Several members of the CLCA family have been reported to have tumor suppressive properties. Human CLCA2 is downregulated in breast cancer and restored expression inhibits tumor cell growth and promotes apoptosis. Functionally, CLCA proteins were first described as a family of plasma membrane Cl- channels that could be activated by calcium. Genetic and electrophysiological studies have supported this view. Structurally, the human CLCA2 protein is expressed as a 943 amino acid precursor that is later cleaved at ∼amino acid 680 to yield 100 kD N-terminal and 40 kD C-terminal glycoprotein products. Earlier investigations of transmembrane geometry suggested five membrane passes. However, our analysis with the more recently derived SMART algorithm predicts that a carboxy terminal 34 amino acid hydrophobic segment comprises the only transmembrane pass. To resolve this question, we raised an antibody against hCLCA2 and investigated the protein’s synthesis, localization, maturation, and topology. We show by a variety of methods that the N-terminal 100 kD product is released from the cell surface following cleavage while the 40 kD C-terminal product remains tightly associated with the plasma membrane. Cell-surface biotinylation and endoglycosidase H analysis revealed a 125 kD precursor confined to the ER and a maturely glycosylated 135 kD precursor at the cell surface by 48 hours post-transfection. By 72 hours, 100 kD N-terminal and 40 kD C-terminal cleavage products were detected at the cell surface but not in the ER. An antibody against the 100 kD product co-immunoprecipitated the 40 kD product, indicating a tight association after cleavage. Surprisingly however, the 100 kD product was shed into the medium or removed by acid washes while the precursor and 40 kD product were retained by the membrane. Ectodomain shedding was also detected for two remotely related CLCA proteins and is likely to be a general property of the family. Together, these data indicate that hCLCA2 is mostly extracellular with only a single transmembrane segment followed by a short cytoplasmic tail and is itself unlikely to form a channel. Thus, the precise function of this tumor suppressor remains open.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]