- Email: [email protected]
Spontaneous and artificial lesions of magnocellular reticular formation of brainstem deteriorate avoidance learning in senescence-accelerated mouse SAM
Brain Research 791 Ž1998. 90–98
Research report
Spontaneous and artificial lesions of magnocellular reticular formation of brainstem deteriorate avoidance learning in senescence-accelerated mouse SAM Hideo Yagi
a, )
, Ichiro Akiguchi a , Akira Ohta b , Noriko Yagi c , Masanori Hosokawa d , Toshio Takeda d a
Department of Neurology, Faculty of Medicine, Kyoto UniÕersity, Kyoto, Japan Department of Psychology, Faculty of Letters, Kyoto UniÕersity, Kyoto, Japan c Faculty of Nutrition, Kohshien UniÕersity, Hyogo, Japan Department of Senescence Biology, Chest Disease Research Institute, Kyoto UniÕersity, Kyoto, Japan b
d
Accepted 13 January 1998
Abstract The role of the magnocellular reticular formation ŽMGRF. of the brainstem on learning and memory was examined in memory-deficient mice with spontaneous spongy degeneration in the brainstem Žsenescence-accelerated mouse, SAMP8. and control mice Žaccelerated-senescence resistant mouse, SAMR 1.. SAMP8 showed spontaneous age-related impairment of learning and memory, as determined by passive and active avoidance responses. The deficits of learning and memory function in passive avoidance performances began at two months of age and increased with ageing. In the brains of SAMP8 at one month of age and older, spongy degeneration was mainly observed in the brainstem, while no vacuoles were evident in SAMR1 control Žnormal ageing mouse. brains in the age range tested Žup to 12 months.. The vacuolization in SAMP8 was marked in the MGRF, especially in the dorsomedial MGRF. Quantitative analysis of the vacuolization showed that the total area and number of vacuoles in the MGRF increased with age, and they were affected by the degree of deficits in learning and memory. The latency 24 h after footshock in passive avoidance tests decreased with the increase in total area and number of vacuoles in MGRF. The number of shocks in active avoidance tests increased with the increase in total number and area of vacuoles. Thus, learning and memory ability in passive and active avoidance responses deteriorated with enlargement in the vacuolated area in MGRF, and it was assumed that MGRF Žespecially, the dorsomedial part. possesses functions related to learning and memory. To confirm this notion, behavior and memory tests Žpassive avoidance and active avoidance tests, open field tests and shock sensitivity measurements. were carried out in SAMR1 mice, whose bilateral dorsomedial MGRF was destroyed electrolytically ŽMGRF-lesioned mice.. The MGRF-lesioned mice showed no difference from sham mice in sensory threshold or open field activity; however, there was severe deterioration in passive avoidance behavior and impairment in the active avoidance performances. From the results in SAMP8 and MGRF-lesioned mice, it was confirmed that MGRF Žespecially the dorsomedial part. has functions related to learning and memory, and is one part in the learning and memory system of the brain. Thus, SAMP8 can serve as a model of RF-lesioned mice with impaired learning and memory functions. q 1998 Elsevier Science B.V. Keywords: Active avoidance; Brainstem; Magnocellular reticular formation; Passive avoidance; Senescence-accelerated mouse SAM; Spongy degeneration
1. Introduction The senescence-accelerated mouse ŽSAM. was established as an animal model of accelerated senescence by Takeda et al. w18x. SAM includes accelerated senescence prone SAMP ŽSAMP1, P2, P3, P6, P7, P8, P9 and P10. and accelerated senescence resistant mouse SAMR ) Corresponding author. Takeda Hospital, Higashi-Shiokoji-cho, 841-5, Shimogyo-ku, Kyoto, Japan. Fax: q81-75-361-7602.
0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 8 . 0 0 0 7 0 - 5
ŽSAMR1, R2 and R4. w6,17x. SAMP shows signs of advanced senescence, such as reduced activity, hair loss, lack of hair glossiness, skin coarseness, periophthalmic lesions, increased lordokyphosis of the spine, and a shortened life span. In the SAMP series, SAMP8 showed more spontaneous age-related deterioration of memory and learning abilities in passive avoidance response tests w10x, especially in the acquisition stage w25x, than did control SAMR1 mice. Impairment of passive avoidance behavior began to occur
H. Yagi et al.r Brain Research 791 (1998) 90–98
at 2 months of age and became increasingly severe with ageing. The SAMP8 also showed age-associated deterioration of active avoidance behavior w13x. Until now, some important neuropathological changes have been found in SAMP8 brains. Remarkable changes of SAMP8 brains were vacuolization mainly in the brainstem w26x and PASpositive granular structures ŽPGS. chiefly in the hippocampus w1x, and not found in the other brain sites as far as we examined histopathologically. Vacuolization begins to appear at 1 month of age in the brainstem reticular formation, and the vacuoles increase with age up to 8 months of age. One of the other pathological changes in SAMP8 brains is the appearance of PAS-positive granular structures ŽPGS., chiefly in the hippocampus w1x. The PGS in SAMP8 appear at 3 months of age in the hippocampus, and their numbers increase rapidly with age from 6 months on. The PGS in control SAMR1 develop only at a very old age and are never seen in young mice. We attempted to determine whether these neuropathological changes in SAMP8 brains are associated with a deteriorating capacity for passive andror active avoidance behavior. Of the neuropathological changes in SAMP8, vacuolization in the brainstem reticular formation seems to be closely related to a deficit in learning and memory abilities, because both the vacuolization and the impairment in learning and memory abilities begin at almost the same age and become severe with ageing. This study was designed to determine whether spongy degeneration in the brainstem of SAMP8 is associated with deterioration of learning and memory ability, and to clarify whether the brainstem reticular formation, especially MGRF, is a locus of learning and memory systems. 2. Materials and methods Senescence-accelerated mouse, SAMP8 and senescence-accelerated resistant mouse, SAMR1 as controls, were used in this experiment. SAMP8 and SAMR1 mice were housed in group cages Žaverage area of 0.01 m2rmouse. under conventional conditions with a temperature- and light-controlled regimen Ž24 " 28C and a 12-h lightrdark cycle with lights on at 7:00 h. and free access to food ŽCE-CLEA Japan. and tap water. All these animals were reared in the Department of Senescence Biology, Chest Disease Research Institute, Kyoto University, Japan. All the tests were performed between 13:00 and 19:00, and the background noise was constant. 2.1. BehaÕioral procedures There were significant differences in a single-trial passive avoidance w25x and active lever press avoidance w13x, but no differences in T-maze avoidance w25x between SAMP8 and SAMR1. Therefore, in this experiment, acquisition and retention abilities and behavior of SAMP8 and SAMR1 were examined using an open field method, the-
91
single trial passive avoidance w25x and the active lever press avoidance tests w13x. In the passive avoidance test, a two-compartment step-through passive avoidance apparatus was used. Once a mouse entered the dark chamber from the illuminated one, an AC 0.5 mA scrambled footshock was applied for 3 s. Retention of the passive avoidance response was examined by replacing the mouse in the illuminated chamber 24 h after the training trial, and the latency of entering the dark chamber was measured. When the mouse continued to stay in the illuminated chamber over 300 s Žcriterion of retention time s 300 s., the test was terminated, and a ceiling score of 300 s was assigned. Mice with a latency of 300 s in the passive avoidance response were classified as passive–intact, and those with a latency less than 300 s as passive–poor. In the open field tests, a floor in the field box was divided into 64 squares measuring 9 = 9 cm. The mouse was placed in the center of the field, and the number of squares touched by all 4 paws was recorded for 3 min. In the active avoidance test, the apparatus was a Skinner box for mice set in a sound-attenuated cubicle. A 0.5 mA scrambled footshock was delivered to the mouse for 0.5 s. The active avoidance response of the mouse was the pressing of a bar. The shock–shock interval was 5 s, and the response–shock interval was 25 s. The experiment, 30 min in duration, was carried out once a day, and 9 sessions were run for each subject. Shock sensitivity thresholds were determined by the flinch-jump threshold method w25x. An ascending series of footshocks with a 30-s interval was administered stepwise Ž0.05 mA. from 0.05 to 0.7 mA. Thresholds for flinch, jump and vocalization were defined as the lowest shock sensitivity eliciting each response. 2.2. Surgical procedure Male and female SAMR1 mice aged 3–4 months were used to produce an artificially brainstem-lesioned mouse model. SAMR1 mice received an injection of atropine sulfate Ž0.1 mgrkg i.p.. and were anesthetized with pentobarbital sodium Ž45 mgrkg i.p... They were then placed in a stereotaxic apparatus according to the method of Montemurro and Dukelow w11x. An incision was made along the midline of the scalp to expose the skull. Small burr holes were drilled in the skull at an appropriate stereotaxic coordinate, and the dura mater was carefully slit to insert electrodes into the brains. A cashew-coated stainless steel electrode Ž0.12 mm diameter, uncoated for 0.1 mm from the tip. was inserted at the coordinate of 2.7–2.8 mm posterior to the interaural line, 0.5 mm lateral to the midline and 4.6 mm ventral to the cortex surface. First, a lesion was produced on the right side of the brainstem by the application of 100 m A-DC for 20 s. After 4 weeks of recovery, a contralateral Žlt side. lesion was made with 100 m A–20 s DC. Sham-operated mice Žsham mice. were prepared in the same manner but without the application of DC. One week after the creation of the bilateral lesions,
92
H. Yagi et al.r Brain Research 791 (1998) 90–98
examine the distributions of electrolytic lesions along the neuroaxis. The sections were stained with hematoxylin and eosin ŽH & E. and luxol fast-blue ŽKluver-Barrera.. Immunohistochemical staining was carried out with anti-cow GFAP Žglial fibrillary acidic protein. antibody ŽDakopatts, Denmark. w26x. For microscopic morphometry of vacuoles in SAMP8 brainstem, a computerized image analyzer LUZEX3U ŽNikon, Japan. was used w15x. The sections of the brainstem stained with H & E were photographed with a light microscope and were printed in A4 size. The vacuoles in the photoprints were carefully traced and filled in with black ink on transparent paper ŽFig. 1a,b.. The number and area of the traced black regions Žs vacuoles. were measured using the LUZEX3U connected to a TV camera. 2.4. Statistics Mann–Whitney U-test, t-test and analysis of variance ŽANOVA. with repeated measures by an unweighted means were used for statistical analysis of the results.
