- Email: [email protected]
Hyperactivity, neuromotor defects, and impaired learning and memory in a mouse model for metachromatic leukodystrophy
Brain Research 907 (2001) 35–43 www.elsevier.com / locate / bres
Research report
Hyperactivity, neuromotor defects, and impaired learning and memory in a mouse model for metachromatic leukodystrophy Rudi D’Hooge a , Debby Van Dam a , Frieda Franck a , Volkmar Gieselmann b , a, Peter P. De Deyn * a
Laboratory of Neurochemistry and Behaviour, Born–Bunge Foundation, and Department of Neurology — Memory Clinic, University of Antwerp, Antwerp, Belgium b ¨ Physiologische Chemie, Rheinische Friedrich-Wilhelms-Universitat ¨ , Bonn, Germany Institut f ur Accepted 27 February 2001
Abstract Deficiency of arylsulfatase A (ASA) causes the autosomal recessive lipidosis, metachromatic leukodystrophy (MLD). Performance on tests of activity, motor ability and learning / memory was assessed in ASA-deficient mice and normal controls at 3, 6 and 12 months-of-age. ASA-deficient mice showed consistently increased cage activity in all age groups, whereas open field activity was increased only in the 3-month-old group. Motor coordination and equilibrium, as tested in the rotarod test, was impaired in 12-month-old ASA-deficient mice. Passive avoidance learning was tested in the step-through box. Performance on this test was impaired in the 12-month-old group only. Spatial learning and memory abilities were tested in the Morris water maze. Six-month-old ASA-deficient mice displayed slightly impaired hidden-platform acquisition performance. Three-month-old animals, on the other hand, did not show any acquisition or retention defect on this task, notwithstanding significantly reduced swimming velocity. Acquisition training, both in the hidden- and visible-platform conditions of the Morris water maze, and retention performance during the probe trials were impaired in 12-month-old ASA-deficient mice. The hyperactivity, motor incoordination and slowing, and the age-related learning / memory defects, reported here in ASA-deficient mice, may relate to the decline of neuromotor and cognitive functions in MLD patients, and could be used as correlative or outcome measures in the study of MLD pathophysiology and treatment. 2001 Elsevier Science B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Degenerative disease: other / genetic models Keywords: Metachromatic leukodystrophy; Lysosomal disorder; Transgenic mouse model; Hyperactivity; Morris water maze
1. Introduction Deficiency of the enzyme arylsulfatase A (ASA; EC 3.1.6.8) causes the autosomal recessive lipidosis, metachromatic leukodystrophy (MLD), which has an estimated frequency of 1 in 40,000 newborns [8]. ASA deficiency leads to lysosomal storage of its substrate, the sphingolipid cerebroside-3-sulfate (sulfatide), and subsequent progressive demyelination in the central nervous system [12]. The *Corresponding author. Laboratory of Neurochemistry and Behaviour, Born–Bunge Foundation, Universiteitsplein 1, B-2610 Antwerp, Belgium. Tel.: 132-3-820-2620; fax: 132-3-820-2618. E-mail address: [email protected] (P.P. De Deyn).
clinical appearance of MLD differs according to the amount of residual enzyme activity, with gene defects that do not allow any functional activity causing the most severe, late infantile form of MLD [6,16]. In all forms of MLD, however, cognitive decline and impaired learning are amongst the first precursory signs of a fatal progression [12]. We have described ASA-deficient knockout mice, presenting histopathological, electrographic and behavioural alterations, reminiscent of the human condition [9]. Neuropathological examination of these animals showed that extensive sulfatide storage occurs during their first year of life within white matter structures like corpus callosum, hippocampal fimbria, internal capsule and optic nerve [9].
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02374-5
36
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Reduced axonal diameter was demonstrated in the optic nerve and corpus callosum of 1-year-old ASA-deficient mice. In two-year-old ASA-deficient animals, sulfatide accumulation was further increased, and white matter tracts like the optic nerve and corpus callosum displayed severe astrogliosis. At this stage, sulfatide storage had become obvious in the cerebellar white matter as well, and a substantial amount of Purkinje cells was lost from the cerebellar cortex [2,9]. In addition, ASA-deficient mice showed severe neuronal damage in the inner ear [9]. Around the age of 6 months sulfatide storage was already obvious, but in mice of 8 and 11 months, the number of acoustic ganglion cells and their surrounding Schwann cells was greatly reduced. Especially early signs of deterioration were considered to be useful in the study of the progression of the disturbances observed, and as possible outcome measures for therapeutic experiments with this MLD mouse model. The neuromotor alterations in ASA-deficient animals included disturbed gait and motor incoordination, which progressed significantly towards the end of their life, and appeared to be related to selective cerebellar degeneration in these animals [2]. In some human cases, motor abnormalities only occurred in the more advanced stages of the disorder or not at all, whereas in other cases, motor symptoms occur first [18]. In late juvenile MLD patients, for example, gait disturbances are preceded by cognitive difficulties [12]. Unfortunately, we were previously unable to positively demonstrate cognitive deficits in our animal model [9]. This could have been due to the relatively small sample of ASA-deficient animals examined or to progressive neuromotor impairments eclipsing in these older animals any slight cognitive deficit we might have been able to detect otherwise. In this study, we have therefore assessed performance on tests of activity, motor ability as well as learning and memory in larger groups of ASAdeficient mice and normal controls, and studied these animals at different ages (3, 6 and 12 months).
2. Materials and methods
2.1. Animals ASA knockout 129 / SvJ mice were generated and bred as previously described [9], and backcrossed to C57BL / 6J. Animals were kept in small groups under standard animal housing conditions with 12 / 12 h dark–light schedule, rodent chow and water ad libitum, and constant room temperature and humidity. Independent groups of transgenic and control mice were tested at 3, 6 and 12 monthsof-age using the small behavioural test battery published before [1]. Behavioural experimenters were unaware of the genetic status of the animals, and genotype was reconfirmed after testing.
2.2. Neuromotor and activity assessment Cage activity was recorded in solitary housed mice during the dusk phase of their activity cycle (2 h recording) and overnight (16 h recording). Animals were put in transparent mouse cages (20325 cm 2 ), which were placed between three infrared photobeams. The number of beam crossings during the recording interval was counted with a microprocessor counter, and used to express (mainly ambulatory) cage activity. Open field activity was recorded for 10 min during the dark phase of their activity in a 50350 cm 2 arena using a video tracking device and assorted software (Chromotrack, San Diego Instruments, USA). Path length, number of entries in the 737 cm 2 corner squares, dwell in the 30 cm centre circle, and entries in this centre circle were used as open field activity measures. Motor coordination and equilibrium were tested in the rotarod test [3]. Testing consisted of five 2 min trials, with 10 min intertrial intervals, during which the animals were placed on a rotating rod (6 r.p.m., constant speed), and the mean time on the rod as well as the number of animals able to stay on the rod was compared between groups.
2.3. Cognitive performance assessment Learning and memory performance was assessed using a step-through box and a Morris-type water maze. The stepthrough box used for passive avoidance learning consisted of a small illuminated compartment and a bigger dark compartment. During training, mice were placed into the small compartment and, 5 s later, the sliding door connecting both compartments was opened. Upon entry, they received a slight electric footshock (0.3 mA, 1 s), and 24 h later they were again placed in the box for a testing trial. Step-through latency was recorded during the training and testing trials with a cut-off time of 5 min (animals not entering within 2 min during training were discarded). The Morris-type water maze consisted of a circular pool (diameter: 150 cm; height: 30 cm) filled with opacified water and maintained at 258C. A round perspex platform (15 cm diameter) was placed inside the pool at the centre of the target (North-East) quadrant, 1 cm beneath water surface. One daily trial block consisted of four swimming trials randomly starting from one of four different positions around the pool with 15 min intertrial intervals. If the animals could not find the platform within the maximum swimming time of 120 s, they were placed on the platform, and had to stay there for 15 s before being allowed to return to their home cage under an infrared lamp to avoid hypothermia. Eight acquisition trial blocks were followed by a probe trial during which mice had to swim for 100 s in the pool without a platform. During acquisition and probe trials, the animals were tracked using a computerised video-tracking system (Chromotrack, San Diego Instruments, USA). During the acquisition trials, performance
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
measures (escape latency, path length and swimming velocity) of the four daily trials were summed, and data were expressed as totals per daily trial block. For statistical analysis of differences between acquisition measures, analysis of variance was used with genotype and trial block as sources of variation. During the probe trial, performance was expressed as percentage of time spent in each of the four quadrants of the pool, and number of crossings through the area where the platform used to be.
