Cellular processing of the nerve growth factor precursor by the mammalian pro-protein convertases

Abstract

In order to define the enzymes responsible for the maturation of the precursor of nerve growth factor (proNGF), its biosynthesis and intracellular processing by the pro-protein convertases furin, PC1, PC2, PACE4, PC5 and the PC5 isoform PC5/6-B were analysed using the vaccinia virus expression system in cells containing a regulated and/or a constitutive secretory pathway. Results demonstrate that in both cell types furin, and to a lesser extent PACE4 and PC5/6-B, are the best candidate proNGF convertases. Furthermore, two processed NGF forms of 16.5 and 13.5 kDa were evident in constitutively secreting cell lines such as LoVo and BSC40 cells, whereas only the 13.5 kDa form was observed in AtT20 cells, which contain secretory granules. Both forms display the same N-terminal sequence as mature NGF, and were also produced following site-directed mutagenesis of the C-terminal Arg-Arg sequence of NGF into Ala-Ala, suggesting that the difference between them is not at the C-terminus. Co-expression of proNGF with furin and either chromogranin B or secretogranin II (but not chromogranin A) in BSC40 cells eliminated the 16.5 kDa form. Data also show that N-glycosylation of the pro-segment of proNGF and trimming of the oligosaccharide chains are necessary for the exit of this precursor from the endoplasmic reticulum and its eventual processing and secretion. Sulphate labelling experiments demonstrated that proNGF is processed into mature NGF following the arrival of the precursor in the trans-Golgi network. This comparative study shows that the three candidate mammalian subtilisin/kexin-like convertases identified process proNGF into NGF and that the nature of the final processed products is dependent on the intracellular environment.

Cellular processing of the neurotrophin precursors of NT3 and BDNF by the mammalian proprotein convertases

“…It is interesting to note that, similar to proNGF, the NT3 precursor contains a single Asn-glycosylation site in the prosegment, 8 amino acids N-terminal to its presumed processing site within the sequence AsnArgThrSerArgArgLysArglTyrAlaGlu. From the relative intensities of the secreted and the remaining precursor forms estimated by scanning the autoradiogram, we propose that the decreasing order of proNT3 processing by the convertases is similar to that for proNGF [18], viz. furin being the best candidate, followed by PACE4; PC5/6-B with PC1 and PC5 being also somewhat effective but PC2 is inactive.…”

“…The VV:hBDNF [21] and VV:hNT3 [10] using the full-length coding regions were obtained essentially as described [18,20]. The full-length cDNAs of hBDNF and hNT3 were generously provided by Regeneron.…”

“…The full-length cDNAs of hBDNF and hNT3 were generously provided by Regeneron. Cellular expression, biosynthetic labeling with [35S](Met + Cys), endoglycosidase H and N-glycanase digestions were performed as described [18,20].…”

“…Accordingly, studies on the colocalization of the neurotrophin mRNAs with those of the convertases and their coregulation revealed that more than one convertase can colocalize with the neurotrophins [17]. Indeed, biochemical analysis of the processing of proNGF by the five possible candidate convertases revealed that furin, PACE4 and PC5/6-B are the best proNGF processing enzymes in both constitutively secreting and regulated cells [18].…”

Functional analysis of human PACE4‐A and PACE4‐C isoforms: identification of a new PACE4‐CS isoform

“…Therefore, although PACE4-A is a secretable ,rotein, PACE4-CS is not detectable in the medium. Further~lore, digestion of the 77 kDa PACE4-CS with either endolycosidase H or N-glycanase F [25] revealed that this form is ,ensitive to both enzymes (not shown). In contrast, the se-,.…”