During extraction and purification, membrane proteins very often undergo denaturation and deactivation. To overcome this problem, the authors have tried to establish a better methodology to make the study at in vivo tissue level, not at the isolated cellular level, possible and easier. This is in vivo direct exposure of animal tissue to the liposome that contains an artificial boundary lipid (D14DPC, 1,2-dimyristamido-1,2,-deoxyphosphatidylcholine). Bullfrog and rat tongues were used. To confirm the reasonableness of this methodology, several different techniques were adopted; the nerve response study, gel electrophoretic analysis, quartz crystal microbalance (QCM) measurement and the affinity gelchromatography. When the tongue was exposed to the D14DPC-containing DMPC liposome, a significant amount of membrane protein was found in the recovered liposome (this was the production of proteoliposome). The nerve response in the neurophysiological measurement to several taste stimuli, such as L-alanine, L-leucine, sucrose and quinine hydrochloride significantly decreased when the tongue was exposed to the same liposome. These phenomena were common to both bullfrog and rat tongues. The nerve response to the stimulation with L-alanine was the most remarkably affected in the liposomal treatment. Therefore, the L-alanine-binding protein was focused upon to confirm the reasonableness of the QCM measurement and the affinity gelchromatography. The D14DPC-containing proteoliposome always showed significant binding to both the L-alanine affinity gel and the L-alanine-conjugated QCM. The results revealed that membrane proteins can be directly and effectively released, even from intact animal tissue epithelium, using the artificial boundary lipid-containing liposome.