37
Both in control and deficient animals, open field activity decreased with increasing age. Motor coordination and equilibrium were tested in the rotarod test (Fig. 1). Rotarod performance tended to decrease with increasing age both in control and deficient mice. More specifically, older mice required more practice to reach the same level of proficiency than younger ones. ASA deficiency significantly affected rotarod performance only in the 12-month-old animals. During all five trials, 12-month-old deficient mice were unable to balance on the rotating rod as long as control animals. Two-way ANOVA revealed a significant effect of genotype on mean time on the rod (F1,180 55.565; P50.019). Passive avoidance learning was tested in the stepthrough box (Fig. 2). Retention performance was represented by step-through latency (and percentage of animals not entering) during the testing trial of this test. Performance on this test clearly did not decrease with age as all control 12-month-old animals reached criterion during testing (i.e., they all waited in the first compartment for 5 min). Only in the 12-month-old group a significant difference in step-through latency was observed between ASA (1 / 1) and (2 / 2) mice. As well, a slightly but significantly lower percentage of 12-month-old animals reached criterion in the ASA (2 / 2) group (not shown in figure). Whereas all 21 control animals stayed in the first compartment during testing, four out of 18 deficient mice (22%) did enter the second compartment despite previous training (P50.037; Fisher exact test). Spatial learning and memory abilities were tested in the hidden-platform Morris water maze. The test consisted of eight daily acquisition trial blocks (Fig. 3) followed by a retention test (probe trial), during which the animals were tracked for 100 s in the pool without platform (Fig. 4). During the acquisition trials, no significant difference in escape latency was found between ASA (1 / 1) and (2 / 2) animals tested at 3 months-of-age (Fig. 3A). Also, path length was not significantly different between deficient mice and controls. Three-month-old ASA-deficient mice did swim consistently slower than ASA (1 / 1) controls, however. For example, the mean swimming velocity (6S.E.M.) on the last acquisition trial block (trial block 8)
2.4. Statistics Significance of differences between mean scores on each of these tests were assessed with two-way analysis of variance (ANOVA) for repeated measures or two-tailed Student’s t-test for comparison between pairs of means; differences between proportions were analysed with Fisher exact test.
3. Results Results of cage and open field activity recording in mice aged 3, 6 and 12 months are shown in Table 1. Cage activity was consistently and significantly higher in ASAdeficient mice than in wild-type control animals, both during dusk and overnight recordings. In all age groups, a highly significant increase in cage activity was found in ASA-deficient mice. However, the relative hyperactivity of ASA (2 / 2) mice did not seem to change with increasing age of the animals. For example, overnight recorded activity was about 1.7 times higher in 12-month-old ASA (2 / 2) mice than in (1 / 1) controls of that age, but deficient animals of 3 months showed a similar 1.6 3 increase in overnight activity. Open field recording, on the other hand, only showed slightly higher activity in the 3-month-old group. Ambulatory measures like path length and corner entries were higher in ASA-deficient mice, whereas time spent in the centre of the field (dwell centre) as well as the number of entries in the centre indicated that their exploration patterns were not significantly altered. Table 1 Cage and open-field activity measures in arylsulfatase A 1 / 1 and 2 / 2 mice 3 months
Cage activity Open field
First 2 h 16 h Path length Corner entries Dwell centre Entries centre
6 months
12 months
ASA(1 / 1)
ASA(2 / 2)
ASA(1 / 1)
ASA(2 / 2)
ASA(1 / 1)
ASA(2 / 2)
18856228 70136473 15296155 3263 2064 17.062.4
31176223*** 11 1776748*** 20766200* 4364* 3668 2364
16236143 80366892 12956156 3063 10.062.5 11.062.2
29056207*** 14 6316847*** 9336253 2165 1565 1063
1522693 76516562 7876120 19.062.8 4.061.1 5.061.3
27746241*** 13 1006841*** 8306141 2264 5.061.4 4.060.9
Cage activity is expressed as beam crossing during 2 h dusk or 16 h overnight recording. Open-field activity is recorded for 10 min in a 50350 cm 2 arena; path length is in cm, dwell in centre in s. All data are means6S.E.M. of 18–21 animals tested. Asterisks indicate significance of differences between means of 1 / 1 and 2 / 2 mice within each age group: *P,0.05, ***P,0.001.
38
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Fig. 2. Retention trial performance in control (black bars) and ASAdeficient (white bars) mice, tested at 3, 6 and 12 months-of-age in the step-through, passive avoidance test. Data are mean step-through latency (with S.E.M. error bars) during the testing trial with 300 s cut-off. The 3-month-old group consisted of 20 ASA (1 / 1) and 21 ASA (2 / 2) animals; the 6-month-old group of 20 ASA (1 / 1) and 20 ASA (2 / 2) animals; and the 12-month-old group of 21 ASA (1 / 1) and 18 ASA (2 / 2) animals. Asterisks indicate significance of differences between ASA (1 / 1) and (2 / 2) values: *P,0.05 (two-tailed Student’s t-test).
Fig. 1. Rotarod performance in control (closed symbols) and ASAdeficient (open symbols) mice, tested at 3, 6 and 12 months-of-age. Data are mean time (and S.E.M.) the animals were able to balance on the rotating rod during each of five trials (trials T0 through T40) with a maximal cut-off time of 120 s / trial. The 3-month-old group consisted of 20 ASA (1 / 1) and 21 ASA (2 / 2) animals (A); the 6-month-old group of 18 ASA (1 / 1) and 20 ASA (2 / 2) animals (B); and the 12-monthold group of 19 ASA (1 / 1) and 19 ASA (2 / 2) animals (C). See text for statistical analysis.
was 19.661.1 cm / s in ASA (1 / 1) mice and 16.160.7 cm / s in ASA (2 / 2) mice (P50.005). Throughout the acquisition period, ASA (1 / 1) mice displayed an average swimming velocity of 22.360.8 cm / s, whereas in ASA (2 / 2) mice this was only 16.860.8 cm / s. Accordingly, two-way repeated measures ANOVA showed a highly
significant effect of genotype on swimming velocity in these animals (F1,36 526.3; P,0.001). In the 6-month-old group, acquisition training did reveal a slight difference in escape latency between ASA (2 / 2) mice and controls (Fig. 3B). Escape latency was longer in deficient mice during the first acquisition trial blocks, but this difference apparently disappeared during subsequent training. Consequently, only the effect of interaction between genotype and trial block on escape latency was significant (F7,336 53.6; P,0.001). Again, swimming velocity was consistently decreased in ASA-deficient animals (effect of genotype on swimming velocity: F1,48 544.9; P,0.001) with an average swimming velocity of 20.860.6 cm / s in ASA (1 / 1) mice and 15.160.6 cm / s in ASA (2 / 2) mice. The most pronounced difference between deficient mice and controls in acquisition training was observed in animals aged 12 months. Escape latency of ASA (2 / 2) mice was consistently increased on all trial blocks (Fig. 3C). Two-way repeated measures ANOVA demonstrated a highly significant effect of genotype on escape latency (F1,39 544.5; P,0.001). The increased escape latency in ASA-deficient mice appeared to be mainly due to the marked decrease in swimming velocity in these animals. Average swimming velocity was only 10.860.5 cm / s in 12-month-old ASA (2 / 2) animals compared to 18.360.5 cm / s in controls, and accordingly, genotype had a highly significant effect on swimming velocity during acquisition training (F1,39 5104.6; P,,0.001). Fig. 4 depicts the results of the probe trials that followed the acquisition training described above. Comparing the percentage of time spent in the different quadrants of the pool between ASA (1 / 1) and (2 / 2) mice at 3 (Fig. 4A) and 6 (Fig. 4B) months-of-age, shows no difference
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
39
significantly different between ASA (1 / 1) and (2 / 2) mice. In 12-month-old ASA (2 / 2) mice, on the other hand, significantly less time was spent in the target quadrant compared to controls (Fig. 4C). Notably, a similar effect was observed in the number of entries in the target area during the probe trials (Fig. 4D). The number of entries was highly significantly reduced in ASA-deficient mice of 12 months, but not in those of 3 or 6 months. Visible-platform Morris water maze again showed the most severe deficit in the 12-month-old animals (Fig. 5). Escape latency was consistently higher in the ASA (2 / 2) animals compared to controls (Fig. 5C), but this difference became much less pronounced towards the end of the training. Statistical analysis of the data of the 12-monthold group indeed showed that genotype significantly affected escape latency (F1,36 562.9; P,0.001), but the effect of the interaction between genotype and trial was also highly significant (F7,252 59.9; P,0.001). Again, swimming velocity was significantly reduced in ASAdeficient mice of all ages studied. At 3 months-of-age, average velocity throughout training was 21.060.7 cm / s in deficient mice compared to 28.360.8 cm / s in controls (F1,29 546.2; P,0.001); at 6 months, 21.160.7 cm / s compared to 27.560.8 cm / s in controls (F1,48 538.9; P, 0.001); and at 12 months, 14.060.7 cm / s compared to 24.960.7 cm / s in controls (F1,36 5125.9; P,,0.001).