3. Results 3.1. PassiÕe aÕoidance performance in SAMP8 and SAMR1 There was no significant difference in the latency before footshock between SAMP8 and SAMR1 except at age
Fig. 1. Example of spongy degeneration in brainstem of 4-month-old SAMP8. Ža. H&E staining. Žb. Vacuoles in the reticular formation of Ža. are traced in black ink.
open field, passive avoidance tests, active avoidance tests and shock sensitivity measurements were carried out, in that order, in the brainstem-lesioned mice Ž7 males, 4 females. and in the sham mice Ž6 males, 5 females.. Further, seven male lesioned and 3 male sham mice were used only in passive avoidance tests, and 3 male lesioned and 2 male sham mice were used only in active avoidance tests. 2.3. Pathological procedure and microscopic morphometry After completion of the behavior and passive and active avoidance tests, the mice were given an overdose of ethyl ether, and the brains were removed immediately and immersed in 10% neutral buffered formalin. The fixed brains were embedded in paraffin, and coronal sections 6 m m thick were cut every 20 m m to demonstrate the area and number of vacuoles in the brainstem, and every 50 m m to
Fig. 2. Age-related changes in one trial passive avoidance response in SAMP8 and SAMR1. Ža. Mean latency before footshock and Žb. median of latency, one day after footshock. Hollow and solid circles mean raw data in SAMR1 and SAMP8, respectively. ) p- 0.05, )) p- 0.01, ))) p- 0.001 ŽMann–Whitney U-test for SAMR1 controls..
H. Yagi et al.r Brain Research 791 (1998) 90–98
2 and 12 months ŽFig. 2a.. After footshock, most of SAMR1 controls at all ages tested reached the criterion Žretention time of 300 s., while impairment of passive avoidance performance began at 2 months of age in SAMP8 ŽFig. 2b.. The number of SAMP8 reaching the criterion and median of latency 24 h after footshock decreased with age Ž p - 0.05 at 4 months; p - 0.01, at 8 months; p 0.001, at 12 months, for SAMR1.. Thus, SAMP8 showed spontaneous age-related deterioration in retention ability of passive avoidance response. 3.2. Distribution and age dependency of spongy degeneration in SAMP8 brains Vacuolization was observed in the brainstem of SAMP8 mice from 1 month of age onward, while no vacuoles were evident in SAMR1 brains in the age range examined Žup to 12 months.. The distribution of spongy degeneration was plotted in the passive–poor SAMP8 mice at 2 and 4 months of age ŽFig. 3.. Vacuolization began to appear at the level of the hypothalamic arcuate nucleus in coronal
93
sections and was marked in the brainstem, especially in the magnocellular reticular formation ŽMGRF. Žpontine reticular nucleus, caudal part ŽPnC. and gigantocellular reticular nucleus ŽGi... The region of spongy degeneration was wider in 4-month-old than in 2-month-old SAMP8 mice. To examine the effect of age on vacuolization in detail, we counted the total number and measured quantitatively the total area of vacuoles at the levels of PnC and Gi ŽFigs. 4 and 5., where spongy degeneration was the most severe, as shown in Fig. 3. Because the distribution of vacuoles was related not only to ageing but also to the degree of impairment in retention abilities in the passive avoidance responses, the total number and area of vacuoles were plotted against age in two representative cases Žpassive–intact: latency ŽLT. s 300 s and passive–poor: LT s 100– 120 s.. The total number of vacuoles in PnC and Gi increased almost linearly with age in both, and it was larger in passive–poor than in passive–intact SAMP8 in the age range tested ŽFig. 4.. The total area of vacuoles in PnC and Gi also increased linearly with ageing, and it was larger in passive–poor SAMP8 ŽFig. 5.. Thus, the number
Fig. 3. Topographic distributions of spongy degeneration in passive–poor SAMP8 at 2 Žrt side. and 4 Žlt side. months of age. Solid circles mean a few or several vacuoles. The latency in the passive avoidance test was 54 s at 2 months and 20 s at 4 months of age in SAMP8 mice. Abbreviations: 4 V, fourth ventricle; 6, abducens nucleus; 7, facial nucleus; 10, dorsal motor nucleus of vagus; 12, hypoglossal nucleus; Arc, arcuate hypothalamic nucleus; CG, central gray; DR, dorsal raphe nucleus; Dtg, dorsal tegmental nucleus; Gi, gigantocellular reticular nucleus; Hi, hippocampus; IC, inferior colliculus; Lc, locus coeruleus; LRt, lateral reticular nucleus; mlf, medial longitudinal fasciculus; Mo5, motor trigeminal nucleus; MP, medial mammillary nucleus, posterior part; PCRt, parvocellular reticular nucleus; PGi, paragigantocellular reticular nucleus; Pn, pontine nucleus; PnC, pontine reticular nucleus, caudal part; PnO, pontine reticular nucleus, oral part; Pr5, principal sensory trigeminal nucleus; PT, pretectal area; py, pyramidal tract; R, red nucleus; SN, substantia nigra; Sol, nucleus solitary tract; Ve, vestibular nucleus; VTg, ventral tegmental nucleus; ZI zona incerta.
94
H. Yagi et al.r Brain Research 791 (1998) 90–98
Fig. 4. Age-related changes in total number of vacuoles in PnC and Gi. Hollow and solid circles mean passive–intact SAMP8 ŽLT s 300 s. and passive–poor SAMP8 ŽLT s100–120 s.. LT means latency one day after footshock.
and area of vacuoles in RF were markedly affected by ageing, and were closely related with degrees of impairment in retention ability. 3.3. Relationship of spongy degeneration to impairment of passiÕe and actiÕe aÕoidance performances To clarify in further detail the effect of vacuolization on acquisition and retention ability, the relationship between
Fig. 5. Age-related changes in total area of vacuoles in PnC and Gi. Hollow and solid circles have the same meaning as in Fig. 4.
Fig. 6. Relationship between latency in the passive avoidance response and Ža. total number and Žb. area of vacuoles in PnC and Gi in 2-month-old SAMP8. Hollow and solid circles represent PnC and Gi, respectively.
the total area or number of vacuoles and the latency Ž24 h after footshock. in the passive avoidance response was examined in PnC and Gi at 2 and 4 months of age in SAMP8 ŽFigs. 6 and 7.. The latency in passive avoidance
Fig. 7. Relationship between latency in the passive avoidance response and Ža. total number and Žb. area of vacuoles in PnC and Gi in 4-month-old SAMP8. Hollow and solid circles represent PnC and Gi, respectively.
