| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Psychiatry and Human Behavior, University of Mississippi Medical Center, Jackson, Mississippi 39216, USA
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSREFERENCES |
|---|
Key Words: animal model • AIDS • dementia • learning • memory • retention
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSREFERENCES |
|---|
Many neurochemical and neuropathological changes accompany HIV-1 disease (8). Several have been noted as possible causal or mediating factors in the development of ADC. These include, but are not limited to, quinolinic acid neurotoxicity (9), increased central nervous system (CNS) cytokine production (10, 11), toxic CNS effects of HIV-1 coat proteins (12), and permeability changes in the blood–brain barrier (13).
Inoculation with the LP-BM5 murine leukemia viruses causes profound immunosuppression (murine acquired immunodeficiency syndrome, or MAIDS) in C57Bl/6 mice (14–16). In later stages of MAIDS, mice show neurological symptoms and behavioral changes with some parallels to those occurring in AIDS (17). Like HIV infection, LP-BM5 infection is accompanied by a host of neurochemical changes. These include increases in brain concentrations of quinolinic acid (18), decreases in striatal met-enkephalin and substance P (19), and increases in brain concentrations of platelet-activating factor (20). The immunological and neurological parallels of LP-BM5 and HIV-1 infection indicate that LP-BM5-infected mice might model some aspects of ADC.
Sei and colleagues (17) reported that LP-BM5 produced deficits in spatial localization and memory in the Morris water maze (21), without gross neurological impairment or neuropathology. This was the first report in a nonprimate species that retroviral infection can produce cognitive or behavioral impairment not secondary to other effects of illness. To further describe the relationship between LP-BM5 infection and learning or behavioral deficits, we tested performance in a signaled two-way shuttle avoidance procedure.
Groups of mice were trained to avoid electric shock in this procedure 7 or 11 wk after LP-BM5 inoculation. The groups trained 7 wk postinoculation were tested again in the same procedure without shock 4 wk later to measure retention of the response. Thus, the experiment measured LP-BM5's effects on shuttle avoidance acquisition 7 and 11 wk postinoculation and retention 11 wk postinoculation.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSREFERENCES |
|---|
Apparatus
Avoidance training was conducted in two identical metal shuttle boxes equipped with wire grid floors for administering shock (Columbus Instruments, Columbus, Ohio). Each shuttle box had two chambers, separated by a metal barrier with an open doorway. A small incandescent light bulb was mounted on the top panel in each chamber. A speaker and sound generator were mounted above the barrier. During the experiment, the light and a 2.8 KHz tone signaled shocks. Shock intensity was set at 0.2 mA, based on values reported by Weinberger et al. (23). Infrared beams on both sides of the barrier registered crosses between chambers. Stimulus programming, shock delivery, and data recording were controlled by computer.
Procedure
Animals were moved to the testing room 2 h before each experimental session. All groups had 100 training trials, run 20 trials per day. At the start of a training session, a mouse was placed in the left chamber of each shuttlebox. The room lights were turned off and the program started. At the start of each trial, the tone and light in the shuttle box were switched on. If the mouse crossed to the other side of the chamber within 10 s, the light and tone were switched off and a 20 s intertrial interval (ITI) began. If the mouse did not cross within 10 s, shock was delivered either until he crossed or for 20 s. Shock was followed by a 20 s ITI. After 20 trials, the mouse was removed and returned to the home cage. Subjects were run in random order each day.
Design
Half the infected animals and half the controls began avoidance training 7 wk after inoculation. The rest of the animals began training 11 wk after inoculation. Animals trained 7 wk after inoculation were tested in retention trials 4 wk after the first block of 100 training trials. Retention trials were identical to training trials, except that shock was not delivered. Subjects trained 11 wk postinoculation received training trials only; substantial morbidity and mortality occurring 12–15 wk after LP-BM5 inoculation would likely have resulted in substantial and differential loss of subjects and made the results of retention trials difficult to interpret. Approximately 12 wk after inoculation, animals were killed and spleens were collected for a measure of LP-BM5-induced disease.
Analysis
To evaluate LP-BM5's effect on response acquisition, we calculated the percentage of subjects avoiding during each trial in the control and virus groups at both 7 and 11 wk. Avoidance responses were used rather than raw latencies, because responses that differ only slightly in latency in the active avoidance procedure can be qualitatively different. For example, a latency of 10.2 s in this experiment is an escape response where a shock is delivered, whereas a latency of 9.9 s indicates an avoidance response with no shock delivery.
Analysis of acquisition for all groups was based on the percentage of trials in which animals made avoidance responses during the first 100 trials. Performance in trials 101–200 was used as the measure of learning retention. We used the t test for paired observations to compare group performance. In this case, the paired observations were the percentages of subjects in each group making an avoidance response on each trial. Spleen weights were analyzed with the Kruskal-Wallis test, a nonparametric counterpart to the one-way analysis of variance. Spleen weight would not be expected to vary between the two groups of control animals; accordingly, control spleen weights were obtained from only one group. The 0.01 level of significance (two-tailed) was used for all tests.
A defect in the design of the chambers allowed subjects to avoid shock by standing on a threshold between the two compartments, which required us to discard the data obtained from the first five subjects in the 7 wk control and virus groups (10 subjects total). Chambers were modified to prevent the remaining subjects from avoiding shock in this manner.