4. Discussion
Fig. 3. Hidden-platform Morris water maze acquisition training in control (closed symbols) and ASA-deficient (open symbols) mice, tested at 3, 6 and 12 months-of-age. Data are mean escape latency (and S.E.M.) during each of eight trial blocks (blocks 1 through 8). Escape latencies with a maximal cut-off time of 120 s / trial were summed for the four trials comprising one trial block. The 3-month-old group consisted of 19 ASA (1 / 1) and 19 ASA (2 / 2) animals (A); the 6-month-old group of 23 ASA (1 / 1) and 27 ASA (2 / 2) animals (B); and the 12-month-old group of 19 ASA (1 / 1) and 22 ASA (2 / 2) animals (C). Asterisks indicate differences between ASA (1 / 1) and (2 / 2) values in case of significant genotype or genotype3trial block effects in repeated-measures ANOVA: *P,0.05, **P,0.01, ***P,0.001 (post-hoc two-tailed Student’s t-test).
between genotypes in probe trial performance. In fact, both deficient and control animals spent most of their time searching the target quadrant, and this percentage was not
Leukodystrophies are a group of heterogeneous central nervous system disorders characterized by demyelination due to pathological lipid accumulation or impaired myelin formation. Mutant jimpy mice constituted the first animal model for inherited sudanophilic leukodystrophy [20]. As a model for human Pelizaeus–Merzbacher disease, these mutants show dysmyelination in the central nervous system due to a single base change in the proteolipid protein gene [14]. Dysmyelination in these mice coincides with a number of alterations in brain histology and biochemistry [13], including the recently reported abnormal expression of neurofilament proteins [7]. Their rapidly progressing condition becomes apparent around 2 weeks-of-age when the animals start to develop a characteristic tremor, ataxia and convulsions, which eventually lead to premature death around 3 to 4 weeks-of-age. The lack of effective therapy for lysosomal storage diseases with neurological complications has also motivated the development of animal models for these disorders. Progress has been relatively slow in this field. However, using the targeted gene disruption technique, induced mouse models for several human lysosomal disorders have recently become available [21,22]. For example, a transgenic mouse model for aspartylglucosaminuria mimicked the brain pathology and
40
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Fig. 4. Probe trial performance following hidden-platform acquisition training in the same control and ASA-deficient mice shown in Fig. 3. Data are mean percentage of time (and S.E.M.) spent searching for the platform in the different quadrants of the pool (see insert) in 3- (A), 6- (B) and 12-month-old (C) mice. Asterisks indicate the significance of the difference between control and deficient animals in mean percentage of time spent in the target quadrant: **P,0.01 (two-tailed Student’s t-test). In addition (D), the mean number of actual entries in the former target area (see insert in D) were compared between the two groups: ***P,0.001 (two-tailed Student’s t-test).
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Fig. 5. Visible-platform Morris water maze acquisition training in control (closed symbols) and ASA-deficient (open symbols) mice, tested at 3, 6 and 12 months-of-age. Data are mean escape latency (and S.E.M.) during each of eight trial blocks (blocks 1 through 8). Escape latencies with a maximal cut-off time of 120 s / trial were summed for the four trials comprising one trial block. The 3-month-old group consisted of 14 ASA (1 / 1) and 17 ASA (2 / 2) animals (A); the 6-month-old group of 22 ASA (1 / 1) and 28 ASA (2 / 2) animals (B); and the 12-month-old group of 21 ASA (1 / 1) and 17 ASA (2 / 2) animals (C). Asterisks indicate differences between ASA (1 / 1) and (2 / 2) values in case of significant genotype or genotype3trial block effects in repeated-measures ANOVA: *P,0.05, **P,0.01, ***P,0.001 (post-hoc two-tailed Student’s t-test).
functional defects of human patients [11]. Impaired acquisition performance in the hidden-platform Morris water maze was found to aggravate with increasing age of these
41
animals, and suggested to relate to the progressive mental impairment seen in aspartylglucosaminuria patients. The inborn lysosomal storage disease MLD is due to deficiency of ASA leading to sulfatide storage, and subsequent demyelination in the central nervous system [12]. An MLD mouse model was previously generated by targeted disruption of the ASA gene [9]. Aged ASAdeficient mice were shown to display some of the biochemical, neuropathological, electrographic and behavioural hallmarks of the human disorder [5]. In the present report, we have compared the performance of 3-, 6- and 12-month-old ASA-deficient mice on several behavioural tests to that of ASA (1 / 1) control animals. We have found that some, but not all behavioural alterations observed in these animals deteriorated with age. Notably, increased cage and open field activity appeared to be unrelated to the increasing age of the animals. Similar degrees of hyperactivity were observed in ASA-deficient mice of 3, 6 and 12 months. Ceiling effects in our cage activity measurements could explain the apparent lack of age effects on cage activity, but not the open field activity profile. As well, since our animals were extensively backcrossed to the C57BL / 6J background, the hyperactivity observed here is unlikely to be due to differences in genetic background between ASA (1 / 1) and (2 / 2) animals. In addition, we have recently confirmed this hyperactivity in 3-month-old inbred ASA (2 / 2) mice with 129 / SvJ background (unpublished observation). It was also found that 12-month-old ASA (2 / 2) animals were impaired on the rotarod test of motor coordination and equilibrium, whereas performance of 3or 6-month-old animals was indistinguishable from controls. Previously, we have reported that the neuromotor and gait impairments of ASA-deficient mice evolved into severe ataxia and tremor during their second year of life [2]. These disturbances may relate to the motor impairments observed in human MLD patients [12]. Early neuromotor defects reported in different types of MLD include general weakness, ataxia, disturbed gait, clumsiness, and postural abnormalities. The deterioration of neuromotor abilities, occurring in ASA-deficient mice at 12 months-of-age and beyond, was found to coincide with selective neurodegeneration within the cerebellum of these animals. Notably, cerebellar lesions were also observed in late infantile MLD [23]. Cognitive decline is a characteristic feature of all forms of MLD. The neuropsychological profile of early-onset MLD patients was reportedly different from that of lateonset patients [18]. Motor and verbal memory deficits are the most prominent symptoms in early-onset patients. Motor symptoms were presented first, followed by perceptual, attentional and memory difficulties. Late-onset patients, on the other hand, often displayed behaviours characteristic of frontal dementia and white matter dysfunction [18,19]. Deficits in new learning capacity often appear much later in late-onset MLD than in other
42
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
kinds of dementia, like e.g., familial Alzheimer disease [17]. Late-onset MLD also showed a relatively slower rate of decline compared to familial Alzheimer disease. Evidence for cognitive decline was found in our animal model as well. Notably, 6-month-old ASA-deficient mice displayed slightly impaired hidden-platform acquisition performance. Three-month-old animals, on the other hand, did not show any acquisition or retention defect on this task, notwithstanding significantly reduced swimming velocity. The slightly impaired acquisition in 6-month-old mice might have foreshadowed the more serious cognitive deficits that occurred at more advanced ages. Learning and memory performance was markedly impaired in 12-monthold ASA-deficient mice. Acquisition training was impaired both in the hidden- and visible-platform conditions of the Morris water maze, and retention performance during the probe trials was significantly impaired. Twelve-month-old ASA-deficient mice showed dramatically reduced swimming velocity, and this may have interfered with the acquisition of the task. However, reduced swimming velocity was observed in all age groups, whereas water maze impairments only appeared in the 6- and 12-monthold animals. Thus, the water maze deficits of ASA-deficient mice could not be explained completely by their reduction in swimming velocity. As well, although motor requirements are much more modest in passive avoidance learning, 12-month-old ASA (2 / 2) mice also showed impaired retention performance in this test. Although hyperactivity could interfere with step-through performance, alterations in cage activity were present in all age groups including those with normal passive avoidance learning, and altered open field activity was only observed in the youngest group, which did not show any passive avoidance deficit. By and large it seems that, apart from obviously impaired motor functions, 12month-old ASA-deficient mice also displayed learning and memory problems, which could not be entirely reduced to neuromotor changes. Specific neuronal damage was observed in the cerebellum of 1- and 2-year-old ASA (2 / 2) mice [2,9]. In view of the role of the cerebellum in water maze learning and open field activity [4,10,15], these cerebellar lesions could have contributed to the motor and cognitive decline described here. However, since widespread neuropathological alterations occur in ASA (2 / 2) animals, it is impossible to ascribe the behavioural changes observed here to any particular brain region. Sulfatide storage was noted in white matter throughout the central nervous system, although mostly without cellular damage or demyelination [9]. In conclusion, hyperactivity, motor incoordination and slowing, and learning / memory defects were observed in ASA-deficient mice. Although signs of slightly impaired spatial information acquisition were already obvious in 6-month-old animals, neuromotor and cognitive functions appeared to be most severely affected in the oldest animals
tested. Hyperactivity appeared to be unrelated to the age of the animals; reduced swimming velocity was present in the 3-month-old deficient animals, but aggravated with increasing age. Like in patients suffering early-onset MLD, neuromotor (and activity) changes preceded the occurrence of cognitive decline in our animal model, and might relate to the slowing of response speed and motor slowing observed in some MLD patients [19]. At least part of the cognitive decline in the ASA-deficient mice could be caused by the cerebellar lesions previously demonstrated in these animals [2,9].