H. Yagi et al.r Brain Research 791 (1998) 90–98
95
increase in total number and area of vacuoles. Thus, acquisition and retention ability in SAMP8 deteriorated with the enlargement of the region of spongy degeneration in the brainstem in both passive and active avoidance responses. Moreover, vacuolization in the dorsomedial part of MGRF was most severe in passiveractive poor SAMP8, as shown in Fig. 3, while it was weak in passiveractive intact SAMP8. 3.4. BehaÕior and acquisition and retention ability in MGRF-lesioned SAMR1 Fig. 10 shows an example of bilateral electrolytic lesions of the brainstem in SAMR1 mice ŽH & E and GFAP
Fig. 8. Sidman active avoidance responses in SAMP8 and SAMR1 Ž4 months old.. Hollow and solid circles mean raw data in SAMR1 and SAMP8, respectively.
behavior decreased with the increase in total area and number of vacuoles in both PnC and Gi at 2 months of age ŽFig. 6.. At 4 months of age ŽFig. 7., the latency decreased with the increase in total number and area of vacuoles in PnC and Gi as at 2 months of age. Active avoidance test result in 4-month-old SAMR1 and SAMP8 was shown in Fig. 8. The total number of shocks the mice received in a session decreased with increasing sessions in both SAMR1 and SAMP8; however, SAMP8 received markedly more shocks than SAMR1 at all sessions tested Ž F Ž7,82. s 2.58, p - 0.05.. The number of shocks were plotted against the total number and area of vacuoles in PnC and Gi of SAMP8 representatively at 8th session in active avoidance tests ŽFig. 9a,b.. The number of shocks increased with the
Fig. 9. Relationship between total number Ža. and total area Žb. of vacuoles in Gi and total number of shocks at 8th session in active avoidance performances. r means Bravais Pearson correlation.
Fig. 10. Example of electrolytic lesionings of the reticular formation of the brainstem of SAMR1. Ža. H&E staining. Žb. anti-GFAP-PAP staining, corresponding to the section in Ža.. Arrows point to lesioned area.
H. Yagi et al.r Brain Research 791 (1998) 90–98
96
Fig. 12. Active avoidance responses in MGRF-lesioned, sham and SAMP8 mice. Fourteen MGRF-lesioned, 13 sham and 8 SAMP8 mice were tested. Vertical lines represent S.E.M.
Fig. 11. Topographic distributions of electrolytic lesions in MGRF of 3–4-month-old SAMR1 mice. The dotted regions represent the lesioned areas in 3 representative MGRF-lesioned mice.
staining.. The lesioned areas were demonstrated more widely and more clearly with GFAP staining ŽFig. 10b. than with H & E staining ŽFig. 10a.. Fig. 11 shows some examples of topographic distributions of electrolytic lesions in MGRF stained with H & E and GFAP staining. Most bilateral dorsomedial parts of MGRF were destroyed bilaterally. Behavior and acquisition and retention ability were examined in the MGRF-lesioned and sham mice. There was no significant difference in behavior or passive and active avoidance tests between male and female of MGRF-lesioned or sham mice. Therefore, the total number of male and female mice tested was used for statistical analysis. The number of squares entered in the open field was almost the same for MGRF-lesioned and sham mice ŽTable 1.. No difference in shock sensitivity Žflinch, jump
and vocalization thresholds. was evident between the two groups. Thus, sensory function and activityrexploratory behavior were not affected by bilateral lesioning of the dorsomedial MGRF. In the passive avoidance test, there was no difference in the latency before shock between MGRF-lesioned and sham mice. Sham mice showed as good retention ability as the non-operated SMRI controls, but the MGRF-lesioned mice were very inferior to the sham mice in retention ability of passive avoidance response. Most of the MGRF-lesioned mice did not reach the criterion after footshock, and the median of latency was significantly shorter in the MGRF-lesioned mice than in the sham ŽTable 1.. The MGRF-lesioned mice also showed inferior ability in active avoidance behavior. Fig. 12 shows the average number of shocks received at each session in the active avoidance tests. In each of MGRF-lesioned, sham and SAMP8 mice, the number of shocks decreased as the session increased. The MGRF-lesioned mice received more shocks than the sham mice in a session of active avoidance tests Ž F Ž8,200. s 2.16, p - 0.05., though they received fewer shocks then SAMP8 mice did.
Table 1 Behavior and passive avoidance responses in MGRF-lesioned and sham mice Behavior
MGRF-lesioned
Sham
Shock sensitivity ŽmA.
Flinch Jump Vocal
0.12 " 0.03 Ž11. 0.30 " 0.09 Ž11. 0.42 " 0.08 Ž11.
0.13 " 0.03 Ž11. 0.31 " 0.10 Ž11. 0.47 " 0.09 Ž11.
Open field
Number of cross Ž3 min.
137 " 34 Ž11.
134 " 42 Ž11.
Step-through latency
Before shock Žmean. After shock Žmedian.
20 " 6 Ž13. 43 ) ) Ž13.
29 " 9 Ž10. 300 Ž10.
Each value represents the mean " S.D. Median value is used only for representing latency after shock. Number in the parenthesis is number of mice tested. )) P - 0.01 ŽMann–Whitney U-test. for sham mice.
H. Yagi et al.r Brain Research 791 (1998) 90–98
Thus, the bilateral dorsomedial MGRF-lesioned mice showed severe impairment of passive avoidance behavior and deterioration in active avoidance behavior. That is, deficits of acquisition and retention ability were caused by bilateral lesions of dorsomedial MGRF of the brainstem, without any disturbances of sensory or motor functions.
4. Discussion The present study showed the following: Ž1. Spongy degeneration was marked in the brainstem Žespecially, the magnocellular reticular formation, MGRF. of SAMP8, and it became severe with ageing. Ž2. The total area and number of vacuoles in MGRF ŽPnC and Gi. increased with age. The latency in passive avoidance behavior decreased with the increase in total area and number of vacuoles in MGRF. The degree of vacuolization in MGRF in SAMP8 was closely related to the deterioration of learning and memory ability in the passive and active avoidance performances. Ž3. Bilateral electrolytic lesioning of the dorsomedial MGRF in SAMR1 mice resulted in impairment of passive and active avoidance behavior without any disturbance of sensory function or open-field activity. The above results indicate that MGRF Žespecially, the dorsomedial part. possesses to some degrees learning and memory functions or that MGRF bears one part in learning and memory systems of brains. There are two methods of examining learning and memory functions in the brainstem: lesioning and electrophysiological studies. One lesioning study showed that learning in the uphill avoidance task was possible in rats devoid of telencephalon and thalamus w7x, suggesting that the brainstem has some degrees of learning and memory ability. Destruction of some nuclei in the brainstem, e.g., pedunculo–pontine nucleus–parabrachial nucleus w2,8x or median raphe nucleus w21x resulted in deterioration of mazelearning and of active avoidance and visual discrimination learning. Midbrain reticular formation-lesioned rats showed impairment in T-maze performance w19,20x. Acquisition of Y-maze and jump avoidance behavior was impossible for rats with bilateral symmetrical lesions of the PnC w12x. These results suggest that some nuclei and reticular formation in the brainstem play important roles in learning and memory function, and that they constitute memory networks even at the level of the brainstem and connect with higher order learning and memory networks such as the limbic system and the temporal cortex through ascending fibers. The present study, for the first time, showed that a caudal part of RF, especially dorsomedial MGRF, also plays an important role in learning and memory functions. Electrophysiological studies have shown that neurons in the RF of the brainstem have learning and memory functions. The activities of neurons or neuronal groups in the brainstem were recorded in animals performing various learning tasks w14,16,27x. Unit responses appeared in the
97
posterior pontine RF of the brainstem in rats during the learning of a classical appetite-related conditioned task w9x, and the development of discriminative neuronal discharges was observed in the pontine RF during acquisition behavior in rabbits w5x. Thus, neurons or neuronal groups in the posterior RF of the brainstem responded to tasks requiring learning and memory, suggesting that neurons in RF possess some degrees of learning and memory functions. On the other hand, RF was reported to affect higher-order learning and memory systems through ascending fibers such as the medial longitudinal fasciculus ŽMLF.. Electric stimulation of MGRF elicited hippocampal rhythmic slow activity ŽRSA or theta rhythm. through ascending fibers w22,23x. A change in frequency of RSA in the hippocampus was related to degrees of learning and memory functions w3x, and loss of theta rhythm resulted in complete deficits of learning and memory w24x. Therefore, bilateral lesions of MGRF might affect RSA in the hippocampus through ascending fibers and result in the deterioration of learning and memory. In the present lesioning study, although the MGRF-lesioned mice showed inferior performances in both passive and active avoidance behavior, as did SAMP8, the deterioration was more severe in passive avoidance than in active avoidance responses. The active avoidance tests need more complicated responses and skills for stimulation Že.g., to press a bar to avoid electric shock. and higher-order learning and memory ability than do the passive avoidance tests. Higher-order or the other learning and memory systems may play more important roles in the learning of active avoidance w19x. The electrolytic lesions in the present study were limited to the dorsomedial MGRF. Wider artificial lesions in the RF may result in more severe deterioration of active avoidance behavior. The electrolytic method used in the present study destroyed fibers of passage as well as neuron cell bodies in MGRF. In order to clarify a role of the neurons in learning and memory functions more clearly, a fiber-sparing method should be used in the future. Another neuropathological change, PGS, appeared rapidly at 6 months of age in SAMP8 w1x, while deterioration of passive and active avoidance behaviors has occurred at 2 and 4 months of age, respectively. Therefore, there seems to be no direct relationship between PGS and impairment of passive and active avoidance behavior. However, the number of PGS increased with age, especially in the hippocampus, so PGS may have a significant effect on the deterioration in the other learning and memory behavior in SAMP8. Flood et al. w4x suggested a correlation between increased PGS and deterioration in appetitive learning in SAMP8. Further studies on the effect of PGS on learning and memory ability in SAMP8 are needed. The serological test showed negative results regarding mouse hepatitis virus, Sendai virus, Mycoplasma pulmonis and Tyzzerr’s organism in both SAMP8 and SAMR1.
