted nerve was sectioned in the PN grafted hamsters. Normal hamsters showed statistically significant increase in the activity during acquisition on the 2nd and 3rd sessions (* p<.05, ** p<.01). PN grafted hamsters showed increase in the activity on the 3rd session, though it is statistically nearly significant (p= .0619). On the other hand, there was no significant change in the blind hamsters. Vertical line shows the time of blind operation in PN grafted and normal groups. Figure 4. Changes of avoidance scores during 10 sessions of Shuttle-box avoidance task in normal hamsters. Shock intensity was either AC 100, 200 or 300V, and number of trials per day was 50 except group N-100-30, which received 30 trials (see Table 1). Size of shuttle box was reduced in groups N-200-50-s and N-300-50-s. Either increased shock intensity, number of trials per day or reduced box size slightly increased avoidance scores, especially evidenced in group N-300-50-s, though there was no statistically significant difference. Each score represents mean value with SE. Figure 5. Changes of avoidance scores in normal (N), blind (B), one-eyed (O) and PN grafted (PN) hamsters. Both PN grafted and one-eyed groups showed significant increase of avoidance scores, though they were less than those in normal controls. On the other hand, no significant increase in avoidance scores was observed in any of blind groups, regardless of US intensity of 100 or 300V, number of trials of 30 or 50 per day, or a reduced box size in group B-300-50-s. Also, a recovery period of 12 weeks has no effect on the score in B-100-50 group. For abbreviation in group names, see Table 1. Figure 6. Percent changes of the avoidance score during 16 sessions in two PN grafted hamsters trained with 300V US and 50 trials per day in a small shuttle box. Although the avoidance scores were low, both of these PN grafted hamsters showed an consistent increase in the avoidance scores especially in the later sessions (sessions 8-16) as compared to those in the early sessions (1-7). Figure 7. Histological data showing fiber connection from the retina to the SC in a PN grafted hamster (T# 181) which showed behavioral recovery of visual function. A: camera lucida drawings of serial coronal sections of the superior colliculus. Anterograde labeling of WGA-HRP injected in the grafted eye were shown as a dot. Rostral is upper left and caudal is lower right. Scale bar shows 1 mm. B: a dark-field microscopy which shows anterogradely labeled WGA-HRP in the superficial layers of SC at the position marked by a square in the section in A. It still contains artifactual non-specific precipitates, because of mild rinsing the preparation with the post-reaction solution to avoid diminution of terminal labelings. - 3517 -