Acknowledgements Financial support was received from Antwerp University, Born–Bunge Foundation, NeuroSearch Antwerp, and the Federal Fund for Scientific Research FWO-Flanders (Grants No. G.0027.97). RD is a postdoctoral fellow and DVD a research assistant of FWO-Flanders.
References [1] R. D’Hooge, F. Franck, L. Mucke, P.P. De Deyn, Age-related behavioural deficits in transgenic mice expressing the HIV-1 coat protein gp120, Eur. J. Neurosci. 11 (1999) 4398–4402. [2] R. D’Hooge, D. Hartmann, J. Manil, F. Colin, V. Gieselmann, P.P. De Deyn, Neuromotor alterations and cerebellar deficits in aged arylsulfatase A-deficient transgenic mice, Neurosci. Lett. 273 (1999) 93–96. [3] N.W. Dunham, T.S. Miya, A note on a simple apparatus for detecting neurological deficit in rats and mice, J. Am. Pharm. Assoc. 46 (1957) 208–209. [4] E. Fransen, R. D’Hooge, G. Van Camp, M. Verhoye, J. Sijbers, E. Reyniers, P. Soriano, H. Kamiguchi, R. Willemsen, S.K.E. Koekkoek, C.I. De Zeeuw, P.P. De Deyn, A.M. Van der Linden, V. Lemmon, R.F. Kooy, P.J. Willems, L1 knockout mice show dilated ventricles, vermis hypoplasia and impaired exploration patterns, Hum. Mol. Genet. 7 (1998) 999–1009. ¨ [5] V. Gieselmann, U. Matzner, B. Hess, R. Lullmann-Rauch, R. Coenen, D. Hartmann, R. D’Hooge, P. De Deyn, G. Nagels, Metachromatic leukodystrophy: molecular genetics and an animal model, J. Inherit. Metab. Dis. 21 (1998) 564–574. [6] V. Gieselmann, J. Zlotogora, A. Harris, D.A. Wenger, C.P. Morris, Molecular genetics of metachromatic leukodystrophy, Hum. Mutat. 4 (1994) 233–242. [7] T. Gotow, J.F. Leterrier, Y. Ohsawa, T. Watanabe, K. Isahara, R. Shibata, K. Ikenaka, Y. Uchiyama, Abnormal expression of neurofilament proteins in dysmyelinating axons located in the central nervous system of jimpy mutant mice, Eur. J. Neurosci. 11 (1999) 3893–3903. [8] K.H. Gustavson, B. Hagberg, The incidence and genetics of metachromatic leukodystrophy in Northern Sweden, Acta Pediatr. Scand. 60 (1971) 585–590. ¨ [9] B. Hess, P. Saftig, D. Hartmann, R. Coenen, R. Lullmann-Rauch, ¨ H.H. Gobel, M. Evers, K. von Figura, R. D’Hooge, G. Nagels, P. De Deyn, C. Peters, V. Gieselmann, Phenotype of arylsulfatase Adeficient mice: relationship to human metachromatic leukodystrophy, Proc. Natl. Acad. Sci. USA 93 (1996) 14821– 14826.
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43 [10] J.C. Houk, J.T. Buckingham, A.G. Barto, Models of the cerebellum and motor learning, Behav. Brain Sci. 19 (1996) 368–383. [11] A. Jalanko, K. Tenhunen, C.E. McKinney, M.E. LaMarca, J. Rapola, T. Autti, R. Joensuu, T. Manninen, I. Sipila, S. Ikonen, P. Riekkinen Jr., E.I. Ginns, L. Peltonen, Mice with an aspartylglucosaminuria mutation similar to humans replicate the pathophysiology in patients, Hum. Mol. Genet. 7 (1998) 265–272. [12] E.H. Kolodny, A.L. Fluharty, Metachromatic leukodystrophy and multiple sulfatase deficiency: sulfatide lipidosis, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic and Molecular Bases of Inherited Disease, McGraw-Hill, New York, 1995, pp. 2693–2739. [13] C. Meier, A. Bischoff, Dysmyelination in ‘jimpy’ mouse, J. Neuropathol. Exp. Neurol. 33 (1974) 343–353. [14] K.A. Nave, F.E. Bloom, R.J. Milner, A single nucleotide difference in the gene for myelin proteolipid protein defines the jimpy mutation in mouse, J. Neurochem. 49 (1987) 1873–1877. [15] L. Petrosini, M. Molinari, O. Delrieu, Cerebellar contribution to spatial processing: Morris water maze and T-maze, Eur. J. Neurosci. 8 (1996) 1882–1896. [16] A. Polten, A.L. Fluharty, C.B. Fluharty, J. Kappler, K. von Figura, V. Gieselmann, Molecular basis of different forms of metachromatic leukodystrophy, New Engl. J. Med. 324 (1991) 18–22. [17] A.D. Sadovnick, H. Tuokko, D.A. Applegarth, J.R. Toone, T.
[18]
[19]
[20]
[21] [22]
[23]
43
Hadjistavropoulos, B.L. Beattie, The differential diagnosis of adult onset metachromatic leukodystrophy and early onset familial Alzheimer disease in an Alzheimer clinic population, Can. J. Neurol. Sci. 20 (1993) 312–318. E.G. Shapiro, L.A. Lockman, M. Balthazor, W. Krivit, Neuropsychological outcomes of several storage diseases with and without bone marrow transplantation, J. Inherit. Metab. Dis. 18 (1995) 413–429. E.G. Shapiro, L.A. Lockman, D. Knopman, W. Krivit, Characteristics of the dementia in late-onset metachromatic leukodystrophy, Neurology 44 (1994) 662–665. R.J. Sidman, M.M. Dickie, S.H. Appel, Mutant mice (quaking and jimpy) with deficient myelination in the central nervous system, Science 144 (1964) 309–311. ˚ K. Suzuki, J.E. Mansson, Animal model of lysosomal disease: an overview, J. Inherit. Metab. Dis. 21 (1998) 540–547. K. Suzuki, M.T. Vanier, K. Suzuki, Lysosomal disorders, in: B. Popko (Ed.), Mouse Models in the Study of Genetic Neurological Disorders, Kluwer Academic / Plenum Publishers, New York, 1999, pp. 245–283. T. Yamano, S. Ohta, M. Shimada, S. Okada, T. Yotaka, T. Sugita, H. Yabuchi, Neuronal depletion of cerebellum in late infantile metachromatic leukodystrophy, Brain Dev. 2 (1980) 359–369.