98
H. Yagi et al.r Brain Research 791 (1998) 90–98
Transmission study and an analysis of the genetic data will be the subjects of the forthcoming reports on the etiology of vacuolization and PGS. In this experiment, we focused on MGRF to explain the cause of deterioration in learning and memory ability of SAMP8. But, the other several learning and memory systems, involving the cerebellum, the hippocampus, the amygdala and the temporal cortex may also play important roles in learning and memory ability of SAMP8. Further study will be required to clarify roles of the other memory systems andror the relation between MGRF and the other memory systems in SAMP8. 5. Conclusion Learning and memory ability consists of a diffuse system including the brainstem, and MGRF ŽPnC and Gi. Žespecially dorsomedial part. possesses some degree of learning and memory functions, and is one site of the diffuse learning and memory networks of brains. SAMP8 can serve as a brainstem-lesioned, especially MGRF-lesioned, model manifesting memory deficiency and may provide new insights into learning and memory functions. Acknowledgements This work was supported by grants from the Ministry of Education, Culture and Science of Japan. We thank T. Matsushita and K. Kogishi for technical assistance. References w1x H. Akiyama, H. Kameyama, I. Akiguchi, H. Sugiyama, T. Kawamata, H. Fukuyama, H. Kimura, M. Matsushita, T. Takeda, Periodic acid-Schiff ŽPAS.-positive, granular structures increase in the brain of senescence-accelerated mouse ŽSAM., Acta Neuropathol. ŽBerlin. 72 Ž1986. 124–129. w2x F. Dellu, W. Mayo, J. Cherkaouri, L. Moal, H. Simon, Learning disturbances following excitotoixic lesion of cholinergic pedunculopontine nucleus in the rat, Brain Res. 544 Ž1991. 126–132. w3x Z. Elazar, W.R. Adey, Spectral analysis of low frequency components in the electrical activity of the hippocampus during learning, Electroenceph. Clin. Neurophysiol. 23 Ž1968. 225–240. w4x J.F. Flood, J.E. Morley, Early onset of age-related impairment of aversive and appetitive learning in the SAM-Pr8 mouse, J. Gerontol. 47 Ž1992. 52–59. w5x M. Gabriel, B. Gregg, A. Clancy, M. Kittrel, W. Dailey, brainstem reticular formation neuronal correlates of stimulus significance and behavior during discriminative avoidance conditioning in rabbits, Behav. Neurosci. 100 Ž1986. 171–184. w6x H. Hosokawa, R. Kasai, K. Higuchi, S. Takeshita, K. Shimizu, H. Hamamoto, A. Honma, M. Irino, K. Toda, A. Matsumura, M. Matsushita, T. Takeda, Grading score system: a method for evaluation of the degree of senescence in senescence-accelerated mouse ŽSAM., Mech. Ageing Dev. 26 Ž1984. 91–102. w7x J.P. Huston, C. Tomaz, I. Fix, Avoidance learning in rats devoid of the telencephalon plus thalamus, Behav. Brain Res. 17 Ž1985. 87–95.
w8x S.F. Ivanova, J. Bures, Acquisition of conditioned taste aversion in rats is prevented by tetrodotoxin blockade of a small midbrain region centered around the parabrachial nuclei, Physiol. Behav. 48 Ž1990. 543–549. w9x G. Kornblith, J. Olds, Unit activity in brainstem reticular formation of the rat during learning, J. Neurophysiol. 36 Ž1973. 489–501. w10x M. Miyamoto, Y. Kiyota, N. Yamazaki, A. Nagaoka, T. Matsuo, Y. Nagawa, T. Takeda, Age-related changes in learning and memory in the senescence accelerated mouse ŽSAM., Physiol. Behav. 38 Ž1986. 399–406. w11x D.G. Montemurro, R.H. Dukelow ŽEds.., A Stereotaxic Atlas of the Diencephalon and Related Structures of the Mouse. Futura Publishing, Mount Kisko, New York, 1972. w12x G. Muller, F. Klinberg, Learning and retention of active avoidance are differently impaired after dorsal and ventral lesions of the nucleus reticularis pontis caudalis of rats, Biomed. Biochim. Acta. 48 Ž1989. 817–827. w13x A. Ohta, T. Hirano, H. Yagi, S. Tanaka, M. Hosokawa, T. Takeda, Behavioral characteristics of the SAM-Pr8 strain in Sidman active avoidance task, Brain Res. 498 Ž1989. 195–198. w14x E.B. Pragay, A.F. Mirsky, C.L. Ray, D.F. Turner, C. Mirsky V, Neuronal activity in the brainstem reticular formation during performance of a ‘go–no go’ visual attention task in the monkey, Exp. Neurol. 60 Ž1978. 83–95. w15x A. Shimada, A. Ohta, I. Akiguchi, T. Takeda, Inbred SAM-Pr10 as a mouse model of spontaneous, inherited brain atrophy, J. Neuropathol. Exp. Neurol. 51 Ž1992. 440–450. w16x D.L. Sparks, R.J. Travis Jr., Patterns of reticular unit activity observed during performance of a discriminative task, Physiol. Behav. 3 Ž1968. 961–967. w17x T. Takeda, M. Hosokawa, K. Higuchi, Senescence-accelerated mouse ŽSAM.: a novel murine model of accelerated senescence, J. Am. Geriatr. Soc. 39 Ž1991. 183–194. w18x T. Takeda, M. Hosokawa, S. Takeshita, M. Irino, K. Higuchi, T. Matsushita, Y. Tomita, K. Yasuhira, H. Hamamoto, K. Shimizu, M. Ishii, T. Yamamuro, A new murine model of accelerated senescence, Mech. Ageing Dev. 17 Ž1981. 183–194. w19x R. Thompson, Localization of a ‘passive avoidance memory system’ in the white rat, Physiol. Psychol. 6 Ž1978. 263–274. w20x R. Thompson, Retention of individual spatial reversal problems in rats with nigral, caudoputamenal and reticular formation lesions, Behav. Neural Biol. 34 Ž1982. 98–103. w21x R. Thompson, A. Ramsay, J. Yu, A generalized learning deficit in albino rats with early median raphe or pontine reticular formation lesions, Physiol. Behav. 32 Ž1984. 107–114. w22x R.P. Vertes, An analysis of ascending brainstem systems involved in hippocampal synchronization and desynchronization, J. Neurophysiol. 46 Ž1981. 1140–1159. w23x R.P. Vertes, Brainstem modulation of the hippocampus: anatomy, physiology and significance, in: R.L. Isaacson, K. Pribram ŽEds.., The Hippocampus, Vol. 4, Plenum, New York, 1986 pp. 41–75. w24x J. Winson, Loss of hippocampal theta rhythm results in spatial memory deficit in the rat, Science 201 Ž1978. 160–163. w25x H. Yagi, S. Katoh, I. Akiguchi, T. Takeda, Age-related deterioration of ability of acquisition in memory and learning in senescence accelerated mouse: SAM-Pr8 as an animal model of disturbance in recent memory, Brain Res. 474 Ž1988. 86–93. w26x H. Yagi, M. Irino, T. Matsushita, S. Katoh, M. Umezawa, T. Tsuboyama, M. Hosokawa, I. Akiguchi, R. Tokunaga, T. Takeda, Spontaneous spongy degeneration of the brainstem in SAM-Pr8 mice, a newly developed memory-deficient strain, J. Neuropathol. Exp. Neurol. 48 Ž1989. 577–590. w27x N. Yoshi, P. Pruvot, H. Gastaut, Electrographic activity of the mesencephalic reticular formation during conditioning in the cat, Clin. Neurophysiol. 9 Ž1957. 595–608.
