A cytosolic enzyme, betaine homocysteine methyltransferase (BHMT), and its partial fragments were discovered as autolysosomal membrane proteins from rat liver in the presence of leupeptin [Ueno et al. (1999) J. Biol. Chem. 274, 15222-15229]. The present study was undertaken to further characterize the transport and processing of BHMT from cytosol to autolysosome and to test if the fragment can be used as an in vitro probe for the maturation step of macroautophagy. Upon subcellular fractionation, BHMT (p44) was found in all fractions, while its 32-kDa fragment (p32) was found only in the mitochondrial-lysosomal (ML) fraction. Incubation of isolated hepatocytes with leupeptin induced time-dependent accumulation of p32 in the ML fraction from 30 to 90 min after the start of incubation. However, chloroquine completely inhibited the appearance of p32, indicating that the processing from p44 to p32 is lysosomal. Incubation with Bafilomycin A(1), a vacuolar H(+)-ATPase inhibitor, together with leupeptin, led to linear accumulation of p44, but not of p32. The p44 accumulation rate was calculated to be 4.9%/h, which was comparable to autophagic sequestration rate. The distribution of p44 within the ML fraction turned out to be dual, i.e., the membrane-surface attached and luminal/sedimentable forms. Amino acids and 3-methyladenine, both of which specifically suppress macroautophagy, inhibited the accumulation of p32 as well as of p44. Finally, energy-dependent appearance of p32 was demonstrated during incubation of postnucler supernatant fractions, making it possible to establish an in vitro assay system. All the results strongly support the idea that BHMT is taken up and degraded to p32 through the macroautophagic pathway, and that p32 could be a novel probe for the maturation of macroautophagy.