Research report
Hyperactivity, neuromotor defects, and impaired learning and memory in a mouse model for metachromatic leukodystrophy Rudi D’Hooge a , Debby Van Dam a , Frieda Franck a , Volkmar Gieselmann b , a, Peter P. De Deyn * a
Laboratory of Neurochemistry and Behaviour, Born–Bunge Foundation, and Department of Neurology — Memory Clinic, University of Antwerp, Antwerp, Belgium b ¨ Physiologische Chemie, Rheinische Friedrich-Wilhelms-Universitat ¨ , Bonn, Germany Institut f ur Accepted 27 February 2001
Abstract Deficiency of arylsulfatase A (ASA) causes the autosomal recessive lipidosis, metachromatic leukodystrophy (MLD). Performance on tests of activity, motor ability and learning / memory was assessed in ASA-deficient mice and normal controls at 3, 6 and 12 months-of-age. ASA-deficient mice showed consistently increased cage activity in all age groups, whereas open field activity was increased only in the 3-month-old group. Motor coordination and equilibrium, as tested in the rotarod test, was impaired in 12-month-old ASA-deficient mice. Passive avoidance learning was tested in the step-through box. Performance on this test was impaired in the 12-month-old group only. Spatial learning and memory abilities were tested in the Morris water maze. Six-month-old ASA-deficient mice displayed slightly impaired hidden-platform acquisition performance. Three-month-old animals, on the other hand, did not show any acquisition or retention defect on this task, notwithstanding significantly reduced swimming velocity. Acquisition training, both in the hidden- and visible-platform conditions of the Morris water maze, and retention performance during the probe trials were impaired in 12-month-old ASA-deficient mice. The hyperactivity, motor incoordination and slowing, and the age-related learning / memory defects, reported here in ASA-deficient mice, may relate to the decline of neuromotor and cognitive functions in MLD patients, and could be used as correlative or outcome measures in the study of MLD pathophysiology and treatment. 2001 Elsevier Science B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Degenerative disease: other / genetic models Keywords: Metachromatic leukodystrophy; Lysosomal disorder; Transgenic mouse model; Hyperactivity; Morris water maze
1. Introduction Deficiency of the enzyme arylsulfatase A (ASA; EC 3.1.6.8) causes the autosomal recessive lipidosis, metachromatic leukodystrophy (MLD), which has an estimated frequency of 1 in 40,000 newborns [8]. ASA deficiency leads to lysosomal storage of its substrate, the sphingolipid cerebroside-3-sulfate (sulfatide), and subsequent progressive demyelination in the central nervous system [12]. The *Corresponding author. Laboratory of Neurochemistry and Behaviour, Born–Bunge Foundation, Universiteitsplein 1, B-2610 Antwerp, Belgium. Tel.: 132-3-820-2620; fax: 132-3-820-2618. E-mail address: [email protected] (P.P. De Deyn).
clinical appearance of MLD differs according to the amount of residual enzyme activity, with gene defects that do not allow any functional activity causing the most severe, late infantile form of MLD [6,16]. In all forms of MLD, however, cognitive decline and impaired learning are amongst the first precursory signs of a fatal progression [12]. We have described ASA-deficient knockout mice, presenting histopathological, electrographic and behavioural alterations, reminiscent of the human condition [9]. Neuropathological examination of these animals showed that extensive sulfatide storage occurs during their first year of life within white matter structures like corpus callosum, hippocampal fimbria, internal capsule and optic nerve [9].
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02374-5
36
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Reduced axonal diameter was demonstrated in the optic nerve and corpus callosum of 1-year-old ASA-deficient mice. In two-year-old ASA-deficient animals, sulfatide accumulation was further increased, and white matter tracts like the optic nerve and corpus callosum displayed severe astrogliosis. At this stage, sulfatide storage had become obvious in the cerebellar white matter as well, and a substantial amount of Purkinje cells was lost from the cerebellar cortex [2,9]. In addition, ASA-deficient mice showed severe neuronal damage in the inner ear [9]. Around the age of 6 months sulfatide storage was already obvious, but in mice of 8 and 11 months, the number of acoustic ganglion cells and their surrounding Schwann cells was greatly reduced. Especially early signs of deterioration were considered to be useful in the study of the progression of the disturbances observed, and as possible outcome measures for therapeutic experiments with this MLD mouse model. The neuromotor alterations in ASA-deficient animals included disturbed gait and motor incoordination, which progressed significantly towards the end of their life, and appeared to be related to selective cerebellar degeneration in these animals [2]. In some human cases, motor abnormalities only occurred in the more advanced stages of the disorder or not at all, whereas in other cases, motor symptoms occur first [18]. In late juvenile MLD patients, for example, gait disturbances are preceded by cognitive difficulties [12]. Unfortunately, we were previously unable to positively demonstrate cognitive deficits in our animal model [9]. This could have been due to the relatively small sample of ASA-deficient animals examined or to progressive neuromotor impairments eclipsing in these older animals any slight cognitive deficit we might have been able to detect otherwise. In this study, we have therefore assessed performance on tests of activity, motor ability as well as learning and memory in larger groups of ASAdeficient mice and normal controls, and studied these animals at different ages (3, 6 and 12 months).
2. Materials and methods
2.1. Animals ASA knockout 129 / SvJ mice were generated and bred as previously described [9], and backcrossed to C57BL / 6J. Animals were kept in small groups under standard animal housing conditions with 12 / 12 h dark–light schedule, rodent chow and water ad libitum, and constant room temperature and humidity. Independent groups of transgenic and control mice were tested at 3, 6 and 12 monthsof-age using the small behavioural test battery published before [1]. Behavioural experimenters were unaware of the genetic status of the animals, and genotype was reconfirmed after testing.
2.2. Neuromotor and activity assessment Cage activity was recorded in solitary housed mice during the dusk phase of their activity cycle (2 h recording) and overnight (16 h recording). Animals were put in transparent mouse cages (20325 cm 2 ), which were placed between three infrared photobeams. The number of beam crossings during the recording interval was counted with a microprocessor counter, and used to express (mainly ambulatory) cage activity. Open field activity was recorded for 10 min during the dark phase of their activity in a 50350 cm 2 arena using a video tracking device and assorted software (Chromotrack, San Diego Instruments, USA). Path length, number of entries in the 737 cm 2 corner squares, dwell in the 30 cm centre circle, and entries in this centre circle were used as open field activity measures. Motor coordination and equilibrium were tested in the rotarod test [3]. Testing consisted of five 2 min trials, with 10 min intertrial intervals, during which the animals were placed on a rotating rod (6 r.p.m., constant speed), and the mean time on the rod as well as the number of animals able to stay on the rod was compared between groups.
2.3. Cognitive performance assessment Learning and memory performance was assessed using a step-through box and a Morris-type water maze. The stepthrough box used for passive avoidance learning consisted of a small illuminated compartment and a bigger dark compartment. During training, mice were placed into the small compartment and, 5 s later, the sliding door connecting both compartments was opened. Upon entry, they received a slight electric footshock (0.3 mA, 1 s), and 24 h later they were again placed in the box for a testing trial. Step-through latency was recorded during the training and testing trials with a cut-off time of 5 min (animals not entering within 2 min during training were discarded). The Morris-type water maze consisted of a circular pool (diameter: 150 cm; height: 30 cm) filled with opacified water and maintained at 258C. A round perspex platform (15 cm diameter) was placed inside the pool at the centre of the target (North-East) quadrant, 1 cm beneath water surface. One daily trial block consisted of four swimming trials randomly starting from one of four different positions around the pool with 15 min intertrial intervals. If the animals could not find the platform within the maximum swimming time of 120 s, they were placed on the platform, and had to stay there for 15 s before being allowed to return to their home cage under an infrared lamp to avoid hypothermia. Eight acquisition trial blocks were followed by a probe trial during which mice had to swim for 100 s in the pool without a platform. During acquisition and probe trials, the animals were tracked using a computerised video-tracking system (Chromotrack, San Diego Instruments, USA). During the acquisition trials, performance
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
measures (escape latency, path length and swimming velocity) of the four daily trials were summed, and data were expressed as totals per daily trial block. For statistical analysis of differences between acquisition measures, analysis of variance was used with genotype and trial block as sources of variation. During the probe trial, performance was expressed as percentage of time spent in each of the four quadrants of the pool, and number of crossings through the area where the platform used to be.