Research report
Spontaneous and artificial lesions of magnocellular reticular formation of brainstem deteriorate avoidance learning in senescence-accelerated mouse SAM Hideo Yagi
a, )
, Ichiro Akiguchi a , Akira Ohta b , Noriko Yagi c , Masanori Hosokawa d , Toshio Takeda d a
Department of Neurology, Faculty of Medicine, Kyoto UniÕersity, Kyoto, Japan Department of Psychology, Faculty of Letters, Kyoto UniÕersity, Kyoto, Japan c Faculty of Nutrition, Kohshien UniÕersity, Hyogo, Japan Department of Senescence Biology, Chest Disease Research Institute, Kyoto UniÕersity, Kyoto, Japan b
d
Accepted 13 January 1998
Abstract The role of the magnocellular reticular formation ŽMGRF. of the brainstem on learning and memory was examined in memory-deficient mice with spontaneous spongy degeneration in the brainstem Žsenescence-accelerated mouse, SAMP8. and control mice Žaccelerated-senescence resistant mouse, SAMR 1.. SAMP8 showed spontaneous age-related impairment of learning and memory, as determined by passive and active avoidance responses. The deficits of learning and memory function in passive avoidance performances began at two months of age and increased with ageing. In the brains of SAMP8 at one month of age and older, spongy degeneration was mainly observed in the brainstem, while no vacuoles were evident in SAMR1 control Žnormal ageing mouse. brains in the age range tested Žup to 12 months.. The vacuolization in SAMP8 was marked in the MGRF, especially in the dorsomedial MGRF. Quantitative analysis of the vacuolization showed that the total area and number of vacuoles in the MGRF increased with age, and they were affected by the degree of deficits in learning and memory. The latency 24 h after footshock in passive avoidance tests decreased with the increase in total area and number of vacuoles in MGRF. The number of shocks in active avoidance tests increased with the increase in total number and area of vacuoles. Thus, learning and memory ability in passive and active avoidance responses deteriorated with enlargement in the vacuolated area in MGRF, and it was assumed that MGRF Žespecially, the dorsomedial part. possesses functions related to learning and memory. To confirm this notion, behavior and memory tests Žpassive avoidance and active avoidance tests, open field tests and shock sensitivity measurements. were carried out in SAMR1 mice, whose bilateral dorsomedial MGRF was destroyed electrolytically ŽMGRF-lesioned mice.. The MGRF-lesioned mice showed no difference from sham mice in sensory threshold or open field activity; however, there was severe deterioration in passive avoidance behavior and impairment in the active avoidance performances. From the results in SAMP8 and MGRF-lesioned mice, it was confirmed that MGRF Žespecially the dorsomedial part. has functions related to learning and memory, and is one part in the learning and memory system of the brain. Thus, SAMP8 can serve as a model of RF-lesioned mice with impaired learning and memory functions. q 1998 Elsevier Science B.V. Keywords: Active avoidance; Brainstem; Magnocellular reticular formation; Passive avoidance; Senescence-accelerated mouse SAM; Spongy degeneration
1. Introduction The senescence-accelerated mouse ŽSAM. was established as an animal model of accelerated senescence by Takeda et al. w18x. SAM includes accelerated senescence prone SAMP ŽSAMP1, P2, P3, P6, P7, P8, P9 and P10. and accelerated senescence resistant mouse SAMR ) Corresponding author. Takeda Hospital, Higashi-Shiokoji-cho, 841-5, Shimogyo-ku, Kyoto, Japan. Fax: q81-75-361-7602.
0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 8 . 0 0 0 7 0 - 5
ŽSAMR1, R2 and R4. w6,17x. SAMP shows signs of advanced senescence, such as reduced activity, hair loss, lack of hair glossiness, skin coarseness, periophthalmic lesions, increased lordokyphosis of the spine, and a shortened life span. In the SAMP series, SAMP8 showed more spontaneous age-related deterioration of memory and learning abilities in passive avoidance response tests w10x, especially in the acquisition stage w25x, than did control SAMR1 mice. Impairment of passive avoidance behavior began to occur
H. Yagi et al.r Brain Research 791 (1998) 90–98
at 2 months of age and became increasingly severe with ageing. The SAMP8 also showed age-associated deterioration of active avoidance behavior w13x. Until now, some important neuropathological changes have been found in SAMP8 brains. Remarkable changes of SAMP8 brains were vacuolization mainly in the brainstem w26x and PASpositive granular structures ŽPGS. chiefly in the hippocampus w1x, and not found in the other brain sites as far as we examined histopathologically. Vacuolization begins to appear at 1 month of age in the brainstem reticular formation, and the vacuoles increase with age up to 8 months of age. One of the other pathological changes in SAMP8 brains is the appearance of PAS-positive granular structures ŽPGS., chiefly in the hippocampus w1x. The PGS in SAMP8 appear at 3 months of age in the hippocampus, and their numbers increase rapidly with age from 6 months on. The PGS in control SAMR1 develop only at a very old age and are never seen in young mice. We attempted to determine whether these neuropathological changes in SAMP8 brains are associated with a deteriorating capacity for passive andror active avoidance behavior. Of the neuropathological changes in SAMP8, vacuolization in the brainstem reticular formation seems to be closely related to a deficit in learning and memory abilities, because both the vacuolization and the impairment in learning and memory abilities begin at almost the same age and become severe with ageing. This study was designed to determine whether spongy degeneration in the brainstem of SAMP8 is associated with deterioration of learning and memory ability, and to clarify whether the brainstem reticular formation, especially MGRF, is a locus of learning and memory systems. 2. Materials and methods Senescence-accelerated mouse, SAMP8 and senescence-accelerated resistant mouse, SAMR1 as controls, were used in this experiment. SAMP8 and SAMR1 mice were housed in group cages Žaverage area of 0.01 m2rmouse. under conventional conditions with a temperature- and light-controlled regimen Ž24 " 28C and a 12-h lightrdark cycle with lights on at 7:00 h. and free access to food ŽCE-CLEA Japan. and tap water. All these animals were reared in the Department of Senescence Biology, Chest Disease Research Institute, Kyoto University, Japan. All the tests were performed between 13:00 and 19:00, and the background noise was constant. 2.1. BehaÕioral procedures There were significant differences in a single-trial passive avoidance w25x and active lever press avoidance w13x, but no differences in T-maze avoidance w25x between SAMP8 and SAMR1. Therefore, in this experiment, acquisition and retention abilities and behavior of SAMP8 and SAMR1 were examined using an open field method, the-
91
single trial passive avoidance w25x and the active lever press avoidance tests w13x. In the passive avoidance test, a two-compartment step-through passive avoidance apparatus was used. Once a mouse entered the dark chamber from the illuminated one, an AC 0.5 mA scrambled footshock was applied for 3 s. Retention of the passive avoidance response was examined by replacing the mouse in the illuminated chamber 24 h after the training trial, and the latency of entering the dark chamber was measured. When the mouse continued to stay in the illuminated chamber over 300 s Žcriterion of retention time s 300 s., the test was terminated, and a ceiling score of 300 s was assigned. Mice with a latency of 300 s in the passive avoidance response were classified as passive–intact, and those with a latency less than 300 s as passive–poor. In the open field tests, a floor in the field box was divided into 64 squares measuring 9 = 9 cm. The mouse was placed in the center of the field, and the number of squares touched by all 4 paws was recorded for 3 min. In the active avoidance test, the apparatus was a Skinner box for mice set in a sound-attenuated cubicle. A 0.5 mA scrambled footshock was delivered to the mouse for 0.5 s. The active avoidance response of the mouse was the pressing of a bar. The shock–shock interval was 5 s, and the response–shock interval was 25 s. The experiment, 30 min in duration, was carried out once a day, and 9 sessions were run for each subject. Shock sensitivity thresholds were determined by the flinch-jump threshold method w25x. An ascending series of footshocks with a 30-s interval was administered stepwise Ž0.05 mA. from 0.05 to 0.7 mA. Thresholds for flinch, jump and vocalization were defined as the lowest shock sensitivity eliciting each response. 2.2. Surgical procedure Male and female SAMR1 mice aged 3–4 months were used to produce an artificially brainstem-lesioned mouse model. SAMR1 mice received an injection of atropine sulfate Ž0.1 mgrkg i.p.. and were anesthetized with pentobarbital sodium Ž45 mgrkg i.p... They were then placed in a stereotaxic apparatus according to the method of Montemurro and Dukelow w11x. An incision was made along the midline of the scalp to expose the skull. Small burr holes were drilled in the skull at an appropriate stereotaxic coordinate, and the dura mater was carefully slit to insert electrodes into the brains. A cashew-coated stainless steel electrode Ž0.12 mm diameter, uncoated for 0.1 mm from the tip. was inserted at the coordinate of 2.7–2.8 mm posterior to the interaural line, 0.5 mm lateral to the midline and 4.6 mm ventral to the cortex surface. First, a lesion was produced on the right side of the brainstem by the application of 100 m A-DC for 20 s. After 4 weeks of recovery, a contralateral Žlt side. lesion was made with 100 m A–20 s DC. Sham-operated mice Žsham mice. were prepared in the same manner but without the application of DC. One week after the creation of the bilateral lesions,
92
H. Yagi et al.r Brain Research 791 (1998) 90–98
examine the distributions of electrolytic lesions along the neuroaxis. The sections were stained with hematoxylin and eosin ŽH & E. and luxol fast-blue ŽKluver-Barrera.. Immunohistochemical staining was carried out with anti-cow GFAP Žglial fibrillary acidic protein. antibody ŽDakopatts, Denmark. w26x. For microscopic morphometry of vacuoles in SAMP8 brainstem, a computerized image analyzer LUZEX3U ŽNikon, Japan. was used w15x. The sections of the brainstem stained with H & E were photographed with a light microscope and were printed in A4 size. The vacuoles in the photoprints were carefully traced and filled in with black ink on transparent paper ŽFig. 1a,b.. The number and area of the traced black regions Žs vacuoles. were measured using the LUZEX3U connected to a TV camera. 2.4. Statistics Mann–Whitney U-test, t-test and analysis of variance ŽANOVA. with repeated measures by an unweighted means were used for statistical analysis of the results.