RESULTS
Analysis of spleen weights showed that median spleen weight was increased approximately 15-fold in LP-BM5-infected animals compared to controls. There was no difference in spleen weight between the two infected groups. Data and statistical details are shown in 
Fig. 1.
|
The number of avoidance responses made by control and LP-BM5-infected animals trained 7 wk after inoculation increased over the 100 training trials. There was a marginally significant difference in the numbers of avoidance responses made by these groups during training (t99=-2.221, P<0.03), with LP-BM5-infected animals making a slightly greater number of avoidance responses. By the fourth day of training, performance in both groups appeared to be near asymptote. Data are shown in Fig. 2A
and in Table 1.
|
|
The numbers of avoidance responses by control and infected animals trained 11 wk after inoculation also increased over the 100 training trials. Although both groups learned to avoid shock, LP-BM5-infected animals made fewer avoidance responses than did controls (t99=3.469, P<0.001). Data are shown in 
Fig. 2B and in Table 1.
When animals trained 7 wk after inoculation were retested at 11 wk, group differences in performance were preserved: virus-infected animals made more avoidance responses than did the control animals (t99=-4.051, P<0.001). Data are shown in Fig. 2C and in Table 1.
Comparison of control groups trained at 7 and 11 wk postinoculation showed that performance did not differ as a function of postinoculation interval (t99=1.914, P<.07). However, virus-inoculated animals trained 11 wk postinoculation made fewer avoidance responses than did the virus-inoculated animals trained 7 wk postinoculation (t99=-3.571, P<0.001).
DISCUSSION
The results show an LP-BM5-induced deficit in active avoidance learning, and support two conclusions. First, LP-BM5 has a time-dependent effect on active avoidance learning. At 7 wk postinoculation, infected animals showed no learning deficit. However, learning was impaired in infected animals trained 11 wk after inoculation. Second, LP-BM5 affects acquisition, but not performance, of a well-learned avoidance response. Retesting of the animals trained 7 wk postinoculation showed that they actually made significantly more avoidance responses than controls.
Although group differences in total numbers of avoidance responses across all trials were significant, they are small when the data from all trials are considered. We believe that this is because the data include trials when performance was essentially at a floor or ceiling level—the first and last day's performances. These data are useful as indices of initial and asymptotic performance, but reveal little about the learning process as such. Examination of the data shows that the greatest group differences occurred during the middle of the training sequence, where one might expect the greatest amount of learning to occur. Future research will be directed toward closer examination of behavior and neurochemistry in this portion of the learning curve.
The performance of LP-BM5-infected animals in the retention trials shows that the learning deficit in the group trained 11 wk after inoculation did not result from general motor or sensory impairment. Such an impairment would also have caused a performance deficit when infected animals were tested in retention trials, because retention trials were also conducted 11 wk after inoculation.
These results have some interesting differences from those reported by Sei et al. (17). In contrast to the results reported here, learning impairment in the water maze was evident 7 wk after viral inoculation. One interpretation is that the water maze is more sensitive to LP-BM5-induced performance disruption than is the active avoidance procedure, detecting subtle deficits present in the early stages of disease. Alternatively, the difference in the time course of deficits in water maze and avoidance may represent selective deficits in specific abilities; for example, spatial learning in the water maze versus associative learning of the tone–light–shock relationship in the active avoidance procedure.
The different results obtained in the acquisition and performance components of the active avoidance task are reminiscent of clinical characteristics of ADC, and dementia more generally. In early stages of dementia, older learning is preserved but memory for recent material is impaired (24). For example, an early-stage patient may forget the location of keys or a wallet; only in later stages will a patient have trouble with such fundamental information as his or her name or occupation. The parallel can be drawn between these symptoms and the difficulty encountered by LP-BM5 infected mice learning active avoidance later in the course of disease in contrast to the well-preserved avoidance in infected animals trained earlier.
Previous reports have shown that quinolinic acid in CNS and plasma is elevated by LP-BM5 infection (18). Quinolinic acid neurotoxicity is a possible mechanism by which learning is impaired in active avoidance; studies in which quinolinic acid was administered i.c.v. showed that it impaired spatial learning in rats (25–27). Similarly, Popoli et al. (28) produced learning impairment in rats by administering quinolinic acid to the striatum 8 wk before testing animals in the Morris water maze. Another possibility is that performance impairment in the active avoidance procedure is related to increased levels of platelet-activating factor (PAF) caused by LP-BM5 infection (20). PAF levels are elevated in the frontal cortex and hippocampus 6 wk after LP-BM5 infection, but are not elevated in the striatum and cerebellum until 12 wk after infection. Thus, if active avoidance learning impairment is caused or mediated by PAF, this would suggest that the striatum and cerebellum are more important in the acquisition of avoidance than in the maintenance or performance of the well-learned avoidance response.
Overall, the results extend the use of the LP-BM5 infected mouse as a model system for the study of ADC. Previously reported data showed a specific deficit in spatial learning. The data reported here demonstrate that LP-BM5 induces a deficit in another type of learning, and show that the deficit is specific to the acquisition or consolidation component of that learning. It will be of interest in future studies to examine the effects of therapeutic interventions on acquisition in active avoidance, and to determine the extent to which LP-BM5-induced learning deficits occur in other tasks or types of learning.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Abbreviations: ADC, AIDS dementia complex; CNS, central nervous system; HIV-1, human immunodeficiency virus type 1; MAIDS, murine acquired immunodeficiency syndrome; DMEM, Dulbecco's modified eagle medium; ITI, intertrial interval; PAF, platelet-activating factor.
Received for publication January 9, 1997. Accepted for publication October 27, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. IIDA, K. SAITO, K. YAMADA, A. S. BASILE, K. SEKIKAWA, M. TAKEMURA, H. FUJII, H. WADA, M. SEISHIMA, and T. NABESHIMA Suppression of neurocognitive damage in LP-BM5-infected mice with a targeted deletion of the TNF-{alpha} gene FASEB J, May 1, 2000; 14(7): 1023 - 1031. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