37
Both in control and deficient animals, open field activity decreased with increasing age. Motor coordination and equilibrium were tested in the rotarod test (Fig. 1). Rotarod performance tended to decrease with increasing age both in control and deficient mice. More specifically, older mice required more practice to reach the same level of proficiency than younger ones. ASA deficiency significantly affected rotarod performance only in the 12-month-old animals. During all five trials, 12-month-old deficient mice were unable to balance on the rotating rod as long as control animals. Two-way ANOVA revealed a significant effect of genotype on mean time on the rod (F1,180 55.565; P50.019). Passive avoidance learning was tested in the stepthrough box (Fig. 2). Retention performance was represented by step-through latency (and percentage of animals not entering) during the testing trial of this test. Performance on this test clearly did not decrease with age as all control 12-month-old animals reached criterion during testing (i.e., they all waited in the first compartment for 5 min). Only in the 12-month-old group a significant difference in step-through latency was observed between ASA (1 / 1) and (2 / 2) mice. As well, a slightly but significantly lower percentage of 12-month-old animals reached criterion in the ASA (2 / 2) group (not shown in figure). Whereas all 21 control animals stayed in the first compartment during testing, four out of 18 deficient mice (22%) did enter the second compartment despite previous training (P50.037; Fisher exact test). Spatial learning and memory abilities were tested in the hidden-platform Morris water maze. The test consisted of eight daily acquisition trial blocks (Fig. 3) followed by a retention test (probe trial), during which the animals were tracked for 100 s in the pool without platform (Fig. 4). During the acquisition trials, no significant difference in escape latency was found between ASA (1 / 1) and (2 / 2) animals tested at 3 months-of-age (Fig. 3A). Also, path length was not significantly different between deficient mice and controls. Three-month-old ASA-deficient mice did swim consistently slower than ASA (1 / 1) controls, however. For example, the mean swimming velocity (6S.E.M.) on the last acquisition trial block (trial block 8)
2.4. Statistics Significance of differences between mean scores on each of these tests were assessed with two-way analysis of variance (ANOVA) for repeated measures or two-tailed Student’s t-test for comparison between pairs of means; differences between proportions were analysed with Fisher exact test.
3. Results Results of cage and open field activity recording in mice aged 3, 6 and 12 months are shown in Table 1. Cage activity was consistently and significantly higher in ASAdeficient mice than in wild-type control animals, both during dusk and overnight recordings. In all age groups, a highly significant increase in cage activity was found in ASA-deficient mice. However, the relative hyperactivity of ASA (2 / 2) mice did not seem to change with increasing age of the animals. For example, overnight recorded activity was about 1.7 times higher in 12-month-old ASA (2 / 2) mice than in (1 / 1) controls of that age, but deficient animals of 3 months showed a similar 1.6 3 increase in overnight activity. Open field recording, on the other hand, only showed slightly higher activity in the 3-month-old group. Ambulatory measures like path length and corner entries were higher in ASA-deficient mice, whereas time spent in the centre of the field (dwell centre) as well as the number of entries in the centre indicated that their exploration patterns were not significantly altered. Table 1 Cage and open-field activity measures in arylsulfatase A 1 / 1 and 2 / 2 mice 3 months
Cage activity Open field
First 2 h 16 h Path length Corner entries Dwell centre Entries centre
6 months
12 months
ASA(1 / 1)
ASA(2 / 2)
ASA(1 / 1)
ASA(2 / 2)
ASA(1 / 1)
ASA(2 / 2)
18856228 70136473 15296155 3263 2064 17.062.4
31176223*** 11 1776748*** 20766200* 4364* 3668 2364
16236143 80366892 12956156 3063 10.062.5 11.062.2
29056207*** 14 6316847*** 9336253 2165 1565 1063
1522693 76516562 7876120 19.062.8 4.061.1 5.061.3
27746241*** 13 1006841*** 8306141 2264 5.061.4 4.060.9
Cage activity is expressed as beam crossing during 2 h dusk or 16 h overnight recording. Open-field activity is recorded for 10 min in a 50350 cm 2 arena; path length is in cm, dwell in centre in s. All data are means6S.E.M. of 18–21 animals tested. Asterisks indicate significance of differences between means of 1 / 1 and 2 / 2 mice within each age group: *P,0.05, ***P,0.001.
38
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Fig. 2. Retention trial performance in control (black bars) and ASAdeficient (white bars) mice, tested at 3, 6 and 12 months-of-age in the step-through, passive avoidance test. Data are mean step-through latency (with S.E.M. error bars) during the testing trial with 300 s cut-off. The 3-month-old group consisted of 20 ASA (1 / 1) and 21 ASA (2 / 2) animals; the 6-month-old group of 20 ASA (1 / 1) and 20 ASA (2 / 2) animals; and the 12-month-old group of 21 ASA (1 / 1) and 18 ASA (2 / 2) animals. Asterisks indicate significance of differences between ASA (1 / 1) and (2 / 2) values: *P,0.05 (two-tailed Student’s t-test).
Fig. 1. Rotarod performance in control (closed symbols) and ASAdeficient (open symbols) mice, tested at 3, 6 and 12 months-of-age. Data are mean time (and S.E.M.) the animals were able to balance on the rotating rod during each of five trials (trials T0 through T40) with a maximal cut-off time of 120 s / trial. The 3-month-old group consisted of 20 ASA (1 / 1) and 21 ASA (2 / 2) animals (A); the 6-month-old group of 18 ASA (1 / 1) and 20 ASA (2 / 2) animals (B); and the 12-monthold group of 19 ASA (1 / 1) and 19 ASA (2 / 2) animals (C). See text for statistical analysis.
was 19.661.1 cm / s in ASA (1 / 1) mice and 16.160.7 cm / s in ASA (2 / 2) mice (P50.005). Throughout the acquisition period, ASA (1 / 1) mice displayed an average swimming velocity of 22.360.8 cm / s, whereas in ASA (2 / 2) mice this was only 16.860.8 cm / s. Accordingly, two-way repeated measures ANOVA showed a highly
significant effect of genotype on swimming velocity in these animals (F1,36 526.3; P,0.001). In the 6-month-old group, acquisition training did reveal a slight difference in escape latency between ASA (2 / 2) mice and controls (Fig. 3B). Escape latency was longer in deficient mice during the first acquisition trial blocks, but this difference apparently disappeared during subsequent training. Consequently, only the effect of interaction between genotype and trial block on escape latency was significant (F7,336 53.6; P,0.001). Again, swimming velocity was consistently decreased in ASA-deficient animals (effect of genotype on swimming velocity: F1,48 544.9; P,0.001) with an average swimming velocity of 20.860.6 cm / s in ASA (1 / 1) mice and 15.160.6 cm / s in ASA (2 / 2) mice. The most pronounced difference between deficient mice and controls in acquisition training was observed in animals aged 12 months. Escape latency of ASA (2 / 2) mice was consistently increased on all trial blocks (Fig. 3C). Two-way repeated measures ANOVA demonstrated a highly significant effect of genotype on escape latency (F1,39 544.5; P,0.001). The increased escape latency in ASA-deficient mice appeared to be mainly due to the marked decrease in swimming velocity in these animals. Average swimming velocity was only 10.860.5 cm / s in 12-month-old ASA (2 / 2) animals compared to 18.360.5 cm / s in controls, and accordingly, genotype had a highly significant effect on swimming velocity during acquisition training (F1,39 5104.6; P,,0.001). Fig. 4 depicts the results of the probe trials that followed the acquisition training described above. Comparing the percentage of time spent in the different quadrants of the pool between ASA (1 / 1) and (2 / 2) mice at 3 (Fig. 4A) and 6 (Fig. 4B) months-of-age, shows no difference
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
39
significantly different between ASA (1 / 1) and (2 / 2) mice. In 12-month-old ASA (2 / 2) mice, on the other hand, significantly less time was spent in the target quadrant compared to controls (Fig. 4C). Notably, a similar effect was observed in the number of entries in the target area during the probe trials (Fig. 4D). The number of entries was highly significantly reduced in ASA-deficient mice of 12 months, but not in those of 3 or 6 months. Visible-platform Morris water maze again showed the most severe deficit in the 12-month-old animals (Fig. 5). Escape latency was consistently higher in the ASA (2 / 2) animals compared to controls (Fig. 5C), but this difference became much less pronounced towards the end of the training. Statistical analysis of the data of the 12-monthold group indeed showed that genotype significantly affected escape latency (F1,36 562.9; P,0.001), but the effect of the interaction between genotype and trial was also highly significant (F7,252 59.9; P,0.001). Again, swimming velocity was significantly reduced in ASAdeficient mice of all ages studied. At 3 months-of-age, average velocity throughout training was 21.060.7 cm / s in deficient mice compared to 28.360.8 cm / s in controls (F1,29 546.2; P,0.001); at 6 months, 21.160.7 cm / s compared to 27.560.8 cm / s in controls (F1,48 538.9; P, 0.001); and at 12 months, 14.060.7 cm / s compared to 24.960.7 cm / s in controls (F1,36 5125.9; P,,0.001).