3. Results 3.1. PassiÕe aÕoidance performance in SAMP8 and SAMR1 There was no significant difference in the latency before footshock between SAMP8 and SAMR1 except at age
Fig. 1. Example of spongy degeneration in brainstem of 4-month-old SAMP8. Ža. H&E staining. Žb. Vacuoles in the reticular formation of Ža. are traced in black ink.
open field, passive avoidance tests, active avoidance tests and shock sensitivity measurements were carried out, in that order, in the brainstem-lesioned mice Ž7 males, 4 females. and in the sham mice Ž6 males, 5 females.. Further, seven male lesioned and 3 male sham mice were used only in passive avoidance tests, and 3 male lesioned and 2 male sham mice were used only in active avoidance tests. 2.3. Pathological procedure and microscopic morphometry After completion of the behavior and passive and active avoidance tests, the mice were given an overdose of ethyl ether, and the brains were removed immediately and immersed in 10% neutral buffered formalin. The fixed brains were embedded in paraffin, and coronal sections 6 m m thick were cut every 20 m m to demonstrate the area and number of vacuoles in the brainstem, and every 50 m m to
Fig. 2. Age-related changes in one trial passive avoidance response in SAMP8 and SAMR1. Ža. Mean latency before footshock and Žb. median of latency, one day after footshock. Hollow and solid circles mean raw data in SAMR1 and SAMP8, respectively. ) p- 0.05, )) p- 0.01, ))) p- 0.001 ŽMann–Whitney U-test for SAMR1 controls..
H. Yagi et al.r Brain Research 791 (1998) 90–98
2 and 12 months ŽFig. 2a.. After footshock, most of SAMR1 controls at all ages tested reached the criterion Žretention time of 300 s., while impairment of passive avoidance performance began at 2 months of age in SAMP8 ŽFig. 2b.. The number of SAMP8 reaching the criterion and median of latency 24 h after footshock decreased with age Ž p - 0.05 at 4 months; p - 0.01, at 8 months; p 0.001, at 12 months, for SAMR1.. Thus, SAMP8 showed spontaneous age-related deterioration in retention ability of passive avoidance response. 3.2. Distribution and age dependency of spongy degeneration in SAMP8 brains Vacuolization was observed in the brainstem of SAMP8 mice from 1 month of age onward, while no vacuoles were evident in SAMR1 brains in the age range examined Žup to 12 months.. The distribution of spongy degeneration was plotted in the passive–poor SAMP8 mice at 2 and 4 months of age ŽFig. 3.. Vacuolization began to appear at the level of the hypothalamic arcuate nucleus in coronal
93
sections and was marked in the brainstem, especially in the magnocellular reticular formation ŽMGRF. Žpontine reticular nucleus, caudal part ŽPnC. and gigantocellular reticular nucleus ŽGi... The region of spongy degeneration was wider in 4-month-old than in 2-month-old SAMP8 mice. To examine the effect of age on vacuolization in detail, we counted the total number and measured quantitatively the total area of vacuoles at the levels of PnC and Gi ŽFigs. 4 and 5., where spongy degeneration was the most severe, as shown in Fig. 3. Because the distribution of vacuoles was related not only to ageing but also to the degree of impairment in retention abilities in the passive avoidance responses, the total number and area of vacuoles were plotted against age in two representative cases Žpassive–intact: latency ŽLT. s 300 s and passive–poor: LT s 100– 120 s.. The total number of vacuoles in PnC and Gi increased almost linearly with age in both, and it was larger in passive–poor than in passive–intact SAMP8 in the age range tested ŽFig. 4.. The total area of vacuoles in PnC and Gi also increased linearly with ageing, and it was larger in passive–poor SAMP8 ŽFig. 5.. Thus, the number
Fig. 3. Topographic distributions of spongy degeneration in passive–poor SAMP8 at 2 Žrt side. and 4 Žlt side. months of age. Solid circles mean a few or several vacuoles. The latency in the passive avoidance test was 54 s at 2 months and 20 s at 4 months of age in SAMP8 mice. Abbreviations: 4 V, fourth ventricle; 6, abducens nucleus; 7, facial nucleus; 10, dorsal motor nucleus of vagus; 12, hypoglossal nucleus; Arc, arcuate hypothalamic nucleus; CG, central gray; DR, dorsal raphe nucleus; Dtg, dorsal tegmental nucleus; Gi, gigantocellular reticular nucleus; Hi, hippocampus; IC, inferior colliculus; Lc, locus coeruleus; LRt, lateral reticular nucleus; mlf, medial longitudinal fasciculus; Mo5, motor trigeminal nucleus; MP, medial mammillary nucleus, posterior part; PCRt, parvocellular reticular nucleus; PGi, paragigantocellular reticular nucleus; Pn, pontine nucleus; PnC, pontine reticular nucleus, caudal part; PnO, pontine reticular nucleus, oral part; Pr5, principal sensory trigeminal nucleus; PT, pretectal area; py, pyramidal tract; R, red nucleus; SN, substantia nigra; Sol, nucleus solitary tract; Ve, vestibular nucleus; VTg, ventral tegmental nucleus; ZI zona incerta.
94
H. Yagi et al.r Brain Research 791 (1998) 90–98
Fig. 4. Age-related changes in total number of vacuoles in PnC and Gi. Hollow and solid circles mean passive–intact SAMP8 ŽLT s 300 s. and passive–poor SAMP8 ŽLT s100–120 s.. LT means latency one day after footshock.
and area of vacuoles in RF were markedly affected by ageing, and were closely related with degrees of impairment in retention ability. 3.3. Relationship of spongy degeneration to impairment of passiÕe and actiÕe aÕoidance performances To clarify in further detail the effect of vacuolization on acquisition and retention ability, the relationship between
Fig. 5. Age-related changes in total area of vacuoles in PnC and Gi. Hollow and solid circles have the same meaning as in Fig. 4.
Fig. 6. Relationship between latency in the passive avoidance response and Ža. total number and Žb. area of vacuoles in PnC and Gi in 2-month-old SAMP8. Hollow and solid circles represent PnC and Gi, respectively.
the total area or number of vacuoles and the latency Ž24 h after footshock. in the passive avoidance response was examined in PnC and Gi at 2 and 4 months of age in SAMP8 ŽFigs. 6 and 7.. The latency in passive avoidance
Fig. 7. Relationship between latency in the passive avoidance response and Ža. total number and Žb. area of vacuoles in PnC and Gi in 4-month-old SAMP8. Hollow and solid circles represent PnC and Gi, respectively.
