ABSTRACT
The process is obscure by which cobalamin (Cbl) in the endocytosed intrinsic factor (IF)-cobalamin (Cbl) complex is released and transferred to transcobalamin II (TCII) within the enterocyte. Using recombinant IF and TCII, binding of Cbl to IF at pH 5.0 was 70% of binding at pH 7.0, whereas for TCII alone, the value was only 12%. TCII binding activity was lost rapidly at lower pH, but this was not due to protease action. TCII incubated at pH 5.0 with cathepsin L was degraded and could not subsequently bind Cbl. Thus, transfer from IF to TCII is unlikely to occur within an acid compartment. Only 13-15% of bound Cbl was released at pH 5.0 and pH 6.0 from either rat IF, human IF, or human TCII. The K(a) of human or rat IF at pH 7.5 was 2.2 nM; for TCII, the value was 0.34 nM. At pH 7.5, Cbl transfers from IF to TCII, but only to a limited extent (21%), as detected by nondenaturing electrophoresis. Transfer of Cbl from IF to TCII could not be demonstrated at pH values of 5.0 or 6.0. Thus, luminal transfer of Cbl between IF and TCII is likely to be limited, but is possible. The most likely mechanism for intracellular transfer of Cbl from IF to TCII involves initial lysosomal proteolysis of IF, with subsequent Cbl binding to TCII in a more neutral cellular compartment.
ABSTRACT
Opossum kidney epithelial cells were shown previously to synthesize and secrete two cobalamin (Cbl)-binding proteins, presumed to be haptocorrin (Hc) and transcobalamin II (TCII). The present study examines the hypothesis that renal tubular cells also produce intrinsic factor (IF), and this production provides an explanation for the presence of IF in urine. By using antisera raised against human IF and against TCII, the presence of TCII was confirmed, and that of IF discovered in the media of opossum kidney (OK) cells in culture. The apparent molecular weight of IF and TCII was 68 and 43 kDa, respectively. Immunoreactivity on Western blot of the putative IF protein was blocked by recombinant human IF. When proteins secreted into the media were separated electrophoretically under nondenaturing conditions after binding with [(57)Co]Cbl, a broad major band migrated at a relative front independently of recombinant IF or TCII, and probably represents Hc, as the Cbl binding is blocked by cobinamide. Small amounts of bound [(57)Co]Cbl migrated in the position of both IF and TCII, when cobinamide was present. The presence of IF and TCII in OK cells was confirmed by immunohistology. Specific reactivity for IF (blocked by recombinant IF) was found in proximal tubules of opossum kidney, but not in other portions of the nephron, confirming the ability of anti-human IF antiserum to detect opossum IF. A 732-bp fragment of IF, nearly identical in sequence to rat IF, was isolated by RT-PCR from opossum kidney mRNA, and Western blot confirmed the presence of IF protein. The presence of IF was also documented in rat kidney by isolation of an RT-PCR fragment, immunocytochemistry, and Western blot. IF should be added to the list of renal (proximal) tubular antigens that are shared by other epithelia.
