Prefrontal Cortex - neuronal networks subserving working memory: The work of Patricia Goldman-Rakic

The prefrontal cortex (PFC) is the most evolved area of the primate brain, subserving our highest order cognitive abilities. The research of Patricia Goldman-Rakic showed that parallel sensory processing streams continue into the PFC, where they interact through extensive interconnections. Thus, there are domains for sensory space more dorsally in PFC, and domains for sensory features more ventrally. The projections are reciprocal,; the PFC projects back to sensory cortices to regulate provide “top-down” control.

The dorsolateral PFC surrounding the principal sulcus (Walker’s area 46) is essential for spatial working memory. Even small lesions to this area produce permanent deficits in visouspatial working memory performance (Goldman PS, Rosvold HE (1970) Localization of function within the dorsolateral prefrontal cortex of the rhesus monkey. Exp Neurol 27: 291-304). This PFC region receives highly processed visuospatial information from the parietal association cortex (area 7a), and the PFC in turn projects back to the parietal association cortex. Selemon and Goldman-Rakic (1988) showed that these two, interconnected regions project to many other brain regions in common, thus forming the networks for visual spatial cognition. These include areas involved in long-term memory formation (parahippocampal gyrus, presubiculum), emotional regulation (orbital and insular cortex), and movement (FEF, premotor cortices). Note that several of these cortical regions are part of what is now called the default network (e.g. cingulate cortices and RSC). There are also extensive shared projections to subcortical areas, e.g. the caudate, that are not shown in this diagram.

Goldman-Rakic described the fundamental ability of the PFC to represent information that is no longer in the environment, and to use these mental representations to guide behavior, thought and emotion. This representational knowledge is used to overcome distraction or prepotent responses, and to maintain goals in the face of interference. Goldman-Rakic discovered the neurobiological basis of mental representation in the dlPFC.

Monkeys performed a spatial working memory task in which a cue briefly appeared at 1of 8 locations while the monkey fixated on a central spot. The monkey had to remember the spatial position over a delay period of several seconds, so that it could move its eyes to the remembered position when the fixation spot disappeared. Recordings from the dlPFC revealed a variety of neurons with task-related firing, including those that fired to the cue, those that maintained firing across the delay period, and the those that fired before during or after the eye movement response. The Delay cells showed spatially-tuned, persistent firing across the delay period, firing for a preferred direction, but not other, nonpreferred directions, thus providing the neural representation of visual space.

The microcircuitryunderlying Delay cells involves two key factors: the persistent firing is generated by the recurrent excitation of deeo layer III pyramidal cells, exciting each other to maintain information “in mind”, while the spatial tuning arises from lateral inhibition from GABAergic interneurons (Basket and Chandelier cells).

The deep layer III microcircuits that generate mental representations are heavily afflicted in schizophrenia. There is reduced thickness of layer III of the dlPFC in patients with schizophrenia, and increased neuronal density corresponding to a loss of neuropil. These findings by Selemon and Goldman-Rakic were reinforced by the work of David Lewis’ lab, who found marked loss of dendritic spines from deep layer III pyramidal cells, and a compensatory weakening in the GABAergic interneurons. Based on the monkey physiology data, these findings suggest that patients with schizophrenia would have weaker ability to maintain information “in mind”, and that the information would be less clear. Indeed, patients with schizophrenia have impaired working memory and reduced dlPFC activity that correlates with symptoms of thought disorder, as predicted by Goldman-Rakic.

If you are interested:

• Goldman-Rakic PS (1994) Working memory dysfunction in schizophrenia. J Neuropsychiatry Clin Neurosci 6: 348-357.
• Selemon LD, Kleinman JE, ., Herman MM, Goldman-Rakic PS (2002) Smaller frontal gray matter volume in postmortem schizophrenic brains. Am J Psychiatry 159: 1983-1991.
• Selemon LD, Rajkowska G, Goldman-Rakic PS (1995) Abnormally high neuronal density in the schizophrenic cortex: A morphometric analysis of prefrontal area 9 and occipital area 17. Arch Gen Psychiatry 52: 805-818.
• Glantz LA, Lewis DA (2000) Decreased dendritic spine density on prefrontal cortical pyramidal neurons in schizophrenia. Archives General Psychiatry 57: 65-73.
• Gonzalez-Burgos G, Hashimoto T, Lewis DA (2010) Alterations of cortical GABA neurons and network oscillations in schizophrenia. Curr Psychiatry Rep 12: 335-344.