4. Discussion
Fig. 3. Hidden-platform Morris water maze acquisition training in control (closed symbols) and ASA-deficient (open symbols) mice, tested at 3, 6 and 12 months-of-age. Data are mean escape latency (and S.E.M.) during each of eight trial blocks (blocks 1 through 8). Escape latencies with a maximal cut-off time of 120 s / trial were summed for the four trials comprising one trial block. The 3-month-old group consisted of 19 ASA (1 / 1) and 19 ASA (2 / 2) animals (A); the 6-month-old group of 23 ASA (1 / 1) and 27 ASA (2 / 2) animals (B); and the 12-month-old group of 19 ASA (1 / 1) and 22 ASA (2 / 2) animals (C). Asterisks indicate differences between ASA (1 / 1) and (2 / 2) values in case of significant genotype or genotype3trial block effects in repeated-measures ANOVA: *P,0.05, **P,0.01, ***P,0.001 (post-hoc two-tailed Student’s t-test).
between genotypes in probe trial performance. In fact, both deficient and control animals spent most of their time searching the target quadrant, and this percentage was not
Leukodystrophies are a group of heterogeneous central nervous system disorders characterized by demyelination due to pathological lipid accumulation or impaired myelin formation. Mutant jimpy mice constituted the first animal model for inherited sudanophilic leukodystrophy [20]. As a model for human Pelizaeus–Merzbacher disease, these mutants show dysmyelination in the central nervous system due to a single base change in the proteolipid protein gene [14]. Dysmyelination in these mice coincides with a number of alterations in brain histology and biochemistry [13], including the recently reported abnormal expression of neurofilament proteins [7]. Their rapidly progressing condition becomes apparent around 2 weeks-of-age when the animals start to develop a characteristic tremor, ataxia and convulsions, which eventually lead to premature death around 3 to 4 weeks-of-age. The lack of effective therapy for lysosomal storage diseases with neurological complications has also motivated the development of animal models for these disorders. Progress has been relatively slow in this field. However, using the targeted gene disruption technique, induced mouse models for several human lysosomal disorders have recently become available [21,22]. For example, a transgenic mouse model for aspartylglucosaminuria mimicked the brain pathology and
40
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Fig. 4. Probe trial performance following hidden-platform acquisition training in the same control and ASA-deficient mice shown in Fig. 3. Data are mean percentage of time (and S.E.M.) spent searching for the platform in the different quadrants of the pool (see insert) in 3- (A), 6- (B) and 12-month-old (C) mice. Asterisks indicate the significance of the difference between control and deficient animals in mean percentage of time spent in the target quadrant: **P,0.01 (two-tailed Student’s t-test). In addition (D), the mean number of actual entries in the former target area (see insert in D) were compared between the two groups: ***P,0.001 (two-tailed Student’s t-test).
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
Fig. 5. Visible-platform Morris water maze acquisition training in control (closed symbols) and ASA-deficient (open symbols) mice, tested at 3, 6 and 12 months-of-age. Data are mean escape latency (and S.E.M.) during each of eight trial blocks (blocks 1 through 8). Escape latencies with a maximal cut-off time of 120 s / trial were summed for the four trials comprising one trial block. The 3-month-old group consisted of 14 ASA (1 / 1) and 17 ASA (2 / 2) animals (A); the 6-month-old group of 22 ASA (1 / 1) and 28 ASA (2 / 2) animals (B); and the 12-month-old group of 21 ASA (1 / 1) and 17 ASA (2 / 2) animals (C). Asterisks indicate differences between ASA (1 / 1) and (2 / 2) values in case of significant genotype or genotype3trial block effects in repeated-measures ANOVA: *P,0.05, **P,0.01, ***P,0.001 (post-hoc two-tailed Student’s t-test).
functional defects of human patients [11]. Impaired acquisition performance in the hidden-platform Morris water maze was found to aggravate with increasing age of these
41
animals, and suggested to relate to the progressive mental impairment seen in aspartylglucosaminuria patients. The inborn lysosomal storage disease MLD is due to deficiency of ASA leading to sulfatide storage, and subsequent demyelination in the central nervous system [12]. An MLD mouse model was previously generated by targeted disruption of the ASA gene [9]. Aged ASAdeficient mice were shown to display some of the biochemical, neuropathological, electrographic and behavioural hallmarks of the human disorder [5]. In the present report, we have compared the performance of 3-, 6- and 12-month-old ASA-deficient mice on several behavioural tests to that of ASA (1 / 1) control animals. We have found that some, but not all behavioural alterations observed in these animals deteriorated with age. Notably, increased cage and open field activity appeared to be unrelated to the increasing age of the animals. Similar degrees of hyperactivity were observed in ASA-deficient mice of 3, 6 and 12 months. Ceiling effects in our cage activity measurements could explain the apparent lack of age effects on cage activity, but not the open field activity profile. As well, since our animals were extensively backcrossed to the C57BL / 6J background, the hyperactivity observed here is unlikely to be due to differences in genetic background between ASA (1 / 1) and (2 / 2) animals. In addition, we have recently confirmed this hyperactivity in 3-month-old inbred ASA (2 / 2) mice with 129 / SvJ background (unpublished observation). It was also found that 12-month-old ASA (2 / 2) animals were impaired on the rotarod test of motor coordination and equilibrium, whereas performance of 3or 6-month-old animals was indistinguishable from controls. Previously, we have reported that the neuromotor and gait impairments of ASA-deficient mice evolved into severe ataxia and tremor during their second year of life [2]. These disturbances may relate to the motor impairments observed in human MLD patients [12]. Early neuromotor defects reported in different types of MLD include general weakness, ataxia, disturbed gait, clumsiness, and postural abnormalities. The deterioration of neuromotor abilities, occurring in ASA-deficient mice at 12 months-of-age and beyond, was found to coincide with selective neurodegeneration within the cerebellum of these animals. Notably, cerebellar lesions were also observed in late infantile MLD [23]. Cognitive decline is a characteristic feature of all forms of MLD. The neuropsychological profile of early-onset MLD patients was reportedly different from that of lateonset patients [18]. Motor and verbal memory deficits are the most prominent symptoms in early-onset patients. Motor symptoms were presented first, followed by perceptual, attentional and memory difficulties. Late-onset patients, on the other hand, often displayed behaviours characteristic of frontal dementia and white matter dysfunction [18,19]. Deficits in new learning capacity often appear much later in late-onset MLD than in other
42
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43
kinds of dementia, like e.g., familial Alzheimer disease [17]. Late-onset MLD also showed a relatively slower rate of decline compared to familial Alzheimer disease. Evidence for cognitive decline was found in our animal model as well. Notably, 6-month-old ASA-deficient mice displayed slightly impaired hidden-platform acquisition performance. Three-month-old animals, on the other hand, did not show any acquisition or retention defect on this task, notwithstanding significantly reduced swimming velocity. The slightly impaired acquisition in 6-month-old mice might have foreshadowed the more serious cognitive deficits that occurred at more advanced ages. Learning and memory performance was markedly impaired in 12-monthold ASA-deficient mice. Acquisition training was impaired both in the hidden- and visible-platform conditions of the Morris water maze, and retention performance during the probe trials was significantly impaired. Twelve-month-old ASA-deficient mice showed dramatically reduced swimming velocity, and this may have interfered with the acquisition of the task. However, reduced swimming velocity was observed in all age groups, whereas water maze impairments only appeared in the 6- and 12-monthold animals. Thus, the water maze deficits of ASA-deficient mice could not be explained completely by their reduction in swimming velocity. As well, although motor requirements are much more modest in passive avoidance learning, 12-month-old ASA (2 / 2) mice also showed impaired retention performance in this test. Although hyperactivity could interfere with step-through performance, alterations in cage activity were present in all age groups including those with normal passive avoidance learning, and altered open field activity was only observed in the youngest group, which did not show any passive avoidance deficit. By and large it seems that, apart from obviously impaired motor functions, 12month-old ASA-deficient mice also displayed learning and memory problems, which could not be entirely reduced to neuromotor changes. Specific neuronal damage was observed in the cerebellum of 1- and 2-year-old ASA (2 / 2) mice [2,9]. In view of the role of the cerebellum in water maze learning and open field activity [4,10,15], these cerebellar lesions could have contributed to the motor and cognitive decline described here. However, since widespread neuropathological alterations occur in ASA (2 / 2) animals, it is impossible to ascribe the behavioural changes observed here to any particular brain region. Sulfatide storage was noted in white matter throughout the central nervous system, although mostly without cellular damage or demyelination [9]. In conclusion, hyperactivity, motor incoordination and slowing, and learning / memory defects were observed in ASA-deficient mice. Although signs of slightly impaired spatial information acquisition were already obvious in 6-month-old animals, neuromotor and cognitive functions appeared to be most severely affected in the oldest animals
tested. Hyperactivity appeared to be unrelated to the age of the animals; reduced swimming velocity was present in the 3-month-old deficient animals, but aggravated with increasing age. Like in patients suffering early-onset MLD, neuromotor (and activity) changes preceded the occurrence of cognitive decline in our animal model, and might relate to the slowing of response speed and motor slowing observed in some MLD patients [19]. At least part of the cognitive decline in the ASA-deficient mice could be caused by the cerebellar lesions previously demonstrated in these animals [2,9].