H. Yagi et al.r Brain Research 791 (1998) 90–98
95
increase in total number and area of vacuoles. Thus, acquisition and retention ability in SAMP8 deteriorated with the enlargement of the region of spongy degeneration in the brainstem in both passive and active avoidance responses. Moreover, vacuolization in the dorsomedial part of MGRF was most severe in passiveractive poor SAMP8, as shown in Fig. 3, while it was weak in passiveractive intact SAMP8. 3.4. BehaÕior and acquisition and retention ability in MGRF-lesioned SAMR1 Fig. 10 shows an example of bilateral electrolytic lesions of the brainstem in SAMR1 mice ŽH & E and GFAP
Fig. 8. Sidman active avoidance responses in SAMP8 and SAMR1 Ž4 months old.. Hollow and solid circles mean raw data in SAMR1 and SAMP8, respectively.
behavior decreased with the increase in total area and number of vacuoles in both PnC and Gi at 2 months of age ŽFig. 6.. At 4 months of age ŽFig. 7., the latency decreased with the increase in total number and area of vacuoles in PnC and Gi as at 2 months of age. Active avoidance test result in 4-month-old SAMR1 and SAMP8 was shown in Fig. 8. The total number of shocks the mice received in a session decreased with increasing sessions in both SAMR1 and SAMP8; however, SAMP8 received markedly more shocks than SAMR1 at all sessions tested Ž F Ž7,82. s 2.58, p - 0.05.. The number of shocks were plotted against the total number and area of vacuoles in PnC and Gi of SAMP8 representatively at 8th session in active avoidance tests ŽFig. 9a,b.. The number of shocks increased with the
Fig. 9. Relationship between total number Ža. and total area Žb. of vacuoles in Gi and total number of shocks at 8th session in active avoidance performances. r means Bravais Pearson correlation.
Fig. 10. Example of electrolytic lesionings of the reticular formation of the brainstem of SAMR1. Ža. H&E staining. Žb. anti-GFAP-PAP staining, corresponding to the section in Ža.. Arrows point to lesioned area.
H. Yagi et al.r Brain Research 791 (1998) 90–98
96
Fig. 12. Active avoidance responses in MGRF-lesioned, sham and SAMP8 mice. Fourteen MGRF-lesioned, 13 sham and 8 SAMP8 mice were tested. Vertical lines represent S.E.M.
Fig. 11. Topographic distributions of electrolytic lesions in MGRF of 3–4-month-old SAMR1 mice. The dotted regions represent the lesioned areas in 3 representative MGRF-lesioned mice.
staining.. The lesioned areas were demonstrated more widely and more clearly with GFAP staining ŽFig. 10b. than with H & E staining ŽFig. 10a.. Fig. 11 shows some examples of topographic distributions of electrolytic lesions in MGRF stained with H & E and GFAP staining. Most bilateral dorsomedial parts of MGRF were destroyed bilaterally. Behavior and acquisition and retention ability were examined in the MGRF-lesioned and sham mice. There was no significant difference in behavior or passive and active avoidance tests between male and female of MGRF-lesioned or sham mice. Therefore, the total number of male and female mice tested was used for statistical analysis. The number of squares entered in the open field was almost the same for MGRF-lesioned and sham mice ŽTable 1.. No difference in shock sensitivity Žflinch, jump
and vocalization thresholds. was evident between the two groups. Thus, sensory function and activityrexploratory behavior were not affected by bilateral lesioning of the dorsomedial MGRF. In the passive avoidance test, there was no difference in the latency before shock between MGRF-lesioned and sham mice. Sham mice showed as good retention ability as the non-operated SMRI controls, but the MGRF-lesioned mice were very inferior to the sham mice in retention ability of passive avoidance response. Most of the MGRF-lesioned mice did not reach the criterion after footshock, and the median of latency was significantly shorter in the MGRF-lesioned mice than in the sham ŽTable 1.. The MGRF-lesioned mice also showed inferior ability in active avoidance behavior. Fig. 12 shows the average number of shocks received at each session in the active avoidance tests. In each of MGRF-lesioned, sham and SAMP8 mice, the number of shocks decreased as the session increased. The MGRF-lesioned mice received more shocks than the sham mice in a session of active avoidance tests Ž F Ž8,200. s 2.16, p - 0.05., though they received fewer shocks then SAMP8 mice did.
Table 1 Behavior and passive avoidance responses in MGRF-lesioned and sham mice Behavior
MGRF-lesioned
Sham
Shock sensitivity ŽmA.
Flinch Jump Vocal
0.12 " 0.03 Ž11. 0.30 " 0.09 Ž11. 0.42 " 0.08 Ž11.
0.13 " 0.03 Ž11. 0.31 " 0.10 Ž11. 0.47 " 0.09 Ž11.
Open field
Number of cross Ž3 min.
137 " 34 Ž11.
134 " 42 Ž11.
Step-through latency
Before shock Žmean. After shock Žmedian.
20 " 6 Ž13. 43 ) ) Ž13.
29 " 9 Ž10. 300 Ž10.
Each value represents the mean " S.D. Median value is used only for representing latency after shock. Number in the parenthesis is number of mice tested. )) P - 0.01 ŽMann–Whitney U-test. for sham mice.
H. Yagi et al.r Brain Research 791 (1998) 90–98
Thus, the bilateral dorsomedial MGRF-lesioned mice showed severe impairment of passive avoidance behavior and deterioration in active avoidance behavior. That is, deficits of acquisition and retention ability were caused by bilateral lesions of dorsomedial MGRF of the brainstem, without any disturbances of sensory or motor functions.
4. Discussion The present study showed the following: Ž1. Spongy degeneration was marked in the brainstem Žespecially, the magnocellular reticular formation, MGRF. of SAMP8, and it became severe with ageing. Ž2. The total area and number of vacuoles in MGRF ŽPnC and Gi. increased with age. The latency in passive avoidance behavior decreased with the increase in total area and number of vacuoles in MGRF. The degree of vacuolization in MGRF in SAMP8 was closely related to the deterioration of learning and memory ability in the passive and active avoidance performances. Ž3. Bilateral electrolytic lesioning of the dorsomedial MGRF in SAMR1 mice resulted in impairment of passive and active avoidance behavior without any disturbance of sensory function or open-field activity. The above results indicate that MGRF Žespecially, the dorsomedial part. possesses to some degrees learning and memory functions or that MGRF bears one part in learning and memory systems of brains. There are two methods of examining learning and memory functions in the brainstem: lesioning and electrophysiological studies. One lesioning study showed that learning in the uphill avoidance task was possible in rats devoid of telencephalon and thalamus w7x, suggesting that the brainstem has some degrees of learning and memory ability. Destruction of some nuclei in the brainstem, e.g., pedunculo–pontine nucleus–parabrachial nucleus w2,8x or median raphe nucleus w21x resulted in deterioration of mazelearning and of active avoidance and visual discrimination learning. Midbrain reticular formation-lesioned rats showed impairment in T-maze performance w19,20x. Acquisition of Y-maze and jump avoidance behavior was impossible for rats with bilateral symmetrical lesions of the PnC w12x. These results suggest that some nuclei and reticular formation in the brainstem play important roles in learning and memory function, and that they constitute memory networks even at the level of the brainstem and connect with higher order learning and memory networks such as the limbic system and the temporal cortex through ascending fibers. The present study, for the first time, showed that a caudal part of RF, especially dorsomedial MGRF, also plays an important role in learning and memory functions. Electrophysiological studies have shown that neurons in the RF of the brainstem have learning and memory functions. The activities of neurons or neuronal groups in the brainstem were recorded in animals performing various learning tasks w14,16,27x. Unit responses appeared in the
97
posterior pontine RF of the brainstem in rats during the learning of a classical appetite-related conditioned task w9x, and the development of discriminative neuronal discharges was observed in the pontine RF during acquisition behavior in rabbits w5x. Thus, neurons or neuronal groups in the posterior RF of the brainstem responded to tasks requiring learning and memory, suggesting that neurons in RF possess some degrees of learning and memory functions. On the other hand, RF was reported to affect higher-order learning and memory systems through ascending fibers such as the medial longitudinal fasciculus ŽMLF.. Electric stimulation of MGRF elicited hippocampal rhythmic slow activity ŽRSA or theta rhythm. through ascending fibers w22,23x. A change in frequency of RSA in the hippocampus was related to degrees of learning and memory functions w3x, and loss of theta rhythm resulted in complete deficits of learning and memory w24x. Therefore, bilateral lesions of MGRF might affect RSA in the hippocampus through ascending fibers and result in the deterioration of learning and memory. In the present lesioning study, although the MGRF-lesioned mice showed inferior performances in both passive and active avoidance behavior, as did SAMP8, the deterioration was more severe in passive avoidance than in active avoidance responses. The active avoidance tests need more complicated responses and skills for stimulation Že.g., to press a bar to avoid electric shock. and higher-order learning and memory ability than do the passive avoidance tests. Higher-order or the other learning and memory systems may play more important roles in the learning of active avoidance w19x. The electrolytic lesions in the present study were limited to the dorsomedial MGRF. Wider artificial lesions in the RF may result in more severe deterioration of active avoidance behavior. The electrolytic method used in the present study destroyed fibers of passage as well as neuron cell bodies in MGRF. In order to clarify a role of the neurons in learning and memory functions more clearly, a fiber-sparing method should be used in the future. Another neuropathological change, PGS, appeared rapidly at 6 months of age in SAMP8 w1x, while deterioration of passive and active avoidance behaviors has occurred at 2 and 4 months of age, respectively. Therefore, there seems to be no direct relationship between PGS and impairment of passive and active avoidance behavior. However, the number of PGS increased with age, especially in the hippocampus, so PGS may have a significant effect on the deterioration in the other learning and memory behavior in SAMP8. Flood et al. w4x suggested a correlation between increased PGS and deterioration in appetitive learning in SAMP8. Further studies on the effect of PGS on learning and memory ability in SAMP8 are needed. The serological test showed negative results regarding mouse hepatitis virus, Sendai virus, Mycoplasma pulmonis and Tyzzerr’s organism in both SAMP8 and SAMR1.
