The number of sites at which a protein can be readily cleaved by a proteolytic enzyme is greatly influenced by its three-dimensional structure. For native, properly-folded proteins both the rate of cleavage and number of sites at which cleavage takes place are usually much less than for the denatured protein. In order to compare the tertiary structure of recombinant HepG2 type glucose transporter with that of its native counterpart in the erythrocyte, the pattern of tryptic cleavage of the protein expressed in insect cell membranes was therefore examined. After 30 minutes digestion, a fragment of approximate Mr 19,000-21,000 was generated. In addition to this, there were two less intensely stained fragments of apparent Mr 28,000 and 17,000. The pattern of labelling was similar up to 2 hours of digestion. However, the fragments of Mr 19,000-21,000 and Mr 17,000 were no longer detectable after 4 hours digestion. The observation of a very similar pattern of fragments yielded by tryptic digestion of the HepG2 type transporter expressed in insect cells suggests that the recombinant protein exhibits a tertiary structure similar if not identical to that of its human counterpart. Also, the endogenous sugar transporter(s) present in Sf21 cells did not bind cytochalasin B, the potent transporter inhibitor. Therefore, the baculovirus/Spodoptera frugiperda (Sf) cell expression system could be very useful for production of large amounts of human glucose transporters, heterologously.