A recent imaging study suggested that damage to the insular cortex (IC) results in cessation of smoking in chronic cigarette users. However, the regions of interest designated as the IC in this imaging study included not only the IC but also the claustrum. The claustrum has been advanced as being involved in multimodal information processing, and as such could plausibly be involved in drug use, which has several components. In addition, the claustrum may be involved in attention. Unfortunately, functional assessment of the claustrum has been impeded by an inability to selectively ablate the claustrum without impinging on adjacent structures, including the IC, endopiriform nucleus, and striatum. In order to circumvent this problem by using a molecular lesioning strategy, we used matrix-assisted laser desorbtion ionization-time of flight imaging mass spectrometry (MALDI-TOF IMS) to search for proteins expressed in the claustrum but not adjacent sites. We identified the G protein gamma 2 subunit (Gng2) as such a protein, with no other significant expression of Gng2 in the forebrain. In subsequent anatomical studies we found that the claustrum of rodents and primates is immediately surrounded by deep layer IC cells, which give rise to the subcortical projections that were previously ascribed to the claustrum. We found that the claustrum itself projects only to cortical sites. Neurons projecting to various cortices were confined to certain regions within the claustrum, with the exception of neurons that innervate the anterior cingulate cortex, consistent with the claustrum being involved in attention. In order to test the hypothesis that the claustrum subserves key processes involved in substance abuse, we propose to generate transgenic mice expressing the toxic polyglutamine repeat protein ataxin-3 under the Gng2 promotor (Gng2-ataxin-3). The strategy of using expression of ataxin-3 to induce lesions has been previously exploited in transgenic mice, in which a gradual loss of neurons expressing the ataxin-3 transgene occurs over the first 12 postnatal weeks. After generating Gng2-ataxin-3 mice we will then confirm that claustral projections to cortical regions are lost but that the IC neurons surrounding the claustrum that project to subcortical sites are intact, indicating a specific lesion of the claustrum. We will screen the animals for a gross behavioral/physiological phenotype. In order to determine if the claustrum is involved in substance abuse, we will first determine if transgenic animals develop a conditioned place preference in response to nicotine or amphetamine. We will then determine if attention is disrupted by amphetamine or nicotine in the transgenic mice, using the 5-choice serial reaction time test. These studies will help clarify the functional attributes of the claustrum, a structure whose function remains largely unknown, and will determine if the claustrum is a key part of the extended circuit involved in certain aspects of substance abuse.