Animal Models of Memory Disorders Give Insight into How Psychological and Neural Systems Interact
Animal models have been key to our understanding of the psychological and physiological underpinnings of many disease states (see Carroll & Overmier, 2001). For example, in the field of learning and memory animals models have been instrumental in shaping our understanding of how the normal and damaged brain processes information. Animal research has taught us that there are multiple memory systems that interact: competitively, cooperatively or in parallel-depending on the cognitive demands and psychological nature of the task (see Gold 2002; Kim & Baxter, 2001; White & McDonald, 2002).
My colleagues and I, using a rodent model of Wernicke-Korsakoff Syndrome (WKS), have asked questions about the changes in psychological and neural processes that occur after specific brain damage. WKS is a nutritional disorder in humans associated with chronic alcoholism that, if left untreated, can cause severe amnesia. The pyrithiamine-induced thiamine deficiency (PTD) model successfully mimics both the neuroanatomical damage and behavioral impairments observed in WKS. However, while WKS is a somewhat heterogeneous disorder, patients vary in the degree and location of neuropathology (Knopelman, 1995). The behavioral symptoms of WKS were critical in establishing the theory of multiple memory systems. It is evident that not all memory functions are impaired by WKS or temporal lobe damage and animal models of these disorders were essential in determining the brain regions that caused certain types of memory impairment (see Squire, 1992). The PTD model has consistently demonstrated that damage to the thalamus (specifically the anterior and midline thalamus), a subcortical structure, is central to the loss of learning and memory function that occurs following thiamine deficiency (see Savage et al 1997; 1998, 1999). Thus, this model, as well as many others, has allowed us to ask and answer pertinent questions regarding changes in function and recover of function that occur under specific neurological disease states.
Interactions Between Memory Systems
Learning and memory are complex phenomenon requiring the coordinated interaction of multiple brain structures. The diencephalon, a subcortical region that includes the thalamus and hypothalamus, has been characterized as an integral connection zone for many memory-related circuits. There are connections between the thalamus and the hippocampus, as well as the amygdala and striatum. All three of those regions (hippocampus, striatum, amygdala) are important for different types of memory (Squire, 1992; White & McDonald, 2002). Although in both the human condition of WKS and the PTD model there appears to be no gross hippocampal neuroanatomical alteration, hippocampal-related behaviors appear to be impaired (Langlais, Savage & Zhang, 1996).
The use of animal models provides us with an unique opportunities to access neural function (i.e., neurotransmitter release) in one brain region after damage to another while the animal is cognitively challenged. Recently, we have used in-vivo microdialysis to observe the interactions between the hippocampus and the diencephalon using the PTD model while the animals are solving a maze task (Savage, Chang & Gold, 2003). This study demonstrated that the behavioral impairment produced by diencephalic damage (i.e., PTD-treatment) is accompanied by decreased release of acetylcholine (ACh) efflux in the hippocampus that is only evident under certain environmental conditions: PTD-treated rats display a reduction in hippocampal ACh efflux when they are navigating a maze and cognitively processing spatial information. In contrast, when PTD-treated rats are in a home cage (not cognitively challenged) they have normal hippocampal ACh levels. This study demonstrated the importance of the nature of the behavioral activity when assessing neurological dysfunction. Psychological challenges drive brain activation and understanding brain activity under different behavioral states will be critical for the development of therapeutics.
The goal of this research and some of our present research is to gain insight into what interconnected brain structures become down-regulated and which ones compensate after brain damage. Understanding the individual and interactive role of brain structures is critical for understanding normal memory function and for treating disorders of memory. Presently, we understand very little about the underlying psychological and neurobiological mechanisms associated with behavioral change and recovery of function after brain damage; however, the continued use of animal models will be key to our gaining further knowledge about such dynamics.
We have found that small manipulations of task contingencies can dramatically change behavioral outcome after brain damage (see Savage, 2001). Using the Differential Outcomes Procedure (DOP), developed by Trapold in 1970, which correlates to-be-remembered events with distinct reinforcement conditions, we have demonstrated in multiple different rodent models of memory disorders (WKS model, advanced age model, Alzhemier's model, temporal-lobe amnesia model) that this slight manipulation reduces or eliminates behavioral impairment (Savage & Langlais, 1995; Savage, Pitkin & Cariei, 1999; Savage, 2001; Savage, Buzzetti, & Ramirez, in press). Rats with damage to the diencephalon, aged rats and rats with pharmacologically-induced amnesia do not show the traditional memory impairments when trained with the DOP, but display impairments when trained with either a common reward or random rewards (a Nondifferential Outcomes Procedure [NOP]; see Savage 2001 for a review). For example, PTD-treated rats will require about twice as many trials as nondamaged control rats to learn a matching-to-position task (200 vs. 100 trials). This task requires rats to remember which spatial location they previously visited.
However, if each to-be-remembered item (i.e., spatial location) in the task is correlated with a unique reward outcome, PTD rats will perform as well as control rats, displaying no learning impairment at all (Savage & Langlais, 1995). Based on the results of Savage & Langlais (1995) a similar procedure was tested in humans. This study that demonstrated that correlating to-be-remembered faces with unique rewards (the DOP) enhances memory performance in WKS patients (Hochhalter, Sweeney, Savage, Bakke, & Overmier, 2001).
