We have previously shown that CD39 undergoes limited cleavage and that inhibition of proteolysis results in a decrease in ATPase activity. The reduction in enzymatic activity correlated with a decrease in the fraction of full-length CD39 present in active membrane raft-localized oligomeric complexes. We exploited N-and C-terminal VP16-and V5-tagged CD39, both transiently and stably expressed in 293 cells, to further elucidate the role of cleavage in the regulation of CD39 processing and activity. To characterize the complexes generated by cross-linking, N-terminal VP16-tagged and C-terminal V5-tagged CD39 were co-expressed in 293 cells. Following crosslinking of membranes with DTSSP and immunoprecipitation with anti-V5, DTT-cleaved species were visualized by Western Blot using VP16 antibody. Interestingly, both VP16-tagged full-length and N-terminal fragments (30 kDa) were immunoprecipitated by anti-V5. This indicates that both full-length CD39 and the N-terminal cleavage fragment are present in raft-localized complexes. The composition of raft-localized CD39 complexes was studied by separating membrane fractions on a discontinuous sucrose gradient using a non-detergent method. When overexpressed, CD39 and its C-terminal fragment distribute across the gradient as visualized by Western with anti-VP16. Importantly, specific activity (expressed as ATPase activity divided by total CD39 content) was 8 times greater in low-density raft-enriched fractions than in high density raft-free fractions. In addition, relative ADPase activity was higher in fractions containing a higher proportion of C-terminal CD39 relative to full-length CD39. Thus, CD39 forms oligomeric complexes and possesses optimal enzyme activity in lipid rafts. The relationship between CD39 cleavage, ATPase activity and raft localization was further studied in 293 cells transfected with C-or N-terminal VP16-tagged CD39. Subcellular fractionation on a discontinuous sucrose gradient yielded membrane fractions enriched in endoplasmic reticulum (ER), early endosomes (EE) and plasma membrane/Golgi (PM-Golgi). Importantly, the EE fraction contained both full-length and C-terminal (or N-terminal) CD39 at the same level as seen in the PM-Golgi fraction, suggesting that near 50% of CD39 resides in the EE compartment. Furthermore, EE-expressed CD39 exhibited an ATPase and ADPase activity equivalent to that seen in Golgi-PM fractions. This led us to examine effects of NH4Cl and bafilomycin (which block acidification of EE), and chloroquine (blocks EE maturation) on CD39 cleavage, activity and raft localization. Each treatment inhibited CD39 cleavage and correspondingly decreased ATPase activity. A shift of ~50% of full-length CD39 from raft fractions to high density membrane fractions was observed upon sucrose gradient fractionation following chloroquine treatment of cells transfected with N-terminal VP16 tagged CD39. This redistribution of CD39 in the membrane correlated with a 40% decrease in ATPase activity and a striking inhibition of CD39 cleavage. Here, at a lower level of expression than cited above, ATPase activity in low-density raft fractions was ~100-fold greater than in high density fractions. Thus, cleavage of a portion of CD39 molecules is required for both raft localization of full-length CD39 and optimal enzyme activity. Regulated proteolytic cleavage of CD39 would allow for rapid upregulation of CD39 activity in response to alterations in cell environment. This would occur via cycling of CD39 between plasma membrane and endosomal compartments, the proposed site of CD39 cleavage and assembly of fully active oligomeric complexes.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author