Acknowledgements Financial support was received from Antwerp University, Born–Bunge Foundation, NeuroSearch Antwerp, and the Federal Fund for Scientific Research FWO-Flanders (Grants No. G.0027.97). RD is a postdoctoral fellow and DVD a research assistant of FWO-Flanders.
References [1] R. D’Hooge, F. Franck, L. Mucke, P.P. De Deyn, Age-related behavioural deficits in transgenic mice expressing the HIV-1 coat protein gp120, Eur. J. Neurosci. 11 (1999) 4398–4402. [2] R. D’Hooge, D. Hartmann, J. Manil, F. Colin, V. Gieselmann, P.P. De Deyn, Neuromotor alterations and cerebellar deficits in aged arylsulfatase A-deficient transgenic mice, Neurosci. Lett. 273 (1999) 93–96. [3] N.W. Dunham, T.S. Miya, A note on a simple apparatus for detecting neurological deficit in rats and mice, J. Am. Pharm. Assoc. 46 (1957) 208–209. [4] E. Fransen, R. D’Hooge, G. Van Camp, M. Verhoye, J. Sijbers, E. Reyniers, P. Soriano, H. Kamiguchi, R. Willemsen, S.K.E. Koekkoek, C.I. De Zeeuw, P.P. De Deyn, A.M. Van der Linden, V. Lemmon, R.F. Kooy, P.J. Willems, L1 knockout mice show dilated ventricles, vermis hypoplasia and impaired exploration patterns, Hum. Mol. Genet. 7 (1998) 999–1009. ¨ [5] V. Gieselmann, U. Matzner, B. Hess, R. Lullmann-Rauch, R. Coenen, D. Hartmann, R. D’Hooge, P. De Deyn, G. Nagels, Metachromatic leukodystrophy: molecular genetics and an animal model, J. Inherit. Metab. Dis. 21 (1998) 564–574. [6] V. Gieselmann, J. Zlotogora, A. Harris, D.A. Wenger, C.P. Morris, Molecular genetics of metachromatic leukodystrophy, Hum. Mutat. 4 (1994) 233–242. [7] T. Gotow, J.F. Leterrier, Y. Ohsawa, T. Watanabe, K. Isahara, R. Shibata, K. Ikenaka, Y. Uchiyama, Abnormal expression of neurofilament proteins in dysmyelinating axons located in the central nervous system of jimpy mutant mice, Eur. J. Neurosci. 11 (1999) 3893–3903. [8] K.H. Gustavson, B. Hagberg, The incidence and genetics of metachromatic leukodystrophy in Northern Sweden, Acta Pediatr. Scand. 60 (1971) 585–590. ¨ [9] B. Hess, P. Saftig, D. Hartmann, R. Coenen, R. Lullmann-Rauch, ¨ H.H. Gobel, M. Evers, K. von Figura, R. D’Hooge, G. Nagels, P. De Deyn, C. Peters, V. Gieselmann, Phenotype of arylsulfatase Adeficient mice: relationship to human metachromatic leukodystrophy, Proc. Natl. Acad. Sci. USA 93 (1996) 14821– 14826.
R. D’ Hooge et al. / Brain Research 907 (2001) 35 – 43 [10] J.C. Houk, J.T. Buckingham, A.G. Barto, Models of the cerebellum and motor learning, Behav. Brain Sci. 19 (1996) 368–383. [11] A. Jalanko, K. Tenhunen, C.E. McKinney, M.E. LaMarca, J. Rapola, T. Autti, R. Joensuu, T. Manninen, I. Sipila, S. Ikonen, P. Riekkinen Jr., E.I. Ginns, L. Peltonen, Mice with an aspartylglucosaminuria mutation similar to humans replicate the pathophysiology in patients, Hum. Mol. Genet. 7 (1998) 265–272. [12] E.H. Kolodny, A.L. Fluharty, Metachromatic leukodystrophy and multiple sulfatase deficiency: sulfatide lipidosis, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic and Molecular Bases of Inherited Disease, McGraw-Hill, New York, 1995, pp. 2693–2739. [13] C. Meier, A. Bischoff, Dysmyelination in ‘jimpy’ mouse, J. Neuropathol. Exp. Neurol. 33 (1974) 343–353. [14] K.A. Nave, F.E. Bloom, R.J. Milner, A single nucleotide difference in the gene for myelin proteolipid protein defines the jimpy mutation in mouse, J. Neurochem. 49 (1987) 1873–1877. [15] L. Petrosini, M. Molinari, O. Delrieu, Cerebellar contribution to spatial processing: Morris water maze and T-maze, Eur. J. Neurosci. 8 (1996) 1882–1896. [16] A. Polten, A.L. Fluharty, C.B. Fluharty, J. Kappler, K. von Figura, V. Gieselmann, Molecular basis of different forms of metachromatic leukodystrophy, New Engl. J. Med. 324 (1991) 18–22. [17] A.D. Sadovnick, H. Tuokko, D.A. Applegarth, J.R. Toone, T.
[18]
[19]
[20]
[21] [22]
[23]
43
Hadjistavropoulos, B.L. Beattie, The differential diagnosis of adult onset metachromatic leukodystrophy and early onset familial Alzheimer disease in an Alzheimer clinic population, Can. J. Neurol. Sci. 20 (1993) 312–318. E.G. Shapiro, L.A. Lockman, M. Balthazor, W. Krivit, Neuropsychological outcomes of several storage diseases with and without bone marrow transplantation, J. Inherit. Metab. Dis. 18 (1995) 413–429. E.G. Shapiro, L.A. Lockman, D. Knopman, W. Krivit, Characteristics of the dementia in late-onset metachromatic leukodystrophy, Neurology 44 (1994) 662–665. R.J. Sidman, M.M. Dickie, S.H. Appel, Mutant mice (quaking and jimpy) with deficient myelination in the central nervous system, Science 144 (1964) 309–311. ˚ K. Suzuki, J.E. Mansson, Animal model of lysosomal disease: an overview, J. Inherit. Metab. Dis. 21 (1998) 540–547. K. Suzuki, M.T. Vanier, K. Suzuki, Lysosomal disorders, in: B. Popko (Ed.), Mouse Models in the Study of Genetic Neurological Disorders, Kluwer Academic / Plenum Publishers, New York, 1999, pp. 245–283. T. Yamano, S. Ohta, M. Shimada, S. Okada, T. Yotaka, T. Sugita, H. Yabuchi, Neuronal depletion of cerebellum in late infantile metachromatic leukodystrophy, Brain Dev. 2 (1980) 359–369.