This functional difference (spared vs. impaired learning/memory performance) produced by manipulating reinforcement contingencies maps onto other multiple memory system approaches. One of the most widely used theoretical perspectives of the dual memory system approach is that of the explicit and implicit memory systems (Squire, 1992). We have applied this perspective to our data produced using the DOP as described below:
Explicit memory is the term used to describe the system that processes the intentional recall and recognition of people, places, things and events. A number of amnestic and demented populations (i.e., temporal lobe and diencephalic amnestics, Alzheimer's patients) are impaired on tasks that activate this memory system. Implicit memory, on the other hand, refers to the process of unintentional learning--one does not have to consciously attend to what one is learning. Classical conditioning and priming are examples of implicit memory processes. Many populations that display explicit memory impairment have normal implicit memory processing
We propose that the system tapped when the DOP (unique rewards correlated with to-be remembered events) is used with a task is the implicit memory system. This implicit memory system is functionally and structurally different from the explicit memory system that is used to solve memory problems when unique rewards are not used ([the NOP]; see Savage, 2001; Savage, Pitkin & Carieri, 1999; Savage, Buzzetti & Ramirez, in press). The explicit memory system is primarily activated when a common reinforcement procedure or the NOP is used within a matching-to-sample paradigm. Under these circumstances (no unique reward information), the subject must rely on its memory of what the sample stimulus was to solve the task successfully. Relative to those data, we hypothesize that the DOP produces reward expectancies-and the reliance on these reward-related processes activates a memory system that differs from the system commonly used to solve conditional discrimination tasks. Our recent data and those of others suggest the amygdala is important for solving a discrimination task when the DOP is used (Blundell, Hall, & Killcross, 2001) whereas the hippocampus is critical when it is not used (Savage et al, in press).
In summary, changing the type of cognitive information an animal can use to solve a task, by manipulating task variables, can recruit different brain regions-demonstrating behavioral and neural plasticity. Using such strategies, brain-damaged subjects can solve complex tasks in new ways without impairment. Understanding this type of behavioral plasticity is important for understanding recovery of function in general. Thus, animal models are critical in understanding cognitive dysfunctions and how brain structures interact in new ways after damage.
Blundell, P., Hall, G. & Killcross, S. (2001). Lesions of the basolateral amygdala disrupt selective aspects of reinforcer representation in rats. Journal of Neuroscience 21, 9018-9025.
Carroll, M.E. & Overmier, J.B. (2001). Animal research and human health: Advancing human welfare through behavioral science. Washington, DC, US: American Psychological Association.
Gold, P.E. (2002). Memory modulation: Regulating interactions between multiple memory systems. In: Squire, L.R. (Ed); Schacter, D.L. (Ed). Neuropsychology of memory (3rd ed.) (pp. 450-462). New York, NY, US: Guilford Press.
Hochhalter, A.K. *, Sweeney, W.A. *, Savage, L.M., Bakke, B.L., & Overmier, J.B. (2001). Using animal models to address the memory deficits of Wernicke-Korsakoff syndrome. In Carroll, M.E., & Overmier, J.B, (Eds). Animal research and human health: Advancing human welfare through behavioral science. (pp. 281-292). Washington, DC, US, American Psychological Association.
Kim, J.J. & Baxter, M.G. (2001). Multiple brain-memory systems: the whole does not equal the sum of its parts. Trends in Neurosciences, 24, 324-30.
Kopelman, M.D. (1995 ). The Korsakoff syndrome. British Journal of Psychiatry, 166, 154-73.
Langlais, P.J., Zhang, S.X., & Savage, L.M. (1996). Neuropathology of Thiamine Deficiency: An Update on the Comparative Analysis Of Human Disorders and Experimental Models. Metabolic Brain Disease, 11, 19-37.
Savage, L.M. (2001). In search of the neurobiological correlates of the differential outcomes effect. Integrative Physiological and Behavioral Science, 36, 182-195.
Savage, L., Buzzetti, R., & Ramirez, D. (in press). The effects of hippocampal lesions on learning, memory and reward expectancies. Neurobiology of Learning and Memory.
Savage, L.M., Castillo, R. & Langlais, P.J. (1998). Effects of thalamic intralaminar nuclei and internal meduallary lamina on spatial memory and object discrimination. Behavioral Neuroscience, 112, 1339-1352.
Savage, L.M., Chang, Q., & Gold, P.E. (2003). Diencephalic damage decreases hippocampal acetylcholine release during spontaneous alternation testing. Learning and Memory, 10, 242-246.
Savage, L.M. & Langlais, P.J. (1995). Differential outcomes attenuate memory impairments on matching-to-position following pyrithiamine-induced thiamine deficiency in rats. Psychobiology, 23, 153-160.
Savage, L.M., Pitkin, S.R., & Careri, J.M. (1999). Memory enhancement in aged rats: The differential outcomes effect. Developmental Psychobiology, 35, 318-327.
Savage, L.M., Pitkin, S. & Knitowski. (1999). Rats exposed to pyrithiamine- induced thiamine deficiency are more sensitive to the amnestic effects of scopolamine and MK-801: Examination of working memory, response selection, and reinforcement contingencies. Behavioural Brain Research, 104, 13-26.
Savage, L.M., Sweet, A.J., Castilo, R. & Langlais, P.J. (1997). The Role of Internal Medullary Lamina Nuclei and Posterior Thalamic Nuclei in Learning, Memory, and Habituation in the Rat. Behavioural Brain Research, 82, 133-147.
Squire, L.R. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans. Psychological Review, 99, 195-231.
White, N.M. & McDonald, R.J. (2002). Multiple parallel memory systems in the brain of the rat. Neurobiology of Learning & Memory, 77, 125-184.
About the Author
Lisa M. Savage received her doctoral degree in Psychology from the University of Minnesota in 1992. She is currently an Associate Professor of Psychology in the Behavioral Neuroscience program at Binghamton University, State University of New York. Her research program uses animal models to ask questions about learning and memory in normal and brain-damaged populations. She was awarded the 2002 APA Distinguished Scientific Award for Early Career Contribution to Psychology in the area of Animal Learning and Behavior/Comparative Psychology.