Several diverse pieces of evidence suggest the possibility that auditory long-term memory may require the the oromotor system. The first set of findings comes from a series of studies carried out on the KE family, half of whose members suffer from an inherited speech and language disorder. The core deficit is in executing orofacial, especially articulatory, movements as a result of a mutation of the FOXP2 gene. The second line of evidence is monkeys are easily able to store visual and other sensory stimuli in long term memory, but are unable to do so with auditory stimuli. Although seemingly unrelated, these pieces of evidence from the human and animal studies suggest the following proposal. Because natural acoustic stimuli such as speech sounds fluctuate rapidly in time, it may be that their neural representations cannot be packaged for long-term storage in the sensory system alone, because the sensory system may not contain an integration time-window that is long enough to represent the full duration of the fluctuating stimulus. Consequently, packaging of such stimuli may require the aid of the oromotor system, which is uniquely organized to chain-link rapid sequences. The neural phenotype of this FOXP2 mutation is primarily characterized by a pronounced volume reduction in the caudate nucleus bilaterally. Since both speech and grammar involve the implicit acquisition of structured sequences, the developmental verbal/orofacial dyspraxia in the affected KE members may extend to impaired grammar learning. To determine whether the structurally abnormal cortico-striatal circuitry of KE members is associated with deficits in implicit acquisition of abstract combinatorial rules, we conducted an fMRI study of artificial grammar learning (AGL). Members of the KE family and controls subjects were trained on an auditory-verbal version of a finite-state grammar and made grammatical classifications on new words during fMRI. Control subjects performed above chance endorsing grammatical more frequently than ungrammatical words and low-similarity items more frequently than high-similarity ones. KE family subjects performed at chance-level, significantly worse than controls, showing reduced Grammaticality but not Similarity effects on endorsement rates. For controls subjects, Grammaticality-related fMRI activations were found in the superior occipital gyri and dorsal striatum; Similarity-related activations activations were seen in the right supramarginal gyrus and left posterior HVIIa Crus I/II. In contrast KE members showed hypoactivations for Grammaticality-related effects in right superior temporal gyrus, left supplementary motor area, bilateral central operculum/anterior insula, left paravermal HVI/HVIIa Crus I and HVIII; no functional abnormalities were observed for Similarity effects. KE subjects showed compromised implicit AGL and associated hypoactivation in structures fundamental to speech processing. Our results support the proposal that the cortico-striatal structural/functional abnormalities of the KE members give rise to a series of impairments that stem from, but go beyond the acquisition of articulate speech and impede the implicit acquisition of verbal structure. The combined behavioral-neural phenotype of KE members is consistent with deficits in the cortico-striatal habit system. While monkeys easily acquire the rules for performing visual and tactile recognition memory, they have extraordinary difficulty acquiring a similar rule in audition. Another striking difference between the modalities is that whereas bilateral ablation of the rhinal cortex (RhC) leads to profound impairment in visual and tactile recognition, the same lesion has no effect on auditory recognition memory. In our previous study, a mild impairment in auditory memory was obtained following bilateral ablation of the entire medial temporal lobe (MTL), including the RhC, and an equally mild effect was observed after bilateral ablation of the auditory cortical areas in the rostral superior temporal gyrus (rSTG). Neural correlates for recognition memory have been observed throughout the auditory and prefrontal cortex, defining a network of areas supporting auditory STM with parallels to that supporting visual STM. In order to test the hypothesis that each of the mild impairments observed was due to partial disconnection of acoustic input with the frontal cortex we examined the effects of a more complete auditory disconnection by combining the removals of both the rSTG and the RhC. We found that the combined lesion led to forgetting thresholds that fell precipitously from the normal retention duration of 30 to 40s to a duration of only 1 to 2s, thus nearly abolishing auditory recognition memory, and leaving behind only a residual echoic memory. In the primate auditory cortex, information flows serially in the mediolateral dimension from core, to belt, to parabelt. In the caudorostral dimension, stepwise serial projections convey information through the primary, rostral, and rostrotemporal (AI, R, and RT) core areas on the supratemporal plane, continuing to the rostrotemporal polar area (RTp) and the temporal pole. In addition to this cascade of corticocortical connections, the auditory cortex receives parallel thalamocortical projections from the medial geniculate nucleus (MGN). Previous studies have examined the projections from MGN to auditory cortex, but most have focused on the caudal core areas AI and R. We investigated the full extent of connections between MGN and AI, R, RT, RTp, and rSTG. Both AI and R received nearly 90% of their thalamic inputs from the ventral subdivision of the MGN (MGv; the primary/lemniscal auditory pathway). By contrast, RT receives only 45% from MGv, and an equal share from the dorsal subdivision (MGd). Area RTp receives 25% of its inputs from MGv, with additional inputs from multisensory areas outside the MGN. The MGN input to RTp distinguished this rostral extension of auditory cortex from the adjacent auditory-related cortex of the rSTG, which received 80% of its thalamic input from multisensory nuclei (primarily medial pulvinar). We identified complementary descending connections by which highly processed auditory information may modulate thalamocortical inputs. In accord with the laminar patterns evident in corticocortical connections these thalamocortical connections support a model in which AI and R lie at the same hierarchical level, but RT and RTp lie at a higher level, perhaps between that of the core and belt. These results demonstrate an expanded hierarchical model with a complexity which may well exceed the complexity of the primates ventral visual stream.
Showing the most recent 10 out of 20 publications