98
H. Yagi et al.r Brain Research 791 (1998) 90–98
Transmission study and an analysis of the genetic data will be the subjects of the forthcoming reports on the etiology of vacuolization and PGS. In this experiment, we focused on MGRF to explain the cause of deterioration in learning and memory ability of SAMP8. But, the other several learning and memory systems, involving the cerebellum, the hippocampus, the amygdala and the temporal cortex may also play important roles in learning and memory ability of SAMP8. Further study will be required to clarify roles of the other memory systems andror the relation between MGRF and the other memory systems in SAMP8. 5. Conclusion Learning and memory ability consists of a diffuse system including the brainstem, and MGRF ŽPnC and Gi. Žespecially dorsomedial part. possesses some degree of learning and memory functions, and is one site of the diffuse learning and memory networks of brains. SAMP8 can serve as a brainstem-lesioned, especially MGRF-lesioned, model manifesting memory deficiency and may provide new insights into learning and memory functions. Acknowledgements This work was supported by grants from the Ministry of Education, Culture and Science of Japan. We thank T. Matsushita and K. Kogishi for technical assistance. References w1x H. Akiyama, H. Kameyama, I. Akiguchi, H. Sugiyama, T. Kawamata, H. Fukuyama, H. Kimura, M. Matsushita, T. Takeda, Periodic acid-Schiff ŽPAS.-positive, granular structures increase in the brain of senescence-accelerated mouse ŽSAM., Acta Neuropathol. ŽBerlin. 72 Ž1986. 124–129. w2x F. Dellu, W. Mayo, J. Cherkaouri, L. Moal, H. Simon, Learning disturbances following excitotoixic lesion of cholinergic pedunculopontine nucleus in the rat, Brain Res. 544 Ž1991. 126–132. w3x Z. Elazar, W.R. Adey, Spectral analysis of low frequency components in the electrical activity of the hippocampus during learning, Electroenceph. Clin. Neurophysiol. 23 Ž1968. 225–240. w4x J.F. Flood, J.E. Morley, Early onset of age-related impairment of aversive and appetitive learning in the SAM-Pr8 mouse, J. Gerontol. 47 Ž1992. 52–59. w5x M. Gabriel, B. Gregg, A. Clancy, M. Kittrel, W. Dailey, brainstem reticular formation neuronal correlates of stimulus significance and behavior during discriminative avoidance conditioning in rabbits, Behav. Neurosci. 100 Ž1986. 171–184. w6x H. Hosokawa, R. Kasai, K. Higuchi, S. Takeshita, K. Shimizu, H. Hamamoto, A. Honma, M. Irino, K. Toda, A. Matsumura, M. Matsushita, T. Takeda, Grading score system: a method for evaluation of the degree of senescence in senescence-accelerated mouse ŽSAM., Mech. Ageing Dev. 26 Ž1984. 91–102. w7x J.P. Huston, C. Tomaz, I. Fix, Avoidance learning in rats devoid of the telencephalon plus thalamus, Behav. Brain Res. 17 Ž1985. 87–95.
w8x S.F. Ivanova, J. Bures, Acquisition of conditioned taste aversion in rats is prevented by tetrodotoxin blockade of a small midbrain region centered around the parabrachial nuclei, Physiol. Behav. 48 Ž1990. 543–549. w9x G. Kornblith, J. Olds, Unit activity in brainstem reticular formation of the rat during learning, J. Neurophysiol. 36 Ž1973. 489–501. w10x M. Miyamoto, Y. Kiyota, N. Yamazaki, A. Nagaoka, T. Matsuo, Y. Nagawa, T. Takeda, Age-related changes in learning and memory in the senescence accelerated mouse ŽSAM., Physiol. Behav. 38 Ž1986. 399–406. w11x D.G. Montemurro, R.H. Dukelow ŽEds.., A Stereotaxic Atlas of the Diencephalon and Related Structures of the Mouse. Futura Publishing, Mount Kisko, New York, 1972. w12x G. Muller, F. Klinberg, Learning and retention of active avoidance are differently impaired after dorsal and ventral lesions of the nucleus reticularis pontis caudalis of rats, Biomed. Biochim. Acta. 48 Ž1989. 817–827. w13x A. Ohta, T. Hirano, H. Yagi, S. Tanaka, M. Hosokawa, T. Takeda, Behavioral characteristics of the SAM-Pr8 strain in Sidman active avoidance task, Brain Res. 498 Ž1989. 195–198. w14x E.B. Pragay, A.F. Mirsky, C.L. Ray, D.F. Turner, C. Mirsky V, Neuronal activity in the brainstem reticular formation during performance of a ‘go–no go’ visual attention task in the monkey, Exp. Neurol. 60 Ž1978. 83–95. w15x A. Shimada, A. Ohta, I. Akiguchi, T. Takeda, Inbred SAM-Pr10 as a mouse model of spontaneous, inherited brain atrophy, J. Neuropathol. Exp. Neurol. 51 Ž1992. 440–450. w16x D.L. Sparks, R.J. Travis Jr., Patterns of reticular unit activity observed during performance of a discriminative task, Physiol. Behav. 3 Ž1968. 961–967. w17x T. Takeda, M. Hosokawa, K. Higuchi, Senescence-accelerated mouse ŽSAM.: a novel murine model of accelerated senescence, J. Am. Geriatr. Soc. 39 Ž1991. 183–194. w18x T. Takeda, M. Hosokawa, S. Takeshita, M. Irino, K. Higuchi, T. Matsushita, Y. Tomita, K. Yasuhira, H. Hamamoto, K. Shimizu, M. Ishii, T. Yamamuro, A new murine model of accelerated senescence, Mech. Ageing Dev. 17 Ž1981. 183–194. w19x R. Thompson, Localization of a ‘passive avoidance memory system’ in the white rat, Physiol. Psychol. 6 Ž1978. 263–274. w20x R. Thompson, Retention of individual spatial reversal problems in rats with nigral, caudoputamenal and reticular formation lesions, Behav. Neural Biol. 34 Ž1982. 98–103. w21x R. Thompson, A. Ramsay, J. Yu, A generalized learning deficit in albino rats with early median raphe or pontine reticular formation lesions, Physiol. Behav. 32 Ž1984. 107–114. w22x R.P. Vertes, An analysis of ascending brainstem systems involved in hippocampal synchronization and desynchronization, J. Neurophysiol. 46 Ž1981. 1140–1159. w23x R.P. Vertes, Brainstem modulation of the hippocampus: anatomy, physiology and significance, in: R.L. Isaacson, K. Pribram ŽEds.., The Hippocampus, Vol. 4, Plenum, New York, 1986 pp. 41–75. w24x J. Winson, Loss of hippocampal theta rhythm results in spatial memory deficit in the rat, Science 201 Ž1978. 160–163. w25x H. Yagi, S. Katoh, I. Akiguchi, T. Takeda, Age-related deterioration of ability of acquisition in memory and learning in senescence accelerated mouse: SAM-Pr8 as an animal model of disturbance in recent memory, Brain Res. 474 Ž1988. 86–93. w26x H. Yagi, M. Irino, T. Matsushita, S. Katoh, M. Umezawa, T. Tsuboyama, M. Hosokawa, I. Akiguchi, R. Tokunaga, T. Takeda, Spontaneous spongy degeneration of the brainstem in SAM-Pr8 mice, a newly developed memory-deficient strain, J. Neuropathol. Exp. Neurol. 48 Ž1989. 577–590. w27x N. Yoshi, P. Pruvot, H. Gastaut, Electrographic activity of the mesencephalic reticular formation during conditioning in the cat, Clin. Neurophysiol. 9 Ž1957